基于改进YOLOv8的汽车门板紧固件检测算法  

Automotive door panel fastener detection algorithm based on improved YOLOv8

作　　者：王晓辉[1] 贾韫硕 郭丰娟[1] WANG Xiao-hui;JIA Yun-shuo;GUO Feng-juan(School of Control and Computer Engineering,North China Electric Power University,Baoding 071003,China)

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003

出　　处：《计算机工程与设计》2025年第1期298-306,共9页Computer Engineering and Design

基　　金：河北省自然科学基金项目(F2022502002);中央引导地方科技发展资金基金项目(236Z1707G)。

摘　　要：针对汽车门板紧固件在复杂场景下存在的检测准确度较低和实时性较差的问题,提出一种小目标改进算法YOLOv8-SOD(small object detection)。在主干网络引入SPD(space-to-depth)模块和自适应权重分配模块,在算法的颈部网络输出位置增加选择性注意力模块,将CIOU损失函数替换为MPDIOU损失函数。实验结果表明,YOLOv8-SOD算法平均检测精度为99.1%,比模板匹配方法和YOLOv8算法分别提高了9.4%、2%,达到了工厂生产流水线的检测标准,具有实用价值。To solve the problems of low detection accuracy and poor real-time performance of automobile door panel fasteners in complex scenes,a small object improvement algorithm YOLOv8-SOD(small object detection)was proposed.The space-to-depth(SPD)module and the adaptive weight allocation module were introduced into the backbone network,and a selective attention module was added to the output position of the neck network of the algorithm,and the CIOU loss function was replaced by the MPDIOU loss function.Experimental results show that the average detection accuracy of the YOLOv8-SOD algorithm is 99.1%,which is 9.4%and 2%higher than that of the template matching method and the YOLOv8 algorithm,respectively,which meets the detection standard of the factory production line and has practical value.

关 键 词：汽车门板紧固件检测 小目标 自适应权重分配 无参注意力 选择性注意力 损失函数 深度学习 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]