We built several machine vision and classification models to detect vehicles in videos captured by the client’s cameras.  Our models were able to detect the vehicles, determine what kind (car, truck, motorcycle, etc), what make (Ford, Toyota, etc.) and what color the vehicles were.   

The models then identified the license plates in the images and extracted the state information and the license plate text. With additional fine tuning, we were able to obtain results in a matter of milliseconds.