Development of a hybrid system based on Cryptography and Steganography techniques

Abstract

The digital media has become more popular than the traditional analog media since it witnessed a rapid decrease in the cost of processing, storage, and bandwidth. The digital data are easy to edit, modify and exploit. Therefore, it is convenient for people to transmit digital data and information through the internet. It is implied that information can be modified easily by unauthorized people. The security and the validity in digital communication have therefore become main issues. Different technologies have been developed to protect the digital data including cryptography and steganography. Cryptography protects the data by converting them into meaningless information; however, it gives an explicit indication about the protection process. Meanwhile, the steganography schemes have been applied to avoid the hint of protecting the data by hiding them in the cover data. In this project, the cryptography and steganography schemes are combined to develop a hybrid security system that can provide a secure communication between the sender and the receiver. This system uses the digital image as a carrier medium to embed the ciphered secret image. Firstly, the secret grayscale image is encrypted by using a popular encryption algorithm called Advance Encryption Standard (AES). Then, the encrypted image is hidden in a cover image (i.e., colour or grayscale) using steganography technique based on the Patch Reference Table Pixel Value Difference (PRT-PVD), which called APP. The APP is evaluated in terms of Peak Signal to Noise Ratio (PSNR), embedding capacity, and the ability to resist the steganalysis attach such as the histogram difference. The experimental results indicated that the proposed approach APP provided stego-images with a good perceptual quality; this was indicated by the obtained high PSNR values in comparison with the PVD, MFPVD, AES-PVD, and AES –LSB techniques. Apart from that, the difference histograms of the cover images and the stego-images are very minimal. For 128×128 grayscale cover images and payload of 20000 bits, the APP achieved an average value of PSNR of 48.7230 dB, while the average PSNR values of PVD, MF-PVD, AES-PVD, AESLSB and PRT-PVD techniques are 42.3399, 43.8579, 42.3639, 47.2249, 48.3877 dB, respectively. Moreover, the APP provides a higher PSNR for all experiments of different cover image (i.e., colour or grayscale) types and sizes (225×225, 512×512) with different payloads. In addition to improvement in performance, this study suggested a method that can automatically calculate the optimal threshold value for the APP system. As a result, using the proposed approach can improve the security of data transmission through the open channels.