基于回归型支持向量机的数字图像水印新算法

以回归型支持向量机（Support Vector Regression,SVR）理论为基础,提出一种数字图像水印新算法.该算法能够结合图像局部相关性,选取稳定的特征向量并获得SVR训练模型,进而利用SVR训练模型嵌入和提取数字水印信息.该算法以保证不可感知性和鲁棒性的良好平衡为前提,实现了数字水印的盲检测.仿真实验表明,本文算法不仅具有较好的不可感知性,而且对叠加噪声、JPEG压缩、锐化、平滑滤波、对比度增强等常规处理及扭曲、剪切等几何攻击均具有较好的鲁棒性,其整体性能明显优于现有SVM图像水印方案.

New Adaptive Digital Image Watermarking Based on SVR

In this paper,a new digital image watermarking algorithm based on support vector regression(SVR) is proposed.Firstly,the steady feature vector can be selected according to the local correlation of digital image,and then the SVR train model can be obtained by applying the support vector regression theory.Finally,the watermark signal can be embedded or extracted by utilizing the above SVR train model.The proposed scheme can extract the digital watermark without the help of the original digital image.Experimental results show that the proposed scheme is not only invisible and robust against common signals processing such as noise adding,JPEG compression,sharpening,smoothing,filtering,and contrast enhancement,but also robust against the geometric attacks such as shearing,distortion.Especially,it performs better than other SVM image watermarking schemes.