Abstract: This paper introduces an improved RDWT-based image steganography with QR decomposition and double entropy system. It demonstrates image steganography method that hides grayscale secret image into grayscale cover image using RDWT, QR decomposition and entropy calculation. The proposed scheme made use of the human visual system (HVS) in the embedding process. Both cover and secret image are being segmented into non-overlapping blocks with identical block size. Then, entropy values generated from every image block will be sorted from the lowest value to the highest value. The embedding process starts by embedding the secret image block with lowest entropy value into the cover image block with lowest entropy value. The process goes on until all image blocks have been embedded. Embedding secret image into cover image according to the entropy values causes differences that HVS can less likely to detect because of the small changes on image texture. By applying the double entropy system, proposed scheme managed to achieve a higher PSNR value of 60.3773 while previous work gave a value of 55.5771. In terms of SSIM value, proposed scheme generated a value of 0.9998 comparing to previous work’s value of 0.9967. The proposed scheme eliminated the false-positive issue and required low computational time of only 0.72 seconds for embedding and 1.14 seconds for extraction process. Also, it has shown better result compared to previous work in terms of imperceptibility.

Ke-Huey Ng, Siau-Chuin Liew and Ferda Ernawan, “An Improved RDWT-based Image Steganography Scheme with QR Decomposition and Double Entropy” International Journal of Advanced Computer Science and Applications(IJACSA), 11(3), 2020. http://dx.doi.org/10.14569/IJACSA.2020.0110327

@article{Ng2020,
title = {An Improved RDWT-based Image Steganography Scheme with QR Decomposition and Double Entropy},
journal = {International Journal of Advanced Computer Science and Applications},
doi = {10.14569/IJACSA.2020.0110327},
url = {http://dx.doi.org/10.14569/IJACSA.2020.0110327},
year = {2020},
publisher = {The Science and Information Organization},
volume = {11},
number = {3},
author = {Ke-Huey Ng and Siau-Chuin Liew and Ferda Ernawan}
}