Research on anti-clogging of ore conveyor belt with static image based on improved Fast-SCNN and U-Net

Abstract

Abstract Conveyor belt is one of the important equipment for ore mining and transportation, which runs smoothly and has high efficiency. However, if the blockage occurs in the transportation process, it will greatly affect the safety of mine production, and even threaten the lives of miners. Therefore, it is difficult to develop a segmentation network algorithm for anti-clogging control that can be applied to the ore conveyor belt. In order to ensure the accuracy of image segmentation task and the speed of network, an improved Fast-SCNN and U-Net method based on channel concern mechanism is proposed in this paper. The accuracy of image segmentation is considered while the speed of network is guaranteed. The improved network of Fast-SCNN and U-Net with better segmentation effect is applied to the segmentation detection system of high-speed and low-speed conveyor belt, and the key frames of the ore conveyor belt in operation are extracted. The material coverage ratio of the conveyor belt in operation is obtained by segmenting the key frame image. Finally, this paper proposes for the first time an anti-clogging method for ore conveyor belt based on static image detection. By judging and predicting the blockage of ore conveyor belt, the fuzzy algorithm is used to control the running speed of the conveyor belt, and the conveyor belt is quickly and accurately slowed down and stopped when it is going to be blocked or has been blocked, so as to avoid serious blockage. Experiments show that the proposed method can improve the safety and efficiency of ore conveyor belt production.