Identification of Wheat Canopy Structure Using Hyperspectral Data
摘要
Some winter wheat varieties were selected in this experiment. The results were as follows: 1) Leaf orientation value (LOV) and leaf area index (LAI) of wheat had different contributions to canopy spectral reflectance (CSR). For example, LOV affected greatly canopy spectra more than LAI did in jointing stage, but LAI had a greater effect on CSR than LOV did after the ground was near to be covered completely. 2) Twenty treatments including different varieties and densities were arranged in this experiment, and the result of cluster analysis showed that all these treatments can be parted into four clusters according to LAI and LOV: varieties with erect leaves and low LAI (denoted as A), varieties with erect leaves and high LAI (denoted as B), varieties with horizontal leaves and low LAI (denoted as C), varieties with horizontal leaves and high LAI (denoted as D). Their CSR had difference in 400-700 nm and 700-1 150 nm at jointing stage, especially in different plant types. 3) There was obvious distribution difference among different clusters in scatter plot (X=ΔR890, Y=R890),ΔR890 was the reflectance increment from jointing to booting stage. It was seen from the Y-axis direction that R890 of horizontal varieties were higher than the erect ones, and seen from the X-axis direction that the greater ΔR890 was, the lower LAI one within the same plant type varieties, which indicted that the combination of plant-type and the population magnitude can be initially identified by this method.