[1]黄心玥,王明泉,耿宇杰,等.基于 YOLOv8 的铝合金铸造轮毂缺陷检测算法研究与分析[J].机械与电子,2025,(02):26-31.
　HUANG Xinyue,WANG Mingquan,GENG Yujie,et al.Research and Analysis of Defect Detection Algorithm for Aluminum Alloy Casting Wheels Based on YOLOv8[J].Machinery & Electronics,2025,(02):26-31.
点击复制

基于 YOLOv8 的铝合金铸造轮毂缺陷检测算法研究与分析()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2025年02期
页码:
26-31
栏目:
自动控制与检测
出版日期:
2025-02-28

文章信息/Info

Title:
Research and Analysis of Defect Detection Algorithm for Aluminum Alloy Casting Wheels Based on YOLOv8
文章编号:
1001-2257 ( 2025 ) 02-0026-06
作者:
黄心玥; 王明泉; 耿宇杰; 谢绍鹏; 商 然
中北大学信息与通信工程学院,山西 太原 030051
Author(s):
HUANG Xinyue ;  WANG Mingquan ;  GENG Yujie ;  XIE Shaopeng ;  SHANG Ran
( School of Information and Communication Engineering , North University of China , Taiyuan 030051 , China )
关键词:
目标检测; 汽车轮毂;  YOLOv8 ; 缺陷检测; 多头自注意力
Keywords:
target detection ;  automobile wheel hub ;  YOLOv8 ;  defect detection ;  MHSA
分类号:
TP391.4
文献标志码:
A
摘要:
为解决以往算法对轮毂内部缺陷进行检测时效率低、精度较低和识别能力差等问题,提出一种基于 YOLOv8 算法优化的铝合金轮毂缺陷检测方法。首先,通过在主干中添加 Focus 层进行切片操作,增加通道防止信息丢失,提高全局感受野。其次,将空间金字塔池化部分替换为 SimSPPF ,减小计算量。最后,在 Backbone 部分加入 BoTNet 模块,通过融合多头自注意力层进一步改善检测效果并减小参数。通过实验证明,改进后的算法可以准确定位并识别出气孔、缩孔、缩松和裂纹 4 类缺陷,平均检测精度( mAP )可以达到 98.80% ,平均检测速度为 75.53 帧/ s ,满足工业检测的实时性要求。
Abstract:
To address the issues of low efficiency , low accuracy and poor recognition ability in the detection of internal defects in wheel hubs , an improved defect detection method for aluminum alloy wheels based on YOLOv8 network is proposed.Firstly , the Focus layer is added to the trunk for slicing operations , and the global receptive field is improved by adding channels to prevent information loss.Secondly , the space pyramid pool is replaced by SimSPPF to reduce the computation.Finally , BoTNet module is added to the Backbone to reduce the parameters and further improve the detection effect by integrating multiple self-attention layers.The experiments show that the improved algorithm can accurately locate and identify 4 types of defects including porosity , shrinkage , porosity and crack.The average detection accuracy ( mAP ) can reach 98.80% , and the average detection speed is 75.53 frame / s , which meets the real-time requirements of industrial detection.

参考文献/References:

[ 1 ] MAIRE E , WITHERS P J.Quantitative X-ray tomography [ J ] .International materials reviews , 2013 , 59( 1 ): 1-43.

