Iterative Optimization Scheme for Image Segmentation

Image Segmentation is an integral part of computer vision. In this paper image segmentation is formulated as label relabeling problem under probability framework. To estimate the label configuration, an iterative optimization scheme is proposed to alternately carry out the maximum a posteriori (MAP) estimation and the maximum likelihood (ML) estimation. This algorithm can automatically partition the image into regions without human intervention. The segmentation obtained is very close to human perception. Comparing to other state-of-the-art algorithms, extensive experiments have shown that this algorithm performs the best. General Terms Pattern Recognition.