Technology Research Based On Wavelet Transformation And PCA For Face Recognition

With the society's development,more and more people from all different areas focus on identity certification and they want easier and faster ways.As the biological characteristics,the face's characteristics are more intrinsic and more stable and are also different from each others,so they are taken as the ideal tools in the area of identity certification,however,they are direct,easy and friendly,which are easily received by people,Therefore,more and more people from both image processing and pattern recognition focus on automatic identity certification.,and more from all the different areas have paid attention to them.First,this paper introduces the background,application,development processing, research situation and the research ways of face recognition;after that,some ways are introduced totally.This paper uses the common PCA(Principle Component Analysis) way.The preprocessing of face recognition is very important,so its ways are introduced in detail,which include:pixel gradation normalization,geometric normalization.Then Wavelet transform is studied in depth.and its application is described in detail.At last,The ways of traditional PCA are studied deeply.Compared with PCA,2DPCA is a better way,which is studied in the thesis.After that,two methods are compared on theory and experimental data,at the same time,it's showed that compared with PCA,2DPCA can take less time on calculation,and it is better than the traditional PCA.