With the development of remote sensing technology, medium and high resolution images are playing an important role in vegetation monitoring. In order to define the advantages of high resolution sensors in extracting vegetation information of different ecosystems, this research select urban and forest as target area in Olunchun autonomous banner, Inner Mongolia. Two kinds of sensor images, GF1-WFV and Landsat8-OLI, were used as comparison dataset to explore the difference of vegetation extracting using different spatial resolution in these two ecosystems. The results showed that: (1) compared with GF-1, Landsat 8 showed opposite overestimation and underestimation of its vegetation index in urban areas and forest areas, while the Normalized Difference Vegetation Index (NDVI) and Soil-adjusted Vegetation (SAVI) of GF-1 in urban areas were 5.69% and 1.41% higher than that of Landsat 8, and 0.77% and 5.86% lower in forest areas. (2) High-resolution images avoided the green vegetation of urban (71.30% and 71.31% of GF-1 vegetation, 58.28% and 58.30% of Landsat 8) and bare land and roads of forest (94.97% and 94.92% of GF-1 vegetation, 95.00% and 94.99% of Landsat 8) were omitted. (3) In terms of graded area, compared with GF-1, Landsat 8 underestimates 6.67% and 6.77% of low coverage levels in urban areas, and overestimates 12.11% and 12.47% of high coverage levels in forest areas. This research reflects it is more necessary to use high-resolution images as vegetation monitoring tools in low green built-up areas and high density forest areas.