Image Segmentation Using Region based Active Contours

Segmentation of images by utilizing fast level set and active contour is proposed. The level set method is a numerical technique for tracking interfaces and shapes. A fast level set method is introduced to decrease the computational labor of the standard level set method for propagating interfaces. The fast approach uses only points close to the curve at every time step. This method utilizes region information to guide the evolution of initial curves. Gaussian smoothing is applied to regularize the evolving curves. This algorithm is able to efficiently segment the objects in complicated images. Natural large images are segmented accurately in very short time using a fast level set implementation. Active contour is a framework of various methods for image and video segmentation. The contour coincides with the real boundary of the segmented object. Contour Initialization is started by the user, by selecting a few pixels in each object to be segmented and a circular contour is generated around each selected pixel. Evolving the contour is done using the fast level set algorithm.