Compressive Embedding Method for Reversible Steganography Using XOR Approach

The need to hide confidential data and the lack of reliable techniques led to the emergence of steganography. This paper aims to bring a new and significant approach to reversible steganography of RGB images. We have devised the algorithm so that irrespective of the size of the payload and the cover images, it will be easily hidden without bringing any change and suspicion. In the proposed method it is ensured that, if the payload is larger than the cover image then it is compressed to the size of the cover image. The traditional LSB technique of steganography has been used but with a little twist to increase the complexity of the algorithm and to prevent its effortless decoding, this will also enhance the security. The MSB bits of the payload have been XORed with the LSB bits of the cover image and then the resultant bits have been stored in place of the LSB bits of the cover image. After the successful execution of the proposed method on the dataset, it has been observed that the recovered cover and payload image after steganalysis are of good quality cover as they have low noise levels which is determined by their PSNR value. Also, they are structurally quite similar to the original cover and payload image used which has been confirmed by their SSIM score. Therefore, the proposed technique gives the freedom to choose the payload and cover image and recovers them with acceptable distortion.