A Contactless Measuring Speed System of Belt Conveyor Based on Optical Flow Techniques

Abstract

In mining operation, measuring the speed of a belt conveyor is important for safe and efficient production. The contact measurement system is the mainstream choice for existing measuring speed systems, but the contact method cannot avoid some potential problems and hidden dangers in the running process of belt conveyor. A novel contactless measurement system based on optical flow techniques is designed in this study, which can calculate running speed of belt conveyor in real time. During the operation of belt conveyor, it can be found that most regions of the images are similar when the camera is set to high-speed continuous shooting mode. These similar image regions have a certain displacement change in the moving direction of the belt conveyor due to the time difference of its shooting, and can be described in a simplified way which is the image feature keypoint. The machine vision algorithm extracts image feature keypoints and acquires the optical flow information of image feature keypoints. Motion displacement information including the displacement change can be extracted by analyzing of optical flow information about captured images to calculate the running speed. The contactless measuring speed system is verified by experiments. Experimental results show that the measuring accuracy can be presented by indicators with RMSE of 0.042 m/s and MAE of 0.029 m/s.