Efficient Interactive Segmentation using Modified Maximal Similarity Region Merging

Abstract

Interactive image segmentation has many applications in image processing, computer vision, computer graphics and medical image analysis. In medical applications, image segmentation is a fundamental process in most systems that support medical diagnosis, surgical planning and treatments. In many editing tasks, the aim is to separate a foreground object from its background. Therefore, we propose a fast and simple interactive image segmentation technique in this paper. The proposed method automatically merges the regions that are initially segmented by mean shift segmentation, and then effectively extracts the object contour by labeling all the non-marker regions as either background or object. Moreover, many experiments are tested and the results show that the proposed method is faster than the existing method. Therefore, the proposed method is effective and can quickly and accurately segment for both medical and natural scene images with ease.