A Content-Based Image Authentication SystemWith Lossless Data Hiding.ppt

A Content-Based Image Authentication SystemWith Lossless Data Hiding Authors: Dekun Zou, Chai Wah Wu, Guorong Xuan, Yun Q. ShiSource : Multimedia and Expo, 2003. ICME 03. Proceedings.  2003 International Conference on ,? Volume: 2 ,?6-9 July 2003 Pages:213 – 216 Speaker: Jui-Yi ChangAdvisor: Chin-Chen ChangDate: 2004/4/28 outline Introduction Generation of information for authentication Tag embedding Verification procedure Experiment results Conclusions Introduction Authentication of multimedia data provides an answer to protect the data integrity of it. Two problem:a. How tolerant is it to minor modifications?b. how hard is it to forge an authentic image? Extend the approach in the other papers to watermarking or data hiding solutions by combining it with a lossless data hiding technique and demonstrate its robustness and effectiveness. Generation of information for authentication Tag Embedding (Cont.) Using circular histogram algorithm. Divided into 8 by 8 blocks. The 64 pixels in each block are pseudo-randomly equally divided into group A and B. For each group, the histogram of the gray scale value is mapped onto circle. The gray level of pixels is quantized into Q ranges. The size of a range is P=256/Q. Tag Embedding (Cont.) Tag Embedding (Cont.) Tag Embedding (Cont.) To embed “1”C’=C+P for pixels in group AC’=C-P for pixels in group B To embed “0”C’=C-P for pixels in group AC’=C+P for pixels in group B Tag Embedding (Cont.) The centers of mass of group A and group B may be so far away that error will occur when extracting. To avoid this problem, such blocks are skipped when embedding information. Verification procedure Verification procedure (Cont.) Result in failure of the verification:(a) if the image is degraded too much so that the embedded tag could not be extracted correctly.(b) When the tag extraction is correct, but the image is corrupted such that the quantized features do not verify with the extracted signature S. Experiment