Robust Mesh-based Hashing for Copy Detection and Tracing of Images Chun-Shien Lu, Chao-Yong Hsu, Shih-Wei S un, and Pao-Chi Chang Proc. IEEE Int. Conf. on Multimedia and Expo: special s ession on Media Identification, Taipei, Taiwan, 2 004 Reporter: Jen-Bang Feng
15
Embed
Robust Mesh-based Hashing for Copy Detection and Tracing of Images Chun-Shien Lu, Chao-Yong Hsu, Shih-Wei Sun, and Pao-Chi Chang Proc. IEEE Int. Conf.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Robust Mesh-based Hashing for Copy Detection and Tracing of Images
Chun-Shien Lu, Chao-Yong Hsu, Shih-Wei Sun, and Pao-Chi Chang
Proc. IEEE Int. Conf. on Multimedia and Expo: special session on Media Identification, Taipei, Taiwan, 2004
Reporter: Jen-Bang Feng
2
Outline
Watermarking and Hashing The Proposed Method Conclusions
3
Watermarking and Hashing Digital Watermarking (Data Hiding)
Content has to be modified (a data hiding technique) Contents to be protected must be watermarked Measures “originality” Stand-along
Media Hashing (Fingerprinting) Content is not modified (a non-hiding technique) Can track the usage of contents already available in
the public domain Measure “similarity” Connection to database required
4
Robust Signal Hashing Problem
Hash(Baboon)= XXX…
Hash(Lena)= YYY…
Hash(Lena 2)= ZZZ…
Should be very different
Should be sufficiently similar
5
Perceptual Hashing The fragility of cryptography
hashing is too restricted Media data permits acceptable
distortions Media hashing needs
Robustness (error-resilience) Collision-free Fast searching (complexity) Scalability
6
Track 1Track 1
Architecture for Robust Identification of Media Content
Global, local warpingGlobal, local transformsJittering
Brute force key searchOracle
Watermark inversionCopy attack
Voloshynovskiy et al. “attacks modeling: towards a second generation watermarking benchmark,” Signal Processing, 2001Kutter and Petitcolas, “A fair benchmark for image watermarking systems,” Proc. SPIE99
13
Harris Detector
where I(x, y) is the grey level intensity and where A represents the integration of A on a given neighborhood. If at a certain point the two eigenvalues of the matrix are large, then a small motion in any direction will cause an important change of grey level. This indicates that the point is a corner.
2
2
,
yI
yI
xI
yI
xI
xI
M
14
Harris Detector The corner response function is given
by:
where k is a parameter set to 0.04 (a suggestion of Harris). Corners are defined as local maxima of the cornerness function. Sub-pixel precision is achieved through a quadratic approximation of the neighborhood of the local maxima.