The invention relates to a
tea tree nitrogen nutrition diagnosis method based on unmanned aerial vehicle multi-source
remote sensing. The method comprises the following steps: S1, setting a gradient
nitrogen application experiment; s2, collecting
tea tree shoot samples, and measuring
biomass and
nitrogen content data of overground parts; collecting and preprocessing
tea tree canopy remote sensing data; s3, constructing a critical nitrogen
dilution curve, and calculating to obtain a nitrogen
nutrition index NNI; s4, constructing a multi-source fusion
feature set; s5, performing feature optimization selection on the multi-source
remote sensing fusion
feature set; s6, taking the optimal subset in S5 as input, taking the NNI in S3 as output, training an NNI
estimation model by utilizing a
machine learning
algorithm, and selecting and determining an
optimal estimation model through model precision; s7, performing inversion and
spatial mapping on the
tea garden NNI, and performing nitrogen
nutrition status diagnosis according to the NNI value; and S8, topdressing the
nitrogen fertilizer in a flat cabinet for the diagnosed nitrogen-deficient area. The method can realize large-area, lossless and rapid diagnosis of the nitrogen condition of the
tea garden, guide variable fertilization, save
fertilizer and increase yield, and has remarkable
environmental protection benefits.