In this section, we evaluate the embedding performance of our method by comparing it with the conventional PEE [12], pairwise PEE [19] and a recently proposed adaptive RDH scheme of Wang et al. [23]. Eight standard 512 × 512 sized gray-scale images are utilized to conduct the comparison, including Lena, Baboon, Barbara, Airplane, Elaine, Lake, Boat, and Peppers. The performance comparison based on capacity-distortion curves is shown in Fig. 10 in which the embedding capacity varies from 10,000 bits to the maximum capacity of the proposed method with a step size of 2000 bits. According to these figures, one can see that the proposed method outperforms the three comparison methods for various capacities. More specifically, the PSNRs for the embedding capacity of 10,000 bits and 20,000 bits with different images for these four methods are presented in Tables 1 and 2.