JayS420 / ANPR

ANPR

Automatic Number Plate Recognition is build using temsorflow object detection models and is being trained on the kaggle dataset and after training opencv is used to detect the liscence plate.

PART 1 | Setup

1. Creating a Virtual Environment
2. Installing Dependencies
3. Installing Tensorflow Object Detection
4. Cloning Pre-Trained Models

PART 2 | Data

1. Kaggle dataset - https://www.kaggle.com/andrewmvd/car-plate-detection
2. Creating a Training and Testing Partition

PART 3 | Training

1. Updating the LabelMap
2. Creating TF Records
3. Updating Transfer Learning Config
4. Training the Model

PART 4 | Detecting Plates

1. Detecting Plates from an Image