基于优化蚁群算法的钢轨轮廓识别

发布时间：2018-10-08 15:52

【摘要】：针对传统蚁群算法在钢轨图像识别中存在的问题,对蚁群算法进行4个方面的优化。初始化过程优化:采用一维Logistic混沌映射序列非线性迭代方程,使蚁群的初始化分布更加均匀,以避免大量的无关运算;搜索过程优化:在蚁群的搜索初期采用随机搜索策略,根据图像灰度梯度值自动设置阈值,初步确定图像中钢轨边缘的像素点,然后建立区域搜索模型,以进行钢轨边缘的精确搜索和描绘;搜索步长优化:在搜索初期,采用大步长随机搜索策略识别钢轨边缘的像素点,然后利用小步长区域搜索策略对钢轨边缘像素点做更精确地识别,从而实现钢轨轮廓的精确识别,并减少了搜索时间和算法的收敛时间;信息素更新策略优化:每完成1次搜索,根据自动设置的信息素最大、最小浓度值更新信息素,防止陷入局部最优。对实际采集到的直线和曲线线路上的钢轨图像分别用Canny边缘检测算子、传统算法和优化算法进行钢轨轮廓识别的对比试验,结果表明:优化算法具有更好的健壮性和识别效率。
[Abstract]:Aiming at the problems of the traditional ant colony algorithm in rail image recognition, the ant colony algorithm is optimized in four aspects. Initialization process optimization: using nonlinear iterative equation of one-dimensional Logistic chaotic mapping sequence to make the initialization distribution of ant colony more uniform to avoid a large number of independent operations; search process optimization: at the initial stage of ant colony search using random search strategy, According to the grayscale gradient value of the image, the threshold value is set up automatically, the pixel points of rail edge in the image are preliminarily determined, and then the region search model is established to accurately search and depict the rail edge; the search step is optimized: at the initial stage of the search, The large step random search strategy is used to identify the pixel points on the rail edge, and the small step size region search strategy is used to identify the rail edge pixels more accurately, so as to realize the accurate recognition of rail contour. The search time and convergence time of the algorithm are reduced, and the pheromone updating strategy is optimized: every time a search is completed, the pheromone is updated according to the maximum and minimum concentration of the pheromone set automatically, so as to prevent from falling into local optimum. Canny edge detection operator, traditional algorithm and optimization algorithm are used to compare rail contour recognition with the actual collected rail images on straight line and curve line. The results show that the optimization algorithm has better robustness and recognition efficiency.
【作者单位】：兰州交通大学自动化与电气工程学院;兰州交通大学光电技术与智能控制教育部重点实验室;兰州交通大学电子与信息工程学院;
【基金】：中国铁路总公司科技研究开发计划项目(2016X003-H) 甘肃省青年科技基金资助项目(1308RJYA096)
【分类号】：TP18;TP391.41

【相似文献】