AUTHOR=Wu Tianao , Zhang Wei , Wu Shuyu , Cheng Minghan , Qi Lushang , Shao Guangcheng , Jiao Xiyun TITLE=Retrieving rice (Oryza sativa L.) net photosynthetic rate from UAV multispectral images based on machine learning methods JOURNAL=Frontiers in Plant Science VOLUME=Volume 13 - 2022 YEAR=2023 URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.1088499 DOI=10.3389/fpls.2022.1088499 ISSN=1664-462X ABSTRACT=Photosynthesis is the key physiological activity in the process of crop growth and plays an irreplaceable role in carbon assimilation and yield formation. This study extracted rice (Oryza sativa L.) canopy reflectance based on the UAV multispectral images and analyzed the correlation between 25 vegetation index (VIs), 3 textural index (TIs) and net photosynthetic rate (Pn) at different growth stages. Linear regression (LR), support vector regression (SVR), gradient boosting decision tree (GBDT), random forest (RF) and multilayer perceptron neural networks (MLP) models were employed for Pn estimation and the modelling accuracy were compared under the input condition of VIs, VIs combined with TIs, and fusion of VIs, TIs with plant height (PH) and SPAD. The results showed that VIs and TIs generally had relatively best correlation with Pn at jointing-booting stage and the number of VIs with significant correlation(P<0.05) was the largest. Therefore, the employed models could achieve the highest overall accuracy (coefficient of determination, R2 0.383-0.938). However, as the growth stage progressed, the correlation gradually weakened and resulted in accuracy decrease (R2 of 0.258-0.928 and 0.125-0.863 at heading-flowering and ripening stages, respectively). Among the tested models, GBDT and RF models could attain the best performance based on only VIs input (with R2 ranging 0.863-0.938 and 0.815-0.872, respectively). Furthermore, the fusion input of VIs, TIs with PH and SPAD could more effectively improve the model accuracy (R2 increased by 0.049-0.249, 0.063-0.470, 0.113-0.471 respectively for 3 growth stages) compared with the input combination of VIs and TIs (R2 increased by 0.015-0.090, 0.001-0.139, 0.023-0.114). Therefore, the GBDT and RF model with fused input could be highly recommended for rice Pn estimating and the methods could also provide reference for Pn monitoring and further yield prediction at field scale.