fast-plate-ocr 0.1.0

Fast & Lightweight OCR for vehicle license plates.

Fast & Lightweight License Plate OCR

---

Introduction

Lightweight and fast OCR models for license plate text recognition. You can train models from scratch or use the trained models for inference.

The idea is to use this after a plate object detector, since the OCR expects the cropped plates.

Features

• Keras 3 Backend Support: Compatible with TensorFlow, JAX, and PyTorch backends 🧠
• Augmentation Variety: Diverse augmentations via Albumentations library 🖼️
• Efficient Execution: Lightweight models that are cheap to run 💰
• ONNX Runtime Inference: Fast and optimized inference with ONNX runtime ⚡
• User-Friendly CLI: Simplified CLI for training and validating OCR models 🛠️
• Model HUB: Access to a collection of pre-trained models ready for inference 🌟

Available Models

Model Name	Time b=1 (ms)[1]	Throughput (plates/second)[1]	Dataset	Accuracy[2]	Dataset
argentinian-plates-cnn-model	2.0964	477	arg_plate_dataset.zip	94.05%	Non-synthetic, plates up to 2020.

^[1]Inference on Mac M1 chip using CPUExecutionProvider. Utilizing CoreMLExecutionProvider accelerates speed by 5x.

^[2] Accuracy is what we refer as plate_acc. See metrics section.

Reproduce results.

• Calculate Inference Time:

  pip install fast_plate_ocr  # CPU
  # or
  pip install fast_plate_ocr[inference_gpu]  # GPU
  

  from fast_plate_ocr import ONNXPlateRecognizer
  
  m = ONNXPlateRecognizer("argentinian-plates-cnn-model")
  m.benchmark()
  

• Calculate Model accuracy

  pip install fast-plate-ocr[train]
  curl -LO https://github.com/ankandrew/fast-plate-ocr/releases/download/v1.0/arg_cnn_ocr_config.yaml
  curl -LO https://github.com/ankandrew/fast-plate-ocr/releases/download/v1.0/arg_cnn_ocr.keras
  curl -LO https://github.com/ankandrew/fast-plate-ocr/releases/download/v1.0/arg_plate_benchmark.zip
  unzip arg_plate_benchmark.zip
  fast_plate_ocr valid \
      -m arg_cnn_ocr.keras \
      --config-file arg_cnn_ocr_config.yaml \
      --annotations benchmark/annotations.csv
  

Inference

For inference only, install:

pip install fast_plate_ocr

For doing inference on GPU, install:

pip install fast_plate_ocr[inference_gpu]

Usage

To predict from disk image:

from fast_plate_ocr import ONNXPlateRecognizer

m = ONNXPlateRecognizer('argentinian-plates-cnn-model')
print(m.run('test_plate.png'))

run demo

To run model benchmark:

from fast_plate_ocr import ONNXPlateRecognizer

m = ONNXPlateRecognizer('argentinian-plates-cnn-model')
m.benchmark()

benchmark demo

CLI

To train or use the CLI tool, you'll need to install:

pip install fast_plate_ocr[train]

Train Model

To train the model you will need:

1. A configuration used for the OCR model. Depending on your use case, you might have more plate slots or different set of characters. Take a look at the config for Argentinian license plate as an example:

   # Config example for Argentinian License Plates
   # The old license plates contain 6 slots/characters (i.e. JUH697)
   # and new 'Mercosur' contain 7 slots/characters (i.e. AB123CD)
   
   # Max number of plate slots supported. This represents the number of model classification heads.
   max_plate_slots: 7
   # All the possible character set for the model output.
   alphabet: '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_'
   # Padding character for plates which length is smaller than MAX_PLATE_SLOTS. It should still be present in the alphabet.
   pad_char: '_'
   # Image height which is fed to the model.
   img_height: 70
   # Image width which is fed to the model.
   img_width: 140
   

2. A labeled dataset, see arg_plate_dataset.zip for the expected data format.
3. Run train script:

   # You can set the backend to either TensorFlow, JAX or PyTorch
   # (just make sure it is installed)
   !KERAS_BACKEND=tensorflow fast_plate_ocr train \
       --annotations path_to_the_train.csv \
       --val-annotations path_to_the_val.csv \
       --batch-size 128 \
       --epochs 750 \
       --dense \
       --early-stopping-patience 100 \
       --reduce-lr-patience 50
   

