基于改进的Snake模型的眼细胞彩色图像边缘检测算法研究

Improved Snake Model Used in the Edge Detection for Eye-cell Color Images

在前人研究的Snake模型上提出了适应彩色眼细胞图像边缘检测的改进的Snake模型.该改进模型选取了有益的能量项,舍去了刚性能量项,并在面积能量项的计算上进行了一些改进.在图像力能量项上采用把单通道灰度图像处理方式推广到真彩色RGB三通道的方法,获取彩色图像力能量项.在轮廓初始化时的受控点逆时针排序上给出了行之有效的算法.最后采用贪婪算法最小化总能量,在彩色眼细胞边缘检测上取得了较为明显的效果.

An improved Snake model is presented,which is adaptable to the eye-cell color image’s real situation.This model is based on traditional and other many improved Snake models.The useful energy terms are selected and the rigid energy term is eliminated.The area energy term is improved based on former researcher.The process of single-channel intensity image in image energy term is extended to three-channel true color images.In this way,the color image energy term is acquired.At the same time,the effective algorithm of anticlockwise sort of controlled points is also given.In the end,the greedy algorithm is adopted to minimize the total energy.So,we successfully get apparent results in the edge detection for eye-cell color images.