首页> 外国专利> IMAGE-PROCESSING BASED ALGORITHM FOR RECOGNIZING TRAIN NUMBER OF URBAN RAIL TRAIN

IMAGE-PROCESSING BASED ALGORITHM FOR RECOGNIZING TRAIN NUMBER OF URBAN RAIL TRAIN

机译:基于图像处理的城市轨道交通列车数量识别算法

摘要

An image-processing based algorithm for recognizing a train number of an urban rail train. The steps of the algorithm are as follows: obtaining an original image by a camera and performing smoothing processing on the original image; highlighting a train number character area according to a background color of the train number character area, and performing binarization on the image; performing morphological processing on the obtained binarized image; positioning train number character area according to a characteristic of a connected region of the train number character area; performing a negation operation on the positioned image of the train number character area, performing analysis on the connected region of the negated image, and segmenting single characters according to characteristics of connected regions of train number characters; performing normalization processing on the segmented single characters and performing binarization processing; and performing train number recognition using an artificial neural network (ANN). The method improves the accuracy of train number information acquisition, reduces costs of a train number acquisition system, and has the advantages of good timeliness and high train number recognition accuracy.
机译:基于图像处理的算法,用于识别城市铁路列车的列车号。该算法的步骤如下:通过摄像机获取原始图像并对原始图像进行平滑处理;根据列车号字符区域的背景颜色突出显示列车号字符区域,并对图像进行二值化处理;对获得的二值化图像进行形态学处理;根据列车号字符区域的连接区域的特征定位列车号字符区域;对列车号字符区域的定位图像进行取反运算,对取反后的图像的连通区域进行分析,并根据列车号字符的连通区域的特征对单个字符进行分割;对分割后的单个字符进行归一化处理并进行二值化处理;并使用人工神经网络(ANN)进行列车号识别。本发明提高了列车号信息获取的准确性,降低了列车号获取系统的成本,具有时效性好,列车号识别精度高的优点。

著录项

  • 公开/公告号WO2020134324A1

    专利类型

  • 公开/公告日2020-07-02

    原文格式PDF

  • 申请/专利权人 NANJING RUISU RAIL TRANSIT TECHNOLOGY CO. LTD.;

    申请/专利号WO2019CN110230

  • 发明设计人 XING ZONGYI;

    申请日2019-10-10

  • 分类号G06K9/20;

  • 国家 WO

  • 入库时间 2022-08-21 11:10:29

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号