Skip to content

Latest commit

 

History

History
51 lines (44 loc) · 1.58 KB

README.md

File metadata and controls

51 lines (44 loc) · 1.58 KB

Automatic Number Plate Recognition

A simple license plate number recognizer/extractor. Based on Nicholas Renotte's ANPR Tutorial

Environment

  • Anaconda (v4.5.11) => for the virtual environment, can use alternatives
  • Ubuntu 18.04 on WSL2

Requirements

  • OpenCV
  • GPU for use in PyTorch
  • Proper GPU drivers (for NVIDIA, see here)
  • (Optional) A dispaly server for running GUI programs designed for the X Window System (e.g. VcXsrv)

Dev Environment

Installation

For new environemnt

conda create --name <env-name> --file requirements.txt

or, for existing (hopefully sparse) environments

conda install --file requirements.txt

That should be it!

Troubleshooting

If PyTorch complains about GPU compatibility, try the following for RTX 3000 series NVIDIA GPUs:

conda install pytorch torchvision cudatoolkit=11 -c pytorch-nightly

Usage

python main.py -i <inputfile>

Where inputfile is the path to your image.

There are some example images in the images directory.

Results

Image 0

alt text alt text

Image 1

alt text alt text

Image 2

alt text alt text Needs some improvement...