vehicle_license_model_recognition

Automatic recognition of license plates, determine if expired, look for hits in police database, recognize vehicle make and model.

General Idea:

Python based ML to locate license plates and make/model of vehicles from pictures and/or video and perform a series of related tasks. In short, make it so that Police officers don't need to manually enter license plate numbers while driving; and automate the task so that all vehicles in a "dash cam" can automatically be scanned. This would be very useful esp in situations like Amber Alerts; the same algorithm/idea can be deployed to as many cameras as possible and try to find the vehicle in question.

Tasks:

1. Get license plate state and unique numbers/letters combination.
2. Read if the registration is expired/current.
3. Check if there are any hits for the license plate/state combo in a simulated database of Police records.
4. Get the make/model of the vehicle based upon rear decals/logo/etc.
5. Validate make/model based upon overall shape and features of the vehicle.
6. Validate that the license plate matches the vehicle's make/model.
7. Auto-scan for infractions or hits and have an easy notification system.
8. Auto-scan for make/model of near by issues when the license plate is unknown or partial.

First Steps:

1. pytesseract to read strings from the images.
2. Simple Django web app?
3. Webcam integration to take the pictures/video?
4. Use django to run the tasks/scans manually for now, and then automate it all later.