自适应权值调整的C-V模型及图像分割

Weights Adjusting C-V Model and Image Segmentation

Chan-Vese模型(简称C-V模型)是基于均质区域能量最小化的曲线演化分割框架。本文提出一种区域相关权重的C-V模型,并对模型的区域权重进行了探讨,定义了自适应权值调整函数,加速曲线收敛过程,得到精确的区域边界。实验表明该算法可行有效。

Chan-Vese model, also called C-V model, is a region-based curve evolving segmentation framework that modeled as an energy minimization function. This paper proposed a region correlative C-V model , discussed the region weights and defined weights adjusting function in order to fasten the curve evolution and got the real object boundary. Our experiments validated our model.