Local feature-based mutual complexity for pixel-value-ordering reversible data hiding

Xinyi Gao, Zhibin Pan, Guojun Fan, Xiaoran Zhang, Hongzhi Yin

Abstract: Highlights•This is the first work that regards the RDH complexity problem as a binary classification problem and comprehensively evaluates the performance.•Neighbor Complexity is designed to measure the distribution of pixel neighbors and can solve the block-dependent embedding problem.•Combining different complexities, Mutual Complexity is proposed to comprehensively describe the local features.•Mutual Complexity adaptively excludes worse predicted pixels in high-dimensional space, contributing to the highest precision score.