摘要
提出了一种新的无损的数据嵌入算法,可应用于军事、法律、医学等需要无损图像的领域。使用了改进的LSB作为嵌入算法并对原图的未修改部分进行了预测、上下文模型与量化、算术编码。预测减少了空间冗余,上下文模型进一步利用了空间相关性和位平面相关。基于上下文模型的算术编码获得了更好的压缩效率和压缩容量。实验证明嵌入数据后的图像损伤小并可完全恢复。
A novel data embedding algorithm is presented, which enables the exact recovery of the original host signal upon extraction of the embedded information. Lossless recovery of the original is achieved by compressing portion of the signal that are susceptible to embedding distortion and transmitting these compressed descriptions as a part of the embedded payload. A prediction-based conditional entropy coder which utilizes unaltered portion of the host signal as side-information improves the compression efficiency and the lossless data embedding capacity.
出处
《计算机应用与软件》
CSCD
北大核心
2007年第7期157-158,186,共3页
Computer Applications and Software
关键词
算术编码
上下文模型
数据隐藏
改进的LSB
Arithmetic coding Context modeling Data hiding Generalized-LSB modification
