冷轧带钢表面缺陷检测系统设计

发布时间：2018-04-24 09:51

本文选题：冷轧带钢 + 表面缺陷　；参考：《仪表技术与传感器》2017年01期

【摘要】：针对现有带钢表面质量检测技术的缺陷检测精度与识别率不高的问题,以及冷轧带钢生产线的实际运行环境,设计了一套带钢表面缺陷实时检测系统,从系统整体设计、硬件结构、软件开发和检测分类算法等方面进行了深入的分析与研究。采用模块化软件设计理念、基于谱残差视觉显著性缺陷快速检测方法以及多分类器融合框架,实现了对带钢表面常见的孔洞、擦伤、氧化、边裂等不同类型和不同等级的缺陷进行精确、实时的检测与分类。实验结果表明:该算法缺陷检测率为94.68%,缺陷识别率达到了93.5%,单幅图像缺陷检测耗时仅为37.6 ms,符合当下高速运动带钢快速检测的需求。
[Abstract]:Aiming at the problem that the defect detection accuracy and recognition rate of the existing strip surface quality detection technology are not high, and the actual operating environment of cold rolled strip production line, a real-time detection system for strip surface defect is designed, which is designed from the whole system. The hardware structure, software development and detection and classification algorithm are deeply analyzed and studied. Based on the concept of modular software design and the fast detection method of spectral residual visual saliency defect and the fusion framework of multi-classifier, the common holes, scratches and oxidation on the steel strip surface are realized. Accurate, real-time detection and classification of defects of different types and levels, such as edge cracks. The experimental results show that the algorithm has a defect detection rate of 94.68, a defect recognition rate of 93.5and a single image defect detection time of only 37.6 Ms, which meets the needs of high-speed moving strip steel rapid detection.
【作者单位】：河北工业大学控制科学与工程学院;河北科技大学;
【基金】：国家自然科学基金项目(61403119、61203275) 河北省自然科学基金项目(F2014202071) 河北省高等学校科学技术研究项目(YQ2013036) 天津市特派员科技计划项目(15JCTPJC55500)
【分类号】：TG335.5;TP391.41

【相似文献】