Multi-Stage Protection Using Pixel Selection Technique for Enhancing Steganography

Abstract

Steganography and data security are extremely important for all organizations. This research introduces a novel stenographic method called multi-stage protection using the pixel selection technique (MPPST). MPPST is developed based on the features of the pixel and analysis technique to extract the pixel's characteristics and distribution of cover-image. A pixel selection technique is proposed for hiding secret messages using the feature selection method. The secret file is distributed and embedded randomly into the stego-image to make the process of the steganalysis complicated.  The attackers not only need to deter which pixel values have been selected to carry the secret file, they also must rearrange the correct sequence of pixels. MPPST generates a complex key that indicates where the encrypted elements of the binary sequence of a secret file are. The analysis stage undergoes four stages, which are the calculation of the peak signal-to-noise ratio, mean squared error, histogram analysis, and relative entropy. These four stages are used to demonstrate the characteristics of the cover image. To evaluate the proposed method, MPPST is compared to the standard technique of Least Significant Bit (LSB) and other algorithms from the literature. The experimental results show that MPPST outperforms other algorithms for all instances and achieves a significant security enhancement.