Research progress on vision based tear detection methods of mining conveyor belts

Conveyor belt tears can significantly impair the efficiency of mineral transportation and even pose safety hazards. Traditional tear detection technologies are often characterized by high costs and low efficiency, while some detection systems require modifications to the conveyor belt production line, making them unsuitable for large-scale, long-distance conveyor belt inspection requirements. In recent years, machine vision technology has been widely applied in industrial inspection, and a series of solutions have been developed for the domain of tears detection in mining conveyor belts. Notably, related technologies that incorporate deep learning algorithms have significantly enhanced feature analysis capabilities, demonstrating even broader prospects for application. Against this backdrop, this paper is dedicated to conducting a comprehensive and in-depth review and analysis of existing visible vision detection methods from multiple dimensions, including the dissection of technical principles, the exploration of application practices, and the rigor of performance evaluations. By comparing the strengths and limitations of various methods, this paper aims to gain insights into the future trends of tear detection in mining conveyor belts and provide valuable references for related research and practice in this field.