Image Steganography using Cryptographic Primitives

Abstract

Information security has received exclusive consideration of numerous researchers in the recent past. Various methods for securing sensitive information have been devised either based on steganography or cryptography. The former method is a form of covert communication in which the information is hidden within some other medium like image, audio, video, or text. Whereas in the latter method, the information is encrypted using cryptographic primitives to generate the random-looking cipher text. Both approaches have their strengths and weaknesses. We devised a smart approach by combining the strengths of both methods; randomness from cryptography and data hiding using steganography. We used the most common technique of steganography in spatial domain called Least Significant Bit (LSB) based on RGB image steganography. Generally, in this method, the secret information bits are hidden in LSB of Red, Green, and Blue (RGB) channels of an image in a sequence. This approach is simple and efficient, however, it is prone to data recovery attacks. In order to destroy the sequential hiding, we used cryptographic primitives (AES-128, RC4 and SHA-256) to produce randomness. The randomness is used to determine the location and channel to hide the data. The random sequence generated by cryptographic primitives is only known to sender and receiver, thus an attacker eavesdropping on the message cannot recover the hidden information. After embedding the secret data in an image, we performed qualitative and quantitative analysis to measure the quality of the stego-image. Quantitative analysis includes Mean Square Error (MSE), Peak signal-to-noise ratio (PSNR) and Normalized Cross-Correlation (NCC). Histogram analysis is also performed on all three channels (Red, Green, and Blue) of both, original and stego images. In the end, the proposed model is analyzed against various attacking scenarios.