基于子图像VCH的文本检测与定位方法研究

Method for Detecting and Locating Text Areas in Images Based on Sub-image VCH

  • 摘要: 提出了一种基于子图像的变异灰度直方图(VCH)的检测与定位文本的方法,对子图像VCH的映射特征及其与图像文本的对应关系作了分析,并给出文本检测与定位的有效算法。该算法通过对行分割子图像VCHC中的凸台和列分割子图像VCHR中的凹谷的识别与定位,实现在图像中快速检测与定位文本。文章给出了计算试验的结果,并进行了分析与评价。

     

    Abstract: It has been attracting wide attention to retrieval,analysis and understanding of huge video data.A number of content based methods for indexing and retrieving have been presented.Text detection and extraction in video images is an effective way for achieving image-content-based retrieval.This paper presents a method for detecting and locating text areas in images based on sub-image VCH.The mapping features of sub-image VCH and the relation between texts and mapping features is analyzed.And the algorithm for detecting and locating text areas in images is described.Some computing experiment results are given and analysis and evaluation are made for it in the end of the paper.

     

/

返回文章
返回