Abstract:

The use of hyperspectral remote sensing to construct an inversion model of agarwood chlorophyll content and leaf area index is the key to accurate diagnosis of agarwood tree growth and health.Based on the experimental plot of agarwood,the spectral reflectance of the canopy of 6-year-old agarwood and its corresponding chlorophyll content and leaf area index were measured.The result shows:1)There was a certain correlation between the spectral reflectance of agarwood canopy and the chlorophyll content and leaf area index of agarwood leaves,and the correlation varied with different parameters;2)Correlation analysis showed that the chlorophyll content was the most closely related to the spectral reflectance in the infrared bands (760,761,759,765,764 nm),and the leaf area index was closely related to the spectral reflectance in the infrared band (778,777,779,776,782 nm);3)Compared with the fitting effect of vegetation index and characteristic band,it was concluded that the regression model of chlorophyll content and leaf area index based on neural network had the best prediction effect.Therefore,it is believed that hyperspectral technology combined with BP neural network method can monitor the dynamic changes of parameters well,such as chlorophyll content and leaf area index in agarwood.

Key words: hyperspectral, BP neural network, chlorophyll content, agarwood