摘要
许多图像中包含的文本信息对于图像高层语义内容的自动理解、图像索引和检索非常有用。复杂背景图像中文本信息的提取一般包括文字的自动检测、定位、提取、分割和识别,由于图像中文本的大小、字体、字形、位置、排列和图像的清晰度、对比度等不定,使文本自动提取非常困难。文章对目前国内外图像中文本提取主要技术和发展方向进行了综述,便于该领域的研究人员了解和借鉴,同时指出了今后研究方向的热点。
Text data contained in many images is greatl useful for automatic understanding of high-level semantics, image indexing and retrieval. Automatic extraction of text information involves automatic detection, location, extraction, divisicn and recognition. However, variations of text due to differences in size, style, orientation, and alignment, as well as low image contrast and definition make the problem of automatic text extraction extremely challenging. In order to make researchers know the academic area more systemically, this paper discusses the current development of the area gives an overview of state of text detection and recognition technique in complex images,and points out promising directions for future research.
出处
《江西教育学院学报》
2008年第3期18-21,共4页
Journal of Jiangxi Institute of Education
关键词
文本信息提取
图像
文字
检测
定位
识别
分割
text information extraction
image
text
de tection
localization
recognition
segmentation
