Reversible Data Hiding of Grayscale Images Using General Distortion Metric

In reversible data hiding (RDH), the use of general distortion metric for improvement of image visual quality is a new issue. In this letter, a new RDH method of grayscale images using general distortion metric is proposed. The proposed method can adaptively compute different distortions for different pixels by considering the influence of surrounding pixels. Due to the fact that the pixel distortion in smooth areas is more obvious than that in texture areas under the same modification, the modifications on pixels are adaptively adjusted by combining a minimum distortion coding to achieve better visual quality. Experimental results show that the proposed method can reduce modifications on smooth areas and achieve better embedding performance for grayscale images in comparison with several classical and advanced RDH methods.