Image segmentation by active contour model using hyperbolic trigonometric formulation

Abstract

The accurate detection of region(s)-of-interest (ROI) via Active Contour Method (ACM) is a well-known and evolving research topic in image segmentation. A novel region-based active contour method is proposed that can segment real and synthetic images with blurred borders more efficiently. Additionally, a new Signed Pressure Force (SPF) function named as Hyperbolic Trigonometric Signed Pressure Force Function (HTSPF) is introduced, that is able to detect the contour of ROI of diverse intensities, even at weak and blurred borders. Our HTSPF utilizes the harmonic mean intensities of the image that result in effective segmentation of low contrast images. Using level set like SBGFRLS method and the harmonic mean intensities of the image like ACMHM method, our HTSPF performs better in cases of images having objects of blurred borders, multiple objects with diverse intensities and objects having low contrast. To regularize the level set function, we utilized the Gaussian filter. It also removes the need of expensive re-initialization technique. The proposed method is tested on synthetic and real images and its segmentation results demonstrate that the proposed method is robust in segmentation of images having objects of blurred borders, objects of low contrast and multiple objects with diverse intensities.