信息网络安全 ›› 2015, Vol. 15 ›› Issue (9): 154-157.doi: 10.3969/j.issn.1671-1122.2015.09.035

• 入选论文 • 上一篇    下一篇

面向网络内容安全的图像识别技术研究

崔鹏飞1(), 裘玥2, 孙瑞2   

  1. 1.郑州金惠计算机系统工程有限公司,河南郑州450001
    2.北京市公安局网络安全保卫总队,北京 100740
  • 收稿日期:2015-07-15 出版日期:2015-09-30 发布日期:2015-11-13
  • 作者简介:

    作者简介: 崔鹏飞(1982-),男,河南,硕士,主要研究方向:图像处理与识别;裘玥(1974-),女,北京,高级工程师,硕士,主要研究方向:信息安全;孙瑞(1984-),男,北京,助理工程师,本科,主要研究方向:软件工程。

Research on Image Recognition Technology for the Network Content Security

Peng-fei CUI1(), Yue QIU2, Rui SUN2   

  1. 1. ZhengZhou Jinhui Computer System Engineering Co., Ltd, Zhengzhou Henan 450001, China
    2.Network Security Defense Corps of Beijing Public Security Bureau, Beijing 100740, China
  • Received:2015-07-15 Online:2015-09-30 Published:2015-11-13

摘要:

随着网络技术的快速发展,大量的违法不良图像在网络中大肆传播,严重危害了网络内容安全。文章从网络安全工作的实际出发,针对传统文本过滤技术的缺点,指出了图像识别技术对于过滤违法不良信息的重要性。文章分析总结了几种违法不良图像识别技术,并探讨了每种技术的适用场景。最后,结合图像识别的本质问题,指出深度学习是未来图像识别技术的发展趋势,讨论了利用卷积神经网络代替传统的方法对违法不良图像进行识别的方法。

关键词: 图像识别, 网络内容安全, 深度学习

Abstract:

With the rapid development of network technology, a variety of illegal images spread suddenly in the network, which endangers the network content security seriously. From the reality of network security in this paper, aiming at the shortcomings of the traditional text filtering technology, this paper point out the importance of the image recognition technology for filtering the illegal and bad information. The paper analyzes and summarizes some kinds of recognition technologies of the illegal images, and discusses the application scene of each technology. Finally, combining with the essence of image recognition problems, the paper indicate that deep learning is the development trend of the image recognition technology in the future, and discuss how to use convolution neural network instead of the traditional method to recognize the illegal images.

Key words: image recognition, network content security, deep learning

中图分类号: