基于opencv人脸识别_基于opencv的人脸识别

作者：IT小白 | 2024-03-21 04:43:27

踩

基于opencv的人脸识别

一、准备

链接：代码+数据集
提取码：led1

环境：win10，vs2013，opencv3.2

注：每个代码都可以在vs中单独运行

该项目所有文件如下：

其中只需要使用的文件如下：

详细代码：


//这个是生成照片的代码
#include <opencv2/opencv.hpp>
#include <iostream>
 
using namespace cv;
using namespace std;
 
//cv库，哈尔检测人脸使用的配置文件
string haar_face_datapath = "E:/opencv/install/etc/haarcascades/haarcascade_frontalface_alt_tree.xml";
 
 
int main(int argc, char** argv) {
	//打开摄像头
	VideoCapture capture(0); 
	if (!capture.isOpened()) {
		printf("could not open camera...\n");
		return -1;
	}
	Size S = Size((int)capture.get(CV_CAP_PROP_FRAME_WIDTH), (int)capture.get(CV_CAP_PROP_FRAME_HEIGHT));
	int fps = capture.get(CV_CAP_PROP_FPS);
 
	//人脸检测库
	CascadeClassifier faceDetector;
	//人脸检测函数
	faceDetector.load(haar_face_datapath);
 
 
	Mat frame;
	namedWindow("camera-demo", CV_WINDOW_AUTOSIZE);
	vector<Rect> faces;
	int count = 0;
	while (capture.read(frame)) {
		flip(frame, frame, 1);
		faceDetector.detectMultiScale(frame, faces, 1.1, 1, 0, Size(100, 120), Size(380, 400));
		for (int i = 0; i < faces.size(); i++) {
			if (count % 10 == 0) {
				Mat dst;
				//重新把照片大小改成模板的大小
				resize(frame(faces[i]), dst, Size(112, 92));
				//保存照片
				imwrite(format("D:/else//led/face_%d.jpg", count), dst);
			}
			rectangle(frame, faces[i], Scalar(0, 0, 255), 2, 8, 0);
		}
		imshow("camera-demo", frame);
		char c = waitKey(10);
		//按键退出截图
		if (c == 27) {
			break;
		}
		count++;
	}
 
	capture.release();
 
	waitKey(0);
	return 0;
}


//这个是生成csv文件的教程，直接用shell命令，不需要写代码
 
 
1. 打开cmd
 
 
2. 输入 F: (因为我的文件在F盘)  注意一定要先进入F盘才可以后续操作，不然会显示找不到文件
 
 
3. 输入cd 文件路径，点回车
 
 
4. 输入dir /b/s/p/w *.jpg>at.csv 注意，我的文件格式为.jpg


#include <opencv2/opencv.hpp>
#include <opencv2/face.hpp>
#include <iostream>
 
using namespace cv;
using namespace cv::face;
using namespace std;
 
int main(int argc, char** argv) {
	
	//加载自己的csv文件
	string filename = string("D:/else/wyz/at.csv");
	ifstream file(filename.c_str(), ifstream::in);
	if (!file) {
		printf("could not load file correctly...\n");
		return -1;
	}
 
	string line, path, classlabel;
	vector<Mat> images;
	vector<int> labels;
	char separator = ';';
	while (getline(file, line)) {
		stringstream liness(line);
		getline(liness, path, separator);
		getline(liness, classlabel);
		if (!path.empty() && !classlabel.empty()) {
			//printf("path : %s\n", path.c_str());
			images.push_back(imread(path, 0));
			labels.push_back(atoi(classlabel.c_str()));
		}
	}
 
	if (images.size() < 1 || labels.size() < 1) {
		printf("invalid image path...\n");
		return -1;
	}
 
	int height = images[0].rows;
	int width = images[0].cols;
	printf("height : %d, width : %d\n", height, width);
 
	Mat testSample = images[images.size() - 1];
	int testLabel = labels[labels.size() - 1];
	images.pop_back();
	labels.pop_back();
 
