信息安全研究

• 物理空间信息安全专题 • 上一篇    下一篇

信息设备电磁泄漏还原图像的文本识别研究

  

  1. 1(中国科学院信息工程研究所北京100093) 2(国家保密科技测评中心北京100044) 3(中国科学院大学北京100049)
  • 收稿日期:2016-01-08 出版日期:2016-02-05 发布日期:2016-02-06
  • 基金资助:
    中国科学院信息工程研究所“优秀青年人才”专项基金项目(1104008004)

Research on Character Recognition of Reconstructed Image from Electromagnetic Emanation of Information Equipment

  • Received:2016-01-08 Online:2016-02-05 Published:2016-02-06

摘要: 信息设备电磁辐射产生的信息泄漏是信息安全中的重要隐患,随着信息设备广泛使用,电磁泄漏信息重建技术也引起了国内外学术界的广泛关注,然而重建信息中文本的自动识别技术并未引起人们的注意.人工识别重建文本的方法耗时耗力,而且图片信息不便于保存及实现检索自动化,因此不能快速及时发现有用信息.基于模板匹配方法,提出一种针对计算机视频电磁泄漏文本“横”缺失特点的自动识别方法,实验证明该方法适用于电磁泄漏文本信息特点,可自动快速地识别电磁泄漏,还原图像的文本信息.

关键词: 电磁泄漏, 重建图像, 字符识别, 阈值投影法, 模板匹配

Abstract: The information leakage caused by electromagnetic emanation of information equipment is a great hidden danger in information security. With information equipment widely used, the technology of information reconstruction from electromagnetic emanation of information equipment has been paid widespread attention. However, the Automatic recognition technology of reconstructing text in the information equipment has not aroused peoples attention. Artificial recognition method of reconstructing the text is timeconsuming, and images are not facilitate to save and realize automatic retrieval. So it is unable to find useful information quickly. This paper proposes an automatic character recognition method based on template matching to fit the "horizontal" lack property of reconstructed image from electromagnetic emanation of computer video. Experiments show that this method is suitable for the characteristics of the electromagnetic leakage text information, which can automatically and quickly identify the text information of the electromagnetic leakage reduction image.

Key words: electromagnetic emanation, reconstructed image, character recognition, threshold projection, template matching