基于图像奇异值的脆弱水印方案

针对精确定位型脆弱水印算法的安全性问题，提出一种基于图像奇异值的脆弱水印方案。该方案选取与图像内容密切相关的图像奇异值作为水印，用混沌系统对其加密来增强水印的安全性。认证时通过水印差值定位图像被篡改的位置，通过奇异值差值反映被篡改区域的篡改强度。实验仿真结果表明该算法不仅提高了精确定位型水印算法的安全性，而且还可以估计出被篡改区域的篡改强度。

Novel fragile watermarking based on image singular value

A novel fragile watermarking based on image singular value was proposed to overcome the security problems of fragile watermarking with precise localization. The embedded watermark was obtained by converting the singular values of image matrix into a binary sequence. To improve watermark security, a chaotic system was applied to encrypt the watermark sequence. The tampered area of the watermarked image could be reflected in the watermarking difference map and the tampered intensity could be reflected in singular values difference. Simulation results show that the proposed method may not only improve the security of fragile watermarking with precise localization, but also point out the intensity of tampered areas.