Multi-Angle Detection of Spatial Differences in Tea Physiological Parameters

Chlorophyll and nitrogen contents were used as leaf physiological parameters. Based on multispectral images from multiple detection angles the stoichiometric data of tea (Camellia sinensis) leaves in different positions plants, spatial differences parameters explored, full channel difference vegetation index was established to effectively remove soil shadow noise. Support vector machine, random forest (RF), partial least square, back-propagation algorithms canopy scales then train parameter model. Finally, effects obtained top, middle, bottom analysed compared. The results revealed distinct individual plants. content lowest at top relatively high middle bottom; highest low bottom. horizontal distribution similar, i.e., values east south high, whereas those west north low. accuracy scale better than that scale; model trained by RF algorithm had comprehensive accuracy. coefficient determination between predicted measured spad-502 plus instrument (R2) = 0.79, root mean square error (RMSE) 0.11. result for value R2 0.36 RMSE 0.03. image taken 60° generally superior other shooting angles. These can guide high-precision remote sensing