一种改进的快速C-V水平集红外图像分割

当红外图像中包含较强噪声时,C-V模型水平集分割方法会产生大量冗余轮廓;同时,C-V水平集采用偏微分方程(PDE)实现,存在计算量大、分割速度慢的缺点.为此,本文提出了改进的快速算法,该算法保留了C-V模型的全局优化特性,并通过窗口滤波整合图像邻域空间信息来构建曲线进化的外部速度,从而提高C-V模型的抗噪性并减少分割中产生的冗余轮廓;采用基于双链表的快速水平集算法来实现曲线的演化,去除了传统算法中的重新初始化和PDE求解的过程,减少了迭代步数,提高了分割的速度.实验结果表明,本文算法对边缘模糊、噪声较大的红外图像能实现快速而有效的分割.

Improved Fast C-V Level Set Infrared Image Segmentation

When the infrared image includes much noise,the C-V model level set segmentation method generates many redundant contours.Meanwhile,the C-V level set is realized by partial differential equation(PDE).It has disadvanges of large amount of calculation and slow segmentation velocity.Therefore,this paper proposes an improved fast algorithm.The improved algorithm preserves the global optimization characteristic of C-V model,and integrates with neighbor space information by the window filter to construct a extern...