Abstract
In this study, we present the idea of using secret bits to repair SMVQ-compressed images. The idea can be viewed as the optimization problem of Using Secret Bits for Image-block Repairing based on Differences under SMVQ (USBIRDS). We propose a novel algorithm named Greedy-USBIRDS to find a near optimal solution for the USBIRDS problem. The experimental results show that, the proposed method provides excellent stego-image quality and large embedding capacity, in particular for complex cover images. While compared with Chen and Lin’s steganographic method, which is the known best method based on VQ/SMVQ, the proposed method achieves about 53% more dB of PSNR of the stego-image quality and about 4.2% more bits of the embedding capacity on average.
Original language | English |
---|---|
Pages (from-to) | 599-612 |
Number of pages | 14 |
Journal | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Volume | 9023 |
DOIs | |
State | Published - 2015 |
Event | 13th International Workshop on Digital-Forensics and Watermarking , IWDW 2014 - Taipei, Taiwan Duration: 1 Oct 2014 → 4 Oct 2014 |
Keywords
- Information hiding
- Side-match vector quantization (SMVQ)
- Steganography