Accuracy Comparison of Estimation on Cotton Leaf and Plant Nitrogen Content Based on UAV Digital Image under Different Nutrition Treatments

The rapid, accurate estimation of leaf nitrogen content (LNC) and plant (PNC) in cotton a non-destructive way is great significance to the nutrient management fields. RGB images fields Shihezi (China) were obtained by using low-cost unmanned aerial vehicle (UAV) with visible-light digital camera. Combined data LNC PNC different growth stages, correlation between N visible light vegetation indices (VIs) was analyzed, then Random Forest (RF), Support Vector Machine (SVM), Back Propagation Neural Network (BP), stepwise multiple linear regression (SMLR) used develop models at stages. accuracy model assessed coefficient determination (R2), root mean squared error (RMSE), relative square (rRMSE), so as determine optimal estimated stage best model. results showed that VIs stronger than PNC, decreased continuously development higher peak squaring stage. Among four algorithms, (R2 = 0.9001, RMSE 1.2309, rRMSE 2.46% for establishment, R2 0.8782, 1.3877, 2.82% validation) when applying RF whole stages could be later due its accuracy. this study there potential an affordable UAV-based system produce predicted maps are representative current field status.