Leveraging sensor technologies for seed phenotyping by genebanks

Genebanks serve as critical repositories for preserving the genetic diversity of plant species, including crops, forages, and their wild relatives, which is essential for adapting to climate change, enhancing food security, and improving agricultural sustainability. Seed phenotyping, the process of evaluating observable seed traits influenced by genetics and environmental factors, plays a pivotal role in characterizing and utilizing this diversity. Traditional phenotyping methods, however, are labor-intensive and inadequate for the vast collections housed in genebanks. This paper explores the transformative potential of high-throughput phenomics technologies, leveraging the electromagnetic spectrum—from gamma rays to radio waves—to enable rapid, precise, and non-invasive assessment of seed traits such as size, shape, biochemical composition, and vigor. We highlight the integration of advanced imaging systems (e.g., hyperspectral, X-ray, and thermal imaging) to enrich genebank datasets, facilitating trait discovery and crop improvement. Despite challenges like cost, scalability, and data standardization, opportunities arise from collaborative initiatives between genebanks and phenomics facilities through organizations such as International Plant Phenotyping Network. Our conclusions underscore how phenomics can revolutionize genebank operations, ensuring the efficient conservation and deployment of genetic resources to address global agricultural demands.

1 Introduction

Genebanks act as repositories for the genetic material of various plant species, including crops, forages, and their wild relatives. This conservation is crucial for maintaining the genetic diversity necessary for the adaptation and evolution of plant species. Genetic diversity within plant populations is vital for resilience against diseases, pests, and changing environmental conditions. Seed phenotyping and genebanks are pivotal components in the field of agriculture and plant science. Correct and reliable seed phenotyping at genebanks plays a crucial role in conserving plant genetic diversity, improving crop varieties, and ensuring food security (Bhattacharya and Mummenhoff, 2024; Ghamkhar et al., 2024).

Seed phenotyping is the process of assessing the observable characteristics of seeds influenced by their genetic makeup and environmental factors. It enables researchers and breeders to identify and select traits beneficial for crop improvement. Traits such as drought tolerance, disease resistance, and enhanced nutritional content are essential for developing improved crop varieties that meet specific agricultural needs. By utilizing the data obtained from seeds stored in genebanks, breeders can assess and estimate some traits that will translate to whole plants in order to enhance agricultural productivity and sustainability (Mascher et al., 2019; Nguyen and Norton, 2020).

The genetic traits preserved in genebanks are indispensable for developing crop varieties resilient to climate change, pests, and diseases. Rapid identification and incorporation of beneficial traits into crops are crucial for adapting to the unpredictable and adverse effects of climate change (Hoffmann and Sgrò, 2011; Chapman et al., 2012). However, the vast amounts of genetic material held in genebanks, often comprising thousands of accessions for a species, present a significant challenge for traditional seed phenotyping methods (Hay and Sershen, 2021). Should a genebank use these traditional phenotyping methods, they would need hours of labor to achieve their phenotyping goals. Fast and high-throughput phenotyping technologies enable the rapid assessment of large numbers of seeds, accelerating the process of identifying valuable traits.

The sheer volume of accessions in genebanks and the need for frequent vitality tests and seed quality assurance required, necessitates efficient and scalable seed phenotyping methods (Ouyang et al., 2024; Wonggasem et al., 2024). High-throughput phenotyping platforms can process thousands of samples simultaneously, providing comprehensive data on seed traits (Colmer et al., 2020; Liu et al., 2024). This capability is crucial for managing and exploiting the extensive genetic diversity stored in genebanks. Advanced phenotyping technologies, such as imaging systems, sensors, and automated data analysis, offer greater precision and accuracy compared to manual methods. These technologies can capture detailed phenotypic information, including seed size, shape, color, and health, with high reproducibility (Zhang et al., 2018; Zhang and Zhang, 2018). Such detail is essential for making informed decisions in breeding and conservation efforts.

The acceleration enabled by high-throughput phenotyping of seeds is crucial for developing new crop varieties that can meet the urgent demands of food production and environmental sustainability. Furthermore, fast phenotyping generates large datasets that can be integrated with genomic information to gain deeper insights into the genetic basis of seed and seedling traits (Colmer et al., 2020). This integration supports data-driven research and innovation in genebanks, leading to more effective strategies for crop improvement and conservation (Ghamkhar et al., 2024).

2 Traditional and current seed phenotyping methods

Traditional seed phenotyping methods encompass a range of manual, visual and/or destructive methods. Among these methods are the 100-seed weight measure, germination tests, tetrazolium staining, seed dimension measurement as well as other methods such as cut-tests and oven drying methods. There are also well established but non-traditional methods, which are early stages of non-destructive approaches.

2.1 100-seed weight

The 100-seed weight method is one of the most widely used traditional seed phenotyping approaches and provides a simple yet informative measure of seed size, density, and overall seed lot quality. In this method, a random sample of one hundred seeds is weighed, usually after equilibrating moisture content, and the resulting value is used as an indicator of seed development, uniformity, and potential vigor (Nautiyal, 2009). Because seed weight correlates with nutrient reserves and early seedling growth (Vaughton and Ramsey, 2001), it is often used in breeding programs and genebank characterization to compare accessions or detect deterioration in storage. Despite its usefulness, the method is labor-intensive, destructive at large scales, and limited in its ability to capture within-sample variability or internal defects.

2.2 Germination tests

The germination test is a cornerstone of traditional seed phenotyping and remains the most widely used method for assessing seed viability and vigor. In this test, seeds are placed under controlled temperature, moisture, and light conditions, and the proportion that successfully produce normal seedlings within a defined period is recorded. Because it directly measures a seed’s ability to initiate growth, the germination test is considered a biological gold standard for viability assessment in both genebanks and seed testing laboratories. However, it is time-consuming, often requiring days to weeks, labor-intensive, and inherently destructive, making it inefficient for large genebank collections (Hay and Whitehouse, 2017). Moreover, germination outcomes can be influenced by dormancy, subtle pathogen loads, or suboptimal test conditions, complicating interpretation (Finch-Savage and Leubner-Metzger, 2006).

2.3 Tetrazolium staining

The tetrazolium (TZ) staining test is a long-established biochemical method used to assess seed viability by detecting the metabolic activity of living tissues (Verma and Majee, 2013). In this procedure, seeds are cut or pierced to expose the embryo and then soaked in a tetrazolium chloride solution, where viable cells reduce the colorless compound to a red formazan pigment. The resulting staining pattern enables trained analysts to distinguish living, damaged, and non-viable tissues, making the method especially valuable for species with deep dormancy or slow germination, where standard germination tests may underestimate viability (França-Neto and Krzyzanowski, 2022).

2.4 Seed dimension measurement

Seed dimension measurement, typically involving manual assessment of seed length, width, and thickness using calipers or simple imaging tools, has long been a fundamental component of traditional seed phenotyping. These measurements provide basic morphological descriptors that are useful for taxonomic identification, detecting varietal purity, monitoring seed lot uniformity, and understanding traits related to seed dispersal or early seedling performance. However, manual dimension measurement is slow, labor-intensive, and often limited to a small number of seeds per batch, which reduces its ability to capture within-lot variability. It also provides only coarse information about external shape and cannot detect internal defects or subtle morphological differences.

2.5 Cut-tests and oven drying

Traditional seed quality assessments also often rely on cut-tests, oven-drying, and manual X-ray evaluation, each providing important but limited insights into seed integrity and composition. Cut-tests involve physically slicing seeds to visually inspect the embryo and endosperm, allowing rapid identification of empty, damaged, or insect-infested seeds, but at the cost of destroying the sample (Bhattacharya and Mummenhoff, 2024). Oven-drying methods, used to determine seed moisture content or to validate fill percentage, require controlled drying conditions and multiple handling steps, making them slow and labor-intensive.

2.6 Non-traditional but well-establised methods

Manual X-ray evaluation, although a more advanced non-destructive technique, still depends heavily on human interpretation of radiographs to detect internal defects such as emptiness, cracks, or malformed embryos, which introduces subjectivity and limits throughput. MRI and X-ray are well-established but not traditional methods; they represent early generations of non-destructive imaging technologies.

As genebanks and breeding programs increasingly require scalable, high-resolution phenotyping, manual dimension measurements are being complemented or replaced by automated imaging systems. These systems extract size, shape, and morphological features rapidly and non-destructively across thousands of seeds, offering far richer datasets. Collectively, traditional approaches remain valuable reference methods in genebanks and seed testing laboratories, but their destructive nature, operator dependency, and low scalability highlight the need for automated, high-throughput phenotyping technologies that can deliver consistent and objective assessments across large seed collections.

3 Using the electromagnetic spectrum for seed phenotyping

The electromagnetic spectrum encompasses a range of electromagnetic waves, classified by their wavelengths/frequencies, from gamma rays to radio waves (Figure 1). Each range of the spectrum interacts uniquely with matter, making it a versatile tool for analyzing seed and plant traits in genebanks. By leveraging various regions of the spectrum, phenotyping technologies can non-invasively assess physical, physiological, and biochemical traits, enhancing the efficiency and precision of seed characterization (Table 1).

Figure 1

Figure 1. The electromagnetic spectrum.

Table 1

Table 1. Electromagnetic range and their use in seed phenotyping.

Shorter wavelengths, such as gamma rays and X-rays, penetrate deep into seeds and are particularly useful for internal structural analysis. Gamma rays have historically been employed to induce mutations for breeding programs (Beyaz and Yildiz, 2017), while X-rays provide high-resolution imaging to evaluate embryo integrity and detect cracks or voids within seeds (Indore et al., 2022; Hamdy et al., 2024). Ultraviolet (UV) light, situated just beyond the visible spectrum, is utilized for sterilization purposes and chemical activation processes, offering applications in seed treatment and pathogen control (Arcos-Limiñana et al., 2025) (Figure 1) as well as general assessment of seed integrity.

In the visible light range, RGB imaging is widely used to capture basic physical traits like color, size, and shape. When the light used is a laser, reflectance imaging can also provide information on the viability of the seeds with a precision of the order of magnitude of the wavelength of the laser used (Zdunek et al., 2014). Expanding into the near-infrared (NIR) and shortwave infrared (SWIR) regions, hyperspectral and multispectral imaging detect subtle variations in biochemical composition, such as moisture, protein, and carbohydrate content (ElMasry et al., 2019), as well as signs of disease or stress in seeds (Vrešak et al., 2016) and structural information. Moving further along the spectrum, thermal infrared imaging detects heat signatures associated with metabolic activity, making it invaluable for assessing seed vigor and germination potential (Kranner et al., 2010) (Figure 1).

Longer wavelengths, including terahertz (THz) (Ghamkhar, 2023; Wang et al., 2024) and radio waves (Zhong et al., 2021), provide emerging possibilities for seed phenotyping at genebanks. Terahertz imaging offers non-invasive analysis of internal seed structures (Hui et al., 2018), moisture distribution (Fu et al., 2024) and early stress detection (Gua et al., 2013; GE et al., 2014; Ghamkhar, 2023) while radio waves, particularly through Wi-Fi imaging, present potential for real-time monitoring of seed health during storage and transport. These technologies are particularly appealing for their ability to penetrate layers and provide data without damaging seeds.

By utilizing the electromagnetic spectrum, genebanks can capture a comprehensive range of seed traits, from physical integrity to genetic and biochemical properties. The diverse applications of these technologies ensure that genebanks can maintain high-quality germplasm collections, enabling efficient conservation and utilization for breeding and agricultural innovation.

4 Sensor technologies for non-destructive seed phenotyping

Phenomics has emerged as a transformative approach for enriching genebank data, enabling precise, high-throughput assessment of seed and plant traits. Traditional genebank evaluation methods often relied on labor-intensive, time-consuming processes that limited the scope of characterization. By integrating advanced imaging and sensor technologies, phenomics provides scalable and automated solutions that generate comprehensive datasets, essential for breeding (Dwivedi et al., 2020; Ghamkhar, 2023), conservation (Lürig et al., 2021; Fernando et al., 2023), and research (Fernando et al., 2023). These advancements allow genebanks to assess large collections efficiently while offering deeper insights into the traits that underpin agricultural productivity and resilience.

4.1 Optical and imaging sensors

One of the most significant contributions of phenomics is the adoption of sensor technologies across the electromagnetic spectrum. Visible light (RGB) imaging is the standard digital imaging in the visible spectrum. It can be used to analyze seed morphology, color, and texture as well as for seed purity monitoring and detection of seed contamination. RGB imaging is Inexpensive and easy to implement and can be integrated into automated systems for high-throughput analysis. The downside of RGB imaging is its limitation to physical traits only. RGB imaging captures high-resolution images of seeds to assess seed physical traits (Lurstwut and Pornpanomchai, 2011; Mahajan et al., 2015; Zhang et al., 2018). By employing advanced algorithms, RGB imaging systems can automate the measurement of individual seed dimensions and classify them based on uniformity and appearance (Punthumast et al., 2012; Morales et al., 2024).

In addition to RGB imaging, multispectral and hyperspectral imaging provide enhanced capabilities for analyzing physical seed traits (ElMasry et al., 2019; Mortensen et al., 2021). These tools capture data across multiple wavelengths, allowing for the detection of subtle variations in surface texture, color intensity, and potential physical damage (Feng et al., 2019; Gaikwad and Tidke, 2022). Multispectral imaging is widely used for assessing physical seed traits such as size, shape, and color, which are critical for determining uniformity and purity (ElMasry et al., 2019, ElMasry et al., 2020; Li et al., 2021; Tu et al., 2021). Hyperspectral imaging, which captures data across a broad range of wavelengths, is particularly valuable for detecting biochemical markers linked to seed health, genetic purity, and quality traits such as protein content (Dumont et al., 2015; Ambrose et al., 2016a) or disease presence (Dumont et al., 2015; Lee et al., 2017). These tools provide unparalleled accuracy and resolution, enhancing the reliability of seed characterization (see section…. below).

