[1]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):10-17.
　ZHENG Zhuowen,WU Panchao,WANG Tingting,et al.Pointer Gauge Reading Recognition Based on Lightweight Target Detection Algorithm[J].Machinery & Electronics,2025,(05):10-17.
点击复制

基于轻量化目标检测算法的指针仪表读数识别()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
期数:
2025年05期
页码:
10-17
栏目:
研究与设计
出版日期:
2025-05-23

文章信息/Info

Title:
Pointer Gauge Reading Recognition Based on Lightweight Target Detection Algorithm
文章编号:
1001-2257 ( 2025 ) 05-0010-08
作者:
郑卓纹; 吴攀超; 王婷婷; 孙 琦
东北石油大学电气信息工程学院,黑龙江 大庆 163318
Author(s):
ZHENG Zhuowen ;  WU Panchao ;  WANG Tingting ;  SUN Qi
( School of Electrical and Information Engineering , Northeast Petroleum University , Daqing 163318 , China )
关键词:
仪表读数识别; 目标检测; 语义分割; 透视变换
Keywords:
instrument reading recognition ;  object detection ;  semantic segmentation ;  perspective transformation
分类号:
TP391.41 ;TH70
文献标志码:
A
摘要:
针对巡检机器人嵌入式系统面临的内存限制和计算能力不足的问题,提出了一种基于轻量化目标检测算法的指针仪表读数识别算法。首先,在 YOLOv8n 的基础上,设计了轻量化特征提取模块 RepC2f 和轻量化检测头 LSD ,这 2 项设计旨在提升检测精度的同时显著降低模型参数量,从而提高检测的实时性。实验结果表明,该算法相比于 YOLOv8n 精度提升了 0.94 百分点,权重模型减少了 2.73 MB ,而帧率则提升了 11.9 帧/ s 。然后,采用 Deeplabv3+ 算法对指针和主刻度线进行精确分割。随后,通过透视变换和最小二乘法实现表盘的几何校正和指针的拟合。最后,依据主刻度线的角度法进行读数的计算。
Abstract:
To address the problems of memory limitation and insufficient computing power faced by the embedded system of inspection robots , this paper proposes a pointer meter reading recognition algorithm based on a lightweight target detection algorithm , named RL YOLO.Firstly , on the basis of YOLOv8n , a lightweight feature extraction module , RepC2f , and a lightweight detection head , LSD , are designed , which aim to improve the detection accuracy , and at the same time significantly reduce the number of model parameters , thus improving the real-time detection.The experimental results show that the algorithm improves the accuracy by 0.94 percentage points compared with YOLOv8n , the weight model is reduced by 2.73 MB , while the frame rate is improved by 11.9 frames / s.Next , the Deeplabv3+ algorithm is used to accurately segment the pointer and the main scale.Subsequently , geometric correction of the dial and fitting of the hands are achieved by perspective transformation and least squares.Finally , the readings are calculated based on the angular method of the main scale line.

参考文献/References:

[ 1 ] BAYOUDH K.A survey of multimodal hybrid deep learning for computer vision : architectures , applications , trends , and challenges [ J ] .Information fusion , 2023 , 105 : 102217.

