• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Novel algorithms for video text extraction with application to license plate recognition.
【24h】

Novel algorithms for video text extraction with application to license plate recognition.

机译：用于视频文本提取的新算法，并应用于车牌识别。

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

With the rapid advances in digital technology, more and more databases are multimedia in nature, containing images and video in addition to the textual information. Many video databases today are manually indexed based on textual annotations. The manual annotation process is often tedious and time consuming. It is therefore desirable to develop effective computer algorithms for the automatic annotation and indexing of digital video. Using computerized approach, indexing and retrieval are performed based on features extracted directly from the video, which directly capture or reflect the content of the video. Text is an attractive feature for video annotations and indexing because it provides rich semantic information about the video. In addition to annotation and indexing, text extracted from a video is also useful for developing computerized methods for video skimming, browsing, summarization, and abstraction, and for other special applications such as car license plate detection and recognition.; In this thesis, two novel video text extraction algorithms are proposed and implemented. The first is based on a scan line approach, which uses the distinctive characteristic of the grayscale waveforms of video scan lines to detect text regions. The second algorithm uses grayscale morphological operations to extract text. These two approaches fit into the broad category of texture-based approaches. Our new text detection algorithms can detect video text of different font styles and sizes in complex background, with improved performance over that of prior work. In addition to applying the scan line algorithm for general video text extraction, we also modify the algorithm to detect and extract license plate characters embedded in a 3-D scene. A novel algorithm using 3-D invariant features was also developed to recognize the extracted license plate characters.
机译：随着数字技术的飞速发展，本质上越来越多的数据库是多媒体的，除了文本信息外还包含图像和视频。如今，许多视频数据库都是基于文本注释手动索引的。手动注释过程通常很繁琐且耗时。因此，需要开发有效的计算机算法，用于数字视频的自动注释和索引。使用计算机化方法，基于直接从视频中提取的特征执行索引和检索，这些特征直接捕获或反映视频的内容。文本是视频注释和索引的吸引人的功能，因为它提供了有关视频的丰富语义信息。除了注释和索引编制之外，从视频中提取的文本还可用于开发计算机化的方法以进行视频浏览，浏览，摘要和抽象，以及用于其他特殊应用，例如车牌检测和识别。本文提出并实现了两种新颖的视频文本提取算法。第一种基于扫描线方法，该方法使用视频扫描线的灰度波形的独特特征来检测文本区域。第二种算法使用灰度形态学运算来提取文本。这两种方法适合基于纹理的方法的大类。我们的新文本检测算法可以在复杂的背景下检测不同字体样式和大小的视频文本，其性能比以前的工作有所提高。除了将扫描线算法应用于一般视频文本提取外，我们还修改了算法以检测和提取嵌入在3D场景中的车牌字符。还开发了一种使用3-D不变特征的新颖算法来识别提取的车牌字符。

著录项

  • 作者

    Chen, Minya.;

  • 作者单位

    Polytechnic University.;

  • 授予单位 Polytechnic University.;
  • 学科 Computer Science.
  • 学位 Ph.D.
  • 年度 2004
  • 页码 107 p.
  • 总页数 107
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 自动化技术、计算机技术;
  • 关键词

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