For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Intelligent Extraction of a Digital Watermark from a Distorted Image
Asifullah KHAN Syed Fahad TAHIR Tae-Sun CHOI
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2008/07/01
Online ISSN: 1745-1361
Print ISSN: 0916-8532
Type of Manuscript: LETTER
Category: Application Information Security
watermark, decoding, transform, copyright protection, robust, blind, Machine Learning,
Full Text: PDF>>
We present a novel approach to developing Machine Learning (ML) based decoding models for extracting a watermark in the presence of attacks. Statistical characterization of the components of various frequency bands is exploited to allow blind extraction of the watermark. Experimental results show that the proposed ML based decoding scheme can adapt to suit the watermark application by learning the alterations in the feature space incurred by the attack employed.