[ 2 ] 虞健,陈凯迪,施展,等 . 基于深度学习的指针式仪表的读数检测方法[ J ] . 自动化与仪表, 2023 , 38 ( 7 ): 91-94 , 114.
[ 3 ] 张平,丁永泽,王致芃,等变电站指针式仪表精准读数及可靠度估计模型[ J ] . 激光与光电子学进展, 2020 , 57( 10 ): 133-142.
[ 4 ] 王明,闫泽恩,王思瑶,等 . 基于 YOLOv5 和数字图像处理的指针式压力表识别方法[ J ] . 组合机床与自动化加工技术,2024 ( 12 ): 135-138 , 144.
[ 5 ] 闫富海,徐望明,黄酋淦,等 . 基于轻量级图像语义分割模型的指针式仪表读数全自动识别[ J ] . 激光与光电子学进展,2022 , 59 ( 24 ): 19-28.
[ 6 ] JING L L , TIAN Y L.Self-supervised visual feature learning with deep neural networks : a survey [ J ] .IEEE Transactions on pattern analysis and machine intelligence , 2021 , 43 ( 11 ): 4037-4058.
[ 7 ] WANG L , WANG P , WU L H , et al.Computer vision based automatic recognition of pointer instruments : dat a set optimization and reading [ J ] .Entropy , 2021 , 23( 3 ): e23030272.
[ 8 ] 沈云青,熊卫华,黄为民,等 . 基于模板匹配与 Hough 圆检测的仪表识别[ J ] . 计算机技术与发展,2021 , 31( 4 ): 69-73.
[ 9 ] 朱斌滨,樊绍胜 . 雨雾环境下的变电站指针式仪表识别方法[ J ] . 激光与光电子学进展, 2021 , 58 ( 24 ): 221-230.
[ 10 ] 邢浩强,杜志岐,苏波 . 变电站指针式仪表检测与识别方法[ J ] . 仪器仪表学报, 2017 , 38 ( 11 ): 2813-2821.
[ 11 ] 王康,陆生华,陈潮,等 . 基于 YOLOv5 的指针式仪表检测与读数识别算法研究[ J ] . 三峡大学学报(自然科学版),2022 , 44 ( 6 ): 42-47.
[ 12 ] 李金红,熊继平,陈泽辉,等 . 基于卷积神经网络的指针式仪表识别[ J ] . 计算机系统应用, 2021 , 30 ( 9 ): 85-91.
[ 13 ] 余永维,韩鑫,杜柳青 . 基于 Inception-SSD 算法的零件识别[ J ] . 光学精密工程, 2020 , 28 ( 8 ): 1799-1809.
[ 14 ] TAN L , LV X Y , LIAN X F , et al.YOLOv4 _ drone : UAV imagetarget detection based on an improved YOLOv4 algorithm [ J ] .Computers and electrical engineering , 2021 , 93 : 107261.
[ 15 ] LIU Y , LIU J , KE Y C.A detection and recognition system of pointer meters in substations based on computer vision [ J ] .Measurement , 2020 , 152 : 107333.
[ 16 ] 胡鑫,欧阳华,侯卓成,等 . 特征增强的 SSD 算法及在仪表读数识别上的应用 [ J ] . 电子测量技术, 2021 , 44( 17 ): 154-159.
[ 17 ] 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.
[ 18 ] 赵伟达,陈海文,郭陆阳,等 . 基于 YOLO E 与改进 OCRNet 图像分割的变电站仪表读数自适应识别方法[ J ] . 电力建设, 2023 , 44 ( 11 ): 75-85.
[ 19 ] 李盛涛,侯立群,董亚松 . 基于 R-YOLOv7 和 MIMO-CTFNet 的指针式仪表自动读数方法[ J ] . 图学学报,2024 , 45 ( 6 ): 1313-1327.
[ 20 ] 李春蕾,阮艺铭,张小明,等 . 基于 YOLOv8 和改进 UNet++ 变电站指针式仪表读数识别[ J ] . 计算机系统应用,2024 , 33 ( 12 ): 170-176.

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,(05):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,(05):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(05):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(05):35.
[5]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.
　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,(05):26.
[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,(05):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(05):35.

备注/Memo

备注/Memo:
收稿日期: 2024-12-31
基金项目:国家自然科学基金资助项目( 52474036 )
作者简介:郑卓纹 ( 2000- ),男,山东济宁人,硕士研究生,研究方向为图像处理;吴攀超 ( 1981- ),男,黑龙江大庆人,博士,讲师,研究方向为计算机视觉,通信作者, E-mail : 36187987@qq.com 。
更新日期/Last Update: 2025-06-11