You will probably want to change the augmentation pipeline to apply to your dataset. In order to do this

Define Albumentations pipeline:

import albumentations as A

transform_pipeline = A.Compose(
    [
        # ...
        A.RandomBrightnessContrast(brightness_limit=0.1, contrast_limit=0.1, p=1),
        A.MotionBlur(blur_limit=(3, 5), p=0.1),
        A.CoarseDropout(max_holes=10, max_height=4, max_width=4, p=0.3),
        # ... and any other augmentation ...
    ]
)

# Export to a file (this resultant YAML can be used by the train script)
A.save(transform_pipeline, "./transform_pipeline.yaml", data_format="yaml")

And then you can train using the custom transformation pipeline with the --augmentation-path option.

Visualize Augmentation

It's useful to visualize the augmentation pipeline before training the model. This helps us to identify if we should apply more heavy augmentation or less, as it can hurt the model.

You might want to see the augmented image next to the original, to see how much it changed:

fast_plate_ocr visualize-augmentation \
    --img-dir benchmark/imgs \
    --columns 2 \
    --show-original \
    --augmentation-path '/transform_pipeline.yaml'

You will see something like:

Validate Model

After finishing training you can validate the model on a labeled test dataset.

Example:

fast_plate_ocr valid \
    --model arg_cnn_ocr.keras \
    --config-file arg_plate_example.yaml \
    --annotations benchmark/annotations.csv

Visualize Predictions

Once you finish training your model, you can view the model predictions on raw data with:

fast_plate_ocr visualize-predictions \
    --model arg_cnn_ocr.keras \
    --img-dir benchmark/imgs \
    --config-file arg_cnn_ocr_config.yaml

You will see something like:

Export as ONNX

Exporting the Keras model to ONNX format might be beneficial to speed-up inference time.

fast_plate_ocr export-onnx \
	--model arg_cnn_ocr.keras \
	--output-path arg_cnn_ocr.onnx \
	--opset 18 \
	--config-file arg_cnn_ocr_config.yaml

Keras Backend

To train the model, you can install the ML Framework you like the most. Keras 3 has support for TensorFlow, JAX and PyTorch backends.

To change the Keras backend you can either:

1. Export KERAS_BACKEND environment variable, i.e. to use JAX for training:

   KERAS_BACKEND=jax fast_plate_ocr train --config-file ...
   

2. Edit your local config file at ~/.keras/keras.json.

Note: You will probably need to install your desired framework for training.

Model Architecture

The current model architecture is quite simple but effective. See cnn_ocr_model the code.

The model output consists of several heads. Each head represents the prediction of a character of the plate. If the plate consists of 7 characters at most (max_plate_slots=7), then the model would have 7 heads.

Example of Argentinian plates:

Each head will output a probability distribution over the vocabulary specified during training. So the output prediction for a single plate will be of shape (max_plate_slots, vocabulary_size).

Model Metrics

During training, you will see the following metrics

• plate_acc: Compute how many plates were correctly classified. For a single plate, if ground truth is ABC123, and the prediction is 'ABC 123', then this would give a score of 1. If the prediction was ABD 123, it would score 0.
• cat_acc: Calculates how many characters of the plate were correctly classified. Example if the correct label is ABC123 and ABC133 is predicted, it will not give a precision of 0% like plate_acc (not completely classified correctly), but 83.3% (5/6).
• top_3_k: Calculates how often the true character is found in the top-3 predictions (the 3 with the highest probability).

Metrics are defined in this custom.py module.

Contributing

Contributions to the repo are greatly appreciated. Whether it's bug fixes, feature enhancements, or new models, your contributions are warmly welcomed.

To start contributing or to begin development, you can follow these steps:

1. Clone repo

   git clone https://github.com/ankandrew/fast-plate-ocr.git
   

2. Install all dependencies using Poetry:

   poetry install --all-extras
   

3. To ensure your changes pass linting and tests before submitting a PR:

   make checks
   

If you want to train a model and share it, we'll add it to the HUB 🚀

TODO

• Expand model zoo.
• Use synthetic image plates.
• Finish and push TorchServe files.