摘要
提出了一种基于小波变换和视觉感知特性的灰度级自适应盲水印算法。该算法利用视觉系统的背景亮度、纹理和边界掩盖特性把载体图像的小波系数分为18类;在对256级灰度水印图像压缩后,依据载体图像小波系数的分类,以不同强度嵌入到载体中。该算法在抽取水印时不需要原始载体图像和原始水印。实验结果表明此算法可有效地隐藏灰度水印图像,并有较好的顽健性,能抵抗裁剪、JPEG压缩、加噪、图像增强、滤波等多种处理。
We propose a gray-level adaptive blind watermarking arithmetic based on wavelet transform and visual perception character. This arithmetic divides the wavelet coefficients of carrier image into 18 classes by the background luminance, the texture and the edge mask character of visual system; then after compressing the 256-level watermark image, it embeds the watermark sequences in the carrier image with different intensity according to the coefficients classes of carrier image. This arithmetic can detect the digital watermark without the original carrier image and the original watermark. Experimental results prove that this arithmetic can hide the gray-level watermark efficiently in the carrier image, and it is robust enough to resist some image degradation processing such as clipping, JPEG compression, adding noise, image enhancing, filtering and so on.
出处
《通信学报》
EI
CSCD
北大核心
2003年第9期132-141,共10页
Journal on Communications
基金
国家自然科学基金(10171033)
留学回国人员科研基金资助项目
广东省"千百十工程"优秀人才基金资助项目
关键词
数字水印
小波变换
人眼视觉特性
刚能鉴别的失真阈值
digital watermarking
wavelet transform
human visual characteristic
just noticeable distortion threshold
