华南理工大学学报(自然科学版) ›› 2015, Vol. 43 ›› Issue (3): 66-71.doi: 10.3969/j.issn.1000-565X.2015.03.010

• 电子、通信与自动控制 • 上一篇    下一篇

基于 ASIFT 和圆投影特征的图像被动取证

赵俊红 哀微 朱学峰   

  1. 华南理工大学 自主系统与网络控制教育部重点实验室,广东 广州 510640
  • 收稿日期:2014-06-06 修回日期:2014-10-13 出版日期:2015-03-25 发布日期:2015-02-10
  • 通信作者: 赵俊红(1976-),女,讲师,博士,主要从事图像取证、图像分析与理解研究. E-mail:jhzhao@scut.edu.cn
  • 作者简介:赵俊红(1976-),女,讲师,博士,主要从事图像取证、图像分析与理解研究.
  • 基金资助:

    工业控制技术国家重点实验室开放课题(ICT1434);广东省自然科学基金博士启动项目(S2013040014995);华南理工大学中央高校基本科研业务费资助项目(2014ZM0044)

Passive Forensics of Images on the Basis of ASIFT and Circle Projection Features

Zhao Jun-hong Ai Wei Zhu Xue-feng   

  1. Key Laboratory of Autonomous Systems and Network Control of the Ministry of Education,South China University of Technology,Guangzhou 510640,Guangdong,China
  • Received:2014-06-06 Revised:2014-10-13 Online:2015-03-25 Published:2015-02-10
  • Contact: 赵俊红(1976-),女,讲师,博士,主要从事图像取证、图像分析与理解研究. E-mail:jhzhao@scut.edu.cn
  • About author:赵俊红(1976-),女,讲师,博士,主要从事图像取证、图像分析与理解研究.
  • Supported by:
    Supported by the Open Research Project of the State Key Laboratory of Industrial Control Technology(ICT1434) and the Natural Science Foundation of Guangdong Province,China(S2013040014995)

摘要: 文中针对图像篡改中最常见的复制粘贴篡改被动取证问题展开图像分析研究,提出了一种基于仿射尺度不变特征变换(ASIFT)和圆投影特征的算法,用于检测图像中是否存在复制粘贴篡改. 实验结果表明:基于仿射尺度不变特征的算法对具有明显角点特征的篡改图像的取证效果较好,而基于改进圆投影特征的算法对不具备明显角点特征的篡改图像(如随机性纹理、平坦区域)的取证效果较好,因此联合使用基于这两种特征的算法对常见的复制粘贴型图像篡改进行被动取证能达到较好效果.

关键词: 复制粘贴篡改, 图像分析, 仿射尺度不变特征变换, 圆投影, 被动取证

Abstract: This paper deals with an image analysis method,namely passive forensics,to detect copy-move forgery,the most popular manipulation in digital tampered images. On the basis of ASIFT (Affine Scale-Invariant Feature Transform) and circle projection features,a new algorithm is proposed to detect copy-move forgery in images. The results show that the adoption of ASIFT is effective only in finding the forgery area in a tampered image with obvious corner information,while the adoption of improved circle projection features is effective only in finding the forgery area in a tampered image without such obvious corner information as random textures and flat areas. Thus,it comes to the conclusion that the combination of ASIFT and improved circle projection features is more effective for the pas-sive forensics of copy-move forgery.

Key words: copy-move forgery, image analysis, affine scale-invariant feature transform, circle projection, passive forensics