2차원 셀룰러오토마트 변환을 기반으로 한 새로운 디지털 워터마킹

Abstract

Cellular Automata are discrete dynamical systems, which provide the basis for the synthesis of complex emergent behavior. Starting from a basic unit, the cell, it not only provides insight into the evolution of life but also opens vistas into the creation of intelligent computer systems. In the first section of this paper, we propose a novel secure watermarking system using Cellular Automata Transform (CAT). Firstly, the original image is decomposed into level-2 CAT algorithm and the sub bands LL2 and HH2 are obtained. And then the original watermarks are embedded into the “low-low” frequency (LL2) and “high-high” frequency (HH2) of the original image respectively. At last, the watermarked image is obtained using the inverse CAT(ICAT) for the transformed image. This algorithm is tested on cover image LENA (512x512, gray scale) with binary watermark of size 128x128. The second section recommends Arnold scrambling algorithm is recommend. At first, the original image is disassembled by level-1 2D CAT transform and is divided into four sub bands. The sub bands HL1 and LH1 are further transformed by 2D CAT, and the obtained sub bands HL2 and LH2 are called the “middle frequency” which are obtained. Then the watermark is scrambled using Arnold’s scrambling algorithm and decomposed into level-1 by CAT transform. At last, level-1 CAT transformed watermark is embedded into the ‘middle frequency’ of the original image. This algorithm is tested on famous images LENA, BOATS and TIGER (512x512, gray scale) with binary watermark of size 128x128. The experiment results show that our CAT-based watermarking system can simultaneously improve security, robustness and image quality of the watermarked image.