Defect detection is a critical task in industrial manufacturing, playing a vital role in achieving automation, improving product quality, and ensuring operational safety. Traditional methods, however, face considerable limitations in terms of accuracy and efficiency. To address these challenges, we propose DCA-YOLO, a lightweight model for steel surface defect detection optimized based on task-specific knowledge. Specifically, our model incorporates a Dynamic Snake Convolution (DSConv) module to capture subtle linear features in challenging defect categories, a context-guided module to leverage contextual information for detecting clustered defects, and an Adaptive Spatial Feature Fusion (ASFF) mechanism to efficiently merge features across scales. The experimental results demonstrate that even with a nanoscale architecture (4.3 million parameters and 9.4 GFLOPs), the enhanced model exhibits marked improvements in detection accuracy and robustness, with mAP50 increasing by 4.6% and mAP50-95 by 7.7%. These findings not only offer a better solution for steel surface defect detection, but also provide new theoretical insights and practical experience for the advancement of industrial inspection technologies. In the future, DCA-YOLO is expected to be applied across a wider range of industrial detection scenarios, further driving progress in the field.

Real-time industrial inspection is a crucial component of production automation, with key challenges lying in enhancing detection accuracy for specific tasks and effectively mitigating the adverse impacts of complex production environments. Addressing these issues, this paper proposes an innovative solution. We introduce the Head_DySnake module in the detection head to significantly bolster the capture and recognition capabilities of defect texture features. Concurrently, at the initial stage of the backbone network, we integrate the Attention Denoising Module(ADConv) module, which employs an attention-guided mechanism for effective noise reduction in production environments, thereby eliminating high-level noise caused by high background similarity. Through these optimizations, our research achieves a 6.3% mAP improvement on the NEU-DET dataset with a computational demand of merely 9.8 GFLOPs, and a 5.9% mAP improvement on the GC10-DET dataset. This study thoroughly explores the impact of steel defect-specific textures on recognition performance and validates the positive role of attention-guided environmental denoising strategies in enhancing model robustness. These findings offer new perspectives for lightweight model design and performance optimization in industrial production inspection and are expected to provide valuable insights for the detection of other types of defects in related research fields.