Estimating leaf chlorophyll content of winter wheat from UAV multispectral images using machine learning algorithms under different species, growth stages, and nitrogen stress conditions

Authors: ZHANG, LIYUAN - Jiangsu University; WANG, AICHEN - Jiangsu University; ZHANG, HUIYUE - Jiangsu University; ZHU, QINGZHEN - Jiangsu University; Zhang, Huihui; SUN, WEIHONG - Jiangsu University; NIU, YAXIAO - Jiangsu University

Submitted to: Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/29/2024
Publication Date: 7/1/2024
Citation: Zhang, L., Wang, A., Zhang, H., Zhu, Q., Zhang, H., Sun, W., Niu, Y. 2024. Estimating leaf chlorophyll content of winter wheat from UAV multispectral images using machine learning algorithms under different species, growth stages, and nitrogen stress conditions. Agriculture. 14(7). Article e1064. https://doi.org/10.3390/agriculture14071064.
DOI: https://doi.org/10.3390/agriculture14071064

Interpretive Summary: This 2023 study on winter wheat sought to develop a reliable model for estimating leaf chlorophyll content (LCC), a vital indicator for assessing crop photosynthesis and nutrient status. Utilizing machine learning algorithms like support vector machine (SVM) and random forest (RF), UAV-derived multispectral data were linked to ground truth LCC, revealing strong correlations with a few vegetative indices. While all algorithms performed well during training, SVM demonstrated superior generalization ability in validation. Notably, the study emphasized the crucial impact of the growth stage on model performance. Ultimately, this research is significant as it showcases the successful application of UAV data for mapping LCC in winter wheat across diverse conditions, offering valuable insights for precision nitrogen fertilization management.

Technical Abstract: The rapid and accurate estimation of leaf chlorophyll content (LCC), an important indicator of crop photosynthetic capacity and nutritional status, is of great significance for precise nitrogen fertilization management. To explore whether a versatile regression model exists that could be successfully used to estimate crop LCC under different species, growth stages, and nitrogen stress conditions, a study was conducted in 2023 across the growing season for winter wheat with five species and five nitrogen application levels. Two machine learning regression algorithms, support vector machine (SVM) and random forest (RF), were used to establish the bridge between UAV-derived multispectral vegetation indices and ground truth LCC (relative chlorophyll content, SPAD), taking the multivariate linear regression (MLR) algorithm as reference. The results show that the VARI, VEG, and NDVI had the highest correlation with ground truth LCC with Pearson’s correlation coefficient of 0.95. All three regression algorithms (MLR, RF, and SVM) successfully estimated the LCC with the averaged R2 varied from 0.932 to 0.944 and averaged RMSE varied from 3.96 to 4.37 which were obtained based on the training dataset. However, when these models were used to estimate LCC based on the stand-alone validation dataset, the best generalization ability was found for SVM with the lowest R2 of 0.60 and highest RSME of 3.86 obtained in the dough stage through the whole validation process. Compared to winter wheat species and nitrogen application levels, the growth stage had the greatest influence on the generalization ability of LCC estimation models. Overall, this study demonstrated that the combination of UAV-derived multispectral VIs and the SVM regression algorithm could be successfully applied to map the LCC of winter wheat under different species, growth stages, and nitrogen stress conditions. Ultimately, this research is significant as it shows the successful application of UAV data for mapping LCC in winter wheat across diverse conditions, offering valuable insights for precision nitrogen fertilization management.