目录

1--前言

2--处理ORL数据集

3--Eigenfaces复现过程

4--Fisherfaces复现过程

5--分析

1--前言

①SYSU模式识别课程作业

②配置：基于Windows11、OpenCV4.5.5、VSCode、CMake（参考OpenCV配置方式）

③原理及源码介绍：Face Recognition with OpenCV

④数据集：ORL Database of Faces

2--处理ORL数据集

①源码：

import sys
import os.pathif __name__ == "__main__":BASE_PATH = './ORL/att_faces/orl_faces/'SEPARATOR = ";"dir_txt = open("./dir.txt", 'w')label = 0for dirname, dirnames, filenames in os.walk(BASE_PATH):# dirname当前路径; dirnames当前路径下所有目录名（不包含子目录）;filenames当前路径下的所有文件名（不包含子目录）for subdirname in dirnames: # 遍历每一个目录subject_path = os.path.join(dirname, subdirname)for filename in os.listdir(subject_path):abs_path = "%s/%s" % (subject_path, filename)print("%s%s%d" % (abs_path, SEPARATOR, label))dir_txt.write(abs_path)dir_txt.write(SEPARATOR)dir_txt.write(str(label))dir_txt.write("\n")label = label + 1dir_txt.close()

②运行及结果：

python create_csv.py

3--Eigenfaces复现过程

①源码：

// 引用依赖
#include "opencv2/core.hpp"
#include "opencv2/face.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/imgproc.hpp"
#include 
#include 
#include // 使用相应的命名空间
using namespace cv;
using namespace cv::face;
using namespace std;// 标准化函数
static Mat norm_0_255(InputArray _src) {Mat src = _src.getMat();// Create and return normalized image:Mat dst;switch(src.channels()) {case 1:cv::normalize(_src, dst, 0, 255, NORM_MINMAX, CV_8UC1);break;case 3:cv::normalize(_src, dst, 0, 255, NORM_MINMAX, CV_8UC3);break;default:src.copyTo(dst);break;}return dst;
}// 读取CSV文件函数
static void read_csv(const string& filename, vector& images, vector& labels, char separator = ';') {std::ifstream file(filename.c_str(), ifstream::in);if (!file) {string error_message = "No valid input file was given, please check the given filename.";CV_Error(Error::StsBadArg, error_message);}string line, path, classlabel;while (getline(file, line)) {stringstream liness(line);getline(liness, path, separator);getline(liness, classlabel);if(!path.empty() && !classlabel.empty()) {images.push_back(imread(path, 0));labels.push_back(atoi(classlabel.c_str()));}}
}
int main(int argc, const char *argv[]) {//检查argc是否符合要求if (argc < 2) {cout << "usage: " << argv[0] << "   " << endl;exit(1);}string output_folder = ".";if (argc == 3) {output_folder = string(argv[2]);}// CSV文件的路径string fn_csv = string(argv[1]);// 初始化存储imgs和labels的向量vector images;vector labels;// 读取CSV文件try {read_csv(fn_csv, images, labels);} catch (const cv::Exception& e) {cerr << "Error opening file \"" << fn_csv << "\". Reason: " << e.msg << endl;exit(1);}// 判断img数目是否符合要求if(images.size() <= 1) {string error_message = "This demo needs at least 2 images to work. Please add more images to your data set!";CV_Error(Error::StsError, error_message);}// images的高度int height = images[0].rows;// 从训练集中选择一张图片作为测试集Mat testSample = images[images.size() - 1];int testLabel = labels[labels.size() - 1];images.pop_back();labels.pop_back();// 创建模型，使用PCA特征脸算法Ptr model = EigenFaceRecognizer::create();model->train(images, labels); // 训练模型int predictedLabel = model->predict(testSample);  // 使用测试集测试模型// 打印准确率string result_message = format("Predicted class = %d / Actual class = %d.", predictedLabel, testLabel);cout << result_message << endl;// 获取模型的特征值Mat eigenvalues = model->getEigenValues();// 展示特征向量Mat W = model->getEigenVectors();// 从训练集中获取样本均值Mat mean = model->getMean();// 根据argc判断进行展示或保存操作if(argc == 2) {imshow("mean", norm_0_255(mean.reshape(1, images[0].rows)));} else {imwrite(format("%s/mean.png", output_folder.c_str()), norm_0_255(mean.reshape(1, images[0].rows)));}// 显示或保存特征脸for (int i = 0; i < min(10, W.cols); i++) {string msg = format("Eigenvalue #%d = %.5f", i, eigenvalues.at(i));cout << msg << endl;// 获取特征向量Mat ev = W.col(i).clone();// resize成原始大小，并归一化到0-255Mat grayscale = norm_0_255(ev.reshape(1, height));// 显示图像并应用Jet颜色图以获得更好的观感。Mat cgrayscale;applyColorMap(grayscale, cgrayscale, COLORMAP_JET);// 根据argc判断进行展示或保存操作if(argc == 2) {imshow(format("eigenface_%d", i), cgrayscale);} else {imwrite(format("%s/eigenface_%d.png", output_folder.c_str(), i), norm_0_255(cgrayscale));}}// 在一些预定义的步骤中显示或保存图像重建的过程：for(int num_components = min(W.cols, 10); num_components < min(W.cols, 300); num_components+=15) {// 从模型中分割特征向量Mat evs = Mat(W, Range::all(), Range(0, num_components));Mat projection = LDA::subspaceProject(evs, mean, images[0].reshape(1,1));Mat reconstruction = LDA::subspaceReconstruct(evs, mean, projection);// 归一化reconstruction = norm_0_255(reconstruction.reshape(1, images[0].rows));// 根据argc判断进行展示或保存操作if(argc == 2) {imshow(format("eigenface_reconstruction_%d", num_components), reconstruction);} else {imwrite(format("%s/eigenface_reconstruction_%d.png", output_folder.c_str(), num_components), reconstruction);}}// 如果没有写入输出文件夹，则等待键盘输入if(argc == 2) {waitKey(0);}return 0;
}

②编译过程：

CMakeLists.txt如下：

cmake_minimum_required(VERSION 3.24)  # 指定 cmake的 最小版本
project(test) # 设置项目名称find_package(Opencv REQUIRED)
INCLUDE_DIRECTORIES(${OpenCV_INCLUDE_DIRS})
add_executable(eigenfaces_demo eigenfaces.cpp) # 生成可执行文件
target_link_libraries(eigenfaces_demo ${OpenCV_LIBS} ) # 设置target需要链接的库

mkdir buildcd buildcmake ..cd ..mingw32-make

③运行及结果展示：

./eigenfaces_demo.exe ./dir.txt ./Engenfaces_Result

特征图：（简单修改源程序生成的文件名，再按顺序进行拼接即可生成拼接图，拼接程序参考）

重建过程：

均值图：

4--Fisherfaces复现过程

①源码：

// 引用依赖
#include "opencv2/core.hpp"
#include "opencv2/face.hpp"
#include "opencv2/highgui.hpp"
#include "opencv2/imgproc.hpp"
#include 
#include 
#include // 使用相应的命名空间
using namespace cv;
using namespace cv::face;
using namespace std;// 标准化函数
static Mat norm_0_255(InputArray _src) {Mat src = _src.getMat();Mat dst;switch(src.channels()) {case 1:cv::normalize(_src, dst, 0, 255, NORM_MINMAX, CV_8UC1);break;case 3:cv::normalize(_src, dst, 0, 255, NORM_MINMAX, CV_8UC3);break;default:src.copyTo(dst);break;}return dst;
}// 读取csv文件函数
static void read_csv(const string& filename, vector& images, vector& labels, char separator = ';') {std::ifstream file(filename.c_str(), ifstream::in);if (!file) {string error_message = "No valid input file was given, please check the given filename.";CV_Error(Error::StsBadArg, error_message);}string line, path, classlabel;while (getline(file, line)) {stringstream liness(line);getline(liness, path, separator);getline(liness, classlabel);if(!path.empty() && !classlabel.empty()) {images.push_back(imread(path, 0));labels.push_back(atoi(classlabel.c_str()));}}
}int main(int argc, const char *argv[]) {//检查argc是否符合要求if (argc < 2) {cout << "usage: " << argv[0] << "   " << endl;exit(1);}string output_folder = ".";if (argc == 3) {output_folder = string(argv[2]);}// CSV文件的路径string fn_csv = string(argv[1]);// 初始化存储imgs和labels的向量vector images;vector labels;// 读取CSV文件try {read_csv(fn_csv, images, labels);} catch (const cv::Exception& e) {cerr << "Error opening file \"" << fn_csv << "\". Reason: " << e.msg << endl;exit(1);}// 判断img数目是否符合要求if(images.size() <= 1) {string error_message = "This demo needs at least 2 images to work. Please add more images to your data set!";CV_Error(Error::StsError, error_message);}// images的高度int height = images[0].rows;// 从训练集中选择一张图片作为测试集Mat testSample = images[images.size() - 1];int testLabel = labels[labels.size() - 1];images.pop_back();labels.pop_back();// 