Detection of image sharpening based on histogram aberration and ringing artifacts

With the wide use of sophisticated photo manipulation, capability of digital images for recording authentic scene information has been addressed with public suspicion. So it is urgent to develop forensics techniques for verifying photographs´ originality and authenticity. In this paper, a blind forensic algorithm is proposed to detect sharpening manipulation in digital images. Gradient aberration of the gray histogram generated from unsaturated luminance regions of an image, is measured and employed to capture trails of sharpening operation. Ringing artifacts around step edges are exploited to provide another complementary clue, especially useful when the histogram-based features are not available. Tests on plenty of photo images show the effecttiveness of our proposed sharpening detection scheme.