摘要
针对现有方法所提取的特征值不能灵活适应图像的纹理变化导致被篡改图像恢复质量不高的问题,提出一种根据图像块纹理特征进行块类型划分的篡改定位和恢复水印算法。首先把宿主图像分3×3大小的子块,再利用方差和Canny边缘检测算子将图像子块划分为简单块、复杂块和边缘块;然后根据图像子块类型,自适应地生成水印信息;最后采用Torus自同构映射方式找到水印嵌入位置,并且对篡改区域进行恢复时根据图像子块类型自适应的恢复。当图像遭到篡改时,采用本文算法恢复图像的平均峰值信噪比(PSNR)较现有方法提5.60%左右。实验表明,本文算法能够对图像纹理复杂区域生成更精确的恢复信息,有效提高了图像篡改区域的恢复质量,同时降低含水印图像失真。算法适用于医学、军事和卫星等领域。
As for the eigen value extracted by the existing methods could not flexibly adapt to the change of image textures,leading to lower recovery quality of tampered images, an image blocks classification watermarking method for tamper detection and recovery is proposed. Firstly of all, the original host im- age is divided into 3 N 3 sub blocks;then the variance of sub blocks and Canny edge detection algorithm are adopted to mark these sub blocks into smooth blocks,rough blocks and edge blocks. According to these sub image types, the eigen value could be adaptively generated. Finally, the Torus automorphismis algorithm is used to find the position of mapping blocks. In the meantime, the tampered images could be recovered adaptively based on sub images type. If the image has not been tampered, the peak signal-to- noise ratio （PSNR） can be improved by 5.60% compared with existing method. Experimental results show that compared with previous similar schemes, the proposed method can more accurately produce recovery watermarking information, make better improvement on the recovery quality of the tampered image and effectively reduce the distortion of watermarking image. The presented method can be applied in many fields, including the areas of medicine, military, and satellites.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2018年第2期197-204,共8页
Journal of Optoelectronics·Laser
基金
国家自然科学基金(61762046
41362015)
江西省自然科学基金(20161bab212048)
江西省教育厅科学技术研究重点(GJJ160599)
江西省教育厅科学技术研究(GJJ151522)
江西省科技厅自然科学基金(20151BAB217008)资助项目
关键词
自适应
篡改恢复
纹理复杂度
数字水印
adaptive
tamper detection and recovery
texture complexity
digital watermarking
