Data embedding in random domain

doi:10.1016/j.sigpro.2014.08.037

Return

Data embedding in random domain

2015-03-01

PRE

Mustafa S. Abdul Karim *

KokSheik Wong

DOI：10.1016/j.sigpro.2014.08.037

Original

Original request for help

Save

Abstract

En 中文

A universal data embedding method based on histogram mapping called DeRand (Data embedding in Random domain) is proposed. DeRand theoretically defines redundancy in any digital signal by applying the universal parser such that high entropy random signals can certainly be utilized for data embedding. First, DeRand recursively parses a random signal into a set of tuples each of certain length until there exist some tuples of zero occurrences in the histogram. Then, tuples that occur in the histogram are associated with those of zero occurrences. Next, a tuple (of non-zero occurrence) is mapped to its corresponding associated tuple to embed 1, while the tuple is left unmodified to embed 0. DeRand is universal, reversible, applicable to any random signal and scalable in terms of embedding capacity and signal quality. Experimental results show that DeRand achieves an embedding capacity up to 4909 bits in random signal of size 256 Kbytes. In addition, the quality of the processed signal ranges from 0.0075 to 395.67 in terms of MSE. (C) 2014 Elsevier B.V. All rights reserved.

Keywords:

DeRand

Random domain

Universal data embedding

Universal parser

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.