Image is processed with existing model - wpod-net.json and are cropped for the area of interest, i.e. to the size of License plate.
Pytesseract Model is used for training and detecting the exact text from the License Plate image.
Anvil Webserver is used to deploy this model to the web with free membership.
Anvil private Hosting link - https://RZNK4SZD7CYXKASZ.anvil.app/63RFTQUZPX5PTBMF6LWLDXDE
Some texts were not detected properly for some images -
Sometimes some extra characters were inserted automatically at start and end of License plate number -
Tesseract!!😂