An active contour model for image segmentation using morphology and nonlinear Poisson’s equation

Abstract

Active contour model (ACM) is considered as a feasible tool to handle image segmentation problems via an unsupervised learning approach. This paper compensates the shortcomings of well-known ACMs to introduce erosion and dilation operations in mathematical morphology to fit the intensity differential functions of contours on both sides of the image. A nonlinear Poisson’s equation is developed to deal with the interference of nonuniform noise within the image, and finally build a new model. This paper develops a new image processing approach using morphology and nonlinear Poisson’s equation, which yields a fast and efficient algorithm among known similar image segmentation methods, and has a high level of accuracy in processing image segmentation.