一种基于模板的档案图像压缩新方法

A New Compression Approach to Document Image Based on Template

在大容量档案图像数据库中,不仅单页档案图像内部存在数据冗余,而且档案图像页之间存在大量集合冗余。本文提出了基于模板的压缩新方法,通过定义相似图像集合的模板,充分利用图像数据的先验知识,对档案图像的内容进行分析和理解,从图像内和图像间以二维模式压缩图像数据。实验表明,该方法能够大幅提高压缩性能。

In super large document image database, data redundancy exists not only in individual image, but also between images. A new compression approach based on template was proposed. Through the definition of template, the approach utilized the image prophetic knowledge sufficiently. After the analysis and understanding of image content, it compressed image from intra-image and inter-image. The experiment showed that the approach could increase compression ratio greatly.