http://docs.opencv.org/trunk/modules/... What does the grid_x and grid_y means ? Suppose I use radius = 2. How can I calculate the neighbor, gird_x, and grid_y?

I successfully compiled and installed the 3.1.0-dev

pkg-config

$pkg-config --libs --cflags opencv
-I/usr/local/include/opencv -I/usr/local/include -L/usr/local/lib -lopencv_shape -lopencv_stitching -lopencv_objdetect -lopencv_superres -lopencv_videostab -lopencv_calib3d -lopencv_features2d -lopencv_highgui -lopencv_videoio -lopencv_imgcodecs -lopencv_video -lopencv_photo -lopencv_ml -lopencv_imgproc -lopencv_flann -lopencv_core

main.cpp:

#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/highgui/highgui.hpp"

using namespace std;
using namespace cv;

Mat src; Mat dst;
char window_name1[] = "Unprocessed Image";
char window_name2[] = "Processed Image";

int main( int argc, char** argv )
{
    /// Load the source image
    src = imread( argv[1], 1 );

    namedWindow( window_name1, WINDOW_AUTOSIZE );
    imshow("Unprocessed Image",src);

    dst = src.clone();
    GaussianBlur( src, dst, Size( 15, 15 ), 0, 0 );

    namedWindow( window_name2, WINDOW_AUTOSIZE );
    imshow("Processed Image",dst);

    waitKey();
    return 0;
}

I can compile and execute main.o by using very simple argument

g++ main.cpp `pkg-config --cflags --libs opencv` -o main.o
./main.o image.jpg

But when I put my include and libs directory to Eclipe

14:49:49 **** Incremental Build of configuration Debug for project aiyara ****
make all 
Building target: aiyara
Invoking: MacOS X C++ Linker
g++ -L/usr/local/lib -o "aiyara"  ./src/aiyara.o   -lopencv_core -lopencv_highgui -lopencv_imgproc
Undefined symbols for architecture x86_64:
  "cv::imread(cv::String const&, int)", referenced from:
      _main in aiyara.o
ld: symbol(s) not found for architecture x86_64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make: *** [aiyara] Error 1

14:49:49 Build Finished (took 126ms)

I am trying to develope my toy project by using OpenCV and Qt
Here is my setup file and simple source from internet.
I install OpenCV3 by brew install opencv3

myproject.pro:

TEMPLATE = app
CONFIG += console c++11 link_pkgconfig
CONFIG -= app_bundle
CONFIG -= qt
QT_CONFIG -= no-pkg-config
PKGCONFIG += opencv

SOURCES += main.cpp
LIBS += -L/usr/local/lib \
     -lopencv_core \
     -lopencv_imgproc \
     -lopencv_features2d\
     -lopencv_highgui

main.cpp:

#include "opencv2/imgproc/imgproc.hpp"
#include "opencv2/highgui/highgui.hpp"

using namespace std;
using namespace cv;

Mat src; Mat dst;
char window_name1[] = "Unprocessed Image";
char window_name2[] = "Processed Image";

int main( int argc, char** argv )
{
    /// Load the source image
    src = imread( argv[1], 1 );

    namedWindow( window_name1, WINDOW_AUTOSIZE );
    imshow("Unprocessed Image",src);

    dst = src.clone();
    GaussianBlur( src, dst, Size( 15, 15 ), 0, 0 );

    namedWindow( window_name2, WINDOW_AUTOSIZE );
    imshow("Processed Image",dst);

    waitKey();
    return 0;
}

Here is my env variables:

FYI: Either PKG_CONFIG_PATH: /usr/local/opt/opencv3/lib/pkgconfig or
/usr/local/Cellar/opencv3/3.1.0_4/lib/pkgconfig are not work

Problem:

18:21:52: Running steps for project aiyara...
18:21:52: Configuration unchanged, skipping qmake step.
18:21:52: Starting: "/usr/bin/make" 
/Users/el/Qt/5.7/clang_64/bin/qmake -spec macx-clang CONFIG+=x86_64 CONFIG+=qml_debug CONFIG+=force_debug_info CONFIG+=separate_debug_info -o Makefile ../aiyara/aiyara.pro
Project ERROR: opencv2 development package not found
make: *** [Makefile] Error 3
18:21:52: The process "/usr/bin/make" exited with code 2.
Error while building/deploying project aiyara (kit: Desktop Qt 5.7.0 clang 64bit)
When executing step "Make"
18:21:52: Elapsed time: 00:00.

