ORIGINAL RESEARCH article
Front. Plant Sci.
Sec. Plant Bioinformatics
Volume 16 - 2025 | doi: 10.3389/fpls.2025.1677883
This article is part of the Research TopicInnovative Techniques for Precision Agriculture and Big DataView all 6 articles
A Single-Seed Uniform Distribution and Spreading Device for the Real-Time Detection of Ambrosia artemisiifolia and Ambrosia trifida Seeds in Imported Soybeans
Provisionally accepted
- 1Northwest A&F University, Xianyang, China
- 2Chinese Academy of Agricultural Sciences Agricultural Genomics Institute at Shenzhen, Shenzhen, China
- 3Shaanxi Agricultural Machinery Research Institute Co., Ltd, Xianyang, China
- 4Guangzhou Customs Technology Center, Guangzhou, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Due to insufficient domestic production to meet overall demand, China heavily relies on imported soybeans, which are frequently contaminated with quarantine weed seeds. The spread of these seeds poses a threat to the ecosystem, underscoring the urgent need for stringent detection measures. In this study, a real-time detection device was developed to ensure that seeds were evenly distributed across the conveyor belt. This approach maximized seed visibility within the camera's field of view while preventing mutual occlusion. As a result, the recognition and counting of Ambrosia artemisiifolia, Ambrosia trifida, and imported soybean seeds were successfully achieved. The detection device comprises a single-seed uniform distribution unit featuring parabolic seed sockets, integrated with an embedded system. After the seeds were arranged in a single layer, a detection camera captured seed images, which were processed by the YOLO_P2 model for seed detection and counting. The device was then optimized using the Taguchi experimental design method. The experimental results were analyzed through signal-to-noise ratio, mean, and variance, revealing that the speed of the seed-spreading roller and the conveyor motor significantly affect the uniform distribution of seeds. Based on these findings, the optimal configuration for the detection device was proposed. Following validation experiments, the average detection accuracies for Ambrosia trifida and Ambrosia artemisiifolia seeds were 95.73% and 94.41%, respectively. The average detection time was 7.6 minutes per sample. These results demonstrate that the detection device offers a practical solution for seed quarantine inspection, combining high-throughput performance with real-time capability and cost-effectiveness.
Keywords: soybean quarantine, Device design, parabolic seed socket, Taguchi experimental design, deep learning
Received: 01 Aug 2025; Accepted: 12 Sep 2025.
Copyright: © 2025 Liu, Qiao, Qiang, Zhang, Tian, Gu and Chen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Shuo Zhang, Northwest A&F University, Xianyang, China
Disclaimer: 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.