DJITelloPy: A new python wrapper for Tello

[Damià Fuentes]

I developed a python library for the Tello drone using the official SDK. I implemented all the functionalities the official SDK has, including the live video. Feel free to try it out.
damiafuentes/DJITelloPy

 

[FlandersNed]

Are you able to dump the video to a file?

 

[Cyke]

I developed a python library for the Tello drone using the official SDK. I implemented all the functionalities the official SDK has, including the live video. Feel free to try it out.
damiafuentes/DJITelloPy

You have implemented the 1.3 SDK version but now there's the 2.0 SDK version !
Could you update your Python library ? ☺️

 

[rojo-amigo]

The SDK 2.0 version will only work with the Tello Edu NOT the Tello.

 

[Cyke]

The SDK 2.0 version will only work with the Tello Edu NOT the Tello.

Could you pinpoint the differences between the Tello SDK and the Tello Edu SDK?

 

[Droner]

Hi,
Great job!

Can your library support shooting stills images and receiving them?

Thanks!

 

[rojo-amigo]

The EDU sdk is at
https://dl-cdn.ryzerobotics.com/downloads/Tello/Tello SDK 2.0 User Guide.pdf
The new commands are:
stop (hovers in air)
ap ssid pass
read commands
-sdk?
-sn?
mission pad commands:
- mon
- moff
- mdirection x
- go x y z speed mid
- curve x1 y1 z1 x2 y2 z2 speed mid
- jump x y z speed yaw mid1 mid2

the read commands that have been removed? are
height?
temp?
attitude?
baro?
acceleration?
tof?

 

[Damià Fuentes]

Hi,
Great job!

Can your library support shooting stills images and receiving them?

Thanks!

Hey Droner. The official SDK does not provide a way to retrieve photos. It only allow us to retrieve photos. But you could retrieve the video stream (which is implemented and tested in the library) and extract frames from there. Then you use the frames as single images.

 

[Damià Fuentes]

Are you able to dump the video to a file?

You just need to retrieve the video stream and go frame by frame saving it in a variable. Then export it to a video file.

 

[nahuel]

This is awesome! Tello and its users impress me more every day. Good job man.

I *so* want to test this but I have a weird problem. My computer won't connect to Tello's wifi (Arch Linux using NetworkManager). Doesn't give me a reason, it just tries for a few seconds then it says it failed :/

 

[nahuel]

Ok, this is so epic. Thank you so much for this, I'm gonna have a lot of fun. It works super well. I can hook this up to OpenCV or something... awesome!

 

[Damià Fuentes]

Ok, this is so epic. Thank you so much for this, I'm gonna have a lot of fun. It works super well. I can hook this up to OpenCV or something... awesome!

View attachment 2215

You are very welcome Nahuel! I actually did an Augmented Reality project for the university using OpenCV, markers and the DJI Tello. That is why I developed this library.

 

[nahuel]

You are very welcome Nahuel! I actually did an Augmented Reality project for the university using OpenCV, markers and the DJI Tello. That is why I developed this library.

Do you have anything we can see? That sounds awesome.

 

[Damià Fuentes]

Do you have anything we can see? That sounds awesome.

I wrote the attached paper for the project. If you are interested in openCV, Augmented Reality or something similar I highly recommend you go through the following book: OpenCV: Computer Vision Projects with Python.

 

[nahuel]

I wrote the attached paper for the project. If you are interested in openCV, Augmented Reality or something similar I highly recommend you go through the following book: OpenCV: Computer Vision Projects with Python.

Thanks a lot!

 

[Paulie]

I wrote the attached paper for the project. If you are interested in openCV, Augmented Reality or something similar I highly recommend you go through the following book: OpenCV: Computer Vision Projects with Python.

After reading through your paper, I was wondering if you could further explain part 4. I've been using openCV to perform certain tasks with images and video feeds taken from my webcam using:

import cv2
cap = cv2.VideoCapture(0)

So I'm curious how to replace the VideoCapture(0) with the input of the drone camera.

 

[nahuel]

After reading through your paper, I was wondering if you could further explain part 4. I've been using openCV to perform certain tasks with images and video feeds taken from my webcam using:

import cv2
cap = cv2.VideoCapture(0)

So I'm curious how to replace the VideoCapture(0) with the input of the drone camera.

Take a look at the source code of the wrapper he made:
damiafuentes/DJITelloPy

Particularly this part:

    def get_udp_video_address(self):
        return 'udp://@' + self.VS_UDP_IP + ':' + str(self.VS_UDP_PORT)  # + '?overrun_nonfatal=1&fifo_size=5000'

    def get_video_capture(self):
        """Get the VideoCapture object from the camera drone
        Returns:
            VideoCapture
        """

        if self.cap is None:
            self.cap = cv2.VideoCapture(self.get_udp_video_address())

        if not self.cap.isOpened():
            self.cap.open(self.get_udp_video_address())

        return self.cap

 

[martinpi]

Thank you so much for writing and sharing this library!
I have not yet done much with it, but it installed and worked smoothly.

Greetings, Martin