References:
http://stackoverflow.com/questions/17...
https://www.learnopencv.com/install-o...
http://stackoverflow.com/questions/31...

I tried (0,1), (1,2), and (0,0) as last comment. But I stop to running the camID since I am so afraid I am going to break something since I have no idea about the returned error at all. I will follow you reply (0,1), (0,2), ...(0,10).

@Tetragramm Thank you for your comments. I had avoided USB overloading. That's why I can run 2 processces of webcam viewer simulataneously. and for the 2nd comment I had tried open calling same index twice. I called zero twice. Then got This
libv4l2: error setting pixformat: Device or resource busy VIDEOIO ERROR: libv4l unable to ioctl S_FMT libv4l2: error setting pixformat: Device or resource busy libv4l1: error setting pixformat: Device or resource busy VIDEOIO ERROR: libv4l unable to ioctl VIDIOCSPICT

Thank you Eduardo. I will follow your given links and then back again. :]

I am struck with face spoofing. Right now I can setup stereo webcam and take images from left and right webcam simulatenously(at least, I am practicing concurrent programming to do this job).
I can extract the eyes, nose, mouth, face center correctly. I had search on face spoofing publication. I usually found texture, skin, or something similar analysis on the single picture. And the prominent technique is 3D face scanner by maser
Question:
I do not know how to let computer distinguish between single picture(spoofing case) and real face(human face).
I heard that is not difficult, but I could not figure out.
Please guide me some clue then I will report back what I found.

VIDEOIO ERROR: V4L: index 2 is not correct!

When I reach my computer I will inform you back here. Put index 1 and 2 is never come up in my mind.

I successfully view 2 webcams parallely by using the multiple processes. As code below. I use dedicate USB bus each for single webcam.

import cv2
from multiprocessing import Process
def f(camID):
        vc = cv2.VideoCapture(camID)
        camIDstr = str(camID)
        cv2.namedWindow("preview" + camIDstr)
        if vc.isOpened(): # try to get the first frame
                rval,  frame  = vc.read()
        else:
                rval  = False

        while rval:
                cv2.imshow("preview" + camIDstr, frame)
                rval, frame = vc.read()
                key = cv2.waitKey(20)
                if key == 27: # exit on ESC
                        break
        cv2.destroyWindow("preview" + camIDstr)

if __name__ == '__main__':
        #f(0)
        '''
        t0 = threading.Thread(target=f, args= (0,))
        t0.start()
        t1 = threading.Thread(target=f, args= (1,))
        t1.start()
        '''
        p0 = Process(target=f, args=(0,))
        p0.start()
        p1 = Process(target=f, args=(1,))
        p1.start()

And here is my simplistic thread version.

import cv2
#from multiprocessing import Process
import threading
def f(camID):
        vc = cv2.VideoCapture(camID)
        camIDstr = str(camID)
        cv2.namedWindow("preview" + camIDstr)
        if vc.isOpened(): # try to get the first frame
                rval,  frame  = vc.read()
        else:
                rval  = False

        while rval:
                cv2.imshow("preview" + camIDstr, frame)
                rval, frame = vc.read()
                key = cv2.waitKey(20)
                if key == 27: # exit on ESC
                        break
        cv2.destroyWindow("preview" + camIDstr)

if __name__ == '__main__':
        #f(0)
        t0 = threading.Thread(target=f, args= (0,))
        t0.start()
        t1 = threading.Thread(target=f, args= (1,))
        t1.start()

Update :

For further information. When I cut all the function and let my main function just this.

   for i in range(0,11):
      vc = cv2.VideoCapture(i)
      if vc.isOpened():
         print i

The available indexes are 0 and 1.

