基于机器视觉的矿车踏面磨耗检测设计

Design of Mining Vehicle Wear Loss Detection Based on Machine Vision

针对矿用轨道车辆传统的人工轮对检测方法的精度低、效率低等问题,设计了一个基于机器视觉的自动检测系统。通过摄像头采集辅助光源照射到轮对上光带的图像,使用Matlab对图像进行处理。算法上根据踏面图像的特征针对细线化方法进行了专门优化,并通过实验确定了轮廓检测的最佳阈值。将处理后的图像与标准图像进行差影比对,使用相机标定的相关参数进行计算得出踏面的磨耗量。实验证明,使用优化后处理算法的系统相比传统算法提高了精度。每个轮对平均检测时间小于2 s,误差小于±0.2 mm,可以快速大量地自动检测轮对踏面的磨损情况,保证了工业现场稳定安全的生产,具有一定的应用意义。

An auto detected system based on machine vision is designed for the wheel set of mine track vehicle, in order to solve the problems of low accuracy and low efficiency of the traditional artificial detection method. The system captured the light band image emitted from the auxiliary light source through the camera and processed images through Matlab. Based on the characteristics of the tread image, the refinement algorithm was specifically optimized, and the optimal threshold for the contour detection was determined by experiments. The processed image was compared with the standard image and the wear loss was calculated by using the parameters of the camera calibration. Experiments show that compared with the traditional algorithm, the accuracy of the optimized algorithm is improved, the system can detect the wear loss quickly and abundantly, the average detection time is less than 2 s, and the error is less than ±0. 2 mm. It has certain application meaning because of ensuring the safety of industrial site and improving the accuracy and speed of detection.