4.2 3D and structural imaging technologies

Tools and approaches such as 3D imaging technologies are used to capture seed structural traits (Jahnke et al., 2016), while CT scanning allows for detailed internal seed analysis (Gargiulo et al., 2020; Kunishima et al., 2020). By combining these technologies, phenomics not only enriches genebank datasets but also facilitates the prioritization of accessions with traits like nutritional quality, ensuring that genebanks remain vital resources in addressing global agricultural challenges.

The 3D imaging technique uses structured light or laser scanning to create 3D models of seeds (Huang et al., 2022; Qin et al., 2022). Genebanks can use this method to measure seed size, shape, and volume accurately. It is also a good tool to assess seed uniformity and detect physical defects. 3D images are suitable for morphological characterization and trait standardization and enable digital archiving of seed phenotypes. If genebanks choose 3D imaging for seed screening, they will require significant data storage and processing capabilities.

4.3 X-ray and wi-fi imaging tools for embryological and structural analysis

X-ray imaging has become an indispensable tool for non-invasive structural analysis of seeds, offering insights into embryological health and potential physical damage (Arkhipov et al., 2019). X-ray imaging complements RGB and spectral imaging methods by enabling non-invasive internal structural analysis, such as identifying cracks, voids, or malformed embryos within seeds (Xia et al., 2019; França-Silva et al., 2023). X-ray imaging further complements spectral imaging techniques by assessing structural integrity, which correlates with physiological performance (Arkhipov et al., 2019; Medeiros et al., 2020). This technology allows genebanks to assess seed viability by visualizing internal structures, such as the embryo and endosperm, without the need for destructive sampling. X-rays are particularly effective in identifying cracks, voids, or malformations within the seed, ensuring the preservation of valuable germplasm. The ability to analyze seed batches rapidly and accurately makes X-ray imaging a vital component of quality control in genebanks, enabling researchers to maintain seed integrity and optimize storage strategies.

X-ray creates detailed images of seed internal structures, helping identify physical damage or abnormalities caused by microbial infestations (Arkhipov et al., 2019; Nadimi et al., 2022). ​It helps the user to evaluate internal seed morphology (e.g., embryo and endosperm development) (Silva et al., 2013; Ahmed et al., 2018). Detecting physical damage, voids, or abnormalities as well as assessment of seed viability and vigor indirectly through structural health are other enablers of genebanks (Whitehouse et al., 2020). Additionally, X-ray imaging is useful for assessment of the embryological status, allowing the identification of seeds with optimal endosperm and embryo development. This non-invasive, reliable and easy to automate technique is very suitable for high-throughput analysis of genebank data. Phase-contrast X-ray provides contrast not only on dry sample but also on wet samples (Rousseau et al., 2015). On the downside, safety protocols are needed for handling X-ray equipment.

Wi-Fi imaging, a more recent development, represents a novel approach to structural and biochemical seed analysis (Roitsch et al., 2019). By utilizing microwave frequencies, Wi-Fi imaging has the potential to penetrate seed layers to detect internal anomalies and moisture distribution. This technology is less resource-intensive than methods like MRI, making it a cost-effective alternative for genebanks with limited budgets. Wi-Fi imaging also holds promise for real-time monitoring of seed health during storage and transport, offering dynamic insights into changes in seed composition over time. While X-ray imaging is a much more mature and ready-to-use technology, together, X-ray and Wi-Fi imaging push the boundaries of what is possible in embryological and structural analysis, providing genebanks with cutting-edge tools to ensure the longevity and quality of their collections.

4.4 Spectroscopy-based sensors

Phenomics tools play a crucial role in ensuring seed genetic purity by providing high-throughput, non-destructive methods for evaluating seed traits. These tools utilize advanced technologies like near-infrared (NIR) spectroscopy, and machine vision to detect subtle phenotypic variations that indicate genetic differences. This spectroscopy tool has been in play for decades now and uses the near-infrared region of the spectrum to assess chemical properties of seeds (Fassio and Cozzolino, 2004; Font et al., 2006). Genebanks can use this method to detect seed quality traits such as protein, oil and starch content as well as seed moisture and contaminants (Agelet and Hurburgh, 2014; Reddy et al., 2022). This rapid and non-destructive technique is suitable for high-throughput seed screening and analysis although it requires calibration models specific to seed types (Baye et al., 2006; Font et al., 2006).

Fourier-transform infrared (FTIR) spectroscopy and Raman spectroscopy are powerful vibrational spectroscopic techniques that enable rapid, non-destructive characterization of water content and chemical composition in seeds (Ambrose et al., 2016b). FTIR measures the absorption of infrared light associated with molecular vibrations, particularly those involving O–H, C–H, C=O, and N–H bonds (Berthomieu and Hienerwadel, 2009). Because water molecules exhibit strong and distinct IR absorption peaks, FTIR provides a sensitive means of quantifying seed moisture and monitoring hydration dynamics during imbibition or storage (Upretee et al., 2024). Additionally, FTIR can detect lipids, proteins, carbohydrates, and secondary metabolites through their characteristic spectral signatures (Hssaini et al., 2021). With appropriate chemometric modelling, FTIR has been used to estimate oil content, protein levels, seed ageing status, and other quality traits across a wide range of crop species (Rahman and Cho, 2016). Its high throughput, minimal sample preparation, and ability to analyze intact seeds make it highly suitable for genebank and breeding applications (Aftab et al., 2025).

Raman spectroscopy, in contrast, measures inelastic scattering of monochromatic light (usually from a laser) to provide a complementary molecular fingerprint that is less affected by water (Smith and Dent, 2019). This makes Raman particularly valuable for studying internal biochemical changes in seeds without interference from moisture, enabling reliable assessment of oil composition, carotenoids, phenolics, and other chemically informative compounds (Sanchez et al., 2019). Raman microscopy can also resolve spatially localized biochemical differences, such as the distribution of storage proteins or lipids within specific seed tissues, offering insights that bulk methods cannot provide (Saletnik et al., 2021). Recent advances in handheld Raman systems and surface-enhanced Raman spectroscopy (SERS) have further expanded its potential for rapid seed screening, including early detection of seed ageing, pathogen contamination, or varietal identity (Weng et al., 2021).

Together, NIR, FTIR and Raman spectroscopy represent versatile tools for modern seed phenotyping by providing rapid, high-resolution chemical profiling without compromising seed viability. These techniques allow researchers and genebanks to move beyond traditional destructive assays toward scalable, predictive, and data-rich approaches to seed quality assessment (Xia et al., 2019). Their integration with machine learning enables automated classification of viability, moisture status, or compositional traits, positioning all three technologies as cornerstones for future high-throughput phenotyping platforms (Kolašinac et al., 2025).

4.5 Hyperspectral imaging

multispectral and hyperspectral imaging provide valuable insights into seed vigor and stress tolerance by detecting biochemical markers such as chlorophyll, carotenoids, and proteins (Wakholi et al., 2018; ElMasry et al., 2019). These markers are indicative of a seed's physiological state and its capacity to withstand adverse environmental conditions (Bewley and Black, 2012; Corbineau, 2012). For example, hyperspectral imaging can detect early signs of stress by identifying spectral changes associated with oxidative damage or nutrient deficiencies. HSI is also used to assess seed composition (e.g., protein, starch, oil, or moisture content) (Dumont et al., 2015; Ambrose et al., 2016a) and potentially help in understanding structure of seeds. It can also be used to detect seed viability, health, and damage (e.g., fungal infections, mechanical damage) (Feng et al., 2019; Ferreira et al., 2024). It can differentiate between seed types and varieties, too (Tan et al., 2019; Zhu et al., 2019). The non-destructive and highly precise nature of this technology along with providing both physical and chemical data in a single scan are the advantages of HSI. The technology requires expertise in data analysis and interpretation and high data storage capacity.

Hyperspectral and multispectral imaging are among the most effective tools for analyzing seed content (Su and Sun, 2018), as they capture spectral data across a wide range of wavelengths, enabling the detection of biochemical markers like proteins, starch, lipids, and antioxidants. For instance, hyperspectral imaging can non-invasively predict amylose content in grains or identify seeds with high oil content (Mahajan et al., 2015; Akewar and Chandak, 2023), providing valuable insights for both breeding and quality control. Similarly, these tools can detect moisture content variations, which are critical for determining seed storability and germination potential (Mahajan et al., 2015; Wang et al., 2021). Multispectral and hyperspectral imaging are widely employed to detect subtle spectral differences associated with seed infections or contamination (Jaillais et al., 2015; Feng et al., 2019). These tools can identify early signs of fungal or bacterial pathogens by analyzing specific wavelengths that correspond to biochemical changes caused by disease (Olesen et al., 2011; da Silva et al., 2021; Singh et al., 2021). HSI also shows biochemical and morphological differences between genetically pure seeds and contaminants (ElMasry et al., 2019; Dhanya et al., 2024).

4.6 Magnetic resonance imaging

This technology uses magnetic fields and radio waves to visualize internal seed structures (Foucat et al., 1993; Tuomainen et al., 2022). MRI would be the perfect tool to study water content and movement in seeds during imbibition and germination (Terskikh et al., 2005; Garnczarska et al., 2007). It can also be used to assess seed dormancy and viability (Gruwel et al., 2002). The advantage of using MRI in genebanks is its high resolution, non-invasive internal imaging capacity but its high cost and complexity make its adoption difficult for genebanks.

4.7 Thermal, fluorescence, and novel sensors

Thermal (Kim et al., 2013; Men et al., 2017) and terahertz imaging (GE et al., 2014; Hu et al., 2024a) extend these capabilities by evaluating physiological qualities like germination capacity, vigor, and moisture content. Technologies like transcriptome analysis and epigenomics complement phenotyping by identifying genetic and functional variations associated with traits such as seed longevity, stress tolerance, and germination. For example, studies on DNA methylation have revealed its role in germination and viability (Mira et al., 2020), particularly in recalcitrant seeds (Ge et al., 2025), while metabolomics has been instrumental in predicting seed quality traits like amylose content and biochemical composition (Wang et al., 2022; Hu et al., 2024b). These advancements bridge the gap between phenotypic observations and underlying genetic mechanisms, providing actionable insights for crop improvement.

Thermal imaging is a key tool for measuring metabolic activity during seed germination, as it detects heat signatures associated with water uptake and respiration processes (Kranner et al., 2010; ElMasry et al., 2020). By analyzing temperature variations, researchers can quickly identify viable seeds and predict germination rates (Men et al., 2017). Infrared imaging (IRI) measures the heat emitted by seeds (Kranner et al., 2010; ElMasry et al., 2020). Using this system enables the genebanks to monitor seed water content and its drying process while also detecting metabolic activity in seeds during germination tests (Xia et al., 2019). The IR cameras also help in assessment of seed viability and stress responses. The non-destructive and sensitive IR helps detect metabolic changes early and at seed stage. It should be considered that the use of IR is only limited to germination and drying studies though.

Thermal imaging also detects temperature variations caused by metabolic activity, which can indicate the presence of pathogens or pests within seeds (Liu et al., 2020; Wen et al., 2023).

Fluorescence imaging technology detects fluorescence emitted by seeds after exposure to specific light wavelengths (Barboza da Silva et al., 2021; Bartolić et al., 2022). It can be used to identify seed viability and vigour by detecting metabolic compounds (e.g., chlorophyll or secondary metabolites) (Galletti et al., 2020; Batista et al., 2022). This imaging technique also enables the genebanks to assess seed health, including fungal or pest contamination (Pasikatan and Dowell, 2001; Bartolić et al., 2022). Fluorescence imaging is highly sensitive to physiological and biochemical changes. This technique may require specialized equipment and expertise. Chlorophyll fluorescence can also reveal metabolic differences linked to genetic variation helping seed genetic purity detection.

Ultrasonic sensors use sound waves and can be used to probe the physical properties of seeds (Sattar and Laila, 2025). These sensors can detect internal voids or fractures in seeds and help genebanks assess seed hardness and physical integrity in a rapid and non-destructive manner. However, this technique is less common and may require development for specific applications.

Terahertz time-domain spectroscopy (THz-TDS) is emerging as a powerful non-destructive tool for probing the internal physical and chemical properties of seeds, offering unique capabilities that complement optical and infrared methods. Terahertz imaging enables the assessment of internal moisture distribution, a critical factor for both germination and longevity (Lei et al., 2022; Hu et al., 2024a). Operating in the 0.1–10 THz frequency range, THz-TDS measures the time-resolved transmission or reflection of ultrafast terahertz pulses as they interact with seed tissues (Rawson and Sunil, 2022). Because terahertz waves are highly sensitive to polar molecules—particularly water—as well as structural discontinuities such as voids, cracks, and embryonic defects, THz-TDS can reveal information related to seed moisture content (Ge et al., 2023), density, and internal integrity (Afsah-Hejri et al., 2019). Its strong response to bound and free water makes it especially suited for characterising hydration dynamics, monitoring desiccation processes (Pal and Chattopadhyay, 2021), and assessing early viability loss associated with membrane deterioration or changes in internal microstructure.

Although still developing as an applied technology in seed science (Jiang et al., 2024), advances in compact terahertz sources, improved signal acquisition, and machine-learning-based spectral interpretation are rapidly increasing its practicality. As a result, THz-TDS is poised to become an important addition to the non-invasive phenotyping toolbox for genebanks and seed researchers aiming to evaluate seed quality at scale.

