Fragile high capacity data hiding in digital images using integer-to-integer DWT and lattice vector quantization

• Published in: Multimedia tools and applications 2020-05, Vol.79 (19-20), p.13427-13447
• Main Authors: Akhtarkavan, Ehsan, Majidi, Babak, Salleh, M. F. M., Patra, Jagdish C.
• Format: Article
• Language: English
• Subjects: Algorithms; Computer Communication Networks; Computer Science; Data Structures and Information Theory; Data transmission; Digital imaging; Discrete Wavelet Transform; Embedding; Fragility; Image quality; Integers; Lattices; Layouts; Multimedia; Multimedia Information Systems; Special Purpose and Application-Based Systems; Vector quantization; Watermarking; Wavelet transforms
• ISSN: 1380-7501; 1573-7721
• DOI: 10.1007/s11042-020-08662-7

Data hiding in digital multimedia has been extensively used for sensitive data transmission and data authentication. An important property of data hiding which makes this method applicable to such applications is its fragility. The fragility is the loss of the embedded authentication credential resulting from any tampering attempt. Another important issue in data hiding and watermarking in digital images is increasing the embedding capacity while keeping the quality of the cover image high enough to avoid any perceptual degradation. In this paper a novel fragile digital image data hiding algorithm based on Lattice Vector Quantization (LVQ) is proposed to solve the above mentioned shortcomings. In the proposed data hiding algorithm after an initial pre-processing stage, the image is transformed into frequency domain using Integer-to-Integer Discrete Wavelet Transform (IIDWT). Then lattice vector quantization of A 4 and Z 4 lattices are used for embedding data into the cover image. The proposed embedding algorithm has the ability to hide the data inside the entire cover image. The experimental results show that the proposed data hiding algorithm performs significantly better than recently proposed algorithms when the embedding capacity is increased. At the high capacity regimes, the proposed algorithm can embed significantly more sensitive data in the cover image while keeping the perceptual quality of the recovered image high.