International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 84 - Number 8 |
Year of Publication: 2013 |
Authors: Archana V. Mire, S. B. Dhok, N. J. Mistry, P. D. Porey |
10.5120/14597-2838 |
Archana V. Mire, S. B. Dhok, N. J. Mistry, P. D. Porey . Resampling Detection in Digital Images: A Survey. International Journal of Computer Applications. 84, 8 ( December 2013), 24-29. DOI=10.5120/14597-2838
Usually digital image forgeries are created by copy-pasting a portion of an image onto some other image. Forged area is often resized & rotated to make it proportional with respect to neighboring unforged area. This is called as resampling operation which changes certain characteristics of the pasted portion. Thus resampling is the default fingerprint present in most of the forged image and resampling detection became a standard tool in digital image forensics. Generally resampling artifacts are not visible to human eye in interpolated images but periodic correlations get introduced in image pixels because of it. These periodic interpolation artifacts present in pixel intensities or other format of data representation such as DFT, wavelet are the features which detectors look for in order to decide if an image, or a segment of image, has undergone a geometrical transformation. JPEG compression process creates its own correlation in image & may confuse resampling detectors. This paper addresses various resampling detection techniques in uncompressed image as well as re-compressed JPEG images.