基于DCT-SVD的双QR码水印防伪算法

目的提出一种基于DCT-SVD的双QR码数字水印算法。方法按照QR编码规则,将文字信息编码生成QR码载体图像和QR码水印图像;对水印QR码图像进行置乱,把置乱后的水印图像用奇异值分解原理嵌入经DCT分块变换的载体QR码图像中,形成嵌入水印后的QR码;运用提出的算法从载体QR码图像中提取QR水印,并用手机扫描打印的载体QR码和提取的QR码。结果提出的算法可以抵抗旋转和剪切等几何攻击,同时对高斯噪声、椒盐噪声、低通滤波、均值滤波和JPEG压缩攻击具有较强的鲁棒性。结论该算法不需要附加设备的参与,不增加印刷成本,能够抵抗打印扫描攻击,是一种高鲁棒性的数字水印算法;进一步可以实现二维码的防伪革新,能够广泛应用于数字作品的版权保护。

Double QR Code Watermarking Algorithm Based on DCT-SVD

The aim of this study was to propose a double QR Code digital watermarking algorithm based on singular value decomposition and discrete cosine transform (DCT-SVD). First of all, the original information and watermark were transformed by the principle of encoding into QR code, then the QR code watermark image was scrambled, the scrambled QR code was embedded in the original QR code through discrete cosine transform by using singular value decomposition principle, then the watermarked QR code was formed. Finally, the watermark QR code was extracted from the original QR image and was scanned through telephone. Experiments showed that this algorithm can resist rotation and cropping attack, which had better robustness to noise, low-pass filtering, and JPEG compression. All the processes did not require any additional equipment, and it can resist printing and scanning. This algorithm can achieve further reform for QR code, which can be widely used to protect the copyright of digital works.