Abstract: With high resolution QuickBird image, based on the analysis of the current methods of measuring vegetation coverage degree(f_c),this paper has developed the dimidiate pixel model for quantifying vegetation fraction from normalized difference vegetation index (NDVI). In the improved model, f_c was calculated by the formula: f_c=(NDVI-NDVI_soil)/(NDVI_veg -NDVI_soil), where NDVI_soiland NDVI_veg represent the NDVI value of pure pixel of barren soil and vegetation, respectively. Using this model to monitor land, the results showed that the monitor effect for vegetation coverage degree of forest land converted from cropland was limited by the size of crown of a tree. It was not good for the small trees whose age were not more than 1 year, but it was good for these trees whose age were more than 2.5 years, of which the average estimated accuracy was more than 83% in the study region. So it will be feasible to use this improved model to monitor vegetation coverage degree of forestland converted from cropland (tree age is more than 2.5 years) from QuickBird data.

Key words: QuickBird image, vegetation coverage degree, forestland converted from cropland, NDVI, the dimidiate pixel model