[ 2 ] 刘军,陶维道 .X 射线数字成像技术在铸造铝合金轮毂检测中的应用[ J ] . 无损探伤, 2017 , 41 ( 3 ): 42-44.
[ 3 ] 焦腾云,王明泉,张俊生,等 . 轮毂高动态 X 射线图像分层色调映射算法[ J ] . 测试技术学报, 2019 , 33 ( 5 ): 412-417.
[ 4 ] CHENG S H , LU J X , YANG M T , et al.Wheel hubdefect detection based on the DS-Cascade RCNN [ J ] . Measurement , 2023 , 206 : 112208.
[ 5 ] LIN J H , YAO Y , MA L , et al.Detection of a casting defect tracked by deep convolution neural network [ J ] . The international journal of advanced manufacturing technology , 2018 , 97 : 573-581.
?[ 6 ] 鲍春生,谢刚,王银,等 . 基于深度学习的铸件缺陷检测[ J ] . 特种铸造及有色合金, 2021 , 41 ( 5 ): 580-584.
[ 7 ] MERY D.Aluminum casting inspection using deep object detection methods and simulated ellipsoidal defects [ J ] .Machine vision and applications , 2021 , 32 ( 3 ): 72.
[ 8 ] 蔡彪,沈宽,付金磊,等 . 基于 Mask R-CNN 的铸件 X 射线 DR 图像缺陷检测研究[ J ] . 仪器仪表学报, 2020 , 41 ( 3 ): 61-69.
[ 9 ] 朱超平,杨永斌 . 基于改进的 Faster-RCNN 模型的汽车轮毂表 面 缺陷 在线 检测 算 法研 究[ J ] . 表 面 技 术,2020 , 49 ( 6 ): 359-365.
[ 10 ] 王陶然,王明泉,张俊生,等 . 基于 Mask R-CNN 的轮毂缺陷 分 割 技 术 [ J ] . 国 外 电 子 测 量 技 术,2021 , 40( 2 ): 1-5.
[ 11 ] JIANG L L , WANG Y X , TANG Z H , et al.Casting defect detection in X-ray images using convolutional neural networks and attention-guided data augmentation [ J ] .Measurement , 2021 , 170 : 108736.
[ 12 ] 范涛,王明泉,张俊生,等 . 基于轻量化 YOLOv4 的轮毂内部缺陷检测算法[ J ] . 机械与电子,2023 , 41 ( 2 ):3-7 , 12.
[ 13 ] 储钰昆,吴磊,杨洪刚,等 . 基于 YOLOv8 的铸件 DR图像小目 标 缺 陷 检 测 [ J ] . 特 种 铸 造 及 有 色 合 金,2024 , 45 ( 1 ): 35-41.
[ 14 ] SRINIVAS A , LIN T Y , PARMAR N , et al.Bottleneck transformers for visual recognition [ C ]// 2021 IEEE / CVF Conference on Computer Vision and Pattern Recognition( CVPR ),2021 : 16514-16524.
[ 15 ] GUO M H , LIU Z N , MU T J , et al.Beyond self at tention : External attention using two linear layers for visual tasks [ J ] .IEEE Transactions on pattern analysis and machine intelligence , 2022 , 45 ( 5 ): 5436-5447.
[ 16 ] 王彦生,曹雪虹,焦良葆,等 . 基于改进 YOLOv5 的电厂人员 吸 烟 检 测 [ J ] . 计 算 机 测 量 与 控 制,2023 , 31( 5 ): 48-55.

相似文献/References:

[1]卞越洋,高晓科,张伟军.机器人带电作业中的视觉定位与优化策略[J].机械与电子,2021,(05):57.
　BIAN Yueyang,GAO Xiaoke,ZHANG Weijun.Vision Positioning and Optimization Strategyin Robot Live Work[J].Machinery & Electronics,2021,(02):57.
[2]刘瑞昊,于振中,孙 强.改进多尺度特征融合的工业现场目标检测算法[J].机械与电子,2022,(11):40.
　LIU Ruihao,YU Zhenzhong,SUN Qiang.Improved Multi-scale Feature Fusion for Industrial Field Object Detection Algorithm[J].Machinery & Electronics,2022,(02):40.
[3]范 涛,王明泉,张俊生,等.基于轻量化 YOLOv4 的轮毂内部缺陷检测算法[J].机械与电子,2023,41(02):3.
　FAN Tao,WANG Mingquan,ZHANG Junsheng,et al.Internal Defect Detection Algorithm of Wheel Hub Based on Lightweight YOLOv4[J].Machinery & Electronics,2023,41(02):3.
[4]张晋钊,刘新妹,等.改进 YOLOv5s 的 PCB 元器件检测技术[J].机械与电子,2024,42(10):35.
　ZHANG Jinzhao,LIU Xinmei,et al.Improved YOLOv5s Component Detection Technology for PCB[J].Machinery & Electronics,2024,42(02):35.
[5]郑卓纹,吴攀超,王婷婷,等.基于轻量化目标检测算法的指针仪表读数识别[J].机械与电子,2025,(05):10.
　ZHENG Zhuowen,WU Panchao,WANG Tingting,et al.Pointer Gauge Reading Recognition Based on Lightweight Target Detection Algorithm[J].Machinery & Electronics,2025,(02):10.
[6]郭 庆,陈 川,季强东,等.基于改进 YOLOv11 的复杂环境下火焰目标检测[J].机械与电子,2025,(11):33.
　GUO Qing,CHEN Chuan,JI Qiangdong,et al.Flame Target Detection in Complex Environments Based on Improved YOLOv11[J].Machinery & Electronics,2025,(02):33.
[7]兰海麟,吉琳娜,杨风暴,等.基于生成对抗网络的红外与可见光视频语义特征驱动融合方法[J].机械与电子,2026,44(01):35.
　LAN Hailin,JI Linna,YANG Fengbao,et al.Semantic Feature Driven Fusion Method for Infrared and Visible Videos Based on Generative Adversarial Networks[J].Machinery & Electronics,2026,44(02):35.

备注/Memo

备注/Memo:
收稿日期: 2024-07-22
基金项目:国家自然科学基金资助项目( 61171177 )
作者简介:黄心玥 ( 1998- ),女,山西长治人,硕士研究生,研究方向为轮毂缺陷检测和图像处理;王明泉 ( 1970- ),男,山西朔州人,博士,教授,博士研究生导师,研究方向为图像处理、工业检测与识别等,通信作者。
更新日期/Last Update: 2025-03-10