Reversible data hiding method based on combining IPVO with bias-added Rhombus predictor by multi-predictor mechanism

•Combining IPVO and Rhombus predictor by the multi-predictor mechanism.•Proposing an adaptive weighted local complexity metric.•Designing a bias-added Rhombus predictor to replace the conventional Rhombus predictor.•Saving the running time by a two-step optimizing strategy. The well-known pixel-value-ordering (PVO) framework based reversible data hiding (RDH) methods like improved pixel-value-ordering (IPVO) take fully advantages of sorting and achieve an outstanding performance. In this paper, we merge the advantages of sorting and averaging by combining IPVO and conventional Rhombus predictor through the multi-predictor mechanism. Furthermore, in order to better utilize the advantages of multi-predictor mechanism, a bias is added to improve the conventional Rhombus predictor. By adding adaptive biases for the pixels with different complexities, our proposed bias-added Rhombus predictor can reduce the embedding distortion significantly. Moreover, a two-step optimization is proposed to save the running time. Experimental results verify that the proposed method outperforms state-of-the-art PVO based RDH methods.