Coal Weight Measurement Method of Belt Conveyor Based on Binocular Stereo Vision

Abstract

Belt conveyors is one of the most used equipment in coal production process. However, the amount of coal on the belt conveyors is fluctuant in practical conditions, and the power output of belt conveyor is often greater than the actual demand, resulting in a large amount of energy waste. The detection of coal transport volume on the belt conveyor has become the key to solve this problem. This paper develops a coal weight detection system for belt conveyor, which integrates a three-dimensional (3D) coal information extraction module based binocular vision, and a coal weight detection calculation module incorporated by 3D scene reconstructing method and T-S fuzzy reasoning. Concretely, the first module combines local entropy transform coal image information and K-means clustering segmentation method to segment coal area, followed by 3D coal information extraction via binocular vision technology; the second module firstly collects 3D information of belt under empty load offline, then calculates the initial volume of coal by reconstructing scene from 3D information of belt and coal; Furthermore, the system calculates the area and perimeter of each coal lump by watershed algorithm. Then the area and perimeter of each coal lump are taken as the inputs of t-s fuzzy reasoning, and the void rate of coal can be estimated. Finally, the volume of coal is modified according to the void rate, and formula for calculating density and weight is used to realize coal weight detection. An experimental study is carried out through actual coal transport images of belt conveyor. The results show that the effectiveness of the proposed measurement method.