An Improved Image Steganography Algorithm Based on PVD

Abstract

In this paper, a modification of PVD (Pixel value differencing) algorithm is used for Image Steganography in spatial domain. It is normalizing secret data value by encoding method to make the new pixel edge difference less among three neighbors (horizontal, vertical and diagonal) and embedding data only to less intensity pixel difference areas or regions. The proposed algorithm shows a good improvemernt for both color and gray-scale images compared to other algorithms. Color images performance are better than gray images. However, in this work the focus is mainly on gray images. The strenght of this scheme is that any random hidden/secret data do not make any shuttle differences to Steg-image compared to original image. The bit plane slicing is used to analyze the maximum payload that has been embeded into the cover image securely. The simulation results show that the proposed algorithm is performing better and showing great consistent results for PSNR, MSE values of any images, also against Steganalysis attack.