Segmentation of images using support vector machines

Support vector machine (SVM) is a general algorithm based on guaranteed risk bounds of statistical learning theory. They have found numerous applications, such as in face recognition, character recognition, face detection and so on. In this paper, we propose the use of SVMs to image segmentation. The keystone that we research is how to choice the feature set for SVMs in this paper. We demonstrate that appropriate feature subset is very important to the generality capability of SVMs.