轻量化YOLOv5s与U2-Net的指针式仪表智能读数方法

doi:10.13438/j.cnki.jdzk.2025.04.007

吉首大学学报（自然科学版） ›› 2025, Vol. 46 ›› Issue (4): 44-51.DOI: 10.13438/j.cnki.jdzk.2025.04.007

轻量化YOLOv5s与U²-Net的指针式仪表智能读数方法

苏盈盈，斯洪云，唐霞，刘兴华，郭利霞

(1.重庆科技大学电气工程学院，重庆 401331；2.中国石化西南石油工程重庆钻井分公司，重庆 400010)

出版日期:2025-07-25 发布日期:2025-08-05
作者简介:苏盈盈（1982—），女，黑龙江伊春人，重庆科技大学电气工程学院教授，博士，硕士生导师，主要从事智能计算与模式识别研究
基金资助:
重庆市教育委员会科学技术研究项目(KJQN202101510)；重庆市自然科学基金面上项目(CSTB2022NSCQ-MSX1425)；重庆科技学院硕士研究生创新计划项目(YKJCX2320403)

Lightweight YOLOv5s with U²-Net for Intelligent Reading of Pointer Meters

SU Yingying,SI Hongyun,TANG Xia,LIU Xinghua,GUO Lixia

(1.College of Electrical Engineering,Chongqing University of Science and Technology,Chongqing 401331,China;2.Sinopec Southwest Petroleum Engineering Chongqing Drilling Branch,Chongqing 400010,China)

Online:2025-07-25 Published:2025-08-05

摘要/Abstract

摘要：针对指针式仪表智能定位速度慢和读数精准度低的问题，设计了一种基于YOLOv5s模型与U²-Net模型的指针式仪表智能读数方法.首先，对YOLOv5s主干与颈部中的部分批量化归一层进行通道剪枝，实现模型轻量化，提升仪表定位速度；然后，针对现场采集图像存在阴影、噪声等问题，对图像进行增强处理，使模型适用于复杂环境；其次，采用U²-Net模型分割指针和刻度，定位指针在刻度上的相对位置，提升读数的准确率；最后，采用距离法进行仪表读数.实验结果表明，轻量化YOLOv5s模型的大小较YOLOv5s模型缩减56.35%，定位时间减少18.18%，轻量化YOLOv5s模型的读数识别测试准确率为93.42%，平均绝对误差为0.032，平均读数误差率为0.605%，每张图片平均测试时间为0.402 s，满足工业检测要求.

关键词： 指针式仪表, YOLOv5s, 通道剪枝, U²-Net, 距离法

Abstract: Aiming at the problem of slow speed of intelligent positioning and low accuracy of reading for pointer meters,a method of intelligent reading for pointer meters based on YOLOv5s and U²-Net is proposed.Firstly,the partial batch normalization layer in the trunk and neck of YOLOv5s is used for channel pruning to realize the model lightweight and improve the speed of meter positioning;secondly,in view of shadow and noisy images collected in the field,the image is enhanced,so as to make the model suitable for complex environments;and then,the U²-Net model is adopted to segment the pointer and the scale,to locate the relative position of pointer on the scale,and to improve the accuracy of readings.Then,the U2-Net model is used to segment the pointer and the scale,locate the relative position of the pointer on the scale,and improve the accuracy of the readings;finally,the distance method is used for meter reading.The results of the experiment and reading test show that the size of the lightweight YOLOv5s model is reduced by 56.35%,the positioning time is reduced by 18.18%,the accuracy of the reading recognition test is 93.42%,the average absolute error is 0.032,the average reading error rate is 0.605%,and the average test time per picture is 0.402 s,which meets the industrial inspection requirements.

Key words: pointer instrument, YOLOv5s, channel pruning, U²-Net, distance method

苏盈盈, 斯洪云, 唐霞, 刘兴华, 郭利霞. 轻量化YOLOv5s与U²-Net的指针式仪表智能读数方法[J]. 吉首大学学报（自然科学版）, 2025, 46(4): 44-51.

