Compile with cmake face_landmark_detection to see landmark numbers

[ManuelTS]

I changed your face_landmark_detection.cpp file to a bare minimum for better overview and I only want to have an output of the first image where not lines or circles are the overlay landmarks but rather the numbers of the landmarks themself. I trained successfully with the 194 features the shape predictor with the helen dataset which works perfectly on my changed version of your webcam_face_pose_ex.cpp. But I'm unable to compile the face_landmark_detection.cpp with this additional dependency:

#include <opencv2/imgproc.hpp>

needed to print numbers instead of lines or circles. My changed face_landmark_detection.cpp:

// The contents of this file are in the public domain. See LICENSE_FOR_EXAMPLE_PROGRAMS.txt

#include <iostream>
#include <opencv2/imgproc.hpp> // <------------ Makes linker problems
#include "dlib/opencv.h"
#include "dlib/image_processing/frontal_face_detector.h"
#include "dlib/image_processing.h"
#include "dlib/gui_widgets.h"
#include "dlib/image_io.h"

using namespace dlib;

int main(int argc, char** argv)
{
    try
    { // ../faces/helen/100032540_1.jpg
        frontal_face_detector detector = get_frontal_face_detector();
        shape_predictor sp;
        deserialize("../faces/helen/helen_shape_predictor_194_face_landmarks.dat") >> sp;

        image_window win;
        for (int i = 0; i < argc; ++i)
        {
            std::cout << "processing image " << argv[i] << std::endl;
            array2d<rgb_pixel> img;
            load_image(img, argv[i]);
            pyramid_up(img);

            std::vector<rectangle> dets = detector(img);
            std::cout << "Number of faces detected: " << dets.size() << std::endl;
            std::vector<full_object_detection> shapes;

            for (unsigned long j = 0; j < dets.size(); ++j)
            {
                full_object_detection shape = sp(img, dets[j]);
                std::cout << "number of parts: "<< shape.num_parts() << std::endl;
                std::cout << "pixel position of first part:  " << shape.part(0) << std::endl;
                std::cout << "pixel position of second part: " << shape.part(1) << std::endl;
                shapes.push_back(shape);

//----------------------Code i need #include <opencv2/imgproc.hpp> for with source:-----------------------------------------------------
                cv::Mat buff_img = dlib::toMat(img); // https://stackoverflow.com/questions/36711905/dlib-train-shape-predictor-ex-cpp/37880450#comment94164497_37880450

                for (int k = 0; k < 194; ++k) {
                  cv::putText(buff_img, std::to_string(k), cv::Point(shape.part(k).x(), shape.part(k).y()), cv::FONT_HERSHEY_COMPLEX_SMALL, 0.5, cv::Scalar(255, 255, 0), 1, cv::LINE_AA);
                }

                save_png(img, "outImage.png");
            }
        }
    }
    catch (std::exception& e)
    {
        std::cout << "\nexception thrown!" << std::endl;
        std::cout << e.what() << std::endl;
    }
}

The compiler output for cmake .. && cmake --build . --config Release is:

Scanning dependencies of target face_landmark_detection_ex
[ 92%] Building CXX object CMakeFiles/face_landmark_detection_ex.dir/face_landmark_detection_ex.cpp.o
[ 93%] Linking CXX executable face_landmark_detection_ex
CMakeFiles/face_landmark_detection_ex.dir/face_landmark_detection_ex.cpp.o: In function `cv::Mat::~Mat()':
face_landmark_detection_ex.cpp:(.text._ZN2cv3MatD2Ev[_ZN2cv3MatD5Ev]+0x81): undefined reference to `cv::Mat::deallocate()'
face_landmark_detection_ex.cpp:(.text._ZN2cv3MatD2Ev[_ZN2cv3MatD5Ev]+0x70): undefined reference to `cv::fastFree(void*)'
CMakeFiles/face_landmark_detection_ex.dir/face_landmark_detection_ex.cpp.o: In function `main':
face_landmark_detection_ex.cpp:(.text.startup+0x26db): undefined reference to `cv::putText(cv::_InputOutputArray const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, cv::Point_<int>, int, double, cv::Scalar_<double>, int, int, bool)'
face_landmark_detection_ex.cpp:(.text.startup+0x29d6): undefined reference to `cv::fastFree(void*)'
face_landmark_detection_ex.cpp:(.text.startup+0x2c13): undefined reference to `cv::Mat::updateContinuityFlag()'
face_landmark_detection_ex.cpp:(.text.startup+0x2f17): undefined reference to `cv::Mat::deallocate()'
face_landmark_detection_ex.cpp:(.text.startup+0x35f8): undefined reference to `cv::error(int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, char const*, char const*, int)'
collect2: error: ld returned 1 exit status
CMakeFiles/face_landmark_detection_ex.dir/build.make:104: recipe for target 'face_landmark_detection_ex' failed
make[2]: *** [face_landmark_detection_ex] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/face_landmark_detection_ex.dir/all' failed
make[1]: *** [CMakeFiles/face_landmark_detection_ex.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2

I know this is a standard problem and I googled a lot to figure this out but I cannot resolve it on my own :(

[goldbattle]

It looks like you are not properly linking the opencv shared library. You need to make sure you add the library .so files in addition to the header files. https://docs.opencv.org/3.1.0/db/df5/tutorial_linux_gcc_cmake.html

I would recommend just using the native functions in dlib for visualization: https://stackoverflow.com/a/37563733 http://dlib.net/imaging.html

[ManuelTS]

Your first link made me realize that instead of

  add_ex_link(face_landmark_detection_ex ${EXTRA_LIBS} ${OpenCV_LIBS})

I had

  add_ex_link(face_landmark_detection_ex ${EXTRA_LIBS})

in my CMakeLists.txt. I'm almost embarrassed how trivial this error was. Thank you very much!