使用dlib 进行人脸识别

编译dlib

从官方网站下载

准备好人脸数据

准备好人脸数据，放入各自的文件夹，然后给文件夹起好名字，这里就是人的名字拼音首字。准备好test数据

test数据也是图片，一次对比一张，其外，还要有dlib的face模型，包含两个文件，一个是dlib_face_recognition_resnet_model_v1.dat和shape_predictor_68_face_landmarks.dat,两个文件，在运行过程中必须加载

//查找文件夹下面的所有文件夹

//人名的文件夹
void face_reco::listfiles(std::string dir,std::string namefolder)
{
	intptr_t handle;
	_finddata_t findData;

	string dirfilter =dir + "/" +  namefolder + "/" + "*.*";
	handle = _findfirst(dirfilter.c_str(), &findData);    // 查找目录中的第一个文件
	if (handle == -1)
	{
		cout << "Failed to find first file!\n";
		return;
	}
	face_desc desc;
	desc.name = namefolder;//子文件夹的名称为人名
	do
	{

		if (!(findData.attrib & _A_SUBDIR))
			desc.files.push_back(findData.name);
	} while (_findnext(handle, &findData) == 0);    // 查找目录中的下一个文件
													//cout << "Done!\n";
	_findclose(handle);    // 关闭搜索句柄
	face_desc_vec.push_back(desc);

}


void face_reco::listfolder(std::string dir, std::vector<string> & names)
{
	intptr_t handle;
	_finddata_t findData;
	
	string dirfilter = dir + "/" + "*.*";
	handle = _findfirst(dirfilter.c_str(), &findData);    // 查找目录中的第一个文件
	if (handle == -1)
	{
		cout << "Failed to find first file!\n";
		return;
	}

	do
	{
		if (findData.attrib & _A_SUBDIR)
			//&& strcmp(findData.name, ".") != 0
			//&& strcmp(findData.name, "..") != 0
			//)    // 是否是子目录并且不为"."或".."
		{
			if (findData.name[0] != '.')
				names.push_back(findData.name);
			//cout << findData.name << "\t\n";
		}
		else //否则是文件
		{
			//子文件
			//face_desc_vec.files.push_back(findData.name);
		}
		//cout << findData.name << "\t" << findData.size << endl;
	} while (_findnext(handle, &findData) == 0);    // 查找目录中的下一个文件

													//cout << "Done!\n";
	_findclose(handle);    // 关闭搜索句柄



}

int face_reco::load_db_faces_prepare(string folder)
{
	
	// We will also use a face landmarking model to align faces to a standard pose:  (see face_landmark_detection_ex.cpp for an introduction)
	deserialize("./shape_predictor_68_face_landmarks.dat") >> sp;

	// And finally we load the DNN responsible for face recognition.
	deserialize("./dlib_face_recognition_resnet_model_v1.dat") >> net;

	std::vector<string> vecs;
	listfolder(folder.c_str(), vecs);
	auto iter = vecs.begin();
	while(iter != vecs.end())
	{
		//文件夹名称就是人名
		string name = *iter;
		cout << name << endl;
		//string mfolder = folder + "/" + name;

		listfiles(folder,name);
		iter++;
	}

	//把所有的文件都放到数据结构中，开始读取并且获取值
	load_faces(folder);


	return 0;
}



void face_reco::load_faces(string folder)
{
	frontal_face_detector detector = get_frontal_face_detector();
	auto iter = face_desc_vec.begin();
	while (iter != face_desc_vec.end())
	{
		face_desc &desc = *iter;
		matrix<rgb_pixel> img;
		string path = folder + "/" + desc.name +"/"+ desc.files[0];
		cout <<"path is " << path << endl;
		//char path[260];
		//sprintf_s(path, "%s/%s/%s", folder.c_str(),desc.name.c_str(),desc.files[0].c_str());
		load_image(img, path);
		//image_window win(img);

		for (auto face : detector(img))
		{
			auto shape = sp(img, face);
			matrix<rgb_pixel> face_chip;
			extract_image_chip(img, get_face_chip_details(shape, 150, 0.25), face_chip);

			//Record the all this face's information
			//FACE_DESC sigle_face;
			desc.face_chip = face_chip;
			//sigle_face.name = fileinfo.name;

			std::vector<matrix<rgb_pixel>> face_chip_vec;
			std::vector<matrix<float, 0, 1>> face_all;

			face_chip_vec.push_back(move(face_chip));

			//Asks the DNN to convert each face image in faces into a 128D vector
			face_all = net(face_chip_vec);