Advanced imaging technologies, such as THz imaging and thermal imaging, further enhance seed content quality analysis (ElMasry et al., 2020; Ge et al., 2021). Terahertz imaging enables non-invasive assessment of storage-related degradation (Han et al., 2024; Li et al., 2025). Further, THz scans detecting anomalies may signify genetic impurities in seeds.

5 Seed quality traits measurable through sensor technologies

Seed quality traits are critical indicators of the viability, vigor, and utility of seeds for agricultural and conservation purposes (De Vitis et al., 2020; Gough, 2020). These traits can be broadly categorized into physical, physiological, health, and content traits, each playing a pivotal role in determining the overall performance of seeds. Physical traits, such as size, shape, color, and uniformity, are essential for ensuring mechanical handling efficiency and market acceptability (McDonald and Panozzo, 2023). For instance, uniformity in seed size facilitates precision sowing and promotes even germination (Finch-Savage and Bassel, 2016). No damage to the external parts of seeds is a crucial physical quality trait (Lemes and Catão, 2024). Analytical purity, or the absence of contamination by weed seeds, other crops, or inert matter, is another vital physical trait, ensuring the integrity of seed lots for planting and trade.

Physiological traits focus on the regenerative capacity of seeds, encompassing germination rate, vigor, and longevity. Germination rate measures the ability of seeds to sprout under optimal conditions, while vigor assesses their potential to thrive under suboptimal or stressful environments (Finch-Savage and Bassel, 2016; Reed et al., 2022). Longevity, or the capacity of seeds to retain viability over extended storage periods, is especially critical for genebanks and seed conservation programs (Walters et al., 2005; Kameswara Rao et al., 2017). Together, these traits ensure seeds not only germinate but also establish healthy plants that can withstand environmental stresses.

In addition, health and content quality traits are paramount for maintaining the productivity and reliability of seed stocks (Kameswara Rao et al., 2017; Sundareswaran et al., 2023). Health traits, such as being disease-free and having optimal moisture content, directly affect storage stability and field performance (Kameswara Rao et al., 2017; Kumar et al., 2023). Content traits, including biochemical and embryological as well as microbiome composition may help the planted seeds produce crops with predictable and desirable characteristics.

Further, genetic traits such as purity and the absence of undesirable mutations are important for resistance to pests, environmental stressors, and enhanced nutritional profiles (Singh et al., 2015) but these quality traits can be screened using genomics tools. The holistic evaluation of these seed quality traits allows genebanks, breeders, and farmers to optimize seed utilization, supporting food security and sustainable agriculture.

5.1 Seed physical quality traits

Phenomics tools have revolutionized the measurement of seed physical quality traits by enabling rapid, precise, and non-destructive assessments (Ghamkhar, 2023; Hacisalihoglu and Armstrong, 2023). Physical traits such as size, shape, density, color, and uniformity are fundamental to seed quality, influencing mechanical handling, market value, and sowing efficiency Phenomics technologies ensure the consistent grading of seed lots, improving quality control processes in both genebanks and commercial seed production.

The integration of these phenomics tools not only enhances the efficiency of physical trait evaluation but also ensures the preservation of seed samples for conservation and research purposes. By combining accuracy, speed, and non-destructive methodologies, phenomics technologies play a critical role in maintaining and enhancing seed physical quality standards (Dell'Aquila, 2007; Mahajan et al., 2015).

Another important physical quality trait is seed analytical purity, i.e. no weeds or other crop seeds in the lot (Powell, 2009; Prasad, 2023). RGB imaging (using different filters) (Granitto et al., 2005), multi- and hyper-spectral imaging and thermal imaging can all help in high throughput phenotyping for this important factor/trait at both genebanks and commercial seed exchange.

5.2 Seed physiological quality traits

Phenomics tools have greatly advanced the assessment of seed physiological quality traits, providing non-invasive and precise methodologies to evaluate attributes like germination potential, moisture content, vigor, and viability (Al-Amery et al., 2018; Feng et al., 2019). These traits are essential for understanding a seed’s regenerative capacity and its ability to establish healthy plants under both optimal and stressful conditions (Finch-Savage and Bassel, 2016; Reed et al., 2022).

By integrating phenomics tools, researchers and genebanks can effectively evaluate the physiological quality of seeds, ensuring the preservation of high-performing germplasm.

5.3 Seed pathology quality traits

Seed pathology quality (infected vs. healthy seeds) is a critical parameter for ensuring the longevity of germplasm, as well as the success of agricultural production (Gebeyaw, 2020). Phenomics tools provide advanced, non-invasive methods to detect and evaluate health-related traits such as the presence of pathogens, disease resistance, and durability during storage (Walcott et al., 1998; Guelpa et al., 2015)– including hard seededness in legumes, which extends seed dormancy. This capability allows for rapid and precise disease detection, reducing the reliance on traditional, labor-intensive diagnostic methods.

Emerging technologies, such as cold plasma treatment, are being coupled with imaging systems to both detect and mitigate seed-borne pathogens (Mravlje et al., 2021; Moumni et al., 2023). These phenomics tools not only enhance the efficiency and accuracy of seed health evaluations but also support better decision-making for storage and planting, ensuring that only healthy, high-quality seeds are used and preserved (Yang et al., 2020; Dhanya et al., 2024).

Magnetic Resonance Imaging (MRI) is an expensive and slow tool and despite providing valuable information, it is not suggested as a phenomics tool for genebanks (Ghamkhar et al., 2024). RGB imaging alone is also not informative enough for seed health assessment as it will only provide information when it is already too late to separate healthy seeds from diseased ones (Vrešak et al., 2016). It also misses the intermediate stages between healthy and diseased seeds.

5.4. Seed biochemical quality traits

Phenomics tools have significantly advanced the evaluation of seed content traits, such as biochemical composition, nutrient profiles, embryological status, anti-nutritional factors and microbiome composition (Singh et al., 2020; Hacisalihoglu and Armstrong, 2023). These traits are critical for assessing seed viability, nutritional value, and suitability for specific agricultural or industrial applications.

Wi-Fi holography technology (Roitsch et al., 2019) and MRI (Plutenko et al., 2025) (more informative but expensive) are also new tools that enable embryo integrity screening while the seed microbiome can be evaluated using muti- hyper-spectral imaging (Reddy et al., 2023) (less informative but more affordable) systems.

By integrating these phenomics tools, genebanks and researchers can perform rapid, precise, and non-destructive evaluations of seed content traits, ensuring the preservation of high-value germplasm and the optimization for specific traits of interest in the seed collection.

5.5 Seed genetic purity traits

For example, hyperspectral imaging captures spectral signatures across various wavelengths, revealing These spectral patterns, combined with machine learning algorithms, allow for accurate classification and sorting of seeds based on purity traits, thereby enhancing quality assurance in breeding programs and seed production.

Meanwhile, Automated phenotyping platforms, integrated with AI and deep learning, can process large datasets to identify and predict genetic purity traits with high accuracy (Tu et al., 2021; Mukasa et al., 2022).

6 Key applications of phenomics in genebanks and seed research

Genebanks, as custodians of plant genetic resources, can significantly enhance their seed phenotyping capabilities by adopting modern phenomics technologies. These technologies allow for non-destructive, high-throughput, and precise assessment of seed traits, ensuring better management, characterization, and utilization of genetic resources. Key sensor and imaging technologies that genebanks should adopt for seed phenotyping will largely depend on a cost/value assessment and whether a quick gains approach best suits genebanks. In general, there are some approaches currently available to genebanks with a range of costs (Figure 2). In the meantime, the aim should be to make efforts for collaboration in order for the availability and use of new and quick gain technologies for genebanks. Below are key sensor and imaging technologies that genebanks could adopt for seed phenotyping:

Figure 2

Figure 2. Estimated cost of a few sensor technologies in genebanks (includes the costs of data preparation but not data processing).

6.1. Seed viability scanning at scale

One of the most critical applications of phenomics at genebanks is in seed quality assessment, where phenomics platforms are used to evaluate physical, physiological, health, and genetic attributes (Volk et al., 2021; Ghamkhar, 2023). Physiological traits such as germination capacity, vigor, and longevity are measured using advanced imaging methods (Hu et al., 2024a; Jin et al., 2025). These tools ensure that only high-quality seeds are preserved and distributed for breeding and research purposes.

6.2 Detecting dormancy, ageing and storability

Detecting seed dormancy and ageing is a critical component of seed phenotyping, as these physiological states directly influence germination performance, storage behavior, and long-term viability within genebanks (Whitehouse et al., 2020). Seed dormancy—whether physiological, morphological, or physical—represents a controlled inhibition of germination (Nonogaki, 2014), and phenotyping efforts increasingly focus on identifying biochemical and structural markers that distinguish dormant from non-dormant seeds before testing. Traditional dormancy detection relies on germination assays under varying temperature, light, or chemical treatments (e.g., gibberellic acid or nitrate), but advances in sensing technologies now allow the detection of traits associated with dormancy-breaking or dormancy-imposing mechanisms. These approaches enable earlier and more accurate identification of dormancy-related traits without the need for extended germination trials.

Seed ageing detection focuses on identifying early biochemical and structural deterioration that precedes visible loss of viability (Xing et al., 2025). During seed ageing—whether natural or accelerated—changes occur in membrane integrity, lipid peroxidation levels, antioxidant capacity, protein carbonylation, and mitochondrial activity (Adetunji et al., 2021). These processes alter the optical, thermal, and dielectric properties of seeds, making them detectable through non-destructive phenotyping technologies. The integration of these approaches allows genebanks and seed scientists to monitor ageing trajectories more effectively, supporting predictive models of seed longevity and improving prioritization for regeneration.

Phenomics enhances the evaluation of seed health and storage performance, critical aspects of genebank operations (Ghamkhar, 2023). This real-time monitoring capability allows genebanks to address potential issues before they compromise seed viability, ensuring long-term preservation (Bhattacharya and Mummenhoff, 2024). These applications make phenomics indispensable for maintaining the quality and longevity of germplasm collections (Nguyen and Norton, 2020; Ghamkhar et al., 2024).

6.3. Pre-breeding characterization

Phenomics tools also play a vital role in trait discovery and association mapping (Hatzig et al., 2015; Xiao et al., 2022). By linking seed phenotypic data with genomic information, researchers can identify markers associated with complex traits such as biochemical composition (Chaudhary et al., 2015; Dwivedi et al., 2021). This capability is particularly valuable for breeding programs targeting the development of high-yield or nutrient rich crops (Gaikwad et al., 2020; Dwivedi et al., 2021). Additionally, trait discovery enables genebanks to identify and preserve rare or underutilized accessions with unique seed characteristics, ensuring their availability for future use (Smale and Jamora, 2020).

6.4 Rapid seed purity testing

Rapid seed purity testing is an essential component of modern seed and genebank management, enabling the quick verification of genetic, physical, and varietal integrity in seed lots destined for breeding, research, or commercial production (Gough, 2020). Traditionally, purity testing relied on manual/visual inspection, examining seeds for off-types, contaminants, or foreign botanical materials. This approach is labor-intensive, subjective, and limited by human detection capabilities. Advances in imaging and machine-learning technologies have transformed purity assessment by enabling high-throughput, objective, and non-destructive analysis of large seed batches. These approaches significantly reduce testing time, increase reproducibility, and enable the detection of subtle phenotypic differences that may not be visible to the human eye.

Beyond visual and spectral methods, molecular approaches such as DNA barcoding, SNP profiling, and PCR-based assays contribute to genetic purity verification. This is specifically the case in hybrid or certified seed production systems where varietal identity is critical. Emerging phenomics technologies offer additional layers of discrimination by detecting biochemical signatures associated with specific genotypes or contaminants.

6.5 Automating regeneration decisions

Automating regeneration decisions is becoming a transformative application of seed phenotyping technologies, offering genebanks a data-driven way to prioritize which accessions require regeneration and when (Nguyen and Norton, 2020). Traditionally, regeneration choices have relied on simple viability thresholds obtained from periodic germination tests, often carried out on fixed schedules (e.g., every 5–10 years). If seed viability falls below 85% of the initial germination percentage in active collections regeneration is required (Dulloo et al., 2008). While effective, this approach can be inefficient as viability may remain high for decades in some species, while others deteriorate rapidly and risk genetic loss before the next scheduled test. Automated decision-support systems integrate phenotyping outputs such as germination behavior, ageing biomarkers, seed moisture profiles, and imaging-derived structural or biochemical indicators to predict viability decline more precisely (Angidi et al., 2025). By combining these measurements with models of seed longevity, storage environment data, and historical regeneration records, genebanks can generate real-time risk assessments and dynamic prioritization lists that minimize unnecessary regeneration while preventing loss of valuable genetic diversity.

Machine-learning algorithms and predictive analytics further strengthen this automation by identifying patterns that link phenotypic traits to longevity outcomes, enabling early detection of ageing, before viability visibly declines (Jin et al., 2025). High-throughput phenotyping tools can screen hundreds of accessions rapidly, feeding multi-dimensional data into longevity prediction models. These systems can alert managers when accessions are approaching critical thresholds or when specific seeds exhibit atypical ageing trajectories. Automating regeneration decision enhances operational efficiency, reduces costs, and ensures the timely regeneration of high-value or vulnerable germplasm. Ultimately, this approach enables genebanks to maintain collections more strategically, preserving genetic integrity while optimizing resource use across large and diverse seed repositories.

