改进的蚁群算法在图像边缘检测中的应用研究

ON APPLYING IMPROVED ANT COLONY ALGORITHM TO IMAGE EDGE DETECTION

针对传统的蚁群边缘检测算法耗时长的问题,提出基于邻域中节点梯度计算启发式信息值的方法。该方法能够更快更好地引导蚂蚁向边缘节点进行移动,减少耗时。同时,还引入模糊C均值算法,用以确定蚁群算法中信息素阈值,使其更加准确合理,更精确地判断边缘节点。实验表明,该改进算法能够减少耗时,有效地抑制噪声,并能更加有效、精确地检测出图像的边缘。

In this paper, in view of the problem of long time consumption the traditional ant colony edge detection algorithm has, we pro- pose a method which calculates the heuristic information value based on gradient of nodes in neighbourhood. The method can guide the ants move to edge nodes faster and better, and spends less time. At the same time, we also introduce fuzzy C-means algorithm to decide the threshold of pheromone in ant colony algorithm, this makes it more exact and reasonable, and judges the edge nodes with higher precision. Experiments show that the improved algorithm can reduce the time consumption, effectively suppress the noise, and can detect the edge of the image more effectively and accurately.