欢迎访问《图学学报》 分享到:

图学学报 ›› 2024, Vol. 45 ›› Issue (6): 1313-1327.DOI: 10.11996/JG.j.2095-302X.2024061313

• 图像处理与计算机视觉 • 上一篇    下一篇

基于R-YOLOv7和MIMO-CTFNet的指针式仪表自动读数方法

李盛涛(), 侯立群(), 董亚松   

  1. 华北电力大学自动化系,河北 保定 071003
  • 收稿日期:2024-07-23 接受日期:2024-09-26 出版日期:2024-12-31 发布日期:2024-12-24
  • 通讯作者:侯立群(1972-),男,副教授,博士。主要研究方向为图像处理、无线传感器网络和设备故障诊断。E-mail:houliqun@ncepu.edu.cn
  • 第一作者:李盛涛(1998-),男,硕士研究生。主要研究方向为图像处理和深度学习。E-mail:lst18315889026@163.com
  • 基金资助:
    河北省自然科学基金(F2016502104)

Automatic reading of pointer meters based on R-YOLOv7 and MIMO-CTFNet

LI Shengtao(), HOU Liqun(), DONG Yasong   

  1. Department of Automation, North China Electric Power University, Baoding Hebei 071003, China
  • Received:2024-07-23 Accepted:2024-09-26 Published:2024-12-31 Online:2024-12-24
  • Contact: HOU Liqun (1972-), associate professor, Ph.D. His main research interests cover image processing, wireless sensor networks, and device fault diagnosis. E-mail:houliqun@ncepu.edu.cn
  • First author:LI Shengtao (1998-), master student. His main research interests cover image processing and deep learning. E-mail:lst18315889026@163.com
  • Supported by:
    Natural Science Foundation of Hebei(F2016502104)

摘要:

针对现有方法中表盘关键信息提取过程繁琐、读数误差较大和相机抖动导致的运动模糊问题,提出了一种基于R-YOLOv7和MIMO-CTFNet的指针式仪表自动读数方法。首先,构建兼顾精度和轻量化的R-YOLOv7算法实现指针式仪表表盘和表盘关键信息检测;然后,设计了MIMO-CTFNet算法以实现运动模糊仪表图像的复原;最后,利用提取的表盘关键信息进行基于小刻度线的角度法读数。实验结果表明改进后的R-YOLOv7在表盘关键信息检测数据集上所需的参数量、FLOPs、ADT和mAP50:95分别为12 M个、60.30 G次、17.04 ms和86.5%;改进后的MIMO-CTFNet算法在采集的运动模糊数据集上的PSNR和SSIM分别达到33.05 dB和0.935 3;该读数方法的读数最大引用误差为0.35%,需要运动模糊处理和无需运动模糊处理的图像读数时间分别为0.561 s和0.128 s,从而验证了该方法的有效性。

关键词: 指针式仪表, R-YOLOv7, MIMO-CTFNet, 自动读数, 轻量化

Abstract:

To solve the problems in current pointer meter reading methods, such as the complicated reading process, significant reading errors, and the motion blur caused by camera shakes, an automatic reading method based on R-YOLOv7 and MIMO-CTFNet (multi-input multi-output CNN-transformer fusion network) was proposed. First, the R-YOLOv7 algorithm was constructed to consider both accuracy and lightweight for detecting the dial and its key information. Then, a MIMO-CTFNet algorithm was designed to recover the motion-blurred meter images. Finally, the angle method based on the extracted small scales was utilized to perform meter reading. The experimental results showed that for the data set of dial key information finding, the parameters, FLOPs, ADT, and mAP50:95 were 12 M, 60.30 G, 17.04 ms, and 86.5%, respectively. The PSNR and SSIM of the improved MIMO-CTFNet algorithm achieved 33.05 dB and 0.935 3, respectively. The maximum fiducial error of the proposed reading method was 0.35%, and the reading time for images requiring and not requiring motion blur was 0.561 s and 0.128 s, respectively, validating the effectiveness of the proposed method.

Key words: pointer meters, R-YOLOv7, MIMO-CTFNet, automatic reading, light weight

中图分类号: