PCA人脸识别算法Matlab版.doc

文档介绍：PCA人脸识别算法Matlab版
2

———————————————————————————————— 作者：
———————————————————————————————— 日期：

个人收集 仅供参考学习 勿做商业用途
PCA人脸识别算法Matlab版
2021-11-18 19:30
function T = CreateDatabase(TrainDatabasePath)
% Align a set of face images (the training set T1, T2, ... , TM )
%
% Description: This function reshapes all 2D images of the training database
% into 1D column vectors. Then, it puts these 1D column vectors in a row to 
% construct 2D matrix 'T'.
%  
% 
% Argument:      TrainDatabasePath      - Path of the training database
%
% Returns:       T                      - A 2D matrix, containing all 1D image vectors.
%                                         Suppose all P images in the training database 
%                                         have the same size of MxN. So the length of 1D 
%                                         column vectors is MN and 'T' will be a MNxP 2D matrix.
%
% See also: STRCMP, STRCAT, RESHAPE
% Original version by Amir Hossein Omidvarnia, October 2007
%                      Email: ******@                  
%%%%%%%%%%%%%%%%%%%%%%%% File management
TrainFiles = dir(TrainDatabasePath);
Train_Number = 0;
for i = 1:size(TrainFiles,1)
    if not(strcmp(TrainFiles(i).name,'.')|strcmp(TrainFiles(i).name,'..')|strcmp(TrainFiles(i).name,''))
         Train_Number = Train_Number + 1; % Number of all images in the training database
     end
end
%%%%%%%%%%%%%%%%%%%%%%%% Construction of 2D matrix from 1D image vectors
T = [];
for i = 1 : Train_Number
    
    % I have chosen the name of each image in databases as a corresponding
    % number. However, it is not mandatory!
     str = int2str(i);
     str = strcat('\',str,'.jpg');
     str = strcat(TrainDatabasePath,str);
    
     img = imread(str);