Coal Mine Belt Conveyor Foreign Object Detection Based on Improved YOLOv8

Belt conveyor is a crucial equipment during the coal mining- and transportation-operation process. Foreign objects, like large gangue, anchor rods, once mixed into the belt, can lead to the belt's damage and will thereby cause safety problems in mines. Since the belt is working in a bad illumination and a rather high-speed condition, accurately detecting foreign objects in a timely manner becomes a challenging task. To deal with such challenges, this paper proposes a conveyor belt foreign object detection method based on YOLOv8. Our method adds an efficient attention mechanism, namely the Coordinate Attention (CA), to strengthen the positional information of the foreign object and modifies the loss function. We trained and tested our model against the dataset collected from the mines in the real word. The experimental results show that our method can effectively improve the accuracy of foreign object detection while shortening the training time.