创建模型，使用LDA线性判别分析Ptr model = FisherFaceRecognizer::create();model->train(images, labels); // 训练模型int predictedLabel = model->predict(testSample); // 使用测试集测试模型// 打印准确率string result_message = format("Predicted class = %d / Actual class = %d.", predictedLabel, testLabel);cout << result_message << endl;// 获取模型的特征值Mat eigenvalues = model->getEigenValues();// 展示特征向量Mat W = model->getEigenVectors();// 从训练集中获取样本均值Mat mean = model->getMean();// 根据argc判断进行展示或保存操作if(argc == 2) {imshow("mean", norm_0_255(mean.reshape(1, images[0].rows)));} else {imwrite(format("%s/mean.png", output_folder.c_str()), norm_0_255(mean.reshape(1, images[0].rows)));}// 显示或保存特征脸for (int i = 0; i < min(16, W.cols); i++) {string msg = format("Eigenvalue #%d = %.5f", i, eigenvalues.at(i));cout << msg << endl;// 获取特征向量Mat ev = W.col(i).clone();// resize成原始大小，并归一化到0-255Mat grayscale = norm_0_255(ev.reshape(1, height));// 显示图像并应用Jet颜色图以获得更好的观感。Mat cgrayscale;applyColorMap(grayscale, cgrayscale, COLORMAP_BONE);// 根据argc判断进行展示或保存操作if(argc == 2) {imshow(format("fisherface_%d", i), cgrayscale);} else {imwrite(format("%s/fisherface_%d.png", output_folder.c_str(), i), norm_0_255(cgrayscale));}}// 在一些预定义的步骤中显示或保存图像重建的过程：for(int num_component = 0; num_component < min(16, W.cols); num_component++) {// 从模型中分割特征向量Mat ev = W.col(num_component);Mat projection = LDA::subspaceProject(ev, mean, images[0].reshape(1,1));Mat reconstruction = LDA::subspaceReconstruct(ev, mean, projection);// 归一化reconstruction = norm_0_255(reconstruction.reshape(1, images[0].rows));// 根据argc判断进行展示或保存操作if(argc == 2) {imshow(format("fisherface_reconstruction_%d", num_component), reconstruction);} else {imwrite(format("%s/fisherface_reconstruction_%d.png", output_folder.c_str(), num_component), reconstruction);}}// 如果没有写入输出文件夹，则等待键盘输入if(argc == 2) {waitKey(0);}return 0;
}

②编译过程：

CMakeLists.txt如下：

cmake_minimum_required(VERSION 3.24)  # 指定 cmake的 最小版本
project(test) # 设置项目名称find_package(Opencv REQUIRED)
INCLUDE_DIRECTORIES(${OpenCV_INCLUDE_DIRS})
#add_executable(eigenfaces_demo eigenfaces.cpp) # 生成可执行文件
#target_link_libraries(eigenfaces_demo ${OpenCV_LIBS} ) # 设置target需要链接的库
add_executable(fisherfaces_demo fisherfaces.cpp) # 生成可执行文件
target_link_libraries(fisherfaces_demo ${OpenCV_LIBS} ) # 设置target需要链接的库

mkdir buildcd buildcmake ..cd ..mingw32-make

③运行及结果展示：

./fisherfaces_demo.exe ./dir.txt ./Fisherfaces_Result

特征图：

重建过程：

均值图：

5--分析

未完待续！