Detection of Vegetation Patch Growth by Absorption Feature Analysis on Tasseled Cap Brightness of Transects from Landsat 7 ETM+ Images

摘要

Vegetation patches are worldwide distributed in arid and semi-arid ecosystems. Mapping vegetation patch dynamics provides valuable information for regional vegetation recovery and re-establishment. The high spatial resolution images may be powerful for the decametric-scale vegetation patch detection. However, the dense and long time series of fine spatial resolution (better than 2.5 m) imagery were not available for large regions until recently due to design considerations on satellite and sensors, satellite data transmission, satellite life and revisited period, and further effects like atmospheric absorption and cloud. For multispectral images (less than 10 m), it was often a challenge for detecting these vegetation patches through visual interpretation and the common image classifications. In this paper, our proposed method based on analysis of absorption features of tasseled cap brightness of Landsat 7 Enhanced Thematic Mapper Plus (ETM+) images along transect provided the acceptable results for vegetation patch recovery detection, which represented a efficient, economical and straightforward procedures for local vegetation management in the Yellow River Delta, China and other similar landscapes.