Authors: Guodong Wang Jingge Lu Zhenkuan Pan Qiguang Miao
Publish Date: 2016/12/05
Volume: 76, Issue: 22, Pages: 24515-24526
Abstract
Color texture image segmentation is one of the fundamental task in image processing For segmenting the color texture we propose a new active contour model based on multichannel nonlocal and Tikhonov regularization In the new model nonlocal operator and Tikhonov regularization are simultaneously contributing to the evolutionary process of the active contour The nonlocal operator which is based on the image slice similarity can direct the contour evolution on the boundary of the object and is not affected by the texture The Tikhonov regularization can diffuse the texture areas and then accelerate the evolution process of the active contour In order to ease the solution of the energy function we also design the SplitBregman algorithm of the proposed method Expeimental color texture image segmentation results demonstrate the validity of the proposed method
Keywords: