Digital Watermarking Using Syndrome Codes
Abstract
The Grantee will conduct research on informed watermarking. Recent analyses proved that the theoretical limit to digital watermarking is much higher than previously believed, however no practical means for accomplishing the theoretical rates are available. As detailed in the technical proposal, the PI will investigate means to achieve increased embedding rates. Specifically, he will: 1. refine the trellis-based coding techniques including (1) studying how variations in the trellis architecture (i.e. number or states and arcs) affect performance and (2) investigating the performance of syndrome coding as a function of the underlying error correction code. 2. investigate approximations to the informed embedding optimizations, including (1) studying how such optimizations can be efficiently computed and (2) examining the choice of symbol alphabet (binary versus n-symbol alphabets) and their corresponding modulation codes. Pending satisfactory results of the initial work and availability of funding, additional research would be performed to: 3. Improve the perceptual modeling required to hide a high data rate embedded signal in a video 4. conduct tests to measure and subsequently model the distortions that occur during the broadcasting of television 5. conduct tests to determine the robustness of our watermarking algorithms to commercial broadcasting.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 04, 2005
- Accession Number
- ADA468319
Entities
People
- Ingemar J. Cox
Organizations
- University of London