Abstract
Watermark detection software is obfuscated using a table to hide embedding and detection algorithms. As the table size is limited, the block size is also limited for watermarking. To address this situation, a new quasi-chirp transform is developed to improve embedding efficiency. The quasi-chirp transform is different from the conventional DCT or Fourier transform. It contains multiple frequency components in a single basis of the transform. It disperses image data rather than compressing it, as the DCT does. The dispersed data increases the range for embedding watermarks. The chirp transform is able to embed even on a flat area of an image. Using this chirp transform, embedding and detection experiments for image data with small block sizes were carried out. A high SNR and robust watermark with an evaluated obfuscation were obtained.
Original language | English |
---|---|
Title of host publication | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
Pages | 57-71 |
Number of pages | 15 |
Volume | 7128 LNCS |
DOIs | |
Publication status | Published - 2012 |
Event | 10th International Workshop on Digital-Forensics and Watermarking, IWDW 2011 - Atlantic City, NJ Duration: 2011 Oct 23 → 2011 Oct 26 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 7128 LNCS |
ISSN (Print) | 03029743 |
ISSN (Electronic) | 16113349 |
Other
Other | 10th International Workshop on Digital-Forensics and Watermarking, IWDW 2011 |
---|---|
City | Atlantic City, NJ |
Period | 11/10/23 → 11/10/26 |
Fingerprint
Keywords
- detection
- embedding
- obfuscation
- orthogonal transform
- quantization
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
A new watermarking method with obfuscated quasi-chirp transform. / Ohzeki, Kazuo; Wei, Yuanyu; Hirakawa, Yutaka; Sato, Kiyotsugu.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7128 LNCS 2012. p. 57-71 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7128 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - A new watermarking method with obfuscated quasi-chirp transform
AU - Ohzeki, Kazuo
AU - Wei, Yuanyu
AU - Hirakawa, Yutaka
AU - Sato, Kiyotsugu
PY - 2012
Y1 - 2012
N2 - Watermark detection software is obfuscated using a table to hide embedding and detection algorithms. As the table size is limited, the block size is also limited for watermarking. To address this situation, a new quasi-chirp transform is developed to improve embedding efficiency. The quasi-chirp transform is different from the conventional DCT or Fourier transform. It contains multiple frequency components in a single basis of the transform. It disperses image data rather than compressing it, as the DCT does. The dispersed data increases the range for embedding watermarks. The chirp transform is able to embed even on a flat area of an image. Using this chirp transform, embedding and detection experiments for image data with small block sizes were carried out. A high SNR and robust watermark with an evaluated obfuscation were obtained.
AB - Watermark detection software is obfuscated using a table to hide embedding and detection algorithms. As the table size is limited, the block size is also limited for watermarking. To address this situation, a new quasi-chirp transform is developed to improve embedding efficiency. The quasi-chirp transform is different from the conventional DCT or Fourier transform. It contains multiple frequency components in a single basis of the transform. It disperses image data rather than compressing it, as the DCT does. The dispersed data increases the range for embedding watermarks. The chirp transform is able to embed even on a flat area of an image. Using this chirp transform, embedding and detection experiments for image data with small block sizes were carried out. A high SNR and robust watermark with an evaluated obfuscation were obtained.
KW - detection
KW - embedding
KW - obfuscation
KW - orthogonal transform
KW - quantization
UR - http://www.scopus.com/inward/record.url?scp=84865682794&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84865682794&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-32205-1_7
DO - 10.1007/978-3-642-32205-1_7
M3 - Conference contribution
AN - SCOPUS:84865682794
SN - 9783642322044
VL - 7128 LNCS
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 57
EP - 71
BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ER -