    0
    1
    VIDEOIO ERROR: V4L: index 2 is not correct!
    VIDEOIO ERROR: V4L: index 3 is not correct!
    VIDEOIO ERROR: V4L: index 4 is not correct!
    VIDEOIO ERROR: V4L: index 5 is not correct!
    VIDEOIO ERROR: V4L: index 6 is not correct!
    VIDEOIO ERROR: V4L: index 7 is not correct!
    VIDEOIO ERROR: V4L: index 8 is not correct!
    VIDEOIO ERROR: V4L: index 9 is not correct!
    VIDEOIO ERROR: V4L: index 10 is not correct!

Problem :
Program retunrs 1 line with error message as follows.

VIDEOIO ERROR: V4L: index 1 is not correct!

Question :
How come V4L says my index is incorrect?
If they are not 0 or 1. What number it will be?

New Question :
Is it possible to create stereo camera by using thread?
Or I need process?

I am successfully run example program in opencv/build/bin
for example opencv_test_core cpp-example-opencv_version. My version is 3.0-dev.
It looks find. But when I try manually without make . I got error. The error comes when I remove libopencv2.4 which is Debian package from my computer and then compiled and installed latest version from github.

 sarit@cmd-center:~/opencv/samples/cpp$ g++ -I/usr/local/include -L/usr/local/lib -lopencv_core opencv_version.cpp 
/tmp/cciphMKR.o: In function `main':
opencv_version.cpp:(.text+0x69): undefined reference to `cv::CommandLineParser::CommandLineParser(int, char const* const*, cv::String const&)'
opencv_version.cpp:(.text+0x99): undefined reference to `cv::CommandLineParser::has(cv::String const&) const'
opencv_version.cpp:(.text+0xb7): undefined reference to `cv::CommandLineParser::printMessage() const'
opencv_version.cpp:(.text+0xc8): undefined reference to `cv::CommandLineParser::check() const'
opencv_version.cpp:(.text+0xdb): undefined reference to `cv::CommandLineParser::printErrors() const'
opencv_version.cpp:(.text+0x104): undefined reference to `cv::CommandLineParser::has(cv::String const&) const'
opencv_version.cpp:(.text+0x11b): undefined reference to `cv::getBuildInformation()'
opencv_version.cpp:(.text+0x171): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
opencv_version.cpp:(.text+0x19e): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
opencv_version.cpp:(.text+0x1da): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
/tmp/cciphMKR.o: In function `cv::String::String(char const*)':
opencv_version.cpp:(.text._ZN2cv6StringC2EPKc[_ZN2cv6StringC5EPKc]+0x4f): undefined reference to `cv::String::allocate(unsigned long)'
/tmp/cciphMKR.o: In function `cv::String::~String()':
opencv_version.cpp:(.text._ZN2cv6StringD2Ev[_ZN2cv6StringD5Ev]+0x14): undefined reference to `cv::String::deallocate()'
collect2: error: ld returned 1 exit status

Here is my pkg-config

sarit@cmd-center:~/opencv/samples/cpp$ pkg-config --cflags --libs opencv
-I/usr/local/include/opencv -I/usr/local/include -L/usr/local/lib -lopencv_cudabgsegm -lopencv_cudaobjdetect -lopencv_cudastereo -lopencv_shape -lopencv_stitching -lopencv_cudafeatures2d -lopencv_superres -lopencv_cudacodec -lopencv_videostab -lopencv_cudaoptflow -lopencv_cudalegacy -lopencv_calib3d -lopencv_features2d -lopencv_objdetect -lopencv_highgui -lopencv_videoio -lopencv_photo -lopencv_imgcodecs -lopencv_cudawarping -lopencv_cudaimgproc -lopencv_cudafilters -lopencv_video -lopencv_ml -lopencv_imgproc -lopencv_flann -lopencv_cudaarithm -lopencv_core -lopencv_hal -lopencv_cudev

