关于水印图像抗攻击隐形优化处理仿真
|
摘要
对水印图像进行抗攻击隐形优化处理,能够有效保护数字图像版权保护,对水印图像进行隐形优化处理,需要对高光谱水印图像中的第一谱段进行预测,并对像素预测参考值进行修正。传统方法建立高光谱水印图像编码块的边信息,并对边信息选择最优的预测阶数,但忽略了修正像素预测参考值,导致处理过程耗时长。提出基于多波段预测的水印图像抗攻击隐形优化处理方法,采用中值滤波器对高光谱水印图像中的第一谱段进行预测,消除谱间冗余;采用线性预测根据第一谱段内像素对当前谱段像素进行预测,获得当前像素预测参考值,建立上下文模型对像素预测参考值进行修正。采用8级查表搜索预测算法对水印图像各谱段像素进行预测,获得最终图像像素预测值,利用熵编码算法对该预测值嵌入编码获取水印图像抗攻击隐形优化处理结果。实验结果表明,所提方法处理过程运算时间短,且处理后的图像质量较优。
Traditional method establishes the edge information of coding block in hyper-spectral watermark image and selects the optimal prediction order of side information, but ignores the corrected pixel prediction reference value. Therefore, a method of invisible optimization for attack-resistant of watermark image based on multiband prediction was presented. First of all, the median filter was used to predict the first spectral coverage in hyper-spectral watermark image and eliminate the inter-spectral redundancy. Then, linear prediction was used to predict the current spectral pixel according to the pixel in the first spectral coverage, and the reference value of current pixel prediction was obtained. In addition, the context model was built to correct the reference value of pixel prediction. Meanwhile, level-8 search table prediction algorithm was used to predict the pixel of each spectral coverage in the watermark image, so as to obtain the final prediction value of image pixel. Finally, the entropy coding algorithm was used to embed the coding into predictive value. Thus, we could obtain the result of invisible optimization of attack-resistant of watermark image. Simulation results show that the operation time of proposed method is short. Meanwhile, the image quality after processing is better.
Traditional method establishes the edge information of coding block in hyper-spectral watermark image and selects the optimal prediction order of side information, but ignores the corrected pixel prediction reference value. Therefore, a method of invisible optimization for attack-resistant of watermark image based on multiband prediction was presented. First of all, the median filter was used to predict the first spectral coverage in hyper-spectral watermark image and eliminate the inter-spectral redundancy. Then, linear prediction was used to predict the current spectral pixel according to the pixel in the first spectral coverage, and the reference value of current pixel prediction was obtained. In addition, the context model was built to correct the reference value of pixel prediction. Meanwhile, level-8 search table prediction algorithm was used to predict the pixel of each spectral coverage in the watermark image, so as to obtain the final prediction value of image pixel. Finally, the entropy coding algorithm was used to embed the coding into predictive value. Thus, we could obtain the result of invisible optimization of attack-resistant of watermark image. Simulation results show that the operation time of proposed method is short. Meanwhile, the image quality after processing is better.
引文
[1] 佟德宇,任娜,朱长青.基于压缩感知的图像水印信息生成算法[J].华中科技大学学报(自然科学版),2017,45(8):38-41.
[2] 王树梅,张文斌.一种基于伪随机的支持大尺寸图像水印算法[J].计算机技术与发展,2017,27(8):121-124.
[3] 贺体刚.基于无损恢复的光学图像可见水印[J].激光杂志,2016,37(5):48-51.
[4] 赵博,秦贵和.高鲁棒性的图像水印算法[J].吉林大学学报(工),2017,47(1):249-254.
[5] 张倩倩,詹永照,林涵阳.基于分块和NSCT-SVD的彩色图像数字水印算法[J].江苏大学学报(自然科学版),2017,38(5):582-589.
[6] 贾园,张贵仓,徐天芝.基于边缘检测和奇异值分解的图像水印算法[J].计算机与数字工程,2017,45(1):138-141.
[7] 廖洪建,黄良永.融合IWT与SVD的鲁棒和隐蔽性数字图像水印方案[J].湘潭大学自科学报,2016,38(3):59-63.
[8] 吴佳楠,等.融合量子密钥真随机性的二值图像水印[J].光学精密工程,2017,25(11):2968-2974.
[9] 李淑芝,等.全方向梯度预测和自适应选块的可逆图像水印[J].小型微型计算机系统,2017,38(1):174-178.
[10] 于广州.数字图像内容保护下双重隐形数字水印仿真[J].计算机仿真,2018,(2).
[2] 王树梅,张文斌.一种基于伪随机的支持大尺寸图像水印算法[J].计算机技术与发展,2017,27(8):121-124.
[3] 贺体刚.基于无损恢复的光学图像可见水印[J].激光杂志,2016,37(5):48-51.
[4] 赵博,秦贵和.高鲁棒性的图像水印算法[J].吉林大学学报(工),2017,47(1):249-254.
[5] 张倩倩,詹永照,林涵阳.基于分块和NSCT-SVD的彩色图像数字水印算法[J].江苏大学学报(自然科学版),2017,38(5):582-589.
[6] 贾园,张贵仓,徐天芝.基于边缘检测和奇异值分解的图像水印算法[J].计算机与数字工程,2017,45(1):138-141.
[7] 廖洪建,黄良永.融合IWT与SVD的鲁棒和隐蔽性数字图像水印方案[J].湘潭大学自科学报,2016,38(3):59-63.
[8] 吴佳楠,等.融合量子密钥真随机性的二值图像水印[J].光学精密工程,2017,25(11):2968-2974.
[9] 李淑芝,等.全方向梯度预测和自适应选块的可逆图像水印[J].小型微型计算机系统,2017,38(1):174-178.
[10] 于广州.数字图像内容保护下双重隐形数字水印仿真[J].计算机仿真,2018,(2).