摘要
由于小波分解在空域和频域上都能提供良好的局部信息,尤其是在小波分解后可以减少图像的分辨率,进而相应地减少计算复杂度,因此小波变换经常被用于图像处理和图像分析中.为探究基于人脸识别的小波变换最佳应用效果,着重讨论了小波变换应用于人脸识别的两个关键问题:小波基的选择和分解层数的确定,并通过实验分析和比较研究得出相应的答案和指导原则.
Wavelet transform(WT) has been a very popular tool for image processing and image analysis,because WT has the advantages of good time and frequency localizations, and the resolutions of the subband images are reduced by decomposing an image using WT. In turn, the computational complexity will be reduced dramatically by working on a lower resolution image.In order to study the effective usage of the WT based on human face recognition, two keys are proposed to the WT's application in human face recognition: the selection of wavelet base and the wavelet decomposition level. Finally,the answer and the guardline are given by experiments.
出处
《浙江大学学报(理学版)》
CAS
CSCD
北大核心
2005年第1期34-38,共5页
Journal of Zhejiang University(Science Edition)
基金
浙江省自然科学基金资助项目(299011).
关键词
小波变换
人脸识别
小波基选择
分解层数
wavelet transform
human face recognition
selection of wavelet base
decomposition level