Installed library. But I don't know why they end with *.a

sarit@cmd-center:~/opencv/samples/cpp$ ls /usr/local/lib
cmake                libflandmark.so.1       libopencv_cudabgsegm.a      libopencv_cudalegacy.a     libopencv_cudev.a       libopencv_imgcodecs.a  libopencv_shape.a      libopencv_videoio.a    python3.4
libclandmark.so      libflandmark.so.1.5     libopencv_cudacodec.a       libopencv_cudaobjdetect.a  libopencv_features2d.a  libopencv_imgproc.a    libopencv_stitching.a  libopencv_videostab.a
libclandmark.so.1    libopencv_calib3d.a     libopencv_cudafeatures2d.a  libopencv_cudaoptflow.a    libopencv_flann.a       libopencv_ml.a         libopencv_superres.a   node_modules
libclandmark.so.1.5  libopencv_core.a        libopencv_cudafilters.a     libopencv_cudastereo.a     libopencv_hal.a         libopencv_objdetect.a  libopencv_ts.a         pkgconfig
libflandmark.so      libopencv_cudaarithm.a  libopencv_cudaimgproc.a     libopencv_cudawarping.a    libopencv_highgui.a     libopencv_photo.a      libopencv_video.a      python2.7

I am not sure if I create a symbolic link and end then by .so for example
$ ln -s libopencv_core.a libopencv_core.so
Will harm my system or not. I am so afriad I am going to break something then I ask you experts in forum here.

OS : GNU/Linux Debian Jessie
Question :
Is this all about linking problem?
Can I simply use ldconfig to get the job done?

I am trying to repeat example to get the current version of opencv in my system. Because I just compiled and installed 3.0-dev.
The compilation seems find. I can run the example.

sarit@cmd-center:~/opencv/build/bin$ ./cpp-example-opencv_version 
Welcome to OpenCV 3.0.0-dev

Howerver, when I try repeat it. I got the error.

sarit@cmd-center:~/opencv/samples/cpp$ g++ `pkg-config --cflags --libs opencv` opencv_version.cpp 
/tmp/ccjiJvDa.o: In function `main':
opencv_version.cpp:(.text+0x69): undefined reference to `cv::CommandLineParser::CommandLineParser(int, char const* const*, cv::String const&)'
opencv_version.cpp:(.text+0x99): undefined reference to `cv::CommandLineParser::has(cv::String const&) const'
opencv_version.cpp:(.text+0xb7): undefined reference to `cv::CommandLineParser::printMessage() const'
opencv_version.cpp:(.text+0xc8): undefined reference to `cv::CommandLineParser::check() const'
opencv_version.cpp:(.text+0xdb): undefined reference to `cv::CommandLineParser::printErrors() const'
opencv_version.cpp:(.text+0x104): undefined reference to `cv::CommandLineParser::has(cv::String const&) const'
opencv_version.cpp:(.text+0x11b): undefined reference to `cv::getBuildInformation()'
opencv_version.cpp:(.text+0x171): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
opencv_version.cpp:(.text+0x19e): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
opencv_version.cpp:(.text+0x1da): undefined reference to `cv::CommandLineParser::~CommandLineParser()'
/tmp/ccjiJvDa.o: In function `cv::String::String(char const*)':
opencv_version.cpp:(.text._ZN2cv6StringC2EPKc[_ZN2cv6StringC5EPKc]+0x4f): undefined reference to `cv::String::allocate(unsigned long)'
/tmp/ccjiJvDa.o: In function `cv::String::~String()':
opencv_version.cpp:(.text._ZN2cv6StringD2Ev[_ZN2cv6StringD5Ev]+0x14): undefined reference to `cv::String::deallocate()'
collect2: error: ld returned 1 exit status

Here is my libs and flags

sarit@cmd-center:~/opencv/build/bin$ pkg-config --libs --cflags opencv
-I/usr/local/include/opencv -I/usr/local/include -L/usr/local/lib -lopencv_cudabgsegm -lopencv_cudaobjdetect -lopencv_cudastereo -lopencv_stitching -lopencv_cudafeatures2d -lopencv_superres -lopencv_cudacodec -lopencv_videostab -lopencv_cudaoptflow -lopencv_cudalegacy -lopencv_cudawarping -lopencv_aruco -lopencv_bgsegm -lopencv_bioinspired -lopencv_ccalib -lopencv_dnn -lopencv_dpm -lopencv_fuzzy -lopencv_line_descriptor -lopencv_optflow -lopencv_reg -lopencv_saliency -lopencv_stereo -lopencv_structured_light -lopencv_rgbd -lopencv_surface_matching -lopencv_tracking -lopencv_datasets -lopencv_text -lopencv_face -lopencv_xfeatures2d -lopencv_shape -lopencv_video -lopencv_ximgproc -lopencv_calib3d -lopencv_features2d -lopencv_flann -lopencv_xobjdetect -lopencv_objdetect -lopencv_ml -lopencv_xphoto -lopencv_highgui -lopencv_videoio -lopencv_photo -lopencv_imgcodecs -lopencv_cudaimgproc -lopencv_cudafilters -lopencv_imgproc -lopencv_cudaarithm -lopencv_core -lopencv_hal -lopencv_cudev

Question :
How to fix this error?

