Study On The Building Contour Extraction From High Resolution Satellite Image

With the rapid development of remote sensing technology, we can get a large quantity of high resolution remote sensing images. However, it's a very important question to automatically process and interpret these image data for us in this information era. As a main means to obtain the geometrical and physical information of the surface, information automatic interpretation of spatial images is an advanced research field in photogrammetry, remote sensing and computer vision community. The automatic recognition and accurate extraction of the building are meaningful for many application such as getting GIS data, understanding image, fine scale mapping and so on.In the paper, the building extraction in high resolution image has been studied, mainly based on the building profile geometry configuration, and the main work are summarized as follows:1. Expatiated the basic theory used in building extraction, studied the essential arithmetic of image processing and select the valid methods of edge detection and line extraction algorithm. Simply introduce the principle and algorithm of the Object-orient classify.2. By concluding and summarizing former methods, proposed two methods to extraction building semi-automatic, one is based on the edge detection combine with area growth, the other is based on object oriented classification, the two methods can extraction building profile semi-automatic.3. Programs, which realize these two methods, are demonstrated in this paper, verifies and appraises the methods, analysis the character and insufficiency.