A Novel Image Scrambling Technique Based On Information Entropy And Quad Tree Decomposition

With the rapid increase of digital data exchange and increased usage of digital images, it is important to protect the confidential image data from illegal access. Digital image scrambling or encryption is the solution which transforms a meaningful image into a meaningless or disordered image in order to enhance the ability to confront attack and in turn improve the security. This paper presents a new scheme for digital image scrambling based on the principle of information entropy. The quad tree decomposition technique is used to hierarchically divide the image into blocks or regions for enforcing security at the block or region level of an image, which thus ensures the security of the entire image. The experimental results show that the proposed algorithm can successfully scramble the images, and the analysis of the algorithm also demonstrate that the scrambled images have good information entropy and low correlation coefficients thereby satisfying the requisite security.