Take a look at my blog post[1] where I use a Jetson Nano with OpenALPR[2] to do license plate recognition from video.
The python program simulates a parking garage AI/ML powered camera that will allow vehicle to enter or exit a simulated parking structure when a registered license plate is detected.
Links -
1 - [url]https://medium.com/@alannewcomer/license-plate-recognition-with-a-jetson-nano-e94c6ff683bc[/url]
2 - [url]https://github.com/openalpr/openalpr[/url]
Tried to follow the instructions on your medium post, before I came here. (Was, actually, searching for another post on compiling ALPR on Nano after unsuccessfully attempting to do so per your instructions on medium.)
What’s the error you are getting? The directions on the post assume that the nano has a clean image.
Thank you for you work.
Couple of questions.
Which version of python did you use… Python 2.7 or python 3 when building from source
I got the default ALPR script that is in the ALPR repo to work with python 2.7. However it does not work with 3. It appears the bindings and the corresponding .so file is built for 2.7. Any suggestions?
Also, how would I capture LPs from an external IP camera rather than through the PI camera?
Thanks
G
You can check out Plate Recognizer https://platerecognizer.com. They have an efficient software for License plate Recognition which works on various devices including the Nvidia Jetson Nano https://platerecognizer.com/blog.
Yea dont fall for the Free service on Plate Recognizer… does not allow you to run the docker from Nano without paying… The free service is just a curl command.
You can always use ready tool License plate recognition - ANPR, LPR
THis is another pay service where they dont tell you how much unti the trail is complete