基于优化的蚁群图像边缘检测算法研究

摘要

图像边缘检测是进行数字图像处理非常重要的环节之一,一直是研究难点.针对蚁群算法检测图像边缘存在一系列的问题(如:边缘不完整,边缘断裂、不清晰等),提出了一种优化的蚁群图像边缘检测算法.通过对基本蚊群算法中信息素矩阵、蚂蚁的信息素更新规则和状态转移规则进行分析,将蚁群算法的信息素更新规则和蚂蚁状态转移规则进行改进,同时对相关的信息素影响因子α、启发函数的影响因子β、蚁群数m、阈值等主要参数进行分析.通过实验选择出更加合理的取值,让蚂蚁具有更强的灵活性.仿真结果表明,优化的蚁群算法可以更好地检测出图像边缘,提高图像的检测精度和检测效果.%Image edge detection is the key for digital image processing and has been difficult in research.Ant colony algorithm can achieve most of the image edge detection,but the effect of which is not very ideal,including incomplete and fractured edge and other defects.In view of the above problems,an optimized ant colony algorithm of image edge detection is presented.Through the analysis of information of basic ant colony algorithm on pheromone matrix,ant pheromone updating rule and state transfer rule,the rules of information pheromone updating and ant state transfer are analyzed and improved,and the main parameters of the impact factors of related pheromone and the heuristic function,the ant colony number and threshold are analyzed.The more reasonable value can be selected by experiments,which let the ant be more flexible.Simulation shows that the proposed algorithm can detect the edge of image and improve the accuracy and efficiency in detection.