Fig. 4From: The Vese-Chan model without redundant parameter estimation for multiphase image segmentationComparison of original Vese-Chan method, TV regularization method, and our method using two label functions. a The input nature image. b The segmentation of original Vese-Chan method. c The segmentation of TV regularization method. d The segmentation of our method. e–h Different regions obtained via the original Vese-Chan method. i–l Different regions obtained via TV regularization method. m–p Different regions obtained via our methodBack to article page