J4

• 计算机科学 • 上一篇    下一篇

车牌识别中的图像提取及分割算法

张禹, 马驷良, 韩笑, 张忠波   

  1. 吉林大学 数学研究所, 长春 130012
  • 收稿日期:2005-11-11 修回日期:1900-01-01 出版日期:2006-05-26 发布日期:2006-05-26
  • 通讯作者: 马驷良

Image Extraction and Segment Arithmetic of License Plates Recognition

ZHANG Yu, MA Si-liang, HAN Xiao, ZHANG Zhong-bo   

  1. Institute of Mathematics, Jilin University, Changchun 130012, China
  • Received:2005-11-11 Revised:1900-01-01 Online:2006-05-26 Published:2006-05-26
  • Contact: MA Si-liang

摘要: 提出一种梯度增强法对图像边缘进行增强, 该方法主要通过对图像本身信息进行简单梯度变换以突出图像边缘, 并辅以数学形态学中的膨胀法实现牌照的快速定位. 在字符分割部分, 给出了最小面积旋转法以确定牌照的最佳旋转角, 并采用垂直投影法分割牌照中的字符.

关键词: 模式识别, 车牌识别, 边缘增强, 膨胀, 包围盒, 分割

Abstract: A grad intensity method is put forward to process edges’ intensity of an image, this measure mostly uses simple grad transform to potentate the image edges, at the same time, a dilation method of mathematical morphology is applied to achieve rapid orientation of the license plates. In the process of characters segmentation, we first proposed least area circumrotation means in order to ascertain the best rotation angle of the license plate, then we adopted the way of perpendicularity projection to segment the characters of the license plate.

Key words: pattern recognition, license plate recognition, edge enhancement, dilatation, bounding box, subdivision

中图分类号: 

  • TP391