For this kind of problem. The forum has a ready made solution here.
http://answers.opencv.org/question/24...

Installation on Linux has 2 links. They are
http://docs.opencv.org/2.4/doc/tutorials/introduction/linux_install/linux_install.html http://docs.opencv.org/master/d7/d9f/tutorial_linux_install.html#gsc.tab=0 I just found the later one. So I have to compile opencv_contrib one by one.
My questions are :
1. Is it safe to follow 2nd link ?
2. Will make install just simply overwrite previous compilations?
3. Do I need to do make uninstall first?
It would be nice for first link let me know how to install opencv_contrib too.

Hi
I am practicing VideoCapture and VideoWriter. I am now following
http://stackoverflow.com/questions/13623394/how-to-write-video-file-in-opencv-2-4-3 Which is supposed to be a brief answer.
However, when I am doing it It got.

$ ./simpleRec.out 0 a.avi
HIGHGUI ERROR: V4L2: Unable to get property <unknown property string>(5) - Invalid argument
HIGHGUI ERROR: V4L2: Unable to get property <unknown property string>(6) - Invalid argument
!!! Output video could not be opened

Here is my code :

#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>

#include <iostream>
#include <iomanip>
#include <string>
#include <stdio.h>

using namespace std;
using namespace cv;

int main (int argc, const char** argv){
  // Load input video
  cv::VideoCapture input_cap(atoi( (argv[1]) ) );
  if (!input_cap.isOpened())
    {
      std::cout << "!!! Input video could not be opened" << std::endl;
      return -1;
    }

  // Setup output video
  //double fps = input_cap.get(CV_CAP_PROP_FPS);
  int width = input_cap.get(CV_CAP_PROP_FRAME_WIDTH);
  int height = input_cap.get(CV_CAP_PROP_FRAME_HEIGHT);
  cv::VideoWriter output_cap(argv[2],
                 CV_FOURCC('M','J','P','G'),
                 20,
                 cv::Size(width,height),
                 true);

  if (!output_cap.isOpened())
    {
      std::cout << "!!! Output video could not be opened" << std::endl;
      return 0;
    }


  // Loop to read from input and write to output
  cv::Mat frame;

  while (true)
    {       
      if (!input_cap.read(frame))             
        break;

      output_cap.write(frame);
    }

  input_cap.release();
  output_cap.release();
}

Update :
I found it is a bug in issue. http://code.opencv.org/issues/1895 it is fps is -1
I had tried 10,20,30 but it still does not work with reply !!! Output video could not be opened
I changed VideoWriter initialization a bit. From

cv::VideoWriter output_cap(argv[2], 
                 input_cap.get(CV_CAP_PROP_FOURCC),
                 input_cap.get(CV_CAP_PROP_FPS),
                 cv::Size(input_cap.get(CV_CAP_PROP_FRAME_WIDTH),
                      input_cap.get(CV_CAP_PROP_FRAME_HEIGHT)),
                 true);

to

  //double fps = input_cap.get(CV_CAP_PROP_FPS);
  int width = input_cap.get(CV_CAP_PROP_FRAME_WIDTH);
  int height = input_cap.get(CV_CAP_PROP_FRAME_HEIGHT);
  cv::VideoWriter output_cap(argv[2],
                 CV_FOURCC('M','J','P','G'),
                 20,
                 cv::Size(width,height),
                 true);

I have to comment fps out otherwise it will thrown an error HIGHGUI ERROR: V4L2: Unable to get property <unknown property string>

Question :
1. What is wrong in my code?
2. How do I solve it?

