[dependancy]

• opencv : http://docs.opencv.org/doc/tutorials/introduction/linux_install/linux_install.html

[example program]

• main function in CV_PJ.cpp

[API usage]

• first, load models for prediction

  CV_PJ_Face::CV_PJ_LoadModel()

    (be sure that data file path is defined in this function)
  

• call prediction function

  CV_PJ_Face::CV_PJ_Detect(const cv::Mat& inputImg, cv::Rect& faceBbox, std::vectorcv::Point2d& landmarkPoints)

  <input>
  
    inputImg : 1 or 3 channel input image to predict the facial landmarks
  
  <output>
  
    faceBbox : output rect object of face bounding box (integer coordinates)
  
    landmarkPoints : output vector of facial landmark points (double precision)