摘要
本文提出了将小波零树编码与神经网络矢量量化相结合实现特征信息保护的虹膜图像压缩思路。该方法利用形状编码与树状映射策略对虹膜重要特征信息区域对应的小波系数进行定位,对包含重要特征信息的小波系数进行近无损压缩,对非特征区域采用树结构矢量组合进行三方向的矢量量化压缩编码。该方法解决了不同区域图像质量要求不同的压缩问题,可以实现渐近解码。仿真实验结果表明本文提出的算法是合理可行的。
A new iris image compression method based on zero-tree wavelet coding and self-organizing neural network for feature information protection is presented. A shape coding and adaptive map method is used for locating the position of most important wavelet coefficients. Then the important wavelet coefficients including important feature information are compressed nearly lossless while the non-feature information area is compressed by lossy vector quantization coding. The experimental simulation results show that this method is feasible and it could solve the problem of different image section with different compression rate.
出处
《电子测量与仪器学报》
CSCD
2005年第3期75-78,共4页
Journal of Electronic Measurement and Instrumentation
