STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Image Processing & Pattern Recognition Progress

Improved Global Region based Chan-Vese Model

  1. Nitesh L. Rudani
  2. Rahul G. Patel

Abstract

Chan-Vese model is one of the fundamental active contour models for image segmentation. In this paper, the authors propose an improved global region based Chan-Vese model which provides solution to some limitations of basic Chan-Vese model. The proposed model avoids re-initialization requirement to increase speed of algorithm by reaction-diffusion method. The stopping criterion to stop segmentation process is also added in the model. To make model automatic, initialization free feature is also added via Poisson’s inverse gradient technique. Experimental results on different images justify the improvement in the basic Chan-Vese model.

Keywords

Support