Computational intelligence: application in digital watermarking

Olanweraju, Rashidah Funke and Khalifa, Othman Omran (2012) Computational intelligence: application in digital watermarking. In: 2nd International Conference on Mathematical Applications in Engineering (ICMAE2012), 3 - 5 July 2012, Seri Pacific Hotel Kuala Lumpur, Malaysia.

PDF (Computational intelligence: application in digital watermarking) - Published Version Restricted to Registered users only Download (489kB) | Request a copy

Abstract

Impersonation and piracy of intellectual properties remain an indispensable problem across the globe. Currently over millions of digital songs, images and videos are copied illegally during file-sharing over the networks, costing loss of revenue to multimedia industries. Digital Watermarking (DW) is an efficient method that embeds an imperceptible message in digital object to protect copyright and authenticate digital media in order to prevent forgery and impersonation. Computational Intelligence (CI); a well-established paradigm is currently gaining attraction in the field of information hiding due to its ability to solve and improve complex problems encountered. CI has been applied in digital watermarking to improve performance in order to curb piracy. This paper brings to view computational intelligence techniques specifically the Artificial Neural Network (ANN); generalizations of mathematical models based on biological nervous systems and its applications in digital watermarking. Mathematical model of ANN in relation to digital watermarking is indentified, elucidated and simplified. Results of image watermarking using both ANN and no ANN were compared. Embedding using NN showed the optimum imperceptibility result as well as enhancement in robustness of the system before and after attacks; IFM equal to 0.9588 of 1.0000 was obtained using NN to embed, meaning that there was almost no loss of fidelity in the watermarked image. It is anticipated that this paper will serve as a launch pad for researchers interested in intelligent watermarking.

Item Type:	Conference or Workshop Item (Full Paper)
Additional Information:	6796/24627
Uncontrolled Keywords:	Artificial Neural Networks, Backpropagation algorithm, Computational intelligence, Digital watermarking, multilayer perceptron, transfer function
Subjects:	Q Science > QA Mathematics
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Engineering > Department of Electrical and Computer Engineering
Depositing User:	Dr. Rashidah Funke Olanrewaju
Date Deposited:	24 Jul 2012 11:55
Last Modified:	18 Dec 2012 10:33
URI:	http://irep.iium.edu.my/id/eprint/24627

Actions (login required)

View Item