Information on vegetation growth can be effectively characterized by the Normalized Difference Vegetation Index (NDVI), in which the long-term NDVI plays a significant role in analyzing global and local vegetation change patterns. The major purpose of this study is to elucidate the characteristics of vegetation coverage changes in the Yarlung Zangbo River Basin, and detect the vegetation pattern changes with a series of Landsat satellite data from Google Earth Engine, applying the LandTrendr time series segmentation algorithm. The results indicated that from 1985 to 2018, (1) with only few regional declines, the overall NDVI showed an upward trend. The intensity of NDVI change increases gradually from upstream to downstream, with the most significant change between 1986 to 1990 and less significant change between 2001 to 2017. (2) 95% of the NDVI disturbance ranged between 0 and 0.42 and the average duration of the disturbance was 4.96 years. 95% of the NDVI recovery changed between 0 and 0.4 and the average duration of the recovery was 12.55 years. (3) The disturbance event had a declined NDVI and the decline rate before 2000 was slower than the decline rate after 2000. The recovery event had an increased NDVI and the increase rate before 2000 was faster than the increase rate after 2000.