LSF evolution with selective local and global segmentation for synthetic and real images

Abstract

The major objective of image segmentation is to extract objects with respect to some input features. One of the impotent methods for image segmentation is Level Set method. In conventional level set function the LSF develops irregularity during its process of evaluation of contour of objects, this destroy the stability and smooth flow of evolution process. A method of selective region-based active contour model (ACM) is suggested in this work, in this first selectively penalizes the level set function to be binary, then uses a Gaussian smoothing kernel to regularize couture evolution. A new region-based signed pressure force (SPF) function is proposed, which can efficiently stop the contours at weak or blurred edges. Other feature is the exterior and interior boundaries can be automatically detected with the initial contour being anywhere in the image. The proposed ACM, it has the property of selective local or global segmentation. The level set function can be easily initialized with a binary function, which is more efficient to construct than the widely used signed distance function (SDF).