基于小波变换的分块PCA人脸识别方法
首发时间:2007-09-21
摘要:提出了一种基于小波变换的分块PCA方法,该方法首先对图像进行小波变换,取得人脸图像的低频子带,在此基础上对图像进行分块,并对每块用PCA方法提取特征向量,然后使用距离分类法进行识别。本方法与传统的PCA方法相比,有着较高的识别率和一定的鲁棒性
For information in English, please click here
Face Recognition of Block-PCA based on Wavelet Transform
Abstract:A face recognition method of block principal component analysis based on wavelet transformation was proposed. First, an image was disposed by wavelet transform and low frequency sub-band of the image was extracted. Then the feature components were extracted with PCA based on the block image. In the process of classification, using distance classifier obtained recognition results. Comparing with the classic PCA, the recognition ration and robustness are greatly improved
Keywords: Face recognition, Wavelet transformation, Block-PCA, Distance classifier
基金:
论文图表:
引用
No.1526314802211903****
同行评议
共计0人参与
勘误表
基于小波变换的分块PCA人脸识别方法
评论
全部评论0/1000