I am new to OpenMV, but seems like you guys built a really cool product and infrastructure. I am in the beginning stage of building a prototype for the Hackster.io tiny ml contest. Part of my desired functionality is to have a license plate number detector. Since I am still in the process of picking parts/platform I kind of wanted to throw the question out there before making any purchases. As far as my understanding goes (not image processing specialist) task such as license plate number recognition is a multistage pipeline/process that typically require steps such as detecting the plate, extracting characters, and performing some type of OPR on the characters. All that functionality is pretty well supported with OpenCV. In theory do you think (i am not looking for someone to derive the algorithm for) OpenMv libraries have enough functionality behind them to be able to build this?
Thanks so much,
Right now we don’t have enough power to do this. Typically you use specialized libraries for this task.
Thanks for such a quick reply!