JPEG Copy Paste Forgery Detection Using BAG Optimized for Complex Images

被引：0

作者：

Ayalneh, Dessalegn Atnafu ^{[1
]}

Kim, Hyoung Joong ^{[1
]}

Choi, Yong Soo ^{[2
]}

机构：

[1] Korea Univ, Ctr Informat Secur Technol, Seoul, South Korea

[2] SungKyul Univ, Div Liberal Arts & Teaching, Anyang Si, Gyeonggi Do, South Korea

来源：

2014 16TH INTERNATIONAL CONFERENCE ON ADVANCED COMMUNICATION TECHNOLOGY (ICACT) | 2014年

关键词：

Copy-paste forgery; JPEG; Block Artifact Grid; Local Effect;

D O I：

暂无

中图分类号：

TN [电子技术、通信技术];

学科分类号：

0809 ;

摘要：

Image forgery detection is one of important activities of digital forensics. Forging an image has become very easy and visually confusing with the real one. Different features of an image can be used in passive forgery detection. Most of lossy compression methods demonstrate some distinct characteristics. JPEG images have a traceable zero valued DCT coefficients in the high frequency regions due to quantization. This appears as a square grid all over the image, known as Block Artifact Grid (BAG). In this paper the BAG based copy-paste forgery detection method is improved by changing the input DCT coefficients for Local Effect computation. The proposed method has shown a better performance especially for complex images.

引用

页码：181 / 185

页数：5