SU Yingying, SI Hongyun, TANG Xia, LIU Xinghua, GUO Lixia. Lightweight YOLOv5s with U²-Net for Intelligent Reading of Pointer Meters[J]. Journal of Jishou University(Natural Sciences Edition), 2025, 46(4): 44-51.

参考文献

[1] 夏臻康,李维刚,田志强.基于YOLOv5的指针式仪表自动读数方法研究[J].仪表技术与传感器,2023(6):44-51.
[2] 刘晏,李玉梅,张涛,等.基于改进的YOLOv5s指针式仪表检测与读数识别[J].北京信息科技大学学报(自然科学版),2023,38(3):9-14.
[3] 彭道刚,黄孟俊,戚尔江,等.基于YOLOv4的火电厂指针仪表识别方法研究[J].计算机应用与软件,2023,40(6):166-172；222.
[4] 王英贤,毕超.基于轻量化深度学习的指针式仪表读数方法[J].自动化与仪器仪表,2023(3):31-34.
[5] 杨诗琪,吴佳仪,陈墨楠,等.基于深度学习的指针式仪表自动读数方法[J].电子测量技术,2023,46(5):149-156.
[6] 万吉林,王慧芳,管敏渊,等.基于Faster R-CNN和U-Net的变电站指针式仪表读数自动识别方法[J].电网技术,2020,44(8):3097-3105.
[7] 宫倩,别必龙,范新南,等.基于关键点检测的指针仪表读数算法[J].电子测量与仪器学报,2023,37(3):66-73.
[8] 延旭博.基于深度学习的巡检机器人指针式仪表识别方法[J].现代计算机,2023,29(22):1-7.
[9] 金爱萍,袁亮,周德勤,等.基于YOLOv5和U-Net的指针式仪表读数识别方法[J].仪表技术与传感器,2022(11):29-33.
[10] 曲海成,张旺.基于YOLOv7+U2-Net的指针式仪表检测与识别[J].计算机系统应用,2023,32(11):276-285.
[11] 林锟煌,李建锋,汪洋,等.改进的YOLOv4肺结节检测算法[J].吉首大学学报(自然科学版),2023,44(1):24-29.
[12] 何亚平,苏盈盈,周能扬,等.一种GSN-YOLOv5s的火焰烟雾检测方法[J].重庆科技学院学报(自然科学版),2022,24(6):55-59；83.
[13] 赵立杰,袁昌彪,黄明忠,等.基于通道剪枝的ACAM-YOLOv5s绝缘子缺陷检测[J].电子测量技术,2023,46(9):108-116.
[14] LIU Zhuang,LI Jianguo,SHEN Zhiqiang,et al.Learning Efficient Convolutional Networks Through Network Slimming[C]∥CUCCHIARA R,MATSUSHITA Y,SEBE N,et al.Proceedings of the IEEE International Conference on Computer Vision (ICCV).New York:IEEE,2017:2736-2744.
[15] QIN Xuebin,ZHANG Zichen,HUANG Chenyang,et al.U2-Net:Going Deeper with Nested U-Structure for Salient Object Detection[J].Pattern Recognition,2020,106:107404.DOI:10.1016/j.patcog.2020.107404.
[16] RONNEBERGER OLAF,FISCHER PHILIPP,BROX THOMAS.U-Net:Convolutional Networks for Biomedical Image Segmentation[C]∥NAVAB NASSIR,HORNEGGER JOACHIM,WELLS WILLIAM M,et al.Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015:18th International Conference.Cham:Springer International Publishing,2015:234-241.
[17] 张建寰,曾亮,刘涛,等.户外巡检机器人指针式仪表识别方法研究[J].仪表技术与传感器,2023(11):51-56;64.
[18] 石孟曦,张文红,张磊,等.基于LSD和改进YOLOv2的压力表识读方法[J].计算机仿真,2023,40(5):284-288;294.

轻量化YOLOv5s与U²-Net的指针式仪表智能读数方法

Lightweight YOLOv5s with U²-Net for Intelligent Reading of Pointer Meters

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics

本文评价