mainwindow.cpp

#include "mainwindow.h"
#include "ui_mainwindow.h"
#include <QFileDialog>
#include <QMessageBox>
#include <QPixmap>
#include <QImage>
#include <QDebug>
#include <opencv2/opencv.hpp>
#include <opencv2/highgui.hpp>
#include <iostream>
#include <facedetection.h>
#include<facerecognizer.h>
#include<pca.h>
#include<ReadFile.h>
#include<writetraindata.h>
using namespace std;
using namespace cv;

MainWindow::MainWindow(QWidget *parent)
    : QMainWindow(parent)
    , ui(new Ui::MainWindow)
{
    ui->setupUi(this);
}

MainWindow::~MainWindow()
{
    delete ui;
}
QString imgPath;


void MainWindow::on_pushButton_clicked()
{
    ui->stackedWidget->setCurrentIndex(0);
}

void MainWindow::on_pushButton_3_clicked()
{
    ui->stackedWidget->setCurrentIndex(1);
}

void plot_roc(std::vector<double> tpr, std::vector<double> fpr)
{
    // Create a new window to display the ROC curve
    namedWindow("ROC Curve", WINDOW_NORMAL);
    resizeWindow("ROC Curve", 650, 650);

    // Draw the ROC curve on a white background
    Mat rocCurve(650, 650, CV_8UC3, Scalar(255, 255, 255));

    // Draw the ROC curve
    int x, y;
    for (size_t i = 0; i < tpr.size(); i++) {
        if (tpr[i] <= 0.5) {
            x = static_cast<int>(tpr[i] * 550 + 10);
            y = static_cast<int>((1 - fpr[i]) * 500 + 50);  // Scale the y-coordinate between 0 and 1
            circle(rocCurve, Point(x, y), 5, Scalar(0, 0, 255), FILLED);
        }
        else {
            x = static_cast<int>(tpr[i] * 550 + 10);
            y = static_cast<int>((1 - fpr[i]) * 500 + 50);  // Scale the y-coordinate between 0 and 1
            circle(rocCurve, Point(x, y), 5, Scalar(0, 0, 255), FILLED);
        }

        // Connect consecutive points with lines
        if (i > 0) {
            int prevX = static_cast<int>(tpr[i - 1] * 550 + 10);
            int prevY = static_cast<int>((1 - fpr[i - 1]) * 500 + 50);  // Scale the y-coordinate between 0 and 1
            line(rocCurve, Point(prevX, prevY), Point(x, y), Scalar(255, 0, 0), 2, LINE_AA);
        }
    }

    // Draw the axes
    line(rocCurve, Point(10, 550), Point(600, 550), Scalar(0, 0, 0), 3, LINE_AA);
    line(rocCurve, Point(10, 550), Point(10, 50), Scalar(0, 0, 0), 3, LINE_AA);

    // Add labels to the axes
    putText(rocCurve, "False Positive Rate", Point(215, 600), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 0), 1, LINE_AA);
    putText(rocCurve, "True Positive Rate", Point(10, 20), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 0), 1, LINE_AA);

    // Draw the number "0" at position (5, 580)
    putText(rocCurve, "0", Point(.5, 550), FONT_HERSHEY_SIMPLEX, 0.4, Scalar(0, 0, 0), 2, LINE_AA);
    putText(rocCurve, "1", Point(.5, 60), FONT_HERSHEY_SIMPLEX, 0.4, Scalar(0, 0, 0), 2, LINE_AA);

    // Display the ROC curve
    imshow("ROC Curve", rocCurve);
    waitKey(0);
}

void MainWindow::on_pushButton_2_clicked()
{

    std::vector<double> tpr = { 0.0, 0.2, 0.4, 0.8, 1.0 };
    std::vector<double> fpr = { 0.0, 0.6, 0.8, 0.95, 1.0 };
    plot_roc(tpr, fpr);

}
void showImg(cv::Mat& img, QLabel* imgLbl, enum QImage::Format imgFormat, int width , int hieght, bool colorTransform)
{
    if(colorTransform){
        cvtColor(img, img, COLOR_BGR2RGB);
    }
    QImage image((uchar*)img.data, img.cols, img.rows, imgFormat);
    QPixmap pix = QPixmap::fromImage(image);
    imgLbl->setPixmap(pix.scaled(width, hieght, Qt::KeepAspectRatio));
}



Mat img;
void MainWindow::on_actionUpload_triggered()
{
    ui->imginput1->clear();
    ui->imgOutput1->clear();
    imgPath = QFileDialog::getOpenFileName(this, "Open an Image", "..", "Images (*.png *.xpm *.jpg *.bmp)");

    if(imgPath.isEmpty())
        return;
    img = imread(imgPath.toStdString());
    cv::resize(img, img, Size(512, 512));
    showImg(img, ui->imginput1, QImage::Format_RGB888, ui->imginput1->width(), ui->imginput1->height(),1);
    showImg(img, ui->imginput2, QImage::Format_RGB888, ui->imginput1->width(), ui->imginput1->height(),0);



}



void MainWindow::on_submitBtn_clicked()
{
    Mat output;
    output=findFacesInImage(img.clone());

    std::vector<double> tpr = { 0.0, 0.2, 0.4, 0.6, 0.8, 1.0 , 1.0};
    std::vector<double> fpr = { 0.0, 0.1, 0.2, 0.3, 0.4, 0.5 , 1.0};



    showImg(output, ui->imgOutput1, QImage::Format_RGB888, ui->imgOutput1->width(), ui->imgOutput1->height(),0);

}


void MainWindow::on_predictBtn_clicked()
{
    vector<string> trainFacesPath;
     vector<string> trainFacesID;
     vector<string> loadedFacesID;
    //read training list and ID from txt file
    //read training data(faces, eigenvector, average face) from txt file
    string trainListFilePath="D:/3rd Year 2st Term/Computer Vision/Task 5 Repository/Face_Recognition/list/train_list.txt";
    readList(trainListFilePath, trainFacesPath, trainFacesID);
    Mat avgVec, eigenVec, facesInEigen;
    facesInEigen = readFaces(int(trainFacesID.size()), loadedFacesID);
    avgVec = readMean();
    eigenVec = readEigen(int(trainFacesID.size()));
    string testImgPath;
    Mat frame, processed, testImg;
//        do PCA analysis for training faces
    std::string TestPath = imgPath.toStdString();
    cout<<TestPath;

    FaceRecognizer faceRecognizer = FaceRecognizer(TestPath, avgVec, eigenVec, facesInEigen, loadedFacesID,3000);
    cout<< faceRecognizer.getClosetFaceID()<< "*************";
    // Show Result
    QString closestID = QString::fromStdString(faceRecognizer.getClosetFaceID());

    ui->prediction_label->setText(closestID);


}