基于计算机视觉的受电弓滑板磨耗图像的识别

Image Recognition of Slide Abrasion for Locomotive Pantograph

根据计算机视觉原理、数字图像处理与识别理论,针对受电弓滑板图片,文章提出边缘提取和上边缘搜索等改进算法,实现滑板磨耗超限故障图像的辨识。将图片进行滤波去噪等预处理,再进行边缘定位检测、图像分割、图像增强、图像复原等处理。综合运用改进的Canny算法、Hough变换、BP神经网络来识别滑板厚度。文章的研究成果对高速铁路受电弓滑板磨耗超限等典型故障的检测与识别具有一定的理论研究价值和工程实用价值。

Based on the theory of digital image processing and recognition,a new recognition methods in slide abrasion of locomotive pantograph are proposed.By comparing different algorithms of the sliding plate image detection,the optimal one to realize the identification of the abrasion image of pantograph slide is chosen.To obtain the best effect of recognition,the edge detection,image segmentation,image enhancement,image restoration and other image analysis and processing is adopted.Meantime,the improved canny algorithm,Hough transform and BP neural network are used to identify the wear image of pantograph slide.The research results have some references for the detection and identification of slide abrasion of locomotive pantograph.