Hi
I am in the early state of my project. I want to proof very simple concept on hardware performance.
Since I have a plenty of USB bus then I attach them individually to avoid bandwidth problem. I do not want to use MPEG because it consume my reading time and it slow down my program. Here is my bus.

$ lsusb
Bus 006 Device 005: ID 0c45:6340 Microdia 
Bus 003 Device 002: ID 0c45:62f1 Microdia

Here is my simple code which is has a delay.

import cv2

cv2.namedWindow("preview")
cv2.namedWindow("preview2")
vc  = cv2.VideoCapture(0)
vc2 = cv2.VideoCapture(1)

if vc.isOpened() and vc2.isOpened(): # try to get the first frame
        rval,  frame  = vc.read()
        rval2, frame2 = vc2.read()
else:
        rval  = False
        rval2 = False

while rval and rval2:
    cv2.imshow("preview", frame)
    cv2.imshow("preview2", frame2)
    rval, frame = vc.read()
    rval2, frame2 =vc2.read()
    key = cv2.waitKey(20)
    if key == 27: # exit on ESC
        break
cv2.destroyWindow("preview")
cv2.destroyWindow("preview2")

My attempt on python thread is as follow :

import cv2
import threading
def f(camID):
        vc = cv2.VideoCapture(camID)
        cv2.namedWindow("preview" + str(camID))
        if vc.isOpened(): # try to get the first frame
                rval,  frame  = vc.read()
        else:
                rval  = False

        while rval:
                cv2.imshow("preview", frame)
                rval, frame = vc.read()
                key = cv2.waitKey(20)
                if key == 27: # exit on ESC
                        break
        cv2.destroyWindow("preview"+str(camID))

if __name__ == '__main__':
        #f(0)
        t0 = threading.Thread(target=f, args= (0,))
        t0.start()
        t1 = threading.Thread(target=f, args= (1,))
        t1.start()

When I use t0 or t1 by comment the others off. It works just fine with the same response. Althought webcams are same model, but older than one 5 years.

My error :

HIGHGUI ERROR: V4L: index 1 is not correct!

(preview1:5154): Gtk-WARNING **: gtk_disable_setlocale() must be called before gtk_init()

(preview1:5154): Gdk-CRITICAL **: IA__gdk_draw_rectangle: assertion 'GDK_IS_GC (gc)' failed
[xcb] Unknown request in queue while dequeuing
[xcb] Most likely this is a multi-threaded client and XInitThreads has not been called
[xcb] Aborting, sorry about that.
python: ../../src/xcb_io.c:179: dequeue_pending_request: Assertion `!xcb_xlib_unknown_req_in_deq' failed.
Aborted

My question :
. 1. How to let program do a computing on captured image from webcam synchronously?
2. How to display the picture in the window without delay?
3. How to use python thread properly in my example?

Thank you berak. sarit@cmd-center:~/clandmark/examples$ g++ pkg-config --cflags opencv static_input.cpp -o static_input.out pkg-config --libs opencv -L/usr/local/include -lclandmark -lflandmark The compilation is done, but got error while loading shared libraries: libclandmark.so.1: cannot open shared object file: No such file or directory. $ ldconfig -v will solve library problem

I am away from my computer. When I reach it I will update to you as fast as I can.

Thank you for your reply. Your given links are my new knowledge for today :)

Hi.
OpenCV is very fasinating. Currently, I want to play aroung with my game it is a FPS with maze game.
How to apply Odometry to help me figure out the maze by advancing the odometry to it?
I have seen odometry in action with online webcam and video file. However, they are not my case.
I want real time I/O from game.

@berak. Thank you so much! I was hitting the book about OO, CMake, and compilation with GCC. What a nightmare without your help.

@berak. Thank you.

I will study CMake later. For now I will use everything in the same directory.
I git clone the repo and put the example which I copy the website, name it as 'sarit.cpp'.
My questions is :
1. Do you think the developer confuses his own function? If he does. How can I fix it?

