改进YOLOv11的轻量级钢板表面缺陷检测算法OA

This paper aims to address the issues of missed detection,high false detection rates,and deployment on the resource-constrained devices in industrial steel plate surface defect detection,and pro-poses a lightweight defect detection algorithm based on an improved YOLOv11.The study consists of tak-ing PP-LCNet with added residual structure as the backbone feature network of the target detection net-work for the reduction of the complexity of model operation and the number of parameters;performing the multi-scale feature fusion by BiFPN to improve the detection ability of objects of different scales;and in-troducing improved SimAM attention mechanism module to enhance the feature extraction capability for better focus on feature image regions.The simulation and experimental results show that the improved model achieves a detection accuracy of 77.1%on the NEU-DET dataset,which is 2.7%higher than that of the YOLOv11n model with the parameters of 2.1 M and a 19.2%reduction compared to the YOLOv11n model.This model becomes more lightweight while enhancing recognition accuracy.