Estimating Chlorophyll Content of Zizania latifolia with Hyperspectral data and Random Forest

The amount of chlorophyll in a plant useful to indicate its physiological activity and then changes content have been used as good indicator disease well nutritional environmental stresses on plants. Chlorophyll estimation is one the most applications hyperspectral remote sensing data. aim this study evaluate dimensionality reduction for estimating contents from reflectance. Random Forest (RF) has applied assess biochemical properties such data; however, an approach integrating with techniques not fully evaluated. A total 200 Zizania latifolia leaves 5 treatments Shizuoka University field were measured reflectance content. then, regression models generated based RF three methods including principal component analysis, kernel analysis independent analysis. This research clarified that PCA best method Latifolia RMSE value 5.65 ± 0.58 μg cm-2.