基于蚁群算法的方向特征分割方法

针对翅脉特征的有效提取问题,根据蝶翅图像具有方向性的特点,提出了方向小波变换与蚁群算法相结合的分割方法.该方法首先采用Morlet方向小波变换对不同方向的翅脉边缘进行检测、提取.通过分析这些边缘点的特征参数,确定初始聚类中心.基于所设置的初始聚类中心,再利用蚁群算法对蝶类翅脉图像进行分割.仿真实验表明,该方法可以将各方向翅脉特征检测出来,是一种有效的方向特征分割方法.

A Detection method of butterfly-vein based on ant colony algorithm

According to the characteristics of the butterfly image which has directional feature, a segmentation algorithm combined the directional wavelet transform with the ant colony algorithm is proposed to extract the butterfly-vein features effectively.The method uses Morlet directional wavelet transform to detect and extract the vein edge in different directions. The feature parameters of these edge points are analyzed to fix the original cluster center parameters. Then the ant colony algorithm is used to segment the butterfly-vein image. Simulation experiments show that the algorithm is an effective method to detect the different directional features.