Extracting Waterbody from Beijing-1 Micro-Satellite Images based on Knowledge Discovery

Extracting waterbody of Chengdu city of China from Beijing-1 micro-satellite images based on knowledge discovery is discussed here. Firstly, the waterbody is extracted from panchromatic image by using suitable threshold value and it is divided into the large waterbody and the small waterbody. Secondly, The shades of the large buildings are removed from the large waterbody by using the spatial relationship knowledge of the buildings and their shades. Thirdly, the shades of buildings are removed from the small waterbody according to the spatial relationship knowledge of the small waterbody and the vegetation area by using the vegetation layer from the multi-spectral images. Finally, the final waterbody is obtained by combining the large waterbody without the shades of buildings and the small waterbody without the shades of buildings. It is shown that the waterbody can be effectively extracted from Beijing-1 micro-satellite images with higher accuracy by the methodology presented here.