Study of Active Contour Modelling for Image Segmentation: A Review

Active contour models are widely used for image segmentation. In this paper the study of various active contour models based on Chan-Vese approach is given. Partial differential equation based method widely used on real and artificial noisy images with different shapes and blurring boundaries and energy minimization has been achieved. The possibilities of extracting exact contour using these methods on artificial and real pictures depends upon the number of iterations and resulted into efficient and accurate segmentation but the dependency of method is on initial points of contour and subject to constraints from a given image. Development of video segmentation, a combination of active contour method and watershed presegmentation for tracking of moving objects is aimed to differentiate ROI and NROI efficiently but uncertainty of the moving micro objects partially deviate the results. The overall techniques can be implemented for the configuration based quality analysis of food products.