Image segmentation plays an important role in a variety of applications such as robot vision, object recognition and medical imaging,…Fuzzy clustering is undoubtedly one of the most widely used methods for image segmentation. In many cases, it happens that some characteristics of image are more significant than the others. Therefore, the introduction of a weight for each feature which defines its relevance is a natural way in image segmentation.
In this paper, we develop an efficient method for image segmentation via feature weighted fuzzy clustering model. Firstly, we formulate the feature weighted fuzzy clustering problem as a DC (Difference of Convex functions) program. DCA (DC Algorithm), an innovative approach in nonconvex programming, is then developed to solve the resulting problem. Experimental results on synthetic and real color images have illustrated the effectiveness of the proposed algorithm and its superiority with respect to the standard feature weighted fuzzy clustering algorithm in both running-time and quality of solutions.