	// train it，训练自己的模型，到这里已经完成训练操作
	Ptr<BasicFaceRecognizer> model = createEigenFaceRecognizer();
	model->train(images, labels);
	model->save("D:/else/wyz/led_wyz.xml");
 
 
	/* 以下两个方法训练模型同样操作
	Ptr<FaceRecognizer> model1 = createFisherFaceRecognizer();
	model1->train(images, labels);
	model1->save("MyFaceFisherModel.xml");
	Ptr<FaceRecognizer> model2 = createLBPHFaceRecognizer();
	model2->train(images, labels);
	model2->save("MyFaceLBPHModel.xml");
	*/
 
 
 
	/* 以下内容用于检测验证自己重建的人脸模，可以直接跳过，本代码全部注释掉了
	// recognition face
	int predictedLabel = model->predict(testSample);
	printf("actual label : %d, predict label :  %d\n", testLabel, predictedLabel);
	Mat eigenvalues = model->getEigenValues();
	Mat W = model->getEigenVectors();
	Mat mean = model->getMean();
	Mat meanFace = mean.reshape(1, height);
	Mat dst;
	if (meanFace.channels() == 1) {
		normalize(meanFace, dst, 0, 255, NORM_MINMAX, CV_8UC1);
	}
	else if (meanFace.channels() == 3) {
		normalize(meanFace, dst, 0, 255, NORM_MINMAX, CV_8UC3);
	}
	imshow("Mean Face", dst);
	// show eigen faces
	for (int i = 0; i < min(25, W.cols); i++) {
		Mat ev = W.col(i).clone();
		Mat grayscale;
		Mat eigenFace = ev.reshape(1, height);
		if (eigenFace.channels() == 1) {
			normalize(eigenFace, grayscale, 0, 255, NORM_MINMAX, CV_8UC1);
		}
		else if (eigenFace.channels() == 3) {
			normalize(eigenFace, grayscale, 0, 255, NORM_MINMAX, CV_8UC3);
		}
		Mat colorface;
		applyColorMap(grayscale, colorface, COLORMAP_JET);
		char* winTitle = new char[128];
		sprintf(winTitle, "eigenface_%d", i);
		imshow(winTitle, colorface);
	}
	// 重建人脸
	for (int num = min(25, W.cols); num < min(W.cols, 300); num += 1) {
		Mat evs = Mat(W, Range::all(), Range(0, num));
		Mat projection = LDA::subspaceProject(evs, mean, images[0].reshape(1, 1));
		Mat reconstruction = LDA::subspaceReconstruct(evs, mean, projection);
		Mat result = reconstruction.reshape(1, height);
		if (result.channels() == 1) {
			normalize(result, reconstruction, 0, 255, NORM_MINMAX, CV_8UC1);
		}
		else if (result.channels() == 3) {
			normalize(result, reconstruction, 0, 255, NORM_MINMAX, CV_8UC3);
		}
		char* winTitle = new char[128];
		sprintf(winTitle, "recon_face_%d", num);
		imshow(winTitle, reconstruction);
	}
	*/
 
	waitKey(0);
	return 0;
}

二、原理

原理1.PCA（主成分分析）原理

通过对高维数据分析发现他们的相同与不同表达为一个低维数据模式

主成分不变

细微损失

高维数据到低维数据

原理2.API调用

主要是使用其中的三个函数，来训练自己的数据，函数如下：


// train it
	Ptr<BasicFaceRecognizer> model = createEigenFaceRecognizer();
	model->train(images, labels);
	model->save("D:/else/wyz/led_wyz.xml");


	
	Ptr<FaceRecognizer> model1 = createFisherFaceRecognizer();
	model1->train(images, labels);
	model1->save("MyFaceFisherModel.xml");
 
	Ptr<FaceRecognizer> model2 = createLBPHFaceRecognizer();
	model2->train(images, labels);
	model2->save("MyFaceLBPHModel.xml");