Face Landmark Detection在疲勞檢測和眼球跟蹤等有着巨大作用。
安裝步驟：
git clone https://github.com/davisking/dlib #下載dlib
cd dlib
cd examples #進入dlib下的examples文件夾
mkdir build #新建build文件夾，存放cmake編譯後的執行文件
cd build #進入新建好的build文件夾
#cmake編譯examples整個文件夾
cmake .. -DDLIB_USE_CUDA=1 -DUSE_AVX_INSTRUCTIONS=1 #選擇是否使用cuda avx
cmake --build . --config Release
攝像頭測試：
進入example/build目錄，執行：
./webcam_face_pose_ex
qt5上c++實現：
配置.pro文件：
TEMPLATE = app
CONFIG += console c++11
CONFIG -= app_bundle
CONFIG -= qt
SOURCES += main.cpp
#QMAKE_CFLAGS = -DDLIB_USE_CUDA=1 -DUSE_AVX_INSTRUCTIONS=1 #這句加不加感覺沒啥用，速度一樣 有沒人知道？
#system
INCLUDEPATH += /usr/local/lib \
/usr/lib/x86_64-linux-gnu
LIBS += -L/usr/local/lib
#opencv
INCLUDEPATH += /usr/include \
/usr/include/opencv \
/usr/include/opencv2/
LIBS += -L /usr/lib/libopencv_*.so
#dlib
SOURCES += /home/bjw/git/dlib/dlib/all/source.cpp #源碼
INCLUDEPATH += /home/bjw/git/dlib #頭文件
LIBS += -L/usr/lib -lpthread -lX11
注意根據自己情況配置。
qt c++代碼：
#include <dlib/opencv.h>
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#include <dlib/image_processing/frontal_face_detector.h>
#include <dlib/image_processing/render_face_detections.h>
#include <dlib/image_processing.h>
#include <dlib/gui_widgets.h>
#include <time.h>
#include <iostream>
using namespace dlib;
using namespace std;
int main()
{
try
{
cv::VideoCapture cap(0);
if (!cap.isOpened())
{
cerr << "Unable to connect to camera" << endl;
return 1;
}
//image_window win;
// Load face detection and pose estimation models.
frontal_face_detector detector = get_frontal_face_detector();
shape_predictor pose_model;
deserialize("/home/bjw/QT_Project/dlib_landmark/dlib_landmark/shape_predictor_68_face_landmarks.dat") >> pose_model;
// Grab and process frames until the main window is closed by the user.
while(cv::waitKey(1) != 27)
{
// Grab a frame
cv::Mat temp;
if (!cap.read(temp))
{
break;
}
// Turn OpenCV's Mat into something dlib can deal with. Note that this just
// wraps the Mat object, it doesn't copy anything. So cimg is only valid as
// long as temp is valid. Also don't do anything to temp that would cause it
// to reallocate the memory which stores the image as that will make cimg
// contain dangling pointers. This basically means you shouldn't modify temp
// while using cimg.
cv_image<bgr_pixel> cimg(temp);
// Detect faces
std::vector<rectangle> faces = detector(cimg);
time_t start = clock();
// Find the pose of each face.
std::vector<full_object_detection> shapes;
for (unsigned long i = 0; i < faces.size(); ++i)
shapes.push_back(pose_model(cimg, faces[i]));
if (!shapes.empty()) {
for (int i = 0; i < 68; i++) {
circle(temp, cvPoint(shapes[0].part(i).x(), shapes[0].part(i).y()), 3, cv::Scalar(0, 0, 255), -1);
// shapes[0].part(i).x();//68個
}
}
// Display it all on the screen
// win.clear_overlay();
// win.set_image(cimg);
// win.add_overlay(render_face_detections(shapes));
//Display it all on the screen
imshow("Dlib特徵點", temp);
time_t end = clock();
double total_time = (double)(end-start)/CLOCKS_PER_SEC*1000;
cout<<"total_time:"<<total_time<<"ms"<<endl;
}
}
catch(serialization_error& e)
{
cout << "You need dlib's default face landmarking model file to run this example." << endl;
cout << "You can get it from the following URL: " << endl;
cout << "http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2" << endl;
cout << endl << e.what() << endl;
}
catch(exception& e)
{
cout << e.what() << endl;
}
}
結論：
（１）我測試了一下耗時，主要還是人臉檢測上，而且不是很準確，人臉檢測40ms左右，而68個特徵點檢測只有幾毫秒。
（２）不直達爲啥開了cuda加速，速度根本沒變化，希望知道的朋友告知一聲。