[1]王琪璇,管声启,胡璐萍.基于改进 YOLOv4 的工业棒料识别算法[J].机械与电子,2022,(01):25-29.
　WANG Qixuan,GUAN Shengqi,HU Luping,et al.Industrial Bar Recognition Algorithm Based on Improved YOLOv4[J].Machinery & Electronics,2022,(01):25-29.
点击复制

基于改进 YOLOv4 的工业棒料识别算法()
分享到：

机械与电子[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2022年01期
页码:
25-29
栏目:
设计与研究
出版日期:
2022-01-20

文章信息/Info

Title:
Industrial Bar Recognition Algorithm Based on Improved YOLOv4
文章编号:
1001-2257 ( 2022 ) 01-0025-05
作者:
王琪璇 1 ; 管声启 1 ; 2 ; 胡璐萍 1
1. 西安工程大学机电工程学院,陕西 西安 710048 ; 2. 绍兴市柯桥区西纺纺织产业创新研究院,浙江 绍兴 312030
Author(s):
WANG Qixuan1 ;  GUAN Shengqi1 ;  2 ;  HU Luping1
(1.School of Mechanical and Electronic Engineering , Xi ’an Polytechnic University , Xi ’an 710048 , China ;2.Shaoxing Keqiao West-Tex Textile Industry Innovative Institute , Shaoxing 312030 , China )
关键词:
棒料识别;  YOLOv4 ;  Mobilenetv3 ;  Repulsion 损失函数
Keywords:
bar recognition ;  YOLOv4 ;  Mobilenetv3 ;  Repulsion loss function
分类号:
TP391.4
文献标志码:
A
摘要:
针对工业棒料存在遮挡干扰时难以快速有效识别的问题,提出了一种基于改进 YOLOv4 的棒料识别算法。首先对 YOLOv4 进行轻量化改进,将改进的 Mobilenetv3 作为 YOLOv4 的主干网络,以减少模型参数量,提高算法的检测速度。然后提出在 YOLOv4 原损失函数基础上串联 Repulsion 损失函数,此新增损失函数包含 2 部分: RepGT 损失和 RepBox 损失,RepGT 损失函数计算目标预测框与相邻真实框所产生的损失值,用来减少棒料误检;RepBox 损失函数计算目标预测框与相邻的其他目标预测框所产生的损失值,用来减少棒料漏检。实验结果表明,改进算法的检测速度为 63 帧/ s ,比原 YOLOv4 算法提升了20 帧/ s ;识别准确率达到 97.85% ,比原 YOLOv4 算法提升了 1.62% 。
Abstract:
Aiming at the problem that it is difficult to recognize the industrial bar quickly and effectively when there is partial occlusion interference , a bar recognition algorithm based on improved YOLOv4 is proposed.Firstly , the lightweight improvement of YOLOv4 is carried out , and the improved Mobilenetv3 is used as the backbone network of YOLOv4 , so as to reduce the amount of model parameters and improve the detection speed of the algorithm.Then the Repulsion loss function is cascaded on the basis of YOLOv4? s original loss function.The improved loss function contains two parts : RepGT loss and RepBox loss.The RepGT loss function calculates the loss value generated by the target prediction box and the adjacent real boxes , which is used to reduce the bar fault detection ; The RepBox loss function calculates the loss value generated by the target prediction box and the adjacent other target prediction boxes , which is used to reduce the missing detection of bars.The experimental results show that the detection speed of the improved algorithm is 63 FPS , which is 20 FPS higher than that of the original YOLOv4 algorithm.The recognition accuracy is 97.85% , which is 1.62% higher than the original YOLOv4 algorithm.

参考文献/References:

[ 1 ] 蒲娟,陈勇 . 棒料切割机自动上料机构设计与实施[ J ] .机械工程师,2019 ( 9 ): 126-127 , 130.

