Coal gangue target detection method based on improved YOLOv5s model

The invention discloses a coal gangue target detection method based on an improved YOLOv5s model. The method comprises the following steps: S1, collecting real-time images of coal and gangue; s2, performing visual identification processing on the acquired real-time image based on an improved YOLOv5s model so as to identify coal and gangue in the real-time image and determine coordinate information of the gangue; and S3, the mechanical arm sorts the gangue out of the coal according to the coordinates of the gangue. On the basis of a YOLOv5s model, a self-correcting convolutional network SCConv is embedded into a Backbone region of the YOLOv5s model, 19 * 19 feature map branches of Neck and Prediction regions in the YOLOv5s model are deleted, linear scaling is performed on an anchor frame obtained through clustering of a K-means algorithm, the improved YOLOv5s model is provided and applied to coal gangue target detection, and the detection speed and the detection precision are effectively improved. 本发明公开了一种基于改进