Improved Fluctuation Derived Block Selection Strategy in Pixel Value Ordering Based Reversible Data Hiding

Xin Tang, Linna Zhou, Guang Tang, Yuchen Wen, Yixuan Cheng

Abstract: Block selection strategy is an effective technique to improve the imperceptibility in pixel value ordering (PVO) based reversible data hiding by selecting blocks with higher smoothness to embed data at first, which largely reduces invalid shifting of pixels before the maximum capacity is reached. As a means to evaluate the smoothness of a certain block selected, fluctuation value is defined accordingly. However, in current solutions, the definition of fluctuation value is not always accurate enough to reflect the true smoothness of blocks, which is usually restrained by the number of pixels participating in calculation as well as those invalid ones introduced. In this paper, we improved the definition of fluctuation in order to achieve a better performance. Take the maximum side for example, context pixels participating in fluctuation calculation are determined at first, considering the second largest pixel as well as neighborhood of the largest pixel simultaneously. Then those invalid ones are eliminated to reduce unnecessary uncertainty introduced. Experimental results show that our proposed scheme outperforms the state-of-the art in accuracy of smoothness evaluation, and achieves a better imperceptibility as a result.