Machine Learning in the Analysis of Multispectral Reads in Maize Canopies Responding to Increased Temperatures and Water Deficit

Abstract

Real-time monitoring of crop responses to environmental deviations represents a neavenue for applications of remote and proximal sensing. Combining the high-throughput devicewith novel machine learning (ML) approaches shows promise in the monitoring of agriculturaproduction. The 3 × 2 multispectral arrays with responses at 610 and 680 nm (red), 730 and 760 nm (red-edge) and 810 and 860 nm (infrared) spectra were used to assess the occurrence of leaf rollin(LR) in 545 experimental maize plots measured four times for calibration dataset (n = 2180) an145 plots measured once for external validation. Multispectral reads were used to calculate 15 simplnormalized vegetation indices. Four ML algorithms were assessed: single and multilayer perceptron (SLP and MLP), convolutional neural network (CNN) and support vector machines (SVM) ithree validation procedures, which were stratified cross-validation, random subset validation anvalidation with external dataset. Leaf rolling occurrence caused visible changes in spectral responseand calculated vegetation indexes. All algorithms showed good performance metrics in stratifiecross-validation (accuracy >80%). SLP was the least efficient in predictions with external datasetwhile MLP, CNN and SVM showed comparable performance. Combining ML with multispectrasensing shows promise in transition towards agriculture based on data-driven decisions especiallconsidering the novel Internet of Things (IoT) avenues.