基于分块DCT变换的HVS模型加权数字水印算法

摘要

ln this paper,a new method was proposed to hide watermark image in images by block-based DCT techniques in conjunction with the visual model for compression.According to the masking information generated by bit decomposition of the watermark image,watermark information was embedded into the mid-frequency DCT coefficient.The simulation experiments show that this robustness method can survive image cropping and sharping,and get better results.%用灰度图象作为水印，并结合应用于图象压缩的人眼视觉模型HVS和基于块分割DCT变换，对原有的基于DCT变换的数字水印算法作了一些改进。实验结果表明，这种改进方法具有良好的鲁棒性和不可见性。