Monitoring Rice Leaves Blast Severity with Hyperspectral Reflectance

The reflectance of rice that infects different severity leaf blasts was measured through artificial inoculation and disease index (DI) of the rice corresponding to the spectra which were acquired in the field. The correlation between DI and the first derivative data was analyzed. The estimation models of DI were built using regression methods, and RMSE were used to test its precision. The result showed that, at the leaf level, rice leaf blasts highly sensitive to 600~700 nm and 720~1000 nm of hyperspectral in the regions of 400~1000 nm, while sensitive to 400~1000 nm of hyperspectral at canopy level. There was significantly negative correlation between DI and the first derivative data in the regions of 700~750 nm. And the 16 regression models were built with leaf hyperspectral index and canopy hyperspectral index. It provided theoretic foundation to further monitor rice leaf blasts at large scale using airborne and airspace remote sensing.