This is machine translation

Translated by Microsoft
Mouseover text to see original. Click the button below to return to the English version of the page.

Note: This page has been translated by MathWorks. Click here to see
To view all translated materials including this page, select Country from the country navigator on the bottom of this page.

Face Detection Using Parrot FPV Drones

This example shows how to use a Parrot® drone to automatically detect human faces captured by the drone's FPV camera.


Use the MATLAB® Support Package for Parrot® Drones to control the drone and capture images from the FPV camera. A cascade object detector uses the Viola-Jones detection algorithm and a trained classification model for face detection. By default, the detector is configured to detect faces, but it can be used to detect other types of objects.

Required MathWorks Products


  • MATLAB® Support Package for Parrot® Drones

  • Computer Vision System Toolbox™


Complete Getting Started with MATLAB® Support Package for Parrot® Drones.

Required Hardware

To run this example you need:

  • A fully charged Parrot FPV drone

  • A computer with a WiFi connection

Task 1 — Create a Connection to the Parrot Drone

Create a parrot object.

parrotObj = parrot;

Task 2 — Create a Cascade Object Detector Instance

Create an instance of the cascade object detector to detect faces using the Viola-Jones algorithm.

detector = vision.CascadeObjectDetector;

Task 3 — Activate FPV Camera

Start the drone flight to activate the FPV camera. Move the drone up to sufficient height to capture faces.


Task 4 — Create a Connection to the Drone's FPV Camera

Use the parrot object from Task 1 to create the connection to the drone's FPV camera.

camObj = camera(parrotObj,'FPV');

Task 5 — Detect Faces While Traversing a Square Path

Detect faces while the drone moves forward for 2 seconds along the edge of a square path.

1 Move the drone forward for the default duration of 0.5 seconds for each forward step, ensuring a nonblocking behaviour. This enables the drone to capture the image and detect faces while in motion.

2 Capture a single frame from the drone's FPV camera.

3 Input the image to the detector, which returns bounding boxes containing the detected objects. The detector performs multiscale object detection on the input image.

4 Display the image with bounding boxes around faces and the title displaying the number of faces detected.

5 Turn the drone by π/2 radians at each square vertex.

tOuter= tic;
while(toc(tOuter)<=30 && parrotObj.BatteryLevel>20)
   tInner = tic;
   % Keep moving the drone for 2 seconds along each square path edge
       moveforward(parrotObj);                                                       % Move the drone forward for default time of 0.5 seconds (nonblocking behaviour)
       picture = snapshot(camObj);                                                   % Capture image from drone's FPV camera
       bbox = detector(picture);                                                     % Detect faces in image
       videoOut = insertShape(picture,'Rectangle',bbox,'Color','r','LineWidth',3);   % Insert bounding box into image
       imshow(picture);                                                              % Show the picture
       title(sprintf(' %d face(s) detected ',size(bbox,1)));
   turn(parrotObj,deg2rad(90));                                                      % Turn the drone by pi/2 radians

6 Execute steps 1–5 for 30 seconds.

This example show two faces detected by the drone's FPV camera.

Task 6 — Land the Drone

Land the drone.


Task 7 — Clean Up

When finished clear the connection to the Parrot drone and the FPV camera.

clear parrotObj;
clear camObj;