CAJ | 학술논문

针对经典量子进化算法及其在图像水印算法中的应用有嵌入容量较小等问题，提出了一种基于改进量子进化算法的图像水印算法，对经典量子进化算法作了两方面的改进，一是将表示量子染色体的量子比特概率幅修改为量子角，并在此基础上对量子旋转门旋转策略作了相应的修改；二是子群优化，每个子群相对独立地执行量子进化算法。实验结果表明：改进后的算法不但简化了量子染色体的表达，还依靠子群优化达到了算法并行性优化的目的，嵌入点的选择与嵌入策略也使得算法有较大的嵌入容量，该算法产生的含水印图像有较高的视觉质量且鲁棒性好。
침대경전양자진화산법급기재도상수인산법중적응용유감입용량교소등문제，제출료일충기우개진양자진화산법적도상수인산법，대경전양자진화산법작료량방면적개진，일시장표시양자염색체적양자비특개솔폭수개위양자각，병재차기출상대양자선전문선전책략작료상응적수개；이시자군우화，매개자군상대독입지집행양자진화산법。실험결과표명：개진후적산법불단간화료양자염색체적표체，환의고자군우화체도료산법병행성우화적목적，감입점적선택여감입책략야사득산법유교대적감입용량，해산법산생적함수인도상유교고적시각질량차로봉성호。
Classical quantum evolutionary algorithm and its application in image watermarking algorithm has the shortcoming such as small embedding capacity etc.In view of these,this paper proposes a image watermarking algorithm based on improved quantum evolutionary algorithm, improves classical quantum evolutionary algorithm in two aspects, the first one is that quantum bit probability amplitude of the quantum chromosome is replaced by quantum angle and corresponding changes to rotation strategy of quantum rotation gate are made; the second one is the optimization of subgroup, quantum evolutionary algorithm are executed relatively independent for each subgroup.Experiment results show that, the improved algorithm simplifies the expression of quantum chromosome and achieves the purpose of algorithm parallel optimization relying on the subgroup optimization, and the choices of strategy and point for embedding bring higher embedding capacity for this algorithm, which produces the watermarked image high visual quality and good robustness.