Open AccessReversible data hiding using B‐tree triangular decomposition based prediction
Author(s) -
Uyyala Ravi,
Pal Rajarshi,
V.N.K. Prasad Munaga
Publication year - 2019
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/iet-ipr.2019.0038
Subject(s) - pixel , mathematics , tree (set theory) , algorithm , image (mathematics) , set (abstract data type) , decomposition , computer science , artificial intelligence , pattern recognition (psychology) , combinatorics , ecology , biology , programming language
A novel reversible data hiding (RDH) technique has been proposed in this study using a B‐tree triangular decomposition‐based prediction of image pixels. The superiority of a prediction error expansion‐based RDH scheme depends on a good prediction strategy for image pixels. In the proposed scheme, the B‐tree triangular decomposition is used to recursively decompose the image into a set of right‐angled triangles. The vertices of such triangles serve as the reference pixels to predict (interpolate) the intensity values of other non‐reference pixels within the triangle. Data bits are reversibly embedded into these non‐reference pixels by expanding the prediction error. Moreover, according to the proposed scheme, the number of bits being embedded in these pixels is varied (either one or two bits) based on an estimated local complexity of the triangle. The local complexity of a triangle is computed from the intensity of vertices of the triangle. The superior performance of the proposed method is verified through extensive experiments.