STUDY AND DEVELOPMENT OF AMBTC BASED DATA HIDING METHODS

Abstract

Recently, Absolute Moment Block Truncation Coding (AMBTC) based data hiding methods have got popularity due to their real-time compression capability while en abling the covert communication. This capability prompt the researchers which in turn resulted in the development of both irreversible and reversible data hiding methods. In this research work, firstly a bibliometric analysis is performed over previous works, and then afterward a new adaptive AMBTC-based RDHEI technique is proposed. For bibliometric analysis over AMBTC-based data hiding methods, documents published between 2005 to 2023 in Web of Science have been considered for evaluation. The analysis is mainly done with the help of VoSViewer software. VoSViewer allows anal ysis based on several indices like co-authorship analysis, keyword occurrence analysis, citation analysis, co-citation analysis, etc. After conducting a bibliometric analysis over previous research works on AMBTC based data hiding methods, it was clear that there was some imbalance between em bedding capacity and image reconstruction. So, this research work proposed a novel reversible data hiding method that utilizes adaptive AMBTC for large payload em bedding. The approach begins by segmenting the image into blocks and classifying these blocks into saved and hiding blocks based on their complexity. The hiding blocks are then encoded using the AMBTC technique and further divided into mul tiple sub-categories based on the resulting quantization levels. The encoding of each hiding block is subsequently refined to minimize the discrepancy between the origi nal and reconstructed images while preserving the compression ratio. Although this refined encoding produces bit-planes and quantization levels of varying sizes, it sig nificantly expands the available space within the cover image, allowing for substan tial payload embedding while ensuring complete recoverability of the original image. Experimental results demonstrate that the proposed method surpasses current state-of the-art AMBTC-based data hiding techniques in both embedding capacity and image quality, while ensuring the original image contents’ privacy.