摘要
针对轨道图像边缘检测效果不佳的问题,提出一种基于改进蚁群算法的边缘检测算法。在蚁群的路径选择中加入搜索热度因子,考虑轨道边缘的连续性和完整性后,结合灰度梯度值、像素点的相位角度和蚂蚁移动的方向性,重新设计了启发函数。动态调整了局部信息素更新公式,并基于混沌扰动改进了全局信息素更新公式。通过仿真实验验证了算法的有效性。
Aiming at poor results for edge detection of railway,an image edge detection method based on improved ant colony algorithm is proposed.Searching heat factor is added in path selection of ant colony,considering continuity and integrity of the railway edge,heuristic function is redesigned combined with gray gradient values,the phase angle of the pixel and the directional of ants moving.The local pheromone update formula is dynamically adjusted and the global pheromone update formula is improved based on chaotic disturbance.Simulation experiment verifies the effectiveness of the algorithm.
出处
《传感器与微系统》
CSCD
北大核心
2013年第6期130-133,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(61164010/F030101
51167009/E0703)
甘肃省高等学校基本科研业务费专项资金资助项目
关键词
轨道
边缘检测
蚁群算法
railway
edge detection
ant colony algorithm
