Visual detection method based on line lasers for the detection of longitudinal tears in conveyor belts

Abstract

This study proposes a novel method based on line lasers for longitudinal tear detection in conveyor belts. Unlike most of the existing methods directly based on the morphological characteristics easily influenced by the change in environmental light, this method transfers the characteristics of the belt to the linear characteristics of the laser line. As a strong light source, the linear laser is not easily affected by the change in light. A new improved gray-gravity center method (IGGM) was proposed to extract the centerline of the laser line, and then detect the longitudinal tear based on the characteristics of the centerline. The experimental results show that this method can remove the influence of the changes in environmental light, the laser line can retain the longitudinal tear features well, and IGGM can extract the centerline more accurately. Compared with other methods, this method has better accuracy and better real-time performance.