Reversible watermarking by prediction-error expansion

Abstract

We pnipme a new reiwrsible (lossless) nutemlarking dgorithnifor digital Oitages. Being reversible, tlle algorirlm enobles (lie recover) qfilie original liosl ii@nlatian tipon the extraction cf the embedded infiinilaiiun.The /imposed techriiqne e.rp1oit.s the inltrrent correlarion on~orig the adjacent pi.els in ai1 image region using a predict<>,: Tlie infiimuirion hits ore ernbedded info the prediction errors, wlticlr enables 11s to einhed a large payload wltile keeping the di.stonion Ion: A Irishigram sliifr ai the encuder enables the decoder io identi& the eriiberlded location.

Optimal Gaussian Weight Predictor and Sorting Using Genetic Algorithm for Reversible Watermarking Based on PEE and HS

“…Place the header's LSB back into the first h pixels of the set. All modified pixels in the sorted sequence can be recovered by using (6). After recovering the dot set, repeat all steps with the cross set.…”

“…Thodi and Rodriquez [6] and [7] developed the histogram shifting method to decrease image distortion by using a threshold value associated with PE. Another advantage of this method is that it allows us to avoid overlapping problem caused by PE expansion.…”

“…Their idea utilized efficient correlation between neighboring pixels to sort data in the image. In 2007, Thodi and Rodriguez [6] introduced another technique called histogram shifting (HS). It was designed to decrease image distortion by shifting the unused pixels instead of expanding them.…”

A Survey on Reversible Watermarking Schemes for Relational Data

“…A novel blind watermarking approach that ensures security of ownership in relational databases is studied. Here, an embedding technique which is reversible called Prediction-error Expansion(PEE) on integers is used [12], [13], [14]. The concept of PEE is modified and more efficient watermarking concept is introduced.…”

Improved Lossless Embedding and Extraction - A Data Hiding Mechanism

“…Reversible data hiding has acknowledged more attention in recent years, where the invisible data (called the payload) is embedded into a host image to form a stego -image in a reversible fashion. In previous work [5], lossless compression techniques are applied to original image in order to create ability for embedding the payload. The payload (compressed data and secret data) is then embedded to form a stego-image.…”