Computational intelligence: application in digital watermarking
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)...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2012
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/24627/ http://irep.iium.edu.my/24627/1/2076C.pdf |
| Summary: | 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. |
|---|