基于模板差分的档案图像集合冗余压缩研究

档案图像信息系统中,页间信息冗余大量存在,对基于页间信息统计特性的压缩方法进行研究具有重要意义。集合冗余压缩正是利用图像之间的相似性降低整个图像集合的嫡。基于模板差分的压缩方法是一种改进的集合冗余压缩技术,它通过模板建立相似档案图像集合,通过在最小一最大差分方法的编解码器中加入模板图像,来提高页间压缩性能。理论分析表明,模板差分压缩方法的压缩性能高于最小一最大差分方法。实验结果表明,模板差分方法和最小最大差分方法均能有效提高图像集合的压缩比,而且模板差分方法比最小一最大差分方法更有利于提高压缩比。

Research on the Set Redundancy Compression of Document Image Based on Template Difference

Much redundant information between document pages exists in the document image information system. The research on the compression method based on the pagcpage statistical features is significant. Set Redundancy Compression(SRC) is such a technique. It reduced the total entropy of the whole image set through utilizing the image page's similarity. Compression based on template differential(CTD) is an improved SRC. The similar image set was constructed by the template. "hhe coding performance was improved by adding the template image to the Min-Max Differential (MMI))coding/decoding model. It was proved theoretically that CTD's coding performance is higher than MMD's. It was demonstrated by experiments that both the CTD and MMI)are benefit to increase the compression ratio of image set, furthermore, and ChD is better than MMD.