Hands-On Robotics Programming with C++

Eye detection using haarcascade_eye

To test the haarcascade_eye output, observe the following steps:

1. Load this classifier inside our program:

CascadeClassifier eyeDetector("haarcascade_eye.xml");

2. To detect the eyes, we need to find the face region (region of interest) in the image (video feed). Inside the face-detection for loop, we will create a Mat variable called faceroi. videofeed(face[f]), which will find faces in videofeed and store them inside the faceroi variable:

Mat faceroi = videofeed(face[f]);

3. Create a vector of the Rect type, called eyes, and then use the detectMultiScale function to detect the eye region:

vector<Rect> eyes;eyeDetector.detectMultiScale(faceroi, eyes, 1.3, 5, 0 |CASCADE_SCALE_IMAGE,Size(30, 30));

In the ...