Title :
Robust watermark detection based on a new spatially weighted total variation based image prior
Author :
Mairgiotis, Antonis ; Galatsanos, Nikolaos ; Yang, Yongyi
Author_Institution :
Dept. of Comput. Sci., Univ. of Ioannina, Ioannina, Greece
Abstract :
In this paper we propose a class of watermark detectors based on Total Variation (TV). A new image prior that uses a spatially weighted version of TV that captures local image properties is introduced for this problem. The resulting detector is compared to previously reported similar detectors and other state-of-the-art detectors for this problem. The proposed detector compares favorably with them both in terms of sensitivity when the watermark is very weak and in terms of robustness against common watermark attacks.
Keywords :
image coding; image recognition; watermarking; image prior; local image properties; robust watermark detection; spatially weighted total variation; state-of-the-art detectors; Computer science; Detectors; Distortion; Noise robustness; Pixel; Random variables; Statistical analysis; TV; Testing; Watermarking; image prior; robustness; spatially weighted; statistical methods; total variation; watermark detection;
Conference_Titel :
ELMAR, 2009. ELMAR '09. International Symposium
Conference_Location :
Zadar
Print_ISBN :
978-953-7044-10-7