Adaptive smoothness evaluation and multiple asymmetric histogram modification for reversible data hiding

In this paper, an adaptive reversible data hiding (RDH) algorithm based on multiple asymmetric histograms is proposed by making full use of the image content. Different from existing multiple prediction error histogram (PEHs) modification methods that directly cluster all the pixels of a cover image into multiple categories, we firstly utilize a smoothness threshold to exclude as many pixels in complex regions as possible for reducing unnecessary pixel shifting, and then exploit fuzzy C-means with multiple deliberately-designed features to construct multiple sharply-distributed categories, which helps in increasing the subsequent embedding performance. Two asymmetric PEHs for each class are generated using a pair of asymmetric predictors, and the short part of each asymmetric PEH is modified to reduce the number of invalid modifications. The improved discrete particle swarm optimization is used to adaptively select the best bin while reducing computational complexity. The experimental results show that the proposed method outperforms several state-of-the-art RDH methods.