[ 2 ] 于小红,程嘉远 . 改进模板匹配的通信目标识别技术[ J ] . 现代防御技术, 2018 , 46 ( 5 ): 69-74.
[ 3 ] 彭玉青,李木,高晴晴,等 . 基于动态模板匹配的移动机器人目标识别[ J ] . 传感技术学报, 2016 , 29 ( 1 ): 58-63.
[ 4 ] 周亦鹏,胡娟,杜军平 . 基于多 Agent 进化计算的图像目标识别[ J ] . 复杂系统与复杂性科学, 2013 , 10 ( 3 ): 55-60.
[ 5 ] 宋曦,周荫清 . 一种基于模板匹配的目标识别方法[ J ] .遥测遥控,2010 , 31 ( 2 ): 51-55.
[ 6 ] 王立中,管声启 . 基于深度学习算法的带钢表面缺陷识别[ J ] . 西安工程大学学报, 2017 , 31 ( 5 ): 669-674.
[ 7 ] LI H , WANG P , SHEN C H.Toward end-to-end car license plate detection and recognition with deep neural networks [ J ] .IEEE Transactions on intelligent transportation systems , 2019 , 20 ( 3 ): 1126-1136.
[ 8 ] 毕松,高峰,陈俊文,等 . 基于深度卷积神经网络的柑橘目标识别方法[ J ] . 农业机械学报, 2019 , 50 ( 5 ): 181-186.
[ 9 ] 黄凤荣,李杨,郭兰申,等 . 基于 Faster R-CNN 的零件表面缺陷检测算法[ J ] . 计算机辅助设计与图形学学报,2020 , 32 ( 6 ): 883-893.
[ 10 ] 周自强,陈强,马必焕,等 . 一种改进的 YOLO 目标检测方法在电缆设备异常状态识别中的应用[ J ] . 电测与仪表,2020 , 57 ( 2 ): 14-20.
[ 11 ] 王伟男,杨朝红 . 基于图像处理技术的目标识别方法综述[ J ] . 电脑与信息技术, 2019 , 27 ( 6 ): 9-15.
[ 12 ] REN S Q , HE K M , GIRSHICK R , et al.Faster R-CNN : towards real-time object detection with region proposal networks [ J ] .IEEE Transactions on pattern analysis and machine intelligence , 2017 , 39 ( 6 ): 1137-1149.
[ 13 ] HE K M , GKIOXARI G , DOLLAR P , et al.Mask R CNN [ C ] ∥2017 IEEE International Conference on Computer Vision ( ICCV ), 2017 : 2980-2988.
[ 14 ] BOCHKOVSKIY A , WANG C Y , LIAO H Y M. YOLOv4 : optimal speed and accuracy of object detection [ EB / OL ] . ( 2020-04-23 )[ 2021-07-23 ] .https : ∥arxiv.org / abs / 2004.10934.
[ 15 ] LIU W , ANGUELOV D , ERHAN D , et al.SSD : single shot multibox detector [ EB / OL ] . ( 2015-1-08 )[ 2021-07-23 ] .https : ∥arxiv.org / abs / 1512.02325.
[ 16 ] 谢斌红,袁帅,龚大立 . 基于 RDB YOLOv4 的煤矿井下有遮挡行人检测[ J / OL ] . 计算机工程与应用: 1-10 [ 2021-07-23 ] .http : ∥kns.cnki.net / kcms / detail / 11.2127.TP.20210419.1501.079.html.
[ 17 ] ZHANG L L , LIN L , LIANG X D , et al.Is faster R-CNN doing well for pedestrian detection [ EB / OL ] .(2016-07-24 )[ 2021-07-23 ] .https : ∥arxiv.org /abs / 1607.07032.
[ 18 ] WANG X L , XIAO T T , JIANG Y N , et al.Repulsion loss : detecting pedestrians in a crowd [ C ] ∥2018 IEEE / CVF Conference on Computer Vision and Pattern Recognition , 2018 : 7774-7783.

相似文献/References:

[1]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(01):3.

备注/Memo

备注/Memo:
收稿日期: 2021-08-06
基金项目:绍兴市柯桥区西纺纺织产业创新研究院 2019 年度产学研协同创新项目( 19KQYB13 )
作者简介:王琪璇 ( 1997- ),女,陕西西安人,硕士研究生,研究方向为机器人视觉;管声启 ( 1971- ),男,安徽安庆人,博士,教授,研究方向为机械零件质量检测、机器人视觉等;胡璐萍 ( 1992- ),女,陕西安康人,硕士研究生,研究方向为机器人视觉。
更新日期/Last Update: 2022-02-28