6.6 Linking phenotype–genotype in digitized gene bank workflows

Historically, genebanks have maintained extensive passport and storage information but comparatively sparse phenotypic and genomic datasets (Anglin et al., 2018). Modern digitization efforts—coupled with advances in high-throughput phenotyping and affordable sequencing—now make it possible to systematically connect seed traits with underlying genetic variation across thousands of accessions. Integrating genomics with phenomics has significantly enhanced data enrichment in genebanks, providing a deeper understanding of seed and plant traits at both the molecular and phenotypic levels (Schmutzer et al., 2017; Ghamkhar et al., 2024). This integration not only improves the characterization of genetic resources but also facilitates the identification of genes associated with specific adaptive or agronomic traits (Volk et al., 2021; Ghamkhar et al., 2024).

The fusion of genomics and phenomics also supports predictive modeling and trait discovery (Togninalli et al., 2023). By correlating genetic data with high-throughput phenotypic measurements, genebanks can generate multidimensional datasets that reveal complex trait networks, such as those related to biochemical composition (Volk et al., 2021; Xiao et al., 2022). These enriched datasets are critical for tackling global agricultural challenges, such as food security. The seamless integration of genomics with phenomics in genebanks not only enhances the depth and reliability of germplasm evaluations but also accelerates the development of crop varieties tailored to future environmental and market demands (Egan and Stiller, 2022; Crossa et al., 2024).

Imaging-derived phenotypes (e.g., seed size, color, density, biochemical signatures), germination and ageing profiles, and environmental metadata can be integrated with genomic information such as SNP markers, whole-genome sequences, or genotyping-by-sequencing outputs. This creates a unified digital framework where each seed lot is represented not only by its identity and origin, but also by quantifiable traits and genetic markers that explain adaptive patterns or functional differences.

Embedding these phenotype–genotype linkages directly into genebank information systems enables powerful new workflows (Arend et al., 2022). Phenotypic screening data from seed imaging or physiological tests can help target genomic analyses toward accessions displaying unique or extreme characteristics. These integrated datasets also support the development of core collections and pre-breeding pipelines, thereby increasing the value and utilization of conserved germplasm.

Ultimately, linking phenotype and genotype in digitized genebank workflows enhances conservation decisions, improves the efficiency of regeneration and characterization strategies, and strengthens the role of genebanks as central hubs for innovation in seed science and plant improvement.

7 Challenges and opportunities in using seed phenomics at genebanks

Genebanks are and will be facing many challenges (and opportunities) if they proceed to integrating seed phenomics to their management system. A list of these challenges and opportunities is itemised below although it is by no means comprehensive.

7.1. Large imaging datasets

Large imaging datasets present both a significant challenge and a transformative opportunity for genebanks integrating seed phenotyping technologies. The vast amount of data generated by high-throughput phenotyping systems poses significant challenges for data storage, processing, and analysis (Shakoor et al., 2017; Yang et al., 2020). Effective data management frameworks and machine learning algorithms are required to ensure that the data generated is actionable and supports decision-making in conservation programs (Yu and Chung, 2024). There is a lot that can be learnt from the biomedical sciences in this space (Alkhatib and Gaede, 2024). Managing these datasets requires robust data infrastructure, including scalable storage solutions, standardised metadata formats, and efficient pipelines for image processing and annotation. Without these systems in place, genebanks risk accumulating unusable data backlogs, where images exist but cannot be analysed, interpreted, or linked to accession records.

At the same time, these large imaging datasets open unprecedented opportunities for enhancing genetic resource conservation and utilisation. These quantified imaging datasets enable machine-learning models to predict traits such as dormancy status, ageing progression, or embryo integrity at scale, dramatically increasing the throughput and precision of seed evaluations. Moreover, image archives can be re-analysed as new algorithms and computational tools emerge, allowing genebanks to extract additional value from historical phenotyping efforts.

7.2 Lack of shared standards

Another challenge lies in the standardisation of phenomics protocols and data management (Houle et al., 2010; Krajewski et al., 2015). Currently, different genebanks use diverse imaging platforms, resolutions, lighting conditions, trait definitions, and analytical pipelines. With a wide array of phenotyping tools and platforms available, there is often inconsistency in how traits are measured and reported across genebanks (Anglin et al., 2018; Nguyen and Norton, 2020). This lack of standardisation can hinder the comparability and integration of data, limiting the utility of phenomics in global collaboration efforts (Großkinsky et al., 2015; Awada et al., 2024). Without standardised trait ontologies, calibration procedures, and reporting guidelines, phenotypic data cannot be reliably integrated into global information systems or used to support large-scale meta-analyses. The absence of common data formats and quality thresholds also complicates long-term data curation and interoperability with existing genebank databases such as GRIN, Genesys, or national PGR information systems. As a result, valuable phenotyping outputs risk becoming siloed, underutilised, or incompatible with the broader plant genetic resources community.

Despite these challenges, the current lack of shared standards also presents a unique opportunity for the genebanking community to take a leadership role in establishing coordinated, future-proof frameworks for digital phenotyping. By collaborating through international networks, such as the Global Crop Diversity Trust, EURISCO, CGIAR genebanks, International Seed Testing Association (ISTA), and the International Plant Phenotyping Network, institutions can co-develop standard operating procedures (SOPs), trait ontologies, metadata requirements, and reference materials tailored specifically to seed phenotyping. Establishing shared benchmarks for image quality, sensor calibration, and data annotation will enable cross-genebank comparability and facilitate the exchange of validated protocols. Over time, this harmonisation will reduce duplication of effort, improve data reusability, and enable federated machine-learning models trained on multi-genebank datasets. Ultimately, addressing the lack of shared standards is not just a technical requirement but a strategic opportunity to build a globally cohesive, interoperable, and scientifically rigorous foundation for next-generation seed phenotyping.

7.3 Cost and instrument complexity

The adoption of phenomics tools in genebanks presents a range of challenges, particularly related to the cost and scalability of these technologies (Nguyen and Norton, 2020; Ghamkhar et al., 2024). Advanced platforms such as hyperspectral scanners, micro-CT systems, terahertz spectroscopy units, and automated imaging chambers require substantial capital investment, ongoing maintenance, and specialised technical expertise. For many genebanks, especially smaller national collections or those in resource-limited regions, these costs are prohibitive, limiting their ability to modernise evaluation workflows or transition from manual to digital phenotyping (Ghamkhar, 2023; Ghamkhar et al., 2024). Additionally, complex instruments often require trained operators, dedicated analysis software, and regular calibration, all of which introduce operational bottlenecks (Nguyen and Norton, 2020; Ghamkhar et al., 2024). This can lead to uneven technological capacity across genebanks, with only a few well-funded centers able to conduct high-throughput or high-resolution phenotyping, ultimately constraining global data integration and interoperability. This highlights the need for affordable and user-friendly solutions that can be deployed across diverse operational contexts.

However, these challenges also create opportunities for innovation, collaboration, and strategic investment. As sensor technologies mature, costs are gradually decreasing, and simpler, more compact instruments, such as portable NIR devices, low-cost multispectral cameras, or benchtop X-ray units, are becoming increasingly accessible (Argirusis et al., 2025). These emerging tools can serve as stepping stones for genebanks to develop digital phenotyping capabilities without immediately committing to high-end systems. Furthermore, shared infrastructure models, such as regional phenotyping hubs, cross-institutional equipment pools, cloud-based analysis environments, and collaborative service agreements, allow multiple genebanks to access advanced technologies without bearing the full financial burden individually. Over time, strategic investment and coordinated capacity building can turn cost and complexity from limiting factors into catalysts for modernisation, equity, and innovation in global seed conservation.

7.4 Calibration drift and transferability

Instruments such as multispectral and hyperspectral imagers, X-ray systems, and terahertz platforms rely heavily on stable calibration to ensure consistent and reproducible measurements. Over time, however, sensor ageing, environmental fluctuations, lamp degradation, and hardware servicing can introduce calibration drift, leading to subtle but significant measurement inconsistencies. For genebanks, where phenotyping may occur over years or decades and where data must remain comparable across regeneration cycles such drift can compromise longitudinal analyses, trait prediction models, and digital records associated with accessions. Similarly, the transferability of calibration models between instruments, laboratories, or organisations is often limited: a predictive model built on one device or platform may fail or perform poorly when applied to another due to differences in optics, illumination, detector sensitivity, or spectral resolution. This limits the scalability of phenotyping workflows and poses a barrier to harmonised global datasets.

Calibration drift and transferability issues also offer opportunities to develop more robust and collaborative phenotyping ecosystems. Regular calibration schedules and the implementation of automated self-calibration routines can improve long-term instrument stability in genebanks. Emerging machine learning approaches, such as domain adaptation, instrument standardisation algorithms, and transfer learning (Chato and Regentova, 2023), can significantly enhance model portability across devices and institutions. Collaborative calibration frameworks, where multiple genebanks collect reference datasets under shared protocols, further enable the creation of cross-instrument models that are more resilient to variation. These developments support the vision of interoperable, globally harmonised phenotyping workflows, where data generated by different genebanks become directly comparable. Therefore, addressing calibration challenges becomes a pathway toward stronger international integration, better model reproducibility, and more reliable digital characterization of conserved germplasm.

Additional challenges arise from the need for long-term data curation, integration with existing databases, and the technical skills required to handle large, complex datasets, especially in organisations where informatics capacity is limited.

Despite all the listed challenges, the opportunities presented by seed phenomics are substantial. By enabling the rapid and precise assessment of a wide range of traits, phenomics allows genebanks to prioritise accessions with specific seed traits. Furthermore, the integration of phenomics with genomic and metabolomic data opens new avenues for trait discovery and functional characterisation, accelerating the development of crops with improved seed yield and nutritional quality of seed. These capabilities position phenomics as a cornerstone of modern genebank operations.

8 Future directions

The future of seed phenomics in genebanks lies in the integration of emerging technologies with scalable and cost-effective solutions that enhance the precision and efficiency of seed characterisation. One promising direction is the development of portable and automated phenotyping tools that can operate seamlessly. These tools, equipped with sensors like hyperspectral cameras, LiDAR, and thermal imagers, will enable the high-throughput assessment of diverse traits across large seed collections. By reducing reliance on fixed infrastructure, these innovations make advanced phenomics accessible to genebanks with limited resources, fostering broader adoption globally.

Another key area for growth is the advancement of data integration and artificial intelligence (AI) applications. The convergence of phenomics with multi-omics (genomics, metabolomics, and other)data will allow genebanks to build comprehensive datasets for complex trait analysis. AI and machine learning algorithms can further enhance this process by identifying patterns and associations within multidimensional datasets, accelerating trait discovery and the prioritisation of accessions. These capabilities are especially relevant for addressing pressing agricultural challenges, such as developing climate-resilient crops or improving nutritional quality. The creation of open-access phenomics databases and collaborative platforms will also facilitate global data sharing and cross-institutional research, enhancing the collective impact of genebanks.

The focus on standardization and community-driven protocols will be essential to ensuring the long-term success of seed phenomics. Such efforts will enable the harmonisation of phenomics practices across genebanks, ensuring data consistency and interoperability. Digital twins for seed collections (virtual, data-rich replicas of physical accessions) represent another transformative future direction for genebanks. By integrating real-time viability data, imaging-derived traits, environmental histories, and predictive models, digital twins would enable proactive management of regeneration needs, precise tracking of ageing dynamics, and more efficient utilization of genetic resources across global conservation networks. By embracing these future directions, seed phenomics will play a pivotal role in enhancing the utility and relevance of genebanks in the face of evolving global challenges.

The best recommendations for genebanks would be to prioritise cost-effective technologies like visible imaging, NIRS and thermal imaging for their routine high-throughput seed phenotyping. Genebank managers, however, should plan to adopt advanced imaging technologies such as HSI and X-ray imaging as their next priority and where possible combine technologies (e.g., hyperspectral and 3D imaging) to generate comprehensive datasets for both physical and chemical traits. An obvious step of preparation would be to invest in data management systems to store, analyze and share phenotyping data efficiently. Finally, they should engage in collaborations with phenomics research centers to access cutting-edge technologies and expertise such as machine learning and data science in order to enable the global network of genebanks adapt to the next wave of phenotyping and sensor technologies for knowledge sharing and advancement.

9 Concluding remarks

Seed phenomics has the potential to transform genebanks into dynamic hubs for genetic resource utilization. By adopting advanced technologies and fostering global collaborations and scalability, genebanks can enhance their role in ensuring food security and agricultural resilience in the face of climate change. This is an urgent need to modernize genebanks for more efficient seed and germplasm management.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

Author contributions

KG: Conceptualization, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing. DR: Validation, Writing – review & editing.

Funding

The author(s) declared that financial support was received for work and/or its publication. This work was supported by the Ministry of Business, Innovation and Employment (MBIE) project number: C10X1701, New Zealand and the SSIF project Funding PRJ0754061 during the preparation of this manuscript.

Conflict of interest

The authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author KG declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Adetunji, A. E., Adetunji, T. L., Varghese, B., Sershen, and Pammenter, N. W. (2021). Oxidative stress, ageing and methods of seed invigoration: an overview and perspectives. Agronomy 11, 2369. doi: 10.3390/agronomy11122369

Crossref Full Text | Google Scholar

Afsah-Hejri, L., Hajeb, P., Ara, P., and Ehsani, R. J. (2019). A comprehensive review on food applications of terahertz spectroscopy and imaging. Compr. Rev. Food Sci. Food Saf. 18, 1563–1621. doi: 10.1111/1541-4337.12490

PubMed Abstract | Crossref Full Text | Google Scholar

Aftab, A., Sivasankarreddy, K., Fahida, P. K., and Joseph, J. (2025). Utilization of modern tools and techniques in seed research. In Advances in Seed Quality Evaluation and Improvement. (Singapore: Springer Nature Singapore) 19, 19–52.

