Adaptive active contour model driven by global and local intensity fitting energy for image segmentation

Abstract

An adaptive active contour model combining global and local information is proposed for image segmentation in this paper. The energy functional consists of global term and local term, the global term is derived from the data fidelity item of CV model, and the local term introduces image local entropy which reflects the grey characteristics. Additionally, an adaptive method which takes the local information of the image into consideration is presented to set the parameters of global term and local term. The global term is able to segment homogeneous image rapidly and robust to segment noisy image; the local term can efficiently segment the intensity inhomogeneous image. By adaptively incorporating the global term with local term, the proposed model retains the advantages of both the global term and local term and widens the range of reasonable parameters. Experimental results on homogeneous images, intensity inhomogeneous images, noisy images and real images have demonstrated the efficiency and robustness of the proposed model.