• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International Conference on Signal Processing(ICSP'06); 20061116-20; Guilin(CN) >A Method of Multi-view Vehicle License Plates Location Based on Rectangle Features
【24h】

A Method of Multi-view Vehicle License Plates Location Based on Rectangle Features

机译：基于矩形特征的多视角车牌定位方法

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In this paper, a method of image processing for Multi-View Vehicle License Plates Location, which is able to process images rapidly and achieve high recognition rates even when images are fairly complex, in poor illumination, and noise, is discussed. This method includes two steps. The first step is to locate the regions of license plates; the second is a transform for each valid area to form an upright image. This paper is distinguished by one key contribution, which is used to find the contour of each valid area including four vertexes, four lines and several characters by using simple rectangle features selected by the AdaBoost algorithm. The other contribution of this work is that a transform based on nonlinear distortion model that is applied to each located license plate to form an upright image convenient for recognition.
机译：本文讨论了一种用于多视图车辆牌照定位的图像处理方法，该方法即使在图像相当复杂，照明不佳且有噪声的情况下，也能够快速处理图像并获得较高的识别率。此方法包括两个步骤。第一步是找到车牌区域。第二个是对每个有效区域进行变换以形成直立图像。本文的主要特点是通过AdaBoost算法选择的简单矩形特征，使用一个关键贡献来查找每个有效区域的轮廓，包括四个顶点，四行和几个字符。这项工作的另一个贡献是，将基于非线性失真模型的变换应用于每个定位的牌照，以形成便于识别的直立图像。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

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