一种汽车玻璃缺陷检测方法——结合潜在扩散模型数据增强的改进YOLOv5s算法  

A Defect Detection Method for Automotive Glass——Improved YOLOv5s Algorithm Enhanced with Latent Diffusion Model Data Augmentation

作　　者：罗志学 黄诗浩 LUO Zhixue;HUANG Shihao(Fujian Provincial Key Laboratory of Automotive Electronics and Electric Drive,Fujian University of Technology,Fuzhou Fujian 350118,China;School of Electronic,Electrical Engineering and Physics,Fujian University of Technology,Fuzhou Fujian 350118,China)

机构地区：[1]福建理工大学福建省汽车电子与电驱动重点实验室,福建福州350118 [2]福建理工大学电子电气与物理学院,福建福州350118

出　　处：《莆田学院学报》2025年第2期79-85,共7页Journal of putian University

基　　金：福建省自然科学基金资助项目(2022J01950)。

摘　　要：针对汽车玻璃缺陷样本稀缺的问题,提出了一种基于潜在扩散模型的数据增强方法。该方法利用潜在扩散模型生成高质量的合成缺陷图像,再结合伪标签标注技术,显著提升了数据集的多样性。为解决原始YOLOv5s算法在跨尺度多目标检测任务中易漏检与错检的问题,通过在颈部区域引入BiFPN结构,实现多尺度特征的高效融合。实验结果表明,改进的算法对缺陷的识别精度明显提高,为汽车玻璃缺陷检测提供了可靠的技术支持与理论参考。To address the issue of scarce defect samples in automotive glass,a data augmentation method based on a latent diffusion model is proposed.This method leverages the latent diffusion model to generate high-quality synthetic defect images,combined with pseudo-labeling techniques,significantly enhancing the diversity of the data sets.To resolve the problem of missed detection and false detection in the original YOLOv5s algorithm during cross-scale multi-object detection tasks,a BiFPN structure is introduced in the neck region to achieve efficient fusion of multi-scale features.Experimental results demonstrate that the improved algorithm significantly enhances defect recognition accuracy,providing reliable technical support and theoretical reference for automotive glass defect detection tasks.

关 键 词：汽车玻璃 潜在扩散模型 数据增强 伪标签标注 YOLOv5s BiFPN 

分 类 号：U463[机械工程—车辆工程]