TY - GEN
T1 - An automatic visible watermark detection method using total variation
AU - Santoyo-Garcia, Hector
AU - Fragoso-Navarro, Eduardo
AU - Reyes-Reyes, Rogelio
AU - Sanchez-Perez, Gabriel
AU - Nakano-Miyatake, Mariko
AU - Perez-Meana, Hector
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/5/26
Y1 - 2017/5/26
N2 - Visible watermarking has been an important research area, widely studied and used for copyright protection over the years. However, new visible watermark removal-oriented attacks and techniques have emerged. Although these methods provide high quality resultant images, the main part of the attack consists of manual detection of the watermark pattern. Considering that visible watermark pattern can be different in each watermarked image, time-consuming human intervention is required, making these attacks inefficient in practice. In this paper, we propose an automatic detection method of visible watermark pattern, which can combine with conventional watermark removal attacks to make them more efficient. To accomplish it, the visible watermarked image is decomposed into structure image and texture image using Total Variation L1 (TV-L1) method. In the structure image, the watermark edge is discriminated from the edges present in the host image to automatically detect the visible watermark pattern.
AB - Visible watermarking has been an important research area, widely studied and used for copyright protection over the years. However, new visible watermark removal-oriented attacks and techniques have emerged. Although these methods provide high quality resultant images, the main part of the attack consists of manual detection of the watermark pattern. Considering that visible watermark pattern can be different in each watermarked image, time-consuming human intervention is required, making these attacks inefficient in practice. In this paper, we propose an automatic detection method of visible watermark pattern, which can combine with conventional watermark removal attacks to make them more efficient. To accomplish it, the visible watermarked image is decomposed into structure image and texture image using Total Variation L1 (TV-L1) method. In the structure image, the watermark edge is discriminated from the edges present in the host image to automatically detect the visible watermark pattern.
KW - Automatic detection
KW - Image decomposition
KW - Image recovery
KW - Total variation
KW - Visible watermarking
KW - Watermark attacks
UR - http://www.scopus.com/inward/record.url?scp=85021822798&partnerID=8YFLogxK
U2 - 10.1109/IWBF.2017.7935109
DO - 10.1109/IWBF.2017.7935109
M3 - Contribución a la conferencia
AN - SCOPUS:85021822798
T3 - Proceedings - 2017 5th International Workshop on Biometrics and Forensics, IWBF 2017
BT - Proceedings - 2017 5th International Workshop on Biometrics and Forensics, IWBF 2017
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 5th International Workshop on Biometrics and Forensics, IWBF 2017
Y2 - 4 April 2017 through 5 April 2017
ER -