Data hiding in AMBTC images using quantization level modification and perturbation technique

A novel data hiding method for Absolute Moment Block Truncation Coding (AMBTC) compressed image based on quantization level modification is proposed. Blocks of AMBTC-compressed image are classified into two categories, namely smooth and complex, according to a predefined threshold. For smooth blocks, the bitmap is replaced by secret data for data embedment. Meanwhile, the corresponding quantization levels are modified to achieve a minimum distortion. If a larger payload is required, the modified quantization levels can be further perturbed for carrying two additional bits. If the blocks are complex, one data bit can be embedded with no distortion by swapping the values of the two quantization levels together with bitmap flipping. In addition, a suppress threshold mechanism is used to prevent from the application of the perturbation technique at low payload to maintain the image quality. The proposed method minimizes the distortion of each stego block while ensuring high payload, thus the embedding efficiency can be enhanced. Experimental results demonstrate the improvement of the proposed method compared with other related state-of-art works.