High-Payload Reversible Data Hiding in Encrypted Images Based on Median-Guided PVO and Enhanced Ternary Mapping

Reversible data hiding in encrypted images (RDHEI) based on pixel value ordering (PVO) and histogram modification has shown exceptional performance, yet there remains potential for improvement. This paper proposes a high-payload RDHEI method with median-guided PVO and enhanced high-frequency-emphasized ternary mapping. Firstly, an extended PVO predictor is introduced that utilizes median encrypted pixels to predict the others in a larger block, increasing efficiency from 0.50 bpp to 0.67 bpp. Secondly, an enhanced ternary mapping is designed to prioritize embedding more data into high-frequency prediction errors, thereby increasing overall payload. Experimental results show that the proposed method outperforms state-of-the-art methods, delivering a 90.96% enhancement in average payload with reversibility and confidentiality.