蚁群算法在图像边缘检测上的应用研究

Application Research of Ant Colony Algorithm in Edge Detection

将蚁群算法应用于图像边缘检测,提出了一种利用灰度梯度引导的蚁群搜索图像边缘的算法,其核心思想是利用图像的灰度梯度作为启发信息,并加入运动方向改变量,引导蚁群搜索边缘,迭代搜索可能的局部边缘曲线。通过对蚂蚁行走过路径上的信息素分布进行更新,使得分布在真实边缘上的信息素逐渐增大,从而使搜索逐渐向真实的边缘收敛,最后根据信息素的遗留结果提取的边缘曲线。此算法可有效抑制噪声干扰,最大限度的保留细节信息,以达到有效提取图像边缘的目的。同时在灰度提取及蚁群的选择上较已有相关算法有较大改进,并分析效果指出不足和进行相关的改进的方向。

As the ant colong algorithm has the characteristics of positive-feedback and thought-provoking, this paper uses it in edge detection, and puts forward a algorithm with gray gradients guide in edge detection. The core idea is to use the gray gradient of image as a enlightening information, and add the change of movement direction, guiding ant colony to search brink, iteratively for the possible local marginal curve. By the distribution of pheromone on the line that the ants passed updates, and the pheromones on the real edge increases, we can extracte the edge curve by final results of the lega- cy of the pheromone. The algorithm can effectively suppress noise interference and preserve the details to a max content, so as to achieve the purpose of extracting the edge of the image. This paper has a greater improvement in the extraction of gray-scale, the choice of the ant colong and the gray gradient guide information and the changement of movement direc- tion. At last, we analysis the test results and pointes out the defects and related improvements in the direction.