/HalloPy

a hand controlled Tello

Primary LanguagePythonMIT LicenseMIT

Hallo - a hand controlled Tello

DJI Tello controller using hand gestures python app

Docker installation in 3 steps

Tested on ubuntu 19.10

step 1

git clone https://github.com/GalBrandwine/HalloPy
cd HalloPy
git submodule update --init --recursive

step 2

docker build --network=host  --tag hallopy:1.3 .

(make sure you have docker installed.)

step 3

(Make sure you're connected to the TELLO wifi)

xhost +  && docker run --rm -it --net=host --ipc=host -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix --privileged -v /dev/video0:/dev/video0 --name hallopy hallopy:1.3

For flags explanation: running docker with gui

Non docker users

Environment:

  • ubuntu 19.10
  • python 3.7

Dependency libraries installation:

  • pip install av==6.1.2
  • pip install opencv-python

Run

From directory /HalloPy/:

python ./hallopy/hallo.py

Controller Usage

  1. Make sure you have all dependency libraries. for great openCV installation tutorial refer to: https://www.pyimagesearch.com/2016/10/24/ubuntu-16-04-how-to-install-opencv/

  2. Turn on Tello drone and connect to it's wifi

    1. there's no setup.py yet, so in order to run this project, open the project in an IDE and run: hallo.py

  3. Application usage:

     press 'b' - to detect palm ( depend on the environment lights, a Thresh Hold slider tuning may be needed)
 center the detected palm at the center of the detection_frame FOR 5 seconds 
 pressing b again will reset calibration.

    After center-circle become GREEN (meaning we are now calibrated):

    press 't' - to take off ( a 3 second hold-up, until drone is in the air)
  after landing - if program is calibrated, press t again to take-off
  
press 'c' - to toggle drone's control between key-board and detected-palm
  toggling back to keyboard will force drone to hover in the same place
  
press 'l' - to land ( at any time)

press 'x' - to adjust background' threshold - for thicker palm recognition
press 'z' - to adjust background' threshold - for thinner palm recognition

press 'esc' to exit program ( only after drone landed)

  4. Video explaining hands movements for controlling the drone can be found here, and here

Thanks to