摘要
针对嵌入水印强度的不确定性,提出了一种利用果蝇优化算法并结合小波变换与奇异值分解的数字图像水印嵌入算法。首先,利用小波变换对载体图像进行二级小波分解,然后对LL2子带进行4×4分块并进行奇异值分解以提高嵌入水印的稳定性,最终将水印信息以一定的强度嵌入到分解的最大奇异值中。果蝇优化算法综合考虑水印嵌入算法的鲁棒性与不可见性之间的矛盾,并结合水印嵌入方案确定最佳嵌入强度,首先根据影响水印性能的指标定义适应度函数,并对该函数的参数进行优化,确定了最佳的适应度函数,然后再应用果蝇优化算法寻找嵌入水印强度的最优解。最后通过仿真实验对该水印方案的性能进行测试与分析,实验结果表明,当嵌入水印图像的峰值信噪比达到40 dB时,提取水印的平均NC值能达到0.98以上。
According to the uncertainty of the embedded watermarking strength,a digital image watermarking algorithm based on fruit fly optimization algorithm combined with wavelet transform and singular value decomposition is proposed.Firstly,the carrier image is decomposed by two-level wavelet transform,and then the LL2 subband is divided into 4*4 blocks and singular value decomposition to improve the stability of the embedded watermarking.Finally,the watermark is embedded into the maximum singular value of the decomposition with certain strength.The fruit fly optimization algorithm considers the contradiction between robustness and invisibility of watermarking algorithm synthetically,and combined with the watermark embedding scheme to determine the best embedding strength.Firstly,the fitness function is defined according to the index affecting the watermarking performance,and the parameters are optimized to determine the best fitness function.Then,the optimal solution is found by applying fruit fly optimization algorithm.Finally,the performance of the watermarking scheme is tested and analyzed through simulation experiments.The experimental results show that when the peak signal-to-noise ratio of the embedded watermarking image reaches 40 dB,the average NC value of the extracted watermarking can reach more than 0.98.
作者
张帅
贾有
杨雪霞
ZHANG Shuai;JIA You;YANG Xue-xia(Department of Teaching,Taiyuan Radio&TV University,Taiyuan 030024,China;School of Applied Sciences,Taiyuan University of Science and Technology,Taiyuan 030024,China)
出处
《太原科技大学学报》
2019年第6期423-429,共7页
Journal of Taiyuan University of Science and Technology
基金
国家自然科学基金(11602157)
关键词
果蝇优化
小波变换
奇异值分解
数字水印
fruit fly optimization
wavelet transform
singular value decomposition
digital watermarking