$ g++ sarit.cpp `pkg-config --cflags --libs opencv`
sarit.cpp: In function ‘int main(int, char**)’:
sarit.cpp:22:57: error: cannot convert ‘float*’ to ‘double*’ for argument ‘4’ to ‘int flandmark_detect(IplImage*, int*, FLANDMARK_Model*, double*, int*)’
   flandmark_detect(img_grayscale, bbox, model, landmarks);

And here is is my pkg-config

$ pkg-config --libs --cflags opencv
-I/usr/include/opencv /usr/lib/x86_64-linux-gnu/libopencv_calib3d.so -lopencv_calib3d /usr/lib/x86_64-linux-gnu/libopencv_contrib.so -lopencv_contrib /usr/lib/x86_64-linux-gnu/libopencv_core.so -lopencv_core /usr/lib/x86_64-linux-gnu/libopencv_features2d.so -lopencv_features2d /usr/lib/x86_64-linux-gnu/libopencv_flann.so -lopencv_flann /usr/lib/x86_64-linux-gnu/libopencv_gpu.so -lopencv_gpu /usr/lib/x86_64-linux-gnu/libopencv_highgui.so -lopencv_highgui /usr/lib/x86_64-linux-gnu/libopencv_imgproc.so -lopencv_imgproc /usr/lib/x86_64-linux-gnu/libopencv_legacy.so -lopencv_legacy /usr/lib/x86_64-linux-gnu/libopencv_ml.so -lopencv_ml /usr/lib/x86_64-linux-gnu/libopencv_objdetect.so -lopencv_objdetect /usr/lib/x86_64-linux-gnu/libopencv_ocl.so -lopencv_ocl /usr/lib/x86_64-linux-gnu/libopencv_photo.so -lopencv_photo /usr/lib/x86_64-linux-gnu/libopencv_stitching.so -lopencv_stitching /usr/lib/x86_64-linux-gnu/libopencv_superres.so -lopencv_superres /usr/lib/x86_64-linux-gnu/libopencv_ts.so -lopencv_ts /usr/lib/x86_64-linux-gnu/libopencv_video.so -lopencv_video /usr/lib/x86_64-linux-gnu/libopencv_videostab.so -lopencv_videostab

My file :

#include "flandmark_detector.h"
#include "cv.h"
#include "opencv/highgui.h"

int main(int argc, char * argv[])
{
  // load flandmark model structure and initialize
  FLANDMARK_Model * model = flandmark_init("flandmark_model.dat");

  // load input image
  IplImage *img = cvLoadImage("photo.jpg");

  // convert image to grayscale
  IplImage *img_grayscale = cvCreateImage(cvSize(img->width, img->height), IPL_DEPTH_8U, 1);
  cvCvtColor(img, img_grayscale, CV_BGR2GRAY);

  // bbox with detected face (format: top_left_col top_left_row bottom_right_col bottom_right_row)
  int bbox[] = {72, 72, 183, 183};

  // detect facial landmarks (output are x, y coordinates of detected landmarks)
  float * landmarks = (float*)malloc(2*model->data.options.M*sizeof(float));
  flandmark_detect(img_grayscale, bbox, model, landmarks);
}

@berak. Thank you berak. Feels like something exciting is waiting for me :o

@berak. Right now, I can detect eyes and mouth of person face. Suppose I use simple affine tranform http://docs.opencv.org/doc/tutorials/... to transform face from first picture to second image by mapping a pair of eyes and mouth. Is the result going to be a wrong one? If it wrong. What's wrong?

@berak. I though your eye answer will rely on face geometry ratio or length of them. I am now observing the variation of them. Let me try and return to forum again. Do you think mouth is impact by face area? Since my Neo detected face has very near edge to the mouth.

@berak. Hello berak. Do you have any idea? Any keyword I can find out on my own?

Hi
I'm using OpenCV 2.4.10. I can detect pair eyes and mouth for alignment with fair accuracy rate. Here is my situation. https://www.youtube.com/watch?v=0ZY5U...

My problems are :
1. Eyes detection return false eye. I got 3rd eye in nearby the correct one.
2. Neo's mouth has been dropped on many frames.

Any help would be appreciated.
Sarit

Hi
My goal is create a program that focus on human face and screen out the duplicate picture.
Now I am stuck with eyes and mouth extraction. They are feature that I use to transform a picture to be a mugshot. I understand it will has a distortion, but that's ok.
I want to know.
1. Do you expert do a transformation from detected face to be frontal face?
2. Does Machine Learning require that?

Thanks Sarit