Google Scholar

Agelet, L. E. and Hurburgh, C. R. (2014). Limitations and current applications of Near Infrared Spectroscopy for single seed analysis. Talanta 121, 288–299. doi: 10.1016/j.talanta.2013.12.038

PubMed Abstract | Crossref Full Text | Google Scholar

Ahmed, M. R., Yasmin, J., Collins, W., and Cho, B.-K. (2018). X-ray CT image analysis for morphology of muskmelon seed in relation to germination. Biosyst. Eng. 175, 183–193. doi: 10.1016/j.biosystemseng.2018.09.015

Crossref Full Text | Google Scholar

Akewar, M. and Chandak, M. (2023). Hyperspectral imaging algorithms and applications: A review. TechRxiv. doi: 10.36227/techrxiv.24743562.v1

Crossref Full Text | Google Scholar

Al-Amery, M., Geneve, R. L., Sanches, M. F., Armstrong, P. R., Maghirang, E. B., Lee, C., et al. (2018). Near-infrared spectroscopy used to predict soybean seed germination and vigour. Seed Sci. Res. 28, 245–252. doi: 10.1017/S0960258518000119

Crossref Full Text | Google Scholar

Alkhatib, R. and Gaede, K. I. (2024). Data management in biobanking: strategies, challenges, and future directions. BioTech 13, 34. doi: 10.3390/biotech13030034

PubMed Abstract | Crossref Full Text | Google Scholar

Ambrose, A., Kandpal, L. M., Kim, M. S., Lee, W.-H., and Cho, B.-K. (2016a). High speed measurement of corn seed viability using hyperspectral imaging. Infrared Phys. Technol. 75, 173–179. doi: 10.1016/j.infrared.2015.12.008

Crossref Full Text | Google Scholar

Ambrose, A., Lohumi, S., Lee, W.-H., and Cho, B. K. (2016b). Comparative nondestructive measurement of corn seed viability using Fourier transform near-infrared (FT-NIR) and Raman spectroscopy. Sensors Actuators B: Chem. 224, 500–506. doi: 10.1016/j.snb.2015.10.082

Crossref Full Text | Google Scholar

Angidi, S., Madankar, K., Tehseen, M. M., and Bhatla, A. (2025). Advanced high-throughput phenotyping techniques for managing abiotic stress in agricultural crops—A comprehensive review. Crops 5, 8. doi: 10.3390/crops5020008

Crossref Full Text | Google Scholar

Anglin, N. L., Amri, A., Kehel, Z., and Ellis, D. (2018). A case of need: Linking traits to genebank accessions. Biopreservation biobanking 16, 337–349. doi: 10.1089/bio.2018.0033

PubMed Abstract | Crossref Full Text | Google Scholar

Arcos-Limiñana, V., Maestre-Pérez, S., and Prats-Moya, M. S. (2025). A comprehensive review on ultraviolet disinfection of spices and culinary seeds and its effect on quality. Compr. Rev. Food Sci. Food Saf. 24, e70076. doi: 10.1111/1541-4337.70076

PubMed Abstract | Crossref Full Text | Google Scholar

Arend, D., Beier, S., König, P., Lange, M., Memon, J. A., Oppermann, M., et al. (2022). “"From genotypes to phenotypes: a plant perspective on current developments in data management and data publication,",” in Integrative bioinformatics: history and future. (Singapore: Springer), 11–43.

Google Scholar

Argirusis, N., Achilleos, A., Alizadeh, N., Argirusis, C., and Sourkouni, G. (2025). IR sensors, related materials, and applications. Sensors 25, 673. doi: 10.3390/s25030673

PubMed Abstract | Crossref Full Text | Google Scholar

Arkhipov, M. V., Priyatkin, N. S., Gusakova, L. P., Potrakhov, N. N., Gryaznov, A. Y., Bessonov, V. B., et al. (2019). X-ray computer methods for studying the structural integrity of seeds and their importance in modern seed science. Tech. Phys. 64, 582–592. doi: 10.1134/S1063784219040030

Crossref Full Text | Google Scholar

Awada, L., Phillips, P. W., and Bodan, A. M. (2024). The evolution of plant phenomics: global insights, trends, and collaborations, (2000-2021). Front. Plant Sci. 15, 1410738. doi: 10.3389/fpls.2024.1410738

PubMed Abstract | Crossref Full Text | Google Scholar

Barboza da Silva, C., Oliveira, N. M., de Carvalho, M. E. A., de Medeiros, A. D., de Lima Nogueira, M., and Dos Reis, A. R. (2021). Autofluorescence-spectral imaging as an innovative method for rapid, non-destructive and reliable assessing of soybean seed quality. Sci. Rep. 11, 17834. doi: 10.1038/s41598-021-97223-5

PubMed Abstract | Crossref Full Text | Google Scholar

Bartolić, D., Mutavdžić, D., Carstensen, J. M., Stanković, S., Nikolić, M., Krstović, S., et al. (2022). Fluorescence spectroscopy and multispectral imaging for fingerprinting of aflatoxin-B1 contaminated (Zea mays L.) seeds: A preliminary study. Sci. Rep. 12, 4849. doi: 10.1038/s41598-022-08352-4

PubMed Abstract | Crossref Full Text | Google Scholar

Batista, T. B., Mastrangelo, C. B., de Medeiros, A. D., Petronilio, A. C. P., Fonseca de Oliveira, G. R., Dos Santos, I. L., et al. (2022). A reliable method to recognize soybean seed maturation stages based on autofluorescence-spectral imaging combined with machine learning algorithms. Front. Plant Sci. 13, 914287. doi: 10.3389/fpls.2022.914287

PubMed Abstract | Crossref Full Text | Google Scholar

Baye, T. M., Pearson, T. C., and Settles, A. M. (2006). Development of a calibration to predict maize seed composition using single kernel near infrared spectroscopy. J. Cereal Sci. 43, 236–243. doi: 10.1016/j.jcs.2005.11.003

Crossref Full Text | Google Scholar

Berthomieu, C. and Hienerwadel, R. (2009). Fourier transform infrared (FTIR) spectroscopy. Photosynthesis Res. 101, 157–170. doi: 10.1007/s11120-009-9439-x

PubMed Abstract | Crossref Full Text | Google Scholar

Bewley, J. D. and Black, M. (2012). Physiology and biochemistry of seeds in relation to germination: volume 2: viability, dormancy, and environmental control. (Berlin: Springer Science & Business Media).

Google Scholar

Beyaz, R. and Yildiz, M. (2017). The use of gamma irradiation in plant mutation breeding. Plant Eng. 17, 32–43. doi: 10.5772/intechopen.69974

Crossref Full Text | Google Scholar

Bhattacharya, S. and Mummenhoff, K. (2024). Effective seedbank management to ensure food security and preserve biodiversity. Plant Systematics Evol. 310, 15. doi: 10.1007/s00606-024-01897-z

Crossref Full Text | Google Scholar

Chapman, S. C., Chakraborty, S., Dreccer, M. F., and Howden, S. M. (2012). Plant adaptation to climate change—opportunities and priorities in breeding. Crop Pasture Sci. 63, 251–268. doi: 10.1071/CP11303

Crossref Full Text | Google Scholar

Chato, L. and Regentova, E. (2023). Survey of transfer learning approaches in the machine learning of digital health sensing data. J. personalized Med. 13, 1703. doi: 10.3390/jpm13121703

PubMed Abstract | Crossref Full Text | Google Scholar

Chaudhary, J., Patil, G. B., Sonah, H., Deshmukh, R. K., Vuong, T. D., Valliyodan, B., et al. (2015). Expanding omics resources for improvement of soybean seed composition traits. Front. Plant Sci. 6, 1021. doi: 10.3389/fpls.2015.01021

PubMed Abstract | Crossref Full Text | Google Scholar

Colmer, J., O'Neill, C. M., Wells, R., Bostrom, A., Reynolds, D., Websdale, D., et al. (2020). SeedGerm: a cost-effective phenotyping platform for automated seed imaging and machine-learning based phenotypic analysis of crop seed germination. New Phytol. 228, 778–793. doi: 10.1111/nph.16736

PubMed Abstract | Crossref Full Text | Google Scholar

Corbineau, F. (2012). Markers of seed quality: from present to future. Seed Sci. Res. 22, S61–S68. doi: 10.1017/S0960258511000419

Crossref Full Text | Google Scholar

Crossa, J., Montesinos-Lopez, O. A., Costa-Neto, G., Vitale, P., Martini, J. W., Runcie, D., et al. (2025). Machine learning algorithms translate big data into predictive breeding accuracy. Trends Plant Science. 30, 167–184. doi: 10.1016/j.tplants.2024.09.011

PubMed Abstract | Crossref Full Text | Google Scholar

da Silva, C. B., Bianchini, V., de Medeiros, A. D., de Moraes, M. H. D., Marassi, A. G., and Tannús, A. (2021). A novel approach for Jatropha curcas seed health analysis based on multispectral and resonance imaging techniques. Ind. Crops Products 161, 113186. doi: 10.1016/j.indcrop.2020.113186

Crossref Full Text | Google Scholar

Dell'Aquila, A. (2007). Towards new computer imaging techniques applied to seed quality testing and sorting. Seed Sci. Technol. 35, 519–538. doi: 10.15258/sst.2007.35.3.01

Crossref Full Text | Google Scholar

De Vitis, M., Hay, F. R., Dickie, J. B., Trivedi, C., Choi, J., and Fiegener, R. (2020). Seed storage: maintaining seed viability and vigor for restoration use. Restor. Ecol. 28, S249–S255. doi: 10.1111/rec.13174

Crossref Full Text | Google Scholar

Dhanya, V., Subeesh, A., Susmita, C., Saji, S. J., Dilsha, C., Keerthi, C., et al. (2024). High throughput phenotyping using hyperspectral imaging for seed quality assurance coupled with machine learning methods: principles and way forward. Plant Physiol. Rep. 29, 1–20. doi: 10.1007/s40502-024-00839-8

Crossref Full Text | Google Scholar

Dulloo, M. E., Hanson, J., Jorge, M., and Thormann, I. (2008).

Google Scholar

Dumont, J., Hirvonen, T., Heikkinen, V., Mistretta, M., Granlund, L., Himanen, K., et al. (2015). Thermal and hyperspectral imaging for Norway spruce (Picea abies) seeds screening. Comput. Electron. Agric. 116, 118–124. doi: 10.1016/j.compag.2015.06.010

Crossref Full Text | Google Scholar

Dwivedi, S. L., Goldman, I., Ceccarelli, S., and Ortiz, R. (2020). Advanced analytics, phenomics and biotechnology approaches to enhance genetic gains in plant breeding. Adv. Agron. 162, 89–142. doi: 10.1016/bs.agron.2020.02.002

Crossref Full Text | Google Scholar

Dwivedi, S. L., Spillane, C., Lopez, F., Ayele, B. T., and Ortiz, R. (2021). First the seed: Genomic advances in seed science for improved crop productivity and food security. Crop Sci. 61, 1501–1526. doi: 10.1002/csc2.20402

Crossref Full Text | Google Scholar

Egan, L. M. and Stiller, W. N. (2022). The past, present, and future of host plant resistance in cotton: An Australian perspective. Front. Plant Sci. 13, 895877. doi: 10.3389/fpls.2022.895877

PubMed Abstract | Crossref Full Text | Google Scholar

ElMasry, G., ElGamal, R., Mandour, N., Gou, P., Al-Rejaie, S., Belin, E., et al. (2020). Emerging thermal imaging techniques for seed quality evaluation: Principles and applications. Food Res. Int. 131, 109025. doi: 10.1016/j.foodres.2020.109025

PubMed Abstract | Crossref Full Text | Google Scholar

ElMasry, G., Mandour, N., Al-Rejaie, S., Belin, E., and Rousseau, D. (2019). Recent applications of multispectral imaging in seed phenotyping and quality monitoring—An overview. Sensors 19, 1090. doi: 10.3390/s19051090

PubMed Abstract | Crossref Full Text | Google Scholar

Fassio, A. and Cozzolino, D. (2004). Non-destructive prediction of chemical composition in sunflower seeds by near infrared spectroscopy. Ind. Crops Products 20, 321–329. doi: 10.1016/j.indcrop.2003.11.004

Crossref Full Text | Google Scholar

Feng, L., Zhu, S., Liu, F., He, Y., Bao, Y., and Zhang, C. (2019). Hyperspectral imaging for seed quality and safety inspection: A review. Plant Methods 15, 1–25. doi: 10.1186/s13007-019-0476-y

PubMed Abstract | Crossref Full Text | Google Scholar

Fernando, E. A. J., Selvaraj, M., and Ghamkhar, K. (2023). The power of phenomics: Improving genebank value and utility. Mol. Plant 16, 1099–1101. doi: 10.1016/j.molp.2023.05.011

PubMed Abstract | Crossref Full Text | Google Scholar

Ferreira, L., Carvalho, I. C. B., Jorge, L., Quezado-Duval, A. M., and Rossato, M. (2024). Hyperspectral imaging for the detection of plant pathogens in seeds: recent developments and challenges. Front. Plant Sci. 15, 1387925. doi: 10.3389/fpls.2024.1387925

PubMed Abstract | Crossref Full Text | Google Scholar

Finch-Savage, W. E. and Bassel, G. W. (2016). Seed vigour and crop establishment: extending performance beyond adaptation. J. Exp. Bot. 67, 567–591. doi: 10.1093/jxb/erv490

