几何变换归一化在盲图像水印算法中的应用

在水印技术的应用中，针对几何攻击的鲁棒性一直是一个难题。这主要是由于即使很小的尺度或旋转攻击，都会使接收图像失去与原水印图像的同步，从而无法检测出水印。提出一种利用几何矩对原始图像和接收图像进行归一化处理的方法，使水印的嵌入和检测都在同步的基础上进行。该方法具有鲁棒性好、校正精度高、运算简单的特点，同时检测端不需要原始图像。实验证明本算法对翻转、旋转和尺度变化等几何攻击具有很好的鲁棒性。

Application of Normalized Geometric Transformation in Blind Image Watermarking Algorithm

In watermarking applications, the robustness of the watermark to geometric manipulations is a critical issue. This is due to the fact that changing the image size or its orientation, even by slight amount, could make the receiver lost synchronization with original watermarked image. This paper proposes a new method which makes use of geometric moments to normalize the original image and the receive image, therefore provides a synchronized environment for watermark embedding and detecting. This method has the advantages of good robustness, precise correction, efficient computation, and the original image does not to be needed to the decoder side. Experimental results show that this algorithm has good performance to such geometric manipulations as flipping, rotation and scaling.