Fast Level Set Image Segmentation Using New Evolution Indicator Operators

Abstract

We propose an effective level set evolution method for robust object segmentation in real images. We construct an effective region indicator and a multiscale edge indicator to adaptively guide the evolution of the level set function. The region indicator is built on the similarity map between image pixels and user specified interest regions, which is computed using Gaussian Mixture Models (GMM). To accurately detect object boundary, we propose to use a multi-scale edge indicator defined in the gradient domain of the multi-scale feature-preserving filtered image. Then, we develop a new mixing edge stop function based on these two indicators, which forces the level set to evolve adaptively based on the image content. Furthermore, we apply an acceleration approach to speed up our evolution process, which yields real time segmentation performance. As the results show, our approach is effective for image segmentation and works well to accurately detect the complex object boundaries in real-time.