基于无人机遥感的茶园多胁迫分层监测方法研究OA

Tea[Camellia sinensis(L.)O.Kuntze]is an important economic crop in China.Its production process is highly susceptible to stresses such as pests and diseases,which subsequently lead to a reduction in yield and quality.Accurate monitoring of stress conditions in tea garden is therefore essential for precision and smart management.This study focused on three typical stresses:tea geometrid(Ectropis obliqua),heat stress and anthracnose(Colletotrichum camelliae),and proposed a stepwise multi-stress monitoring method based on unmanned aerial vehicle(UAV)remote sensing.The research first focused on the characteristics of tea garden ridge-and-furrow structures.By combining a decision tree and edge detection(DT-ED)algorithm,which utilizes the RedEdge band,high-precision extraction of tea rows was achieved.Subsequently,considering the spatial distribution differences of stress within tea garden plots,a plot type discrimination model was constructed based on the coefficient of variation(CV)of the plot's spectrum and linear discriminant analysis(LDA).This model successfully categorized plots into entirely healthy plot(EHTP),entirely stressed plot(ESTP),and partially stressed plot(PSTP),achieving an overall accuracy of 94.7％.Based on this classification,a differentiated strategy was applied:UAV five-point sampling was used for stress assessment and health validation in ESTP and EHTP plots,while a two-step approach of"abnormal zone detection-stress type identification"was applied to PSTP plots.The abnormal zones were delineated using two-stage clustering strategy.Stress type classification was then carried out using algorithms such as support vector machine(SVM),k-nearest neighbors(KNN),and multilayer perceptron(MLP).The results show that the MLP achieved the best performance,with an overall accuracy of 92.3％.The findings demonstrate that the proposed multi-step monitoring method can effectively improve the accuracy and efficiency of multi-stress identification in tea garden,providing technical support for smart tea garden management and offering a methodological reference for other economic crops.