Steel surface defect detection algorithm based on improved YOLOv8

Steel surface defect detection is an important step to ensure the quality of steel production. Aiming at the problems such as low detection accuracy and insufficient feature extraction ability in steel surface defect detection, an improved YOLOv8n steel defect detection algorithm FCV-YOLO was proposed. Firstly, the fast feature extraction module Faster-C2f replaces C2f in the backbone extraction network to enhance the feature extraction capability of the model. Secondly, the coordinate attention mechanism is introduced to improve the model's attention to important features. Finally, the parameter number and computation amount of the model are reduced through feature compression and simplification in the Neck structure, so that the model is more suitable for industrial scenarios with limited resources. Experiments on NEU-DET data set show that the average detection accuracy of the improved FCV-YOLO model is 77.6%, which can better meet the needs of steel surface defect detection.