Extraction of Diseases and Insect Pests for Tobacco Based on Hyperspectral Remote Sensing

To study the feasibility of monitoring the diseases and insect pests in tobacco using hyperspectral remote sensing, leaf spectrum of tobacco infected with diseases and insect pests at different severity levels was measured by using ASD hand-held spectroradiometers. The reflectance data was transformed by the method of the first differential coefficient. Meanwhile, the correlation between severity levels and spectral data was analyzed. The results suggested that the wavelength bands between 631 nm and 638 nm, 696 nm and 733 nm as well as 864 nm were selected out as sensitive bands region to the severity levels. The leaf spectral reflectance de- creased due to the damage of diseases and insect pests. Moreover, the spectrum of tobacco leaf infected diseases and insect pests moved to the direction of long wave. This research is the basis to monitor the diseases and insect pests in tobacco, and it has a practical significance for applying remote sensing monitoring and determi- ning the appropriate control time.