作者: Zemin Ren
DOI: 10.1016/J.SIGPRO.2015.05.009
关键词: Active contour model 、 Regularization (mathematics) 、 Variational method 、 Algorithm 、 Energy functional 、 Balanced flow 、 Mathematics 、 Image segmentation 、 Real image 、 Mathematical optimization 、 Level set method
摘要: In this paper, a new adaptive active contour model is proposed for image segmentation, which built based on fractional order differentiation, level set method and curve evolution. The energy functional the consists of three terms: fitting term, regularization tern penalty term. By incorporating novel term can describe original more accurately, be robustness to noise. ensure stable evolution function, added into model. results function gradient flow that minimizes overall functional. Experimental both synthetic real show desirable performance our method. We an differentiation.The accurately.Adaptive length used smooth function.