DocumentCode
264765
Title
Fast and robust passive copy-move forgery detection using SURF and SIFT image features
Author
Pandey, Ramesh Chand ; Singh, Sanjay Kumar ; Shukla, K.K. ; Agrawal, Rishabh
Author_Institution
Dept. of Comput. Sci. &Eng., Indian Inst. of Technol., Varanasi, Varanasi, India
fYear
2014
fDate
15-17 Dec. 2014
Firstpage
1
Lastpage
6
Abstract
With the advent of sophisticated image editing softwares, it has become easy to forge digital images. Copy-move is a very common technique for image forgery; retouching tools are applied along with it which makes copy-move invisible to the naked eye. An image can be forged using copy-move technique to duplicate or hide undesirable objects. When some object is copy-moved with the help of geometrical and illumination transform, it becomes difficult to detect that object. Speed up Robust Feature (SURF) and Scale Invariant Feature Transform (SIFT) are invariant with respect to geometrical and illumination transform. In this paper we have proposed a novel method to detect copy-move forgery in an image based on passive forensic scheme. The proposed method use SURF and SIFT, which make it very fast and robust in detecting copy-moved regions. Experimental results demonstrate commendable performance in image copy-move forgery detection.
Keywords
feature extraction; image forensics; transforms; SIFT image features; SURF image features; forensic scheme; geometrical transform; illumination transform; image editing softwares; image forgery; passive copy-move forgery detection; retouching tools; scale invariant feature transform; speed up robust feature; Accuracy; Feature extraction; Forensics; Forgery; Lighting; Robustness; Transforms; SIFT; SURF; copy-move forgery detection; geometrical transform; illumination transform; passive forensic;
fLanguage
English
Publisher
ieee
Conference_Titel
Industrial and Information Systems (ICIIS), 2014 9th International Conference on
Conference_Location
Gwalior
Print_ISBN
978-1-4799-6499-4
Type
conf
DOI
10.1109/ICIINFS.2014.7036519
Filename
7036519
Link To Document