A Contourlet-Domain Image Signature for Content Authentication
Xiaofeng Wang, Lixin Zhang, Kemu Pang · 2013
An image signature approach for content authentication is proposed. In the proposal, Contour let transformation is used to extract image feature, and the statistics of Contour let transformation coefficients are used to generate image signature. Then a tampering detection and tampering localization algorithm are developed by using generated image signature. Proposed method is robust for perceptually insignificant distortions introduced by content-preserving manipulations such as JPEG compression, adding noise, filtering, etc. It is sensitive to the changes caused by malicious attacks, and it possesses the functionality to detect the changed image regions.