Closed Calimerorulez closed 2 years ago
you could use "double take" to use "deepstack.ai" as a detector and in "deepstack" use a custom-model for license-plate-> https://docs.deepstack.cc/custom-models-samples/
longer road to go..
I used to do something similar with 'motion' - whenever motion was detected, I sent the snapshot off to be recognised (and then open the gate if I knew them) -- Of course every frame of motion was detected so it was expensive in terms of CPU. Since Frigate can already detect a car and send it via MQTT to Home Assistant, I would have thought this is already possible with a bit of plumbing?
I've found a YouTube video, explaining the use of Plate Recognizer, with Frigate.
I've managed to get it to work, sort of. Only the resolution of my cam is a little to low for long distance recognition :)
i would love this. would make it so easy to do presence detection on my vehicles
you could use "double take" to use "deepstack.ai" as a detector and in "deepstack" use a custom-model for license-plate-> https://docs.deepstack.cc/custom-models-samples/
longer road to go..
As far as I am aware double-take is only exposing the face end point at this time.
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
https://www.opensourceagenda.com/tags/license-plate-recognition https://github.com/search?q=License+Plate
detections/recognition ...
lots of frameworks available. would still love to see something coming up. but not all is open source at the end...
Hi,
First of all, thank you for your amazing work.
Is it feasible for Frigate to recognize the front side or the back side of a car, e.g. the image containing a license plate? I could then send the snapshot via MQTT to an app like OpenALPR to 'decypher' the plate.