摘要
网络高度发达的信息时代,防止涉密信息被泄露是一件非常重要的任务,尤其是对于政府、军队、公安等重点单位。传统的涉密信息监测系统往往是安装在主机等终端中,无法对于通过手机等智能移动终端偷拍涉密图片或者通过聊天软件上传涉密图片的行为无法进行有效的制止。针对这个问题,设计了一种将CTPN文本检测算法、光学字符识别技术(optical character recognition,OCR)与场景识别、图片传输监控相结合的智能移动终端涉密信息监测系统,可广泛应用于Android移动平台中。该系统通过全局扫描,实时相机监察,社交管控三防一体对失泄密行为进行监控监察,有效防止失泄密事故案件的发生。测试结果显示,该系统不仅可以准确识别涉密图片、监测涉密行为并且处理速度快、占用内存空间小,可以满足涉密单位的基本需求。
Preventing classified information from being leaked is a very important task in the highly developed information age,especially for key units such as government,military,and public security.Traditional monitoring systems for classified information installed in terminals such as mainframes cannot effectively stop the act of secretly taking classified pictures through smart mobile terminals such as mobile phones or uploading classified pictures through chatting software.A system that combines the CTPN text detection algorithm was proposed to address this problem,optical character recognition(OCR)technology with scene recognition and image transmission monitoring.The system can be widely used in the Android mobile platform.The system can monitor and supervise the breach of confidentiality through global scanning,real-time camera monitoring,and social control,effectively preventing the occurrence of a breach of confidentiality cases.The test results show that the system can not only accurately identify classified images and monitor classified acts,but also process them quickly and take up little memory space,which can meet the basic needs of classified units.
作者
王本钰
顾益军
彭舒凡
WANG Ben-yu;GU Yi-jun;PENG Shu-fan(College of information network security, People's Public Security University of China, Beijing 102600, China;Key Laboratory of Security & Risk Assessment, Ministry of Public Security, Beijing 102600, China)
出处
《科学技术与工程》
北大核心
2022年第6期2317-2325,共9页
Science Technology and Engineering
基金
中国人民公安大学基本科研业务费(2021JKF420)
公安部科技强警基础工作专项(2020GABJC02)。