论文部分内容阅读
Hyper spectrum remote sensing with fine spectrum information is an efficient method to estimate the verticillium wilt of cotton. The research was conducted in Xinjiang, the largest cotton plant region of China, by using the data which were collected both by canopy spectrum infected with verticillium wilt and severity level (SL) in the year 2005-2006. The quantitative correlation was analyzed between SL and canopy of reflectance spectrum or derivative spectrum reflectance.The results indicated that spectrum characteristics of cotton canopy infected with verticilliurn wilt changed regularly with the increase of SL in different periods and varieties. Spectrum reflectance increased in the visible light region (620-700 nm)with the increase of the SL, which inverted in near-infrared region and was extremely significant in the region of (780-1300nm). When SL attained b2 (DI = 25), cotton canopy infected with verticillium wilt was used as a watershed and diagnosed index in the beginning stages of the disease. The results also indicated that there were marked different characteristics of the first derivative spectrum in these SL, it changed significantly in the red edge ranges (680-760 nm) with different SL,i.e., red edge swing decreased, and red edge position equally moved to the blue. In this study 1 001-1 110 nm and 1 205-1 320 nm were selected out as sensitive bands for SL of canopy. Inversion models established for estimating cotton canopy infected with verticillium wilt reached the most significant level. Finally, the different spectrum characteristics of cotton canopy infected with verticillium wilt were marked, some inversion models were established, which could estimate SL of canopy infected with verticillium wilt. The best recognized model was the first derivative spectra at (FD731 nm-FD 1317nm), and it might be used to forecast the position of cotton canopy infected with verticillium wilt quantitatively.