For the conventional PEE, there are many different implementations, and they mainly differ in the prediction method. In this paper, we take the work [12] as the conventional PEE for comparison. The method [12] employed rhombus prediction and then proposed the double-layered embedding mechanism. This method performs well and can be viewed as a bench marked 1D-PEH based reversible embedding. According to Fig. 10, compared with [12], our proposed one can provide a significant improvement with a larger PSNR whatever the test image or capacity is. Specifically, according to Tables 1 and 2, our method outperforms the conventional PEE [12] with an average increase of PSNR by 2.86 dB and 2.10 dB for an embedding capacity of 10,000 bits and 20,000 bits, respectively.