PubMed Abstract | Crossref Full Text | Google Scholar

Finch-Savage, W. E. and Leubner-Metzger, G. (2006). Seed dormancy and the control of germination. New Phytol. 171, 501–523. doi: 10.1111/j.1469-8137.2006.01787.x

PubMed Abstract | Crossref Full Text | Google Scholar

Font, R., del Río-Celestino, M., and de Haro-Bailón, A. (2006). The use of near-infrared spectroscopy (NIRS) in the study of seed quality components in plant breeding programs. Ind. Crops Products 24, 307–313. doi: 10.1016/j.indcrop.2006.06.012

Crossref Full Text | Google Scholar

Foucat, L., Chavagnat, A., and Renou, J.-P. (1993). Nuclear magnetic resonance micro-imaging and X-radiography as possible techniques to study seed germination. Scientia Hortic. 55, 323–331. doi: 10.1016/0304-4238(93)90043-P

Crossref Full Text | Google Scholar

França-Neto, J. and Krzyzanowski, F. C. (2022). Use of the tetrazolium test for estimating the physiological quality of seeds. Seed Sci. Technol. 50, 31–44. doi: 10.15258/sst.2022.50.1.s.03

Crossref Full Text | Google Scholar

França-Silva, F., Gomes-Junior, F. G., Rego, C. H. Q., Marassi, A. G., and Tannús, A. (2023). Advances in imaging technologies for soybean seed analysis. J. Seed Sci. 45, e202345022. doi: 10.1590/2317-1545v45274098

Crossref Full Text | Google Scholar

Fu, Y., Ren, Y., and Sun, D.-W. (2024). Novel analysis of food processes by terahertz spectral imaging: A review of recent research findings. Trends Food Sci. Technol. 147, 104463. doi: 10.1016/j.tifs.2024.104463

Crossref Full Text | Google Scholar

Gaikwad, K. B., Rani, S., Kumar, M., Gupta, V., Babu, P. H., Bainsla, N. K., et al. (2020). Enhancing the nutritional quality of major food crops through conventional and genomics-assisted breeding. Front. Nutr. 7, 533453. doi: 10.3389/fnut.2020.533453

PubMed Abstract | Crossref Full Text | Google Scholar

Gaikwad, S. and Tidke, S. (2022). Multi-spectral imaging for fruits and vegetables. Int. J. Advanced Comput. Sci. Appl. 13, 743–760. doi: 10.14569/IJACSA.2022.0130287

Crossref Full Text | Google Scholar

Galletti, P. A., Carvalho, M. E., Hirai, W. Y., Brancaglioni, V. A., Arthur, V., and Barboza da Silva, C. (2020). Integrating optical imaging tools for rapid and non-invasive characterization of seed quality: Tomato (Solanum lycopersicum L.) and carrot (Daucus carota L.) as study cases. Front. Plant Sci. 11, 577851. doi: 10.3389/fpls.2020.577851

PubMed Abstract | Crossref Full Text | Google Scholar

Gargiulo, L., Leonarduzzi, C., and Mele, G. (2020). Micro-CT imaging of tomato seeds: Predictive potential of 3D morphometry on germination. Biosyst. Eng. 200, 112–122. doi: 10.1016/j.biosystemseng.2020.09.003

Crossref Full Text | Google Scholar

Garnczarska, M., Zalewski, T., and Kempka, M. (2007). Water uptake and distribution in germinating lupine seeds studied by magnetic resonance imaging and NMR spectroscopy. Physiologia Plantarum 130, 23–32. doi: 10.1111/j.1399-3054.2007.00883.x

Crossref Full Text | Google Scholar

Ge, N., Jia, J.-S., Wang, Q.-Y., Li, C.-L., Huang, M., and Chen, J.-W. (2025). A genome-wide survey of DNA methylation reveals hyper-methylation regulates after-ripening and dormancy of recalcitrant Panax notoginseng seeds. J. Exp. Bot. 76, erae508. doi: 10.1093/jxb/erae508

PubMed Abstract | Crossref Full Text | Google Scholar

Ge, H.-y., Jiang, Y.-y., Lian, F.-y., Zhang, Y., and Xia, S.-h. (2014). Nondestructive evaluation of wheat quality using terahertz time domain spectroscopy. Spectrosc. Spectral Anal. 34, 2897–2900. doi: 10.3964/j.issn.1000-0593(2014)11-2897-04

PubMed Abstract | Crossref Full Text | Google Scholar

Ge, H., Lv, M., Lu, X., Jiang, Y., Wu, G., Li, G., et al. (2021).Applications of THz spectral imaging in the detection of agricultural products.

Google Scholar

Ge, H., Sun, Z., Jiang, Y., Wu, X., Jia, Z., Cui, G., et al. (2023). Recent advances in THz detection of water. Int. J. Mol. Sci. 24, 10936. doi: 10.3390/ijms241310936

PubMed Abstract | Crossref Full Text | Google Scholar

Gebeyaw, M. (2020). Review on: Impact of seed-borne pathogens on seed quality. Am. J. Plant Biol. 5, 77–81. doi: 10.11648/j.ajpb.20200504.11

Crossref Full Text | Google Scholar

Ghamkhar, K. (2023). “Phenomics for the improvement of crop adaptation,” in Plant genetic resources for the 21st century. (New York: Apple Academic Press), 115–129.

Google Scholar

Ghamkhar, K., Hay, F. R., Engbers, M., Dempewolf, H., and Schurr, U. (2024). Realizing the potential of plant genetic resources: the use of phenomics for genebanks. Plants People Planet. 7, 23–32. doi: 10.1002/ppp3.10570

Crossref Full Text | Google Scholar

Gough, R. E. (2020). Seed quality: basic mechanisms and agricultural implications. (Binghampton, NY: CRC Press).

Google Scholar

Granitto, P. M., Verdes, P. F., and Ceccatto, H. A. (2005). Large-scale investigation of weed seed identification by machine vision. Comput. Electron. Agric. 47, 15–24. doi: 10.1016/j.compag.2004.10.003

Crossref Full Text | Google Scholar

Großkinsky, D. K., Svensgaard, J., Christensen, S., and Roitsch, T. (2015). Plant phenomics and the need for physiological phenotyping across scales to narrow the genotype-to-phenotype knowledge gap. J. Exp. Bot. 66, 5429–5440. doi: 10.1093/jxb/erv345

PubMed Abstract | Crossref Full Text | Google Scholar

Gruwel, M. L., Yin, X. S., Edney, M. J., Schroeder, S. W., MacGregor, A. W., and Abrams, S. (2002). Barley viability during storage: use of magnetic resonance as a potential tool to study viability loss. J. Agric. Food Chem. 50, 667–676. doi: 10.1021/jf0108617

PubMed Abstract | Crossref Full Text | Google Scholar

Gua, T., Ding, Z., Zhang, D., Zhou, J., Yang, Z., Wang, X., et al. (2013Evaluation of wheat seeds by terahertz imaging, in 2013 6th UK, europe, China millimeter waves and THz technology workshop (UCMMT) (Rome, Italy: IEEE), 1–2. doi: 10.1109/UCMMT.2013.6641513

Crossref Full Text | Google Scholar

Guelpa, A., du Plessis, A., Kidd, M., and Manley, M. (2015). Non-destructive estimation of maize (Zea mays L.) kernel hardness by means of an X-ray micro-computed tomography (μCT) density calibration. Food Bioprocess Technol. 8, 1419–1429. doi: 10.1007/s11947-015-1502-3

Crossref Full Text | Google Scholar

Hacisalihoglu, G. and Armstrong, P. (2023). Crop seed phenomics: focus on non-destructive functional trait phenotyping methods and applications. Plants 12, 1177. doi: 10.3390/plants12051177

PubMed Abstract | Crossref Full Text | Google Scholar

Hamdy, S., Charrier, A., Corre, L. L., Rasti, P., and Rousseau, D. (2024). Toward robust and high-throughput detection of seed defects in X-ray images via deep learning. Plant Methods 20, 63. doi: 10.1186/s13007-024-01195-2

PubMed Abstract | Crossref Full Text | Google Scholar

Han, C., Qu, F., Wang, X., Zhai, X., Li, J., Yu, K., et al. (2024). Terahertz spectroscopy and imaging techniques for herbal medicinal plants detection: a comprehensive review. Crit. Rev. Analytical Chem. 54, 2485–2499. doi: 10.1080/10408347.2023.2183077

PubMed Abstract | Crossref Full Text | Google Scholar

Hatzig, S. V., Frisch, M., Breuer, F., Nesi, N., Ducournau, S., Wagner, M.-H., et al. (2015). Genome-wide association mapping unravels the genetic control of seed germination and vigor in Brassica napus. Front. Plant Sci. 6, 221. doi: 10.3389/fpls.2015.00221

PubMed Abstract | Crossref Full Text | Google Scholar

Hay, F. R. and Sershen, D. (2021).

Google Scholar

Hay, F. R. and Whitehouse, K. J. (2017). Rethinking the approach to viability monitoring in seed genebanks. Conserv. Physiol. 5, cox009. doi: 10.1093/conphys/cox009

PubMed Abstract | Crossref Full Text | Google Scholar

Hoffmann, A. A. and Sgrò, C. M. (2011). Climate change and evolutionary adaptation. Nature 470, 479–485. doi: 10.1038/nature09670

PubMed Abstract | Crossref Full Text | Google Scholar

Houle, D., Govindaraju, D. R., and Omholt, S. (2010). Phenomics: the next challenge. Nat. Rev. Genet. 11, 855–866. doi: 10.1038/nrg2897

PubMed Abstract | Crossref Full Text | Google Scholar

Hssaini, L., Elfazazi, K., Razouk, R., Ouaabou, R., Hernandez, F., Hanine, H., et al. (2021). Combined effect of cultivar and peel chromaticity on figs’ Primary and secondary metabolites: preliminary study using biochemical and FTIR fingerprinting coupled to chemometrics. Biology 10, 573. doi: 10.3390/biology10070573

PubMed Abstract | Crossref Full Text | Google Scholar

Hu, Q., Liu, J., Chen, X., Guzmán, C., Xu, Q., Zhang, Y., et al. (2024b). Multi-omic analysis reveals the effects of interspecific hybridization on the synthesis of seed reserve polymers in a Triticum turgidum ssp. durum× Aegilops sharonensis amphidiploid. BMC Genomics 25, 626. doi: 10.1186/s12864-024-10352-9

PubMed Abstract | Crossref Full Text | Google Scholar

Hu, J., Xu, S., Huang, Z., Liu, W., Zheng, J., and Liao, Y. (2024a). Rapid non-destructive detection of rice seed vigor via terahertz spectroscopy. Agriculture 15, 34. doi: 10.3390/agriculture15010034

Crossref Full Text | Google Scholar

Huang, X., Zheng, S., and Zhu, N. (2022). High-throughput legume seed phenotyping using a handheld 3D laser scanner. Remote Sens. 14, 431. doi: 10.3390/rs14020431

Crossref Full Text | Google Scholar

Hui, L., Jingzhu, W., Cuiling, L., and Xiaorong, S. (2018). Study on pretreatment methods of terahertz time domain spectral image for maize seeds. IFAC-PapersOnLine 51, 206–210. doi: 10.1016/j.ifacol.2018.08.142

Crossref Full Text | Google Scholar

Indore, N. S., Karunakaran, C., and Jayas, D. S. (2022). Synchrotron tomography applications in agriculture and food sciences research: a review. Plant Methods 18, 101. doi: 10.1186/s13007-022-00932-9

PubMed Abstract | Crossref Full Text | Google Scholar

Jahnke, S., Roussel, J., Hombach, T., Kochs, J., Fischbach, A., Huber, G., et al. (2016). Pheno seeder-a robot system for automated handling and phenotyping of individual seeds. Plant Physiol. 172, 1358–1370. doi: 10.1104/pp.16.01122

PubMed Abstract | Crossref Full Text | Google Scholar

Jaillais, B., Roumet, P., Pinson-Gadais, L., and Bertrand, D. (2015). Detection of Fusarium head blight contamination in wheat kernels by multivariate imaging. Food Control 54, 250–258. doi: 10.1016/j.foodcont.2015.01.048

Crossref Full Text | Google Scholar

Jiang, W., Wang, J., Lin, R., Chen, R., Chen, W., Xie, X., et al. (2024). Machine learning-based non-destructive terahertz detection of seed quality in peanut. Food Chemistry: X 23, 101675. doi: 10.1016/j.fochx.2024.101675

PubMed Abstract | Crossref Full Text | Google Scholar

Jin, C., Zhou, L., Pu, Y., Zhang, C., Qi, H., and Zhao, Y. (2025). Application of deep learning for high-throughput phenotyping of seed: a review. Artif. Intell. Rev. 58, 76. doi: 10.1007/s10462-024-11079-5

Crossref Full Text | Google Scholar

Kameswara Rao, N., Dulloo, M. E., and Engels, J. M. (2017). A review of factors that influence the production of quality seed for long-term conservation in genebanks. Genet. Resour. Crop Evol. 64, 1061–1074. doi: 10.1007/s10722-016-0425-9

Crossref Full Text | Google Scholar

Kim, G., Kim, G. H., Ahn, C.-K., Yoo, Y., and Cho, B.-K. (2013). Mid-infrared lifetime imaging for viability evaluation of lettuce seeds based on time-dependent thermal decay characterization. Sensors 13, 2986–2996. doi: 10.3390/s130302986

