Classification of landsat remote sensing images by a fuzzy unsupervised clustering algorithm

The classification of each pixel in a Landsat image to one of the land cover types by conventional clustering techniques is highly inappropriate due to the low resolution of Landsat images and the multiplicity of terrain. The concept of fuzzy logic provides a flexible solution to this problem. This paper presents a new two-pass unsupervised clustering algorithm incorporated the fuzzy theory. In the first pass the mean vectors of different land cover types representing their geographic attributes are derived. In the second pass the membership grade of a pixel belonging to different land cover types is computed based on the distance between its gray-value vector and the mean vector of each type. Experimental results show that the developed fuzzy clustering algorithm produces more reasonable phenomenon interpretation than the traditional hard partition techniques.