李伟;王飒;丁健刚;陈昊;肖力炀;

• 1:长安大学信息工程学院
• 2:西安热工研究院有限公司

摘要(Abstract)：

针对石油化工厂中人工抄表导致的低效、高误差和成本高等弊端，以及仪表图像拍摄条件场景复杂等问题，提出了一种基于改进Faster RCNN模型的工业数字表检测方法。首先，在特征提取网络阶段对卷积层低层和高层的网络特征进行融合，提高模型对细粒度细节和小目标的敏感度；其次，结合SENet网络结构，使模型关注不同通道的重要程度，通过分配不同的学习权重来强化对目标的关注度；最后，利用RPN网络进行最后处理，提取出数字表图像的边界框位置信息。结果表明，本文提出的模型检测精度为97.3%,相对于传统目标检测算法来说能更精准地识别出数字表。

关键词(KeyWords)： Faster RCNN;特征融合;SENet;数字表检测

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金资助项目(51978071);; 中央高校基本科研业务费专项资金资助项目(300102249301,300102249306)

作者(Authors): 李伟;王飒;丁健刚;陈昊;肖力炀;

参考文献(References)：

• [1] 中国报告大厅.2021年1-2月全国电工仪器仪表产量统计分析[EB/OL].(2021-03-24)[2022-03-20].http://www.chinabgao.com/chanliang/60255.html
• [2] 刘雯雯.基于数字图像处理的家用电表识别系统研究[D].哈尔滨：东北林业大学，2009.
• [3] 陈广华，苗鹏，张宗健，等.字轮电能表读数系统的数字识别算法[J].北京交通大学学报，2012,36(3):87-91.CHEN Guanghua,MIAO Peng,ZHANG Zongjian,et al.Algorithm of digit recognition of wheel-power-meter reading system[J].Journal of Beijing Jiaotong University,2012,36(3):87-91.
• [4] 杨娟.基于数字图像处理的电能表图像识别技术研究与实现[D].南京：南京理工大学，2012.
• [5] ZHANG Yunzhou,YANG Shanbao,SU Xiaolin,et al.Automatic reading of domestic electric meter:an intelligent device based on image processing and ZigBee/Ethernet communication[J].Journal of Real Time Image Processing,2016,12(1):133-143
• [6] REDMON J,DIVVALA S,GIRSHICK R,et al.You only look once:unified,real-time object detection[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas,NV,USA,2016:779-778.
• [7] LIU W,ANGUELOV D,ERHAN D,et al.SSD:single shot MultiBox detector[C]//Proceedings of the 14th European Conference on Computer Vision.Amsterdam,The Netherlands,2016:1-17.
• [8] GIRSHIC R.Faster R-CNN[C]//Proceedings of 2015 IEEE International Conference on Computer Vision.Santiago Chile,2015:1440-1448.
• [9] REN Shaoqing,HE Kaiming,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.
• [10]ZUO Lin,HE Peilin,ZHANG Changhua,et al.A robust approach to reading recognition of pointer meters based on improved mask-RCNN[J].Neurocomputing,2020,388:90-101.
• [11] ZUO Lin,HE Peilin,ZHANG Changhua,et al.A value recognition algorithm for pointer meter based on improved Mask-RCNN[C]//9th International Conference on Information Science and Technology(ICIST).IEEE,2019:108-113.
• [12] WANG Zheng,CHEN Kai,HE Peilin,et al.Multi-scale aggregation network for temporal action proposals[J].Pattern Recognition Letters,2019,122:60-65.
• [13]林刚，王波，彭辉，等.基于改进Faster-RCNN的输电线巡检图像多目标检测及定位[J].电力自动化设备，2019,39(5):213-218.LIN Gang,WANG Bo,PENG Hui,et al.Multiple-target detection and localization of transmission line inspection image based on improved Fast-RCNN[J].Electric Power Automation Equipment,2019,39(5):213-218.
• [14]刘召，张黎明，耿美晓，等.基于改进的Faster R-CNN高压线缆目标检测方法[J].智能系统学报，2019,14(4):627-634.LIU Zhao,ZHANG Liming,GENG Meixiao,et al.Object detection of high-voltage cable based on improved Faster R-CNN[J].CAAI Transaction on Intelligent Systems,2019,14(4):627-634.
• [15]赵文清，程幸福，赵振兵，等.注意力机制和Faster RCNN相结合的绝缘子识别[J].智能系统学报，2020,15(1):92-98.ZHAO Wenqing,CHENG Xingfu,ZHAO Zhenbing,et al.Insulator recognition based on attention mechanism and Faster RCNN[J].CAAI Transactions on Intelligent Systems,2020,15(1):92-98.
• [16]陈英，蒋文峰，杨丰玉，等.智能电能表数字识别算法研究[J].信息通信，2020(1):17-21,26.CHEN Ying,JIANG Wenfeng,YANG Fengyu,et al.Research on digital recognition algorithm of intelligent electric meter[J].Information and Communication,2020(1):17-21,26.
• [17] 骆文莉，吴秦.多层次特征融合与注意力机制的文本检测[J/OL].小型微型计算机系统：1-9[2021-04-23].http://kns.cnki.net/kcms/detail/21.1106.TP.20210420.1106.017.html.
• [18] LUO Huiyuan,HAN Guangliang,WU Xiaotian,et al.LF3Net:leader-follower feature fusing network for fast saliency detection[J].Neurocomputing,2021,449:24-37.
• [19]彭昆福，王子磊，王磊，等.基于深度回归的指针仪表读数识别方法[J].信息技术与网络安全，2020,39(12):37-43.PENG Kunfu,WANG Zilei,WANG Lei,et al.Pointer meter reading recognition by deep regression[J].Information Technology and Net Security,2020,39(12):37-43.
• [20]MA He,HAN Guoqing,PENG Long,et al.Rock thin sections identification based on improved squeeze-and-Excitation Networks model[J].Computers and Geosciences,2021,152:104780.