PubMed Abstract | Crossref Full Text | Google Scholar

Kolašinac, S. M., Mladenović, M., Pećinar, I., Šoštarić, I., Nedović, V., Miladinović, V., et al. (2025). Raman and FT-IR spectroscopy coupled with machine learning for the discrimination of different vegetable crop seed varieties. Plants 14, 1304. doi: 10.3390/plants14091304

PubMed Abstract | Crossref Full Text | Google Scholar

Krajewski, P., Chen, D., Ćwiek, H., van Dijk, A. D., Fiorani, F., Kersey, P., et al. (2015). Towards recommendations for metadata and data handling in plant phenotyping. J. Exp. Bot. 66, 5417–5427. doi: 10.1093/jxb/erv271

PubMed Abstract | Crossref Full Text | Google Scholar

Kranner, I., Kastberger, G., Hartbauer, M., and Pritchard, H. W. (2010). Noninvasive diagnosis of seed viability using infrared thermography. Proc. Natl. Acad. Sci. 107, 3912–3917. doi: 10.1073/pnas.0914197107

PubMed Abstract | Crossref Full Text | Google Scholar

Kumar, S., Sripathy, K., Bhaskar, K. U., and Vinesh, B. (2023). Principles of quality seed production. In Seed Science and Technology: Biology, Production, Quality (Singapore: Springer) 109–130. doi: 10.1007/978-981-19-5888-5_6

Crossref Full Text | Google Scholar

Kunishima, N., Takeda, Y., Hirose, R., Kalasová, D., Šalplachta, J., and Omote, K. (2020). Visualization of internal 3D structure of small live seed on germination by laboratory-based X-ray microscopy with phase contrast computed tomography. Plant Methods 16, 1–10. doi: 10.1186/s13007-020-0557-y

PubMed Abstract | Crossref Full Text | Google Scholar

Lee, H., Kim, M. S., Song, Y. R., Oh, C. S., Lim, H. S., Lee, W. H., et al. (2017). Non-destructive evaluation of bacteria-infected watermelon seeds using visible/near-infrared hyperspectral imaging. J. Sci. Food Agric. 97, 1084–1092. doi: 10.1002/jsfa.7832

PubMed Abstract | Crossref Full Text | Google Scholar

Lei, T., Li, Q., and Sun, D.-W. (2022). A dual AE-GAN guided THz spectral dehulling model for mapping energy and moisture distribution on sunflower seed kernels. Food Chem. 380, 131971. doi: 10.1016/j.foodchem.2021.131971

PubMed Abstract | Crossref Full Text | Google Scholar

Lemes, E. M. and Catão, H. C. R. M. (2024). Soybean seed Coat cracks and Green seeds—predisposing conditions, identification and management. Seeds 3, 133–148. doi: 10.3390/seeds3010011

Crossref Full Text | Google Scholar

Li, G., Ge, H., Jiang, Y., Zhang, Y., and Jin, X. (2025). Non-destructive detection of early wheat germination via deep learning-optimized terahertz imaging. Plant Methods 21, 1–18. doi: 10.1186/s13007-025-01393-6

PubMed Abstract | Crossref Full Text | Google Scholar

Li, C., Li, H., Liu, Z., Li, B., and Huang, Y. (2021). SeedSortNet: a rapid and highly effificient lightweight CNN based on visual attention for seed sorting. PeerJ Comput. Sci. 7, e639. doi: 10.7717/peerj-cs.639

PubMed Abstract | Crossref Full Text | Google Scholar

Liu, L., Wang, Z., Li, J., Zhang, X., and Wang, R. (2020). A non-invasive analysis of seed vigor by infrared thermography. Plants 9, 768. doi: 10.3390/plants9060768

PubMed Abstract | Crossref Full Text | Google Scholar

Liu, F., Yang, R., Chen, R., Guindo, M. L., He, Y., Zhou, J., et al. (2024). Digital techniques and trends for seed phenotyping using optical sensors. J. Advanced Res. 63, 1–16. doi: 10.1016/j.jare.2023.11.010

PubMed Abstract | Crossref Full Text | Google Scholar

Lürig, M. D., Donoughe, S., Svensson, E. I., Porto, A., and Tsuboi, M. (2021). Computer vision, machine learning, and the promise of phenomics in ecology and evolutionary biology. Front. Ecol. Evol. 9, 642774. doi: 10.3389/fevo.2021.642774

Crossref Full Text | Google Scholar

Lurstwut, B. and Pornpanomchai, C. (2011). Plant seed image recognition system (PSIRS). Int. J. Eng. Technol. 3, 600. doi: 10.7763/IJET.2011.V3.292

Crossref Full Text | Google Scholar

Mahajan, S., Das, A., and Sardana, H. K. (2015). Image acquisition techniques for assessment of legume quality. Trends Food Sci. Technol. 42, 116–133. doi: 10.1016/j.tifs.2015.01.001

Crossref Full Text | Google Scholar

Mascher, M., Schreiber, M., Scholz, U., Graner, A., Reif, J. C., and Stein, N. (2019). Genebank genomics bridges the gap between the conservation of crop diversity and plant breeding. Nat. Genet. 51, 1076–1081. doi: 10.1038/s41588-019-0443-6

PubMed Abstract | Crossref Full Text | Google Scholar

McDonald, L. and Panozzo, J. (2023). A review of the opportunities for spectral-based technologies in post-harvest testing of pulse grains. Legume Sci. 5, e175. doi: 10.1002/leg3.175

Crossref Full Text | Google Scholar

Medeiros, A., Martins, M. S., Silva, L., Pereira, M. D., León, M. J. Z., and Dias, D. (2020). X-ray imaging and digital processing application in non-destructive assessing of melon seed quality. J. Seed Sci. 42, e202042005. doi: 10.1590/2317-1545v42229761

Crossref Full Text | Google Scholar

Men, S., Yan, L., Liu, J., Qian, H., and Luo, Q. (2017). A classification method for seed viability assessment with infrared thermography. Sensors 17, 845. doi: 10.3390/s17040845

PubMed Abstract | Crossref Full Text | Google Scholar

Mira, S., Pirredda, M., Martín-Sánchez, M., Marchessi, J. E., and Martín, C. (2020). DNA methylation and integrity in aged seeds and regenerated plants. Seed Sci. Res. 30, 92–100. doi: 10.1017/S0960258520000021

Crossref Full Text | Google Scholar

Morales, M. A., Worral, H., Piche, L., Adeniyi, A. S., Dariva, F., Ramos, C., et al. (2024). High-Throughput phenotyping of seed quality traits using imaging and deep learning in dry pea. bioRxiv. 2003, 2024.03.05.583564. doi: 10.1101/2024.03.05.583564

Crossref Full Text | Google Scholar

Mortensen, A. K., Gislum, R., Jørgensen, J. R., and Boelt, B. (2021). The use of multispectral imaging and single seed and bulk near-infrared spectroscopy to characterize seed covering structures: methods and applications in seed testing and research. Agriculture 11, 301. doi: 10.3390/agriculture11040301

Crossref Full Text | Google Scholar

Moumni, M., Brodal, G., and Romanazzi, G. (2023). Recent innovative seed treatment methods in the management of seedborne pathogens. Food Secur. 15, 1365–1382. doi: 10.1007/s12571-023-01384-2

Crossref Full Text | Google Scholar

Mravlje, J., Regvar, M., and Vogel-Mikuš, K. (2021). Development of cold plasma technologies for surface decontamination of seed fungal pathogens: Present status and perspectives. J. Fungi 7, 650. doi: 10.3390/jof7080650

PubMed Abstract | Crossref Full Text | Google Scholar

Mukasa, P., Wakholi, C., Faqeerzada, M. A., Amanah, H. Z., Kim, H., Joshi, R., et al. (2022). Nondestructive discrimination of seedless from seeded watermelon seeds by using multivariate and deep learning image analysis. Comput. Electron. Agric. 194, 106799. doi: 10.1016/j.compag.2022.106799

Crossref Full Text | Google Scholar

Nadimi, M., Loewen, G., and Paliwal, J. (2022). Assessment of mechanical damage to flaxseeds using radiographic imaging and tomography. Smart Agric. Technol. 2, 100057. doi: 10.1016/j.atech.2022.100057

Crossref Full Text | Google Scholar

Nautiyal, P. (2009). Seed and seedling vigour traits in groundnut (Arachis hypogaea L.). Seed Sci. Technol. 37, 721–735. doi: 10.15258/sst.2009.37.3.19

Crossref Full Text | Google Scholar

Nguyen, G. N. and Norton, S. L. (2020). Genebank phenomics: A strategic approach to enhance value and utilization of crop germplasm. Plants 9, 817. doi: 10.3390/plants9070817

PubMed Abstract | Crossref Full Text | Google Scholar

Nonogaki, H. (2014). Seed dormancy and germination—emerging mechanisms and new hypotheses. Front. Plant Sci. 5, 233. doi: 10.3389/fpls.2014.00233

PubMed Abstract | Crossref Full Text | Google Scholar

Olesen, M. H., Carstensen, J. M., and Boelt, B. (2011). Multispectral imaging as a potential tool for seed health testing of spinach (Spinacia oleracea L.). Seed Sci. Technol. 39, 140–150. doi: 10.15258/sst.2011.39.1.12

Crossref Full Text | Google Scholar

Ouyang, Z., Fu, X., Zhong, Z., Bai, R., Cheng, Q., Gao, G., et al. (2024). An exploration of the influence of ZnO NPs treatment on germination of radish seeds under salt stress based on the YOLOv8-R lightweight model. Plant Methods 20, 110. doi: 10.1186/s13007-024-01238-8

PubMed Abstract | Crossref Full Text | Google Scholar

Pal, S. and Chattopadhyay, A. (2021). Hydration dynamics in biological membranes: Emerging applications of terahertz spectroscopy. J. Phys. Chem. Lett. 12, 9697–9709. doi: 10.1021/acs.jpclett.1c02576

PubMed Abstract | Crossref Full Text | Google Scholar

Pasikatan, M. and Dowell, F. (2001). Sorting systems based on optical methods for detecting and removing seeds infested internally by insects or fungi: a review. Applied Spectroscopy Reviews. 36, 399–416. doi: 10.1081/ASR-100107719

Crossref Full Text | Google Scholar

Plutenko, I., Radchuk, V., Mayer, S., Keil, P., Ortleb, S., Wagner, S., et al. (2025). MRI-Seed-Wizard: combining deep learning algorithms with magnetic resonance imaging enables advanced seed phenotyping. J. Exp. Bot. 76, 393–410. doi: 10.1093/jxb/erae408

PubMed Abstract | Crossref Full Text | Google Scholar

Powell, A. (2009)What is seed quality and how to measure it,” in Responding to the challenges of a changing world: The role of new plant varieties and high quality seed in agriculture. Proceedings of the Second World Seed Conference. (Rome) 8–10.

Google Scholar

Prasad, S. R. (2023). Testing seed for quality. In Seed Science and Technology: Biology, Production, Quality (Singapore: Springer) 299–335. doi: 10.1007/978-981-19-5888-5_13

Crossref Full Text | Google Scholar

Punthumast, P., Auttawaitkul, Y., Chiracharit, W., and Chamnongthai, K. (2012)Non-destructive Identification of unmilled rice using digital image analysis, in 2012 9th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology. (Phetchaburi, Thailand: IEEE) 1–4.

Google Scholar

Qin, Z., Zhang, Z., Hua, X., Yang, W., Liang, X., Zhai, R., et al. (2022). Cereal grain 3D point cloud analysis method for shape extraction and filled/unfilled grain identification based on structured light imaging. Sci. Rep. 12, 3145. doi: 10.1038/s41598-022-07221-4

PubMed Abstract | Crossref Full Text | Google Scholar

Rahman, A. and Cho, B.-K. (2016). Assessment of seed quality using non-destructive measurement techniques: a review. Seed Sci. Res. 26, 285–305. doi: 10.1017/S0960258516000234

Crossref Full Text | Google Scholar

Rawson, A. and Sunil, C. K. (2022). Recent advances in terahertz time-domain spectroscopy and imaging techniques for automation in agriculture and food sector. Food Analytical Methods 15, 498–526. doi: 10.1007/s12161-021-02132-y

Crossref Full Text | Google Scholar

Reddy, P., Guthridge, K. M., Panozzo, J., Ludlow, E. J., Spangenberg, G. C., and Rochfort, S. J. (2022). Near-infrared hyperspectral imaging pipelines for pasture seed quality evaluation: An overview. Sensors 22, 1981. doi: 10.3390/s22051981

PubMed Abstract | Crossref Full Text | Google Scholar

Reddy, P., Panozzo, J., Guthridge, K. M., Spangenberg, G. C., and Rochfort, S. J. (2023). Single seed near-infrared hyperspectral imaging for classification of perennial ryegrass seed. Sensors 23, 1820. doi: 10.3390/s23041820

PubMed Abstract | Crossref Full Text | Google Scholar

Reed, R. C., Bradford, K. J., and Khanday, I. (2022). Seed germination and vigor: ensuring crop sustainability in a changing climate. Heredity 128, 450–459. doi: 10.1038/s41437-022-00497-2

PubMed Abstract | Crossref Full Text | Google Scholar

Roitsch, T., Cabrera-Bosquet, L., Fournier, A., Ghamkhar, K., Jiménez-Berni, J., Pinto, F., et al. (2019). New sensors and data-driven approaches—A path to next generation phenomics. Plant Sci. 282, 2–10. doi: 10.1016/j.plantsci.2019.01.011

PubMed Abstract | Crossref Full Text | Google Scholar

Rousseau, D., Widiez, T., Di Tommaso, S., Rositi, H., Adrien, J., Maire, E., et al. (2015). Fast virtual histology using X-ray in-line phase tomography: application to the 3D anatomy of maize developing seeds. Plant Methods 11, 1–10. doi: 10.1186/s13007-015-0098-y

PubMed Abstract | Crossref Full Text | Google Scholar

Saletnik, A., Saletnik, B., and Puchalski, C. (2021). Overview of popular techniques of Raman spectroscopy and their potential in the study of plant tissues. Molecules 26, 1537. doi: 10.3390/molecules26061537

PubMed Abstract | Crossref Full Text | Google Scholar

Sanchez, L., Farber, C., Lei, J., Zhu-Salzman, K., and Kurouski, D. (2019). Noninvasive and nondestructive detection of cowpea bruchid within cowpea seeds with a hand-held Raman spectrometer. Analytical Chem. 91, 1733–1737. doi: 10.1021/acs.analchem.8b05555

PubMed Abstract | Crossref Full Text | Google Scholar

Sattar, M. A. and Laila, D. S. (2025). A review of ultrasound monitoring applications in agriculture. Front. Plant Sci. 16, 1620868. doi: 10.3389/fpls.2025.1620868

PubMed Abstract | Crossref Full Text | Google Scholar

Schmutzer, T., Bolger, M. E., Rudd, S., Chen, J., Gundlach, H., Arend, D., et al. (2017). Bioinformatics in the plant genomic and phenomic domain: The German contribution to resources, services and perspectives. J. Biotechnol. 261, 37–45. doi: 10.1016/j.jbiotec.2017.07.006

PubMed Abstract | Crossref Full Text | Google Scholar

Shakoor, N., Lee, S., and Mockler, T. C. (2017). High throughput phenotyping to accelerate crop breeding and monitoring of diseases in the field. Curr. Opin. Plant Biol. 38, 184–192. doi: 10.1016/j.pbi.2017.05.006

PubMed Abstract | Crossref Full Text | Google Scholar

Silva, V., Cicero, S. M., and Bennett, M. (2013). Associations between X-ray visualised internal tomato seed morphology and germination. Seed Sci. Technol. 41, 225–234. doi: 10.15258/sst.2013.41.2.05

Crossref Full Text | Google Scholar

Singh, P., Chatterjee, A., Rajput, L. S., Rana, S., Kumar, S., Nataraj, V., et al. (2021). Development of an intelligent laser biospeckle system for early detection and classification of soybean seeds infected with seed-borne fungal pathogen (Colletotrichum truncatum). Biosyst. Eng. 212, 442–457. doi: 10.1016/j.biosystemseng.2021.11.002

Crossref Full Text | Google Scholar

Singh, R. P., Prasad, P. V. V., and Reddy, K. R. (2015). Climate change: implications for stakeholders in genetic resources and seed sector. Adv. Agron. 129, 117–180. doi: 10.1016/bs.agron.2014.09.002

Crossref Full Text | Google Scholar

Singh, N., Rai, V., and Singh, N. K. (2020). Multi-omics strategies and prospects to enhance seed quality and nutritional traits in pigeonpea. Nucleus 63, 249–256. doi: 10.1007/s13237-020-00341-0

Crossref Full Text | Google Scholar

Smale, M. and Jamora, N. (2020). Valuing genebanks. Food Secur. 12, 905–918. doi: 10.1007/s12571-020-01034-x

Crossref Full Text | Google Scholar

Smith, E. and Dent, G. (2019). Modern Raman spectroscopy: a practical approach. (West Sussex, UK: John Wiley & Sons).

Google Scholar

Su, W. H. and Sun, D. W. (2018). Multispectral imaging for plant food quality analysis and visualization. Compr. Rev. Food Sci. Food Saf. 17, 220–239. doi: 10.1111/1541-4337.12317

PubMed Abstract | Crossref Full Text | Google Scholar

Sundareswaran, S., Ray Choudhury, P., Vanitha, C., and Yadava, D. K. (2023). Seed quality: Variety development to planting—An overview. In Seed Science and Technology: Biology, Production, Quality (Singapore: Springer) 1–16. doi: 10.1007/978-981-19-5888-5_1

Crossref Full Text | Google Scholar

Tan, K., Wang, R., Li, M., and Gong, Z. (2019). Discriminating soybean seed varieties using hyperspectral imaging and machine learning. J. Comput. Methods Sci. Eng. 19, 1001–1015. doi: 10.3233/JCM-193562

Crossref Full Text | Google Scholar

Terskikh, V. V., Feurtado, J. A., Ren, C., Abrams, S. R., and Kermode, A. R. (2005). Water uptake and oil distribution during imbibition of seeds of western white pine (Pinus monticola Dougl. ex D. Don) monitored in vivo using magnetic resonance imaging. Planta 221, 17–27. doi: 10.1007/s00425-004-1426-z

PubMed Abstract | Crossref Full Text | Google Scholar

Togninalli, M., Wang, X., Kucera, T., Shrestha, S., Juliana, P., Mondal, S., et al. (2023). Multi-modal deep learning improves grain yield prediction in wheat breeding by fusing genomics and phenomics. Bioinformatics 39, btad336. doi: 10.1093/bioinformatics/btad336

PubMed Abstract | Crossref Full Text | Google Scholar

Tu, K., Wen, S., Cheng, Y., Zhang, T., Pan, T., Wang, J., et al. (2021). A non-destructive and highly efficient model for detecting the genuineness of maize variety'JINGKE 968′ using machine vision combined with deep learning. Comput. Electron. Agric. 182, 106002. doi: 10.1016/j.compag.2021.106002

Crossref Full Text | Google Scholar

Tuomainen, T. V., Himanen, K., Helenius, P., Kettunen, M. I., and Nissi, M. J. (2022). Quantitative magnetic resonance imaging of Scots pine seeds and the assessment of germination potential. Can. J. For. Res. 52, 685–695. doi: 10.1139/cjfr-2021-0273

Crossref Full Text | Google Scholar

Upretee, P., Bandara, M. S., and Tanino, K. K. (2024). The role of seed characteristics on water uptake preceding germination. Seeds 3, 559–574. doi: 10.3390/seeds3040038

Crossref Full Text | Google Scholar

Vaughton, G. and Ramsey, M. (2001). Relationships between seed mass, seed nutrients, and seedling growth in Banksia cunninghamii (Proteaceae). Int. J. Plant Sci. 162, 599–606. doi: 10.1086/320133

Crossref Full Text | Google Scholar

Verma, P. and Majee, M. (2013). Seed germination and viability test in tetrazolium (TZ) assay. Bio-protocol 3, e884–e884. doi: 10.21769/BioProtoc.884

Crossref Full Text | Google Scholar

Volk, G. M., Byrne, P. F., Coyne, C. J., Flint-Garcia, S., Reeves, P. A., and Richards, C. (2021). Integrating genomic and phenomic approaches to support plant genetic resources conservation and use. Plants 10, 2260. doi: 10.3390/plants10112260

PubMed Abstract | Crossref Full Text | Google Scholar

Vrešak, M., Halkjaer Olesen, M., Gislum, R., Bavec, F., and Ravn Jørgensen, J. (2016). The use of image-spectroscopy technology as a diagnostic method for seed health testing and variety identification. PloS One 11, e0152011. doi: 10.1371/journal.pone.0152011

PubMed Abstract | Crossref Full Text | Google Scholar

Wakholi, C., Kandpal, L. M., Lee, H., Bae, H., Park, E., Kim, M. S., et al. (2018). Rapid assessment of corn seed viability using short wave infrared line-scan hyperspectral imaging and chemometrics. Sensors Actuators B: Chem. 255, 498–507. doi: 10.1016/j.snb.2017.08.036

Crossref Full Text | Google Scholar

Walcott, R., McGee, D., and Misra, M. (1998). Detection of asymptomatic fungal infections of soybean seeds by ultrasound analysis. Plant Dis. 82, 584–589. doi: 10.1094/PDIS.1998.82.5.584

PubMed Abstract | Crossref Full Text | Google Scholar

Walters, C., Wheeler, L. M., and Grotenhuis, J. M. (2005). Longevity of seeds stored in a genebank: species characteristics. Seed Sci. Res. 15, 1–20. doi: 10.1079/SSR2004195

Crossref Full Text | Google Scholar

Wang, Z., Fan, S., Wu, J., Zhang, C., Xu, F., Yang, X., et al. (2021). Application of long-wave near infrared hyperspectral imaging for determination of moisture content of single maize seed. Spectrochimica Acta Part A: Mol. Biomolecular Spectrosc. 254, 119666. doi: 10.1016/j.saa.2021.119666

PubMed Abstract | Crossref Full Text | Google Scholar

Wang, Q., Zhang, D., Zhao, L., Liu, J., Shang, B., Yang, W., et al. (2022). Metabolomic analysis reveals insights into deterioration of rice quality during storage. Foods 11, 1729. doi: 10.3390/foods11121729

PubMed Abstract | Crossref Full Text | Google Scholar

Wang, D., Zheng, J., Sarsaiya, S., Jin, L., and Chen, J. (2024). Unveiling terahertz wave stress effects and mechanisms in Pinellia ternata: Challenges, insights, and future directions. Physiologia Plantarum 176, e14195. doi: 10.1111/ppl.14195

PubMed Abstract | Crossref Full Text | Google Scholar

Wen, T., Li, J.-H., Wang, Q., Gao, Y.-Y., Hao, G.-F., and Song, B.-A. (2023). Thermal imaging: The digital eye facilitates high-throughput phenotyping traits of plant growth and stress responses. Sci. Total Environ. 899, 165626. doi: 10.1016/j.scitotenv.2023.165626

PubMed Abstract | Crossref Full Text | Google Scholar

Weng, S., Hu, X., Wang, J., Tang, L., Li, P., Zheng, S., et al. (2021). Advanced application of Raman spectroscopy and surface-enhanced Raman spectroscopy in plant disease diagnostics: a review. J. Agric. Food Chem. 69, 2950–2964. doi: 10.1021/acs.jafc.0c07205

PubMed Abstract | Crossref Full Text | Google Scholar

Whitehouse, K. J., Hay, F. R., and Lusty, C. (2020). Why seed physiology is important for genebanking. Plants 9, 584. doi: 10.3390/plants9050584

PubMed Abstract | Crossref Full Text | Google Scholar

Wonggasem, K., Wongchaisuwat, P., Chakranon, P., and Onwimol, D. (2024). Utilization of machine learning and hyperspectral imaging technologies for classifying coated maize seed vigor: A case study on the assessment of seed DNA repair capability. Agronomy 14, 1991. doi: 10.3390/agronomy14091991

Crossref Full Text | Google Scholar

Xia, Y., Xu, Y., Li, J., Zhang, C., and Fan, S. (2019). Recent advances in emerging techniques for non-destructive detection of seed viability: A review. Artif. Intell. Agric. 1, 35–47. doi: 10.1016/j.aiia.2019.05.001

Crossref Full Text | Google Scholar

Xiao, Q., Bai, X., Zhang, C., and He, Y. (2022). Advanced high-throughput plant phenotyping techniques for genome-wide association studies: A review. J. advanced Res. 35, 215–230. doi: 10.1016/j.jare.2021.05.002

PubMed Abstract | Crossref Full Text | Google Scholar

Xing, W., Li, Y., Zhou, L., Hong, H., Liu, Y., Luo, S., et al. (2025). Deciphering seed deterioration: molecular insights and priming strategies for revitalizing aged seeds. Plants 14, 1730. doi: 10.3390/plants14111730

PubMed Abstract | Crossref Full Text | Google Scholar

Yang, W., Feng, H., Zhang, X., Zhang, J., Doonan, J. H., Batchelor, W. D., et al. (2020). Crop phenomics and high-throughput phenotyping: past decades, current challenges, and future perspectives. Mol. Plant 13, 187–214. doi: 10.1016/j.molp.2020.01.008

PubMed Abstract | Crossref Full Text | Google Scholar

Yu, J.-K. and Chung, Y. S. (2024). Image-based digitalization of germplasm stock: overcoming its limitations. Korean J. Breed. Sci. 56, 293–300. doi: 10.9787/KJBS.2024.56.3.293

Crossref Full Text | Google Scholar

Zdunek, A., Adamiak, A., Pieczywek, P. M., and Kurenda, A. (2014). The biospeckle method for the investigation of agricultural crops: A review. Optics Lasers Eng. 52, 276–285. doi: 10.1016/j.optlaseng.2013.06.017

Crossref Full Text | Google Scholar

Zhang, C., Si, Y., Lamkey, J., Boydston, R. A., Garland-Campbell, K. A., and Sankaran, S. (2018). High-throughput phenotyping of seed/seedling evaluation using digital image analysis. Agronomy 8, 63. doi: 10.3390/agronomy8050063

Crossref Full Text | Google Scholar

Zhang, Y. and Zhang, N. (2018). Imaging technologies for plant high-throughput phenotyping: a review. Front. Agric. Sci. Eng. 5, 406–419. doi: 10.15302/J-FASE-2018242

Crossref Full Text | Google Scholar

Zhong, Z., Wang, X., Yin, X., Tian, J., and Komatsu, S. (2021). Morphophysiological and proteomic responses on plants of irradiation with electromagnetic waves. Int. J. Mol. Sci. 22, 12239. doi: 10.3390/ijms222212239

PubMed Abstract | Crossref Full Text | Google Scholar

Zhu, S., Chao, M., Zhang, J., Xu, X., Song, P., Zhang, J., et al. (2019). Identification of soybean seed varieties based on hyperspectral imaging technology. Sensors 19, 5225. doi: 10.3390/s19235225

PubMed Abstract | Crossref Full Text | Google Scholar