Open kiwimato opened 1 year ago
Thanks for the kind words, and your support!
Sounds like a good idea, i have one question tho. What is the output of your custom model when running it using the DeepStack API?
At this point it's just a model which detects if it's a license plate, so it doesn't actually read the plate itself, however there's also a tesseract model which can actually read it, but I don't know yet how to chain all of them together, because this is all new to me. I assume I would also be able to train Deepstack to include this model as an alternative, but I have to figure that one out myself.
I think even just ability to trigger a custom python script would be interesting because I can write it myself, or I can also take a look at the actual code after work, I could also contribute code if that's OK with you.
Sounds great! There is an open feature request on license plate recognition as a post processor (face recognition is a post processor).
The available post processors can be found here: https://viseron.netlify.app/docs/documentation/configuration#post-processors
Adding one for license plates would probably be pretty easy, takes a bit of extra work tho when its a completely new category of post processors tho.
I think even just ability to trigger a custom python script would be interesting because I can write it myself, or I can also take a look at the actual code after work, I could also contribute code if that's OK with you.
A custom one also sounds like a great idea. Would love some contributions! Right now its pretty much only me working on Viseron. I dont really have any good developer docs apart from the Contributing guidelines, but i am more than happy to answer any questions you might have
Hi,
First of all, super nice project! I'm using Deepstack as a object detector, and I'm interested if for example after it detects a
car
if i would be able to trigger another Deepstack custom model which is able to detect license plates? That would be in the same way face detection works, but a bit more flexible.Thank you