A Statistical and Iterative Method for Data Hiding in Palette-Based Images

Abstract

This paper proposes an improved iterative method for data hiding in palette-based images, taking into account the statistics of the data that need to be embedded in an image. In particular, the proposed method considers the distribution of the number of zeroes and ones in the input message, as well as how the message bits are distributed over the colors in the image palette. First, according to the statistics of the input message, the proposed method modifies the pixel indexes and the color palette using an enhanced version of an iterative image pre-processing method, replacing less frequent colors by colors that are close to frequently used colors. In a next step, the actual message bits are embedded into the parity bits of the remaining colors. Finally, the proposed method applies a post-processing step, adjusting particular colors in order to further reduce the amount of image distortion. Experimental results show that the proposed method preserves image quality better than previously proposed techniques for data hiding in palette-based images.Keywordsdata hidinginformation hidingpalette-based imagessteganography