节点文献

多尺度Retinex算法在简牍图像增强中的应用

Application of Multiscale Retinex in JianDu Image Enhancement

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 汪刘艳王绪本

【Author】 WANG Liu-yan WANG Xu-ben (College of Information Engineering,Chengdu University of Technology,Chengdu 610059,China)

【机构】 四川成都理工大学信息工程学院

【摘要】 简牍图像增强是简牍文字修复和文字信息提取最为关键的一个步骤。以长沙简牍博物馆收藏的简牍文物为实物模型,以该馆提供的简牍数字图像为研究对象,提出了基于多尺度Retinex(MSR)算法的简牍图像增强方法,论述了本算法的原理及实现方法。采用MSR算法来消除光照对简牍图像的退化影响,以达到增强图像暗区细节信息的目的;同时针对MSR输出图像偏暗,提出了直方图拉伸算法。试验结果表明:本算法能够提高简牍图像对比度的同时,能显著地增强图像暗区的细节信息并且使简牍图像具有较好、较高的色彩保真度。本方法克服了目前常规的图像增强方法的不足,能够满足简牍考古人员的要求。

【Abstract】 Enhancement of the JianDu Image is a key procedure in character reconstruction and text extraction.We made the JianDu relics from Changsha Jiandu museum as a model and taken the JianDu images as our research objects.An enhancement algorithm of JianDu image based on multiscale Retinex was presented.Based on the theory of Retinex,the multi-scale Retinex(MSR) algorithm was employed to remove the luminance degradation of the JianDu image.The information of hidden area of the JianDu image was obviously enhanced.For the intensity of the MSR processing output is weak,one normal histogram stretching method was deployed to improve the output brightness.The results of experiment show that MSR can improve effectively not only the contrast but also can observably enhance the information of hidden area in the JianDu image and provided better tonal rendition.The algorithm can overcome the lack of enhancement of traditional enhancement methods and satisfy the archaeologist’s demand.

【基金】 国家“十五”科技攻关项目“文物保护关键技术研究”(2004BA810B)资助
  • 【文献出处】 计算机科学 ,Computer Science , 编辑部邮箱 ,2009年03期
  • 【分类号】TP391.41
  • 【被引频次】8
  • 【下载频次】291
节点文献中: 

本文链接的文献网络图示:

本文的引文网络