基于YOLOv7-TSCR的连铸坯表面缺陷检测  

Surface defect detection of continuous casting billets based on YOLOv7-TSCR

作　　者：曾凯 陈波 马智华 肖鹏程 王雁 朱立光 ZENG Kai;CHEN Bo;MA Zhihua;XIAO Pengcheng;WANG Yan;ZHU Liguang(College of Electrical Engineering,North China University of Science and Technology,Tangshan,Hebei 063210,China;College of Metallurgy and Energy,North China University of Science and Technology,Tangshan,Hebei 063210,China;Hebei High Quality Steel Continuous Casting Engineering Technology Research Center,Tangshan,Hebei 063000,China;School of Materials Science and Engineering,Hebei University of Science and Technology,Shijiazhuang,Hebei 050018,China)

机构地区：[1]华北理工大学电气工程学院,河北唐山063210 [2]华北理工大学冶金与能源学院,河北唐山063210 [3]河北省高品质钢连铸工程技术研究中心,河北唐山063000 [4]河北科技大学材料科学与工程学院,河北石家庄050018

出　　处：《河北科技大学学报》2024年第4期351-361,共11页Journal of Hebei University of Science and Technology

基　　金：国家自然科学基金(51904107);中央引导地方科技发展资金项目(236Z1017G);河北省博士研究生创新资助项目(CXZZBS2021096);唐山市市级科技计划项目(22130220G)。

摘　　要：为解决连铸生产过程中铸坯表面缺陷检测准确率低、检测速度慢、模型参数量大难以部署等问题,提出一种融合重参数化和注意力机制的轻量化铸坯表面缺陷检测算法YOLOv7-TSCR。首先,利用Mish和SiLU激活函数、SimAM注意力机制,构建了改进的高效层聚合模块ELAN-S,以有效增强对多尺度缺陷特征的提取;其次,设计了C2f_RG模块改进特征融合网络,减少参数量的同时获得更丰富的梯度流信息,增强特征融合能力;最后,根据采集实际生产中的缺陷图像,构建铸坯缺陷数据集并进行验证。结果表明:YOLOv7-TSCR相较其他网络模型检测效果显著提升,在模型参数量减小的情况下,精确率达93.5%,平均精度均值提高了2.8%,检测速度可达120 FPS;在NEU-DET公开数据集上进行的泛化性对比实验证明了算法具有较强的泛化性。改进算法在保证较高检测精度的基础上,具有较快的检测速度和较小的参数量,为铸坯表面缺陷的高效检测提供了技术参考。To solve the problems of low accuracy,slow detection speed,and difficulty in deploying model parameters in surface defect detection of continuous casting production process,a lightweight surface defect detection algorithm YOLOv7-TSCR that integrates heavy parameterization and attention mechanism was proposed.Firstly,based on the Mish and SiLU activation functions and the SimAM attention mechanism,an improved high-efficiency layer aggregation module ELAN-S was constructed to effectively enhance the extraction of multi-scale defect features.Secondly,the C2f_RG module was designed to improve the feature fusion network,reducing the number of parameters while obtaining richer gradient flow information and enhancing feature fusion capabilities.Finally,based on the collected defect images from actual production,a dataset of casting defects was constructed and validated.The results show that YOLOv7-TSCR has significantly improved detection performance compared to other network models;With a reduced number of model parameters,the accuracy reaches 93.5%,the average accuracy increasesby 2.8%,and the detection speed reaches 120 FPS;The generalization comparison experiment on the NEUDET public dataset proves that the algorithm has strong generalization.On the basis of ensuring high detection accuracy,the improved algorithm has a fast detection speed and a small number of parameters,which provides a technical reference for the efficient detection of surface defects in casting billets.

关 键 词：炼钢 铸坯表面缺陷 注意力机制 多尺度特征 YOLOv7 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]