- Joined
- Aug 20, 2019
- Messages
- 5
- Reaction score
- 0
Ill connect it to an esp module with a magnet connectorHow would the Home Assistant power on the Tello? It can't sit around powered on 24/7
Do you have the skills required?
You will need excellent programming skills (I mean real programming with bits and bytes, computer vision, UDP networking, video decoding, etc - _not_ scripting ) and some electronics skills (for power management).
Oh, OK. Then just do it!Its simpler then you think...Home Assistant support node-red ?![]()
Oh, OK. Then just do it!
I just wonder how node red will decode h264 video from a UDP stream. And will find the intruder for you using computer vision and machine learning. It will also direct Tello to the center of a room without knowning Tello's true direction or position. And it will also flash a new Tello firmware that allows Tello to charge while being powered up - or powers up Tello while being charged. And it will cool Tello reliably while being powered up.
No. Node Red is completely useless for this task. Its not programming - its not even scripting. Node Red is like drone blocks for simple automation tasks. You can use it for automating the tasks that happen _after_ the Tello surveillance program has done it's job. Something like "interuder found => turn on lights and play soundfile".
This project is not simpler than I think - I have a pretty sound idea whats involved. But it's far more complex than you can imagine.
If you want to proceed: have a look at any existing tello control library (or the mother of all: tellolib), then get wireshark up and running and work from there.
That's what I was thinking at the first of this thread. A Wi-Fi camera with motion detector and you could see a live video feed on your smartphone... but:i could have made it with a regular cam...?![]()
Now it's beginning to sound like something I would do LOL... and a lot of other folks here who like to play with electronics, modify drones, build computers, root phones, build DIY security systems, etc. etc.ofcourse its all just a challenge
Well, you wrote "and search for intruder" which implies actively searching for it. Not just taking random photos. In that case ignore the part about computer vision and machine learning.
If you want photos you can't use the SDK, you have to use binary communication. Or you have to do streaming and grab still images from the video feed, but these will be of greatly reduced quality compared to real photos.
We use essential cookies to make this site work, and optional cookies to enhance your experience.