Abstract | ||
---|---|---|
Combined average encoding with vector quantization (VQ) encoding, a new self-embedding fragile watermarking scheme is proposed. To take into account watermark payload, localization accuracy and recovery quality, the 6-bit average-watermark of a 2×2 original block and the 8-bit VQ-watermark of a 4×4 block of image high-frequency component are generated and hidden in the corresponding mapping blocks of them based on secret key, respectively. To improve the tamper detection performance, the validity of a 2×2 block is determined by combining the average-watermark with the VQ-watermark. The average, VQ and inpainting recovery operations are executed in sequence to improve the recovery quality especially for a larger tampering ratio. Simulation results demonstrate that the proposed scheme not only provides a better invisibility and security against the known counterfeiting attacks, but also allows image recovery with an acceptable visual quality up to 70% tampering ratios. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1007/978-3-642-40099-5_11 | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Keywords | Field | DocType |
combined average encoding,6-bit average-watermark,inpainting recovery operation,fragile watermarking scheme,vq encoding,original block,corresponding mapping block,acceptable visual quality,8-bit vq-watermark,image recovery,recovery quality,image high-frequency component | Computer vision,Digital watermarking,Computer science,Watermark,Inpainting,Vector quantization,Artificial intelligence,Self-embedding,Invisibility,Encoding (memory),Payload | Conference |
Volume | Issue | ISSN |
7809 LNCS | null | 16113349 |
Citations | PageRank | References |
2 | 0.40 | 14 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hongjie He | 1 | 238 | 20.34 |
Fan Chen | 2 | 137 | 11.05 |
Yaoran Huo | 3 | 20 | 2.80 |