Answers are grounded in this paper and cite their sources.
Steel is a critical material in civil engineering manufacturing, and its surface quality directly affects product performance and service life. However, existing detection algorithms often face challenges such as diverse defect types, significant size differences, and complex background interference. Therefore, this paper proposes a steel surface defect detection algorithm CFL-YOLO (Convolution Focused Lightweight-You Only Look Once) based on YOLO11 for civil engineering applications. A new CF-CGLU (ConvFormer-Convolutional Gated Linear Unit) architecture, FPSC (Feature Pyramid Shared Conv) module, FDD-ASI (Focusing Diffusion Dimension-Aware Selective Integration) pyramid network, and DEC (Detail-Enhanced Convolution) detection head were developed to optimize the original structure. Experimental results show that CFL-YOLO improves the F1 Score by 4.4% and mAP by 3.4% on GC10-DET. At the same time, the calculation parameters are reduced by 6.2%, and the running speed is increased by 25.2%. In addition, CFL-YOLO improves the F1 Score by 3.5% and mAP by 1.5% on NEU-DET. At the same time, the calculation parameters are reduced by 6.2%, and the running speed is increased by 32.9%. CFL-YOLO realizes efficient and accurate detection of steel surface defects in actual civil engineering scenarios, providing reliable support for quality control of intelligent production.
No takes yet. Share an insight, caveat, or question.
Xie et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: