Extract non-point pollution source information from a TM image by hierarchical classification
利用TM图像逐级分类提取非点源信息

The issue of the article is hierarchical classification that is used to extract land-use information relating to non-point source pollution from TM image. The result shows that it performs better than traditional maximum likelihood method. It is useful to extract information quickly and effectively with advantages of simplifying complex problem, giving prominence to key point and having more flexibility.