An efficient reversible data hiding method for AMBTC compressed images

Analyzing multimedia data in mobile devices is often constrained by limited computing capacity and power storage. Therefore, more and more studies are trying to investigate methods with algorithm efficiency. Sun et al. proposed a low computing cost reversible data hiding method for absolute moment block truncation coding (AMBTC) images with excellent embedding performance. Their method predicts quantization values and uses encrypted data bits, division information, and prediction errors to construct the stego codes. This method successfully embeds data while providing a comparable bit-rate; however, it does not fully exploit the correlation of neighboring pixels and division of prediction error for better embedment. Therefore, the payload and bit-rate are penalized because the embedding performance directly depends on the prediction accuracy and division efficiency. In this paper, we use median edge detection predictor to better predict the quantization values. We also employ an alternative prediction technique to increase the prediction accuracy by narrowing the range of prediction values. Besides, an efficient centralized error diversion technique is proposed to further decrease the bit-rate. The experimental results show that the proposed method offers 8 % higher payload with 5 % lower bit-rate on average if compared to Sun et al.'s method and has better embedding performance than prior related works.