			//Get the feature of this person
			std::vector<matrix<float, 0, 1>>::iterator iter_begin = face_all.begin(),
				iter_end = face_all.end();
			if (face_all.size() > 1) break;
			desc.face_feature = *iter_begin;

			//all the person description into vector
			//face_desc_vec.push_back(sigle_face);
		}
		iter++;
	}

}

以上过程就是程序拿到取走文件夹中所有文件进行学习，实际上dlib本身是使用的dnn方式来训练网络数据的。

调用过程

#include "face_recognition_dlib.h"
#include "Sort_method.h"
#include 
#include 
using namespace std;
#include 

typedef struct face_pics
{
	string name;
	std::vector<string> pics;
}face_pics;

int load_folder_jpg()
{
	return 0;
}

face_reco face_recognize;


int compareone(face_desc &desc)
{
	return 0;

}

int main(int argc, char** argv)
{
	//std::vector folder;
	//listFiles("H:/git/dlib/qb_recognize/qbrecog/bin/face/*.*",folder, true);
	//FACE_RECOGNITION dlib;
	if (argc == 1)
	{
		cout << "Give some image files as arguments to this program." << endl;
		return 0;
	}

	cout << "processing image " << argv[1] << endl;

	std::string ff = "face";
	face_recognize.load_db_faces_prepare(ff);

	matrix<rgb_pixel> face_cap;
	//save the capture in the project directory
	load_image(face_cap, argv[1]);
	image_window win(face_cap);
	//Display the raw image on the screen
	//image_window win1(face_cap);

	frontal_face_detector detector = get_frontal_face_detector();
	std::vector<matrix<rgb_pixel>> vect_faces;

	for (auto face : detector(face_cap))
	{
		auto shape = face_recognize.sp(face_cap, face);
		matrix<rgb_pixel> face_chip;
		extract_image_chip(face_cap, get_face_chip_details(shape, 150, 0.25), face_chip);
		vect_faces.push_back(move(face_chip));
		//win1.add_overlay(face);
	}

	if (vect_faces.size() != 1)
	{
		cout << "Capture face error! face number " << vect_faces.size() << endl;
		return -1;
	}

	//Use DNN and get the capture face's feature with 128D vector
	std::vector<matrix<float, 0, 1>> face_cap_desc = face_recognize.net(vect_faces);
	//Browse the face feature from the database, and find the match one
	std::pair<double, std::string> candidate_face;
	std::vector<double> len_vec;

	std::vector<std::pair<double, std::string>> candi_face_vec;
	candi_face_vec.reserve(256);
	cout << "face find " << face_recognize.face_desc_vec.size() << endl;
	for (size_t i = 0; i < face_recognize.face_desc_vec.size(); ++i)
	{
		auto len = length(face_cap_desc[0] - face_recognize.face_desc_vec[i].face_feature);
		cout << "len is " << len << endl;
		if (len < 0.4)
		{
			len_vec.push_back(len);
			candidate_face.first = len;
			candidate_face.second = face_recognize.face_desc_vec[i].name.c_str();
			candi_face_vec.push_back(candidate_face);

//#ifdef _FACE_RECOGNIZE_DEBUG
//			char buffer[256] = { 0 };
//			sprintf_s(buffer, "Candidate face %s Euclid length %f",
//				face_recognize.face_desc_vec[i].name.c_str(),
//				len);
//			MessageBox(CString(buffer), NULL, MB_YESNO);
//#endif
		}
		else
		{
			cout << "This face from database is not match the face, continue!" << endl;
		}
	}

	//Find the most similar face
	if (len_vec.size() != 0)
	{
		shellSort(len_vec);

		int i(0);
		for (i = 0; i != len_vec.size(); i++)
		{
			if (len_vec[0] == candi_face_vec[i].first)
				break;
		}

		char buffer[256] = { 0 };
		sprintf_s(buffer, "The face is %s -- Euclid length %f",
			candi_face_vec[i].second.c_str(), candi_face_vec[i].first);
		cout << buffer << endl;
		/*if (MessageBox(CString(buffer), NULL, MB_YESNO) == IDNO)
		{
			face_record();
		}*/
	}
	else
	{
		cout << "not found" << endl;
		/*if (MessageBox(CString("Not the similar face been found"), NULL, MB_YESNO) == IDYES)
		{
			face_record();
		}*/
	}
	//char a;
	getchar();
    return 0;
}

以上就是整个训练和使用过程，比较简单，读者可以继续深入研究，比如人脸图片曝光过强去除，人脸对齐等等，都是需要调整的，上面代码并不包含这些。