【摘要】 针对噪声图像边缘检测问题,提出了一种基于改进蚁群算法的边缘检测方法。算法对蚁群算法收敛速度慢,易收敛于局部最优解的缺点进行了优化,引入了改进的蚂蚁生命周期策略,综合考虑像素邻域差和图像边缘曲线连续性等因素来确定启发式引导函数,在蚂蚁搜索起始点的选取、蚂蚁路径选择策略、信息素更新策略、启发因子的选择等方面提出了优化,实验证明,算法在收敛速度和边缘检测效果上相比传统蚁群算法有了较明显的改善,是一种较为有效的边缘检测方法。更多还原

【Abstract】 For noisy image edge detection problem, the study propose a method of edge detection based on improved ant colony algorithm. The disadvantage of slowly convergence and easy to converge to local optimal solution has been optimized, the introduction of improved ant lifecycle strategy, considering the neighboring difference of pixels and the image edge curves and other factors to determine the continuity heuristic guide function. The essay presents some optimizations on selection of the starting point in the search of ants, ant path selection policy, pheromone update strategy and inspiring factors. The experimental results indicate that the algorithm compared to the traditional ant colony algorithm on convergence speed and edge detection results have obvious improvement and it is a more effective edge detection method.更多还原