基于提升小波变换与学习矢量量化网络的鉴别分析方法

提出了一种基于提升小波变换（LWT）和学习矢量量化网络（LVQ）相结合的鉴别分析方法。提升小波又叫作第二代小波，比传统的第一代小波变换更为快速有效，利用它的多分辨率特性，可以获取输入图像的低频信息并使图像降维。LVQ算法是在有教师状态下对竞争层进行训练的一种学习算法。LVQ网络结构简单，但却表现出比BP网络更强的有效性和鲁棒性。在ORL标准人脸库及现实人脸图像上的实验结果表明该方法具有很好的鉴别分析能力。

Discriminant analysis based on lifting wavelet transform and learning vector quantization

A new discriminant analysis method based on LWT(Lifting Wavelet Transform) and LVQ(Learning Vector Quantization) Network was proposed in this paper.LWT is faster and more efficient than the first generation wavelet transform,but it also has the multi-resolution characteristics.LWT can be used to extract the low frequency coefficients and reduce the dimension of an image.LVQ is an effective learning algorithm that trains the competitive layer under supervision.It has a simple network structure,but it also has good discriminant analysis ability.The experimental results on ORL face database show that the method proposed has very good classification capability and high recognition rate.