项目名称: 笔迹图像中关键词语过滤技术研究
项目编号: No.61462080
项目类型: 地区科学基金项目
立项/批准年度: 2015
项目学科: 计算机科学学科
项目作者: 卡米力·木依丁
作者单位: 新疆大学
项目金额: 47万元
中文摘要: 本项目在前期笔迹鉴别技术研究工作基础上,将深入研究笔迹图像中直接搜索是否有感兴趣的词语(统称为关键词语或敏感词语)方面的基础理论与关键技术。具体研究内容包括较大规模笔迹图像库建设,笔迹图像预处理,文本行检测与提取,文本行图像分割后的倾斜歪曲校正,以及基于整词变形模板匹配的关键词语检测与过滤技术等。这种采用现代科技手段加以监控不良信息的传播、定位和取证技术不仅在基于内容的信息检索领域具有广泛的应用价值,而且在我区民族团结、社会稳定、以及国家的长治久安、维稳工作中具有重要的实用价值。
中文关键词: 关键词过滤;倾斜校正;整词匹配;笔迹图像;维吾尔文
英文摘要: based on the previous research results on handwritten writer identification rechnology, this project will conduct the further study on the keyword spotting issues directly from the handwritten document images, mainly deal with the problem of wether the keyword or sensitive word exist in those text lines.the research contents include gathering large scale handwritten documents, preprocessing, text line detection and extraction, registering those segmented text line images, and the variable template matching based algorithms for keyword detection etc. these kind of technologies have a great realistic applications in maintaining ethnic unity, social stability, as well as the country's long-term stability, as well as in the civil applications like content based information retrieval.
英文关键词: Keyword Spotting;tilt correction;Wholeword Matching;Handwritten Document Image;Uyghur Script