摘要
为了解决当前数字水印技术不可见性和鲁棒性不能较好兼顾的问题,提出了一种基于QR(quick response)码和Schur分解的感兴趣区域(ROI)盲提取水印算法。算法基于时间复杂性较低的Schur分解和符合人眼视觉感知的显著图,通过将显著图归一取反作为嵌入强度,自适应地将加密水印信息嵌入到载体图像经小波变换和分块Schur分解的系数中。实验结果表明,本文算法不仅具有极高的不可见性,而且对裁切、任意角度旋转、高斯噪声、椒盐噪声、斑纹噪声、高斯滤波、中值滤波、对比度增强和JPEG压缩等攻击具有较强的鲁棒性;由于显著图作为公开密钥,所以算法在提取过程中不需要原载体图像和水印图像,属于盲提取水印算法。因此,本文算法在数字产品版权保护中具有较高的实际应用价值。
In order to solve the problem that the invisibility and robustness of the current digital watermarking technology cannot be well balanced, one region of interest （ROI） based watermarking algorithm is proposed based on quick response （QR） code and Schur decomposition. In consideration of the lower time complexity of the Schur decomposition and the saliency map of the carrier image which accords with the human visual perception,the encrypted watermark information is embedded adaptively into the coefficients of the carrier image through the discrete wavelet transform and the block Schur decomposition. During the embedding process, the inverse of normalized saliency map is used as the embedding strength to improve the watermark invisibility. Experimental results show that the algorithm not only has high invisibility, and but also has strong robustness, for some common image attacks, such as cutting, arbitrary angle rotation, Gaussian noise, salt and pepper noise, speckle noise, Gaussian filtering, median filtering, contrast enhanced and JPEG compression,Due to the saliency map is used as the public key, the algorithm don＇t need the original carrier image and the watermark image in extraction process,so it belongs to the blind watermarking algorithm. Therefore, this algorithm has high practical application value in digital product copyright protection.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2017年第4期419-426,共8页
Journal of Optoelectronics·Laser
关键词
QR码
SCHUR分解
显著图
感兴趣区域（ROI)
盲水印
quick response （QR） code
Schur decomposition
saliency map
region on interest （ROI）
blind watermarking
作者简介
E-mail：wang_keyan@163．com.王晓红（1971-），女，陕西人，博士，教授，主要从事颜色科学、数字图像处理方面的研究．
