Extraction of vegetation fraction based on the dimidiate pixel model and vegetation index transform plan

In order to derive high accuracy vegetation fraction, the article combines QuickBird image data and TM image data, based on normalized difference vegetation index (NDVI) of dimidiate pixel model and vegetation index transform plan, then derive vegetation fraction of desertification area, take an example of half-shrub Artemisia ordosica Krasch. Lingwu in Ningxia. By field surveys, the result of estimation of vegetation fraction is proved to be valid, the correlation coefficient between field survey data and extracted values from the vegetation coverage imagery was 0. 881, especially in fixed sandy land. The results show that using this improved model to estimate vegetation fraction is feasible. The model can be used to many scales remote sensing data, and Northwest Territories in China.