文本分类在工程图中的信息隐藏应用

工程图的信息隐藏在CAD领域中具有重大的意义。为了技术保密,工程图在发布时往往需安将其中的关键信息(如尺寸和材料)隐藏起来,其中大部分需要隐藏的信息是以文本方式存在的。传统的信息隐藏是通过人工的方式进行关键文本抽取并加以隐藏而实现的。本文提出了用线性分类的方法,来实现工程图关键文本的自动抽取,从而达到工程图信息隐藏的目的。

A New Way of Information Hiding in Drawings-text Classification

The drawings contain a lot of information confidential. Nowadays, as one format of most frequently transferred archive, drawings may possibly lose its confidentiality partly or entirely. To avoid the potential lost of confidentiality, in traditional ways, engineers have to hide the confidential part of drawings manually. However, almost all the confidential information of drawings is text for- mat, and in recent research we found that this information is probably well-regulated. In this paper, we propose a method of extracting the confidential part from drawings automatically by text classification. The text classification takes a simple but effective way, Linear Classification. Text classification leads to the Extraction of the confidential text, and brings on the hiding of confidentiality finally.