小波包域灰度级双水印技术  

Wavelet Packet Domain Gray-Level Bi-watermarking Technique

作　　者：仝黎[1] 程正兴[1] 

机构地区：[1]西安交通大学理学院,西安710049

出　　处：《计算机工程与应用》2004年第33期78-82,共5页Computer Engineering and Applications

摘　　要：数字水印由于在数字产品版权保护方面的潜在作用而引起人们的重视。目前广泛采用的是在图像的时空域或变换域中感知重要的部分嵌入一个水印。然而感知重要的部分通常是图像的低频部分,当图像经历诸如对比度调整、亮度调整等操作时,从低频提取的水印视觉效果会很差,某些恶意的攻击者会采用这些操作使得正版的数字产品由于提取不到水印而遭到拒绝。文章提出了一种基于小波包域的双水印技术,结合人类视觉系统同时在低频部分和高频部分分别嵌入一个水印。试验结果证明结合人类视觉系统的双水印技术可以有更广泛的适应性和更强的稳健性。Gray-level digital watermarks are more attractive by people because of its potential effects on copyright protection of digital products,which are often embedded in the perceptual parts of spatial domain or transform domain now.However,we know that perceptual parts are usually low frequency of the image.Some authorized copyrights of digital products undergoing malicious attacks including contrast adjust and brightness adjust which degenerate low frequency parts will be considered to be illegal.This paper presents a so-called gray-level bi-watermarking technique which inserts a watermark into low frequency and high frequency of the wavelet packet domain respectively together with human visual system.Experimental results show that the proposed technique has stronger robustness to common image processing and can adapt to much more situations.

关 键 词：小波包 灰度级双水印 人类视觉系统 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]