基于改进C-V模型的图像分割算法

提出一种新的模型——Chan-Vese模型,该模型是基于曲线演化、水平集方法、局部的统计信息,新模型包括两个方面:局部核心函数和惩罚项.引入局部统计信息后的新模型可以对非同质图像进行有效的分割.另外,核心函数中加入惩罚项,可以有效避免水平集函数初始化,缩短模型演化时间.通过实验的仿真结果发现,新模型在对非同质图像进行分割时得到了良好的结果.

Image segmentation algorithm based on improved C-V model

Aiming at the problem of the ineffective segmentation results of heterogeneous image for C-V model, this paper proposes an improved C-V model, and this kind of model based on C-V model accelerates and makes an accurate C-V mode effect by introducing the concept of the local core function. At the same time, penalty term is put in the C-V model energy function in order to avoid the re-initialization in the process of evolution and further improve the segmentation speed. The experiment simulation results show that the improved C-V model has better segmentation effect than the original model in the heterogeneous images.