摘要
针对不同光照条件对轨道扣件定位影响的问题,提出一种基于形态学处理的扣件定位算法。首先使用Canny自适应算法提取轨道图像的边缘特征,交替采用腐蚀、膨胀等操作以凸显钢轨、轨枕和枕肩的边缘特征,再对其边缘特征进行投影,结合相关空间关系和边界长度可以消除光照条件对扣件定位的影响,从而实现轨道扣件的准确定位。通过实验表明,在不同光照条件下,基于形态学处理的轨道扣件定位算法的准确率大大提高,可满足轨道扣件检测与定位的实际工程应用需求。
Aiming at the influence of different illumination conditions on track fastener positioning, a fastener positioning algorithm based on morphological processing is proposed. Firstly, Canny adaptive algorithm was used to extract the edge features of track images. Corrosion and expansion operations wee used alternately to highlight the edge features of rail, sleeper and sleeper shoulder. Then, the edge features were projected. The influence of illumination conditions on fastener positioning can be eliminated by combining the spatial relationship and boundary length To realize accurate positioning of fasteners. Experiments show that under different illumination conditions, the accuracy of track fastener location algorithm based on morphological processing is greatly improved, which can meet the practical engineering application requirements of track fastener detection and location.
作者
孟建军
程思柳
李德仓
MENG Jian-jun;CHENG Si-liu;LI De-cang(School of Mechatronic Engineering,Lanzhou Jiaotong University,Lanzhou Gansu730070,China;Mechatronics T&R Institute,Lanzhou Jiaotong University,Lanzhou Gansu 730070,China;Engineering Technology Center for Informatization of Logistics&Transport Equipment,Lanzhou Gansu 730070,China)
出处
《计算机仿真》
北大核心
2019年第11期105-109,170,共6页
Computer Simulation
关键词
扣件定位
形态学处理
边缘检测
投影法
Railway fastener positioning
Morphological processing
Edge-detection
Projection method
