基于Tesseract-OCR的燃气表自动识别研究  

Research on automatic identification of gas meter based on Tesseract-OCR

作　　者：冯玉田[1] 侯玖廷 顾乐易 Feng Yutian;Hou Jiuting;Gu Leyi(School of Communication and Information Engineering,Shanghai University,Shanghai 200444,China)

机构地区：[1]上海大学通信与信息工程学院

出　　处：《电子测量技术》2019年第21期82-86,共5页Electronic Measurement Technology

摘　　要：设计了摄像直读式燃气表自动抄表系统,通过图像处理和OCR技术对燃气表读数进行识别。采用在图像HSV空间进行直方图均衡和中值滤波对燃气表图像进行预处理,提高了适应性。在用Tesseract光学字符识别引擎进行字符特征提取中,提出了基于颜色和轮廓的定位算法分割燃气表读数区域,使得系统具有更好的稳定性。识别中针对燃气表字体训练Tesseract,从而实现了较高的识别正确率。实验中,通过对多个品牌的燃气表进行测试,识别准确率达到98.7%。本文设计的燃气表自动识别系统具有识别准确率高、鲁棒性好、识别速度快等特点,可为燃气表自动抄表或自助抄表提供有效的技术途径。An automatic meter reading system of direct reading gas meter by camera is designed,and the gas meter reading is recognized by image processing and OCR technology.Histogram equalization and median filtering in the image HSV space are used to preprocess the gas meter image to improve the adaptability.In the Tesseract optical character recognition engine for character feature extraction,a positioning algorithm based on color and contour is proposed to segment the gas meter reading area,so that the system has better stability.Tesseract is trained in gas meter font recognition,thus achieving high recognition accuracy.In the experiment,the recognition accuracy reached 98.7%by testing the gas meters of several brands.The automatic gas meter recognition system designed in this paper has the characteristics of high recognition accuracy,good robustness and fast recognition speed,etc.,which can provide an effective technical way for automatic gas meter reading or self-reading.

关 键 词：燃气表自动抄表 图像处理 图像识别 Tesseract-OCR 

分 类 号：TN911.73[电子电信—通信与信息系统]