Automatic Colorization of Black and White Images

Description

This project implements an automatic colorization algorithm for black and white images using a pre-trained deep learning model. The model, based on the Caffe framework, predicts the 'a' and 'b' channels of the Lab color space, given the 'L' channel of an image.

Table of Contents

• Installation
• Usage
• Project Structure
• Contributing
• License
• Acknowledgments

Installation

Prerequisites

• Python 3.x
• OpenCV
• NumPy
• Matplotlib

Steps

1. Clone the repository:

   git clone https://github.com/kalainilavan314/Automatic_Colorization.git
   cd Automatic_Colorization

2. Install the required Python packages:

   pip install -r requirements.txt

Usage

1. Download the pre-trained models and place them in the models directory:

   • colorization_deploy_v2.prototxt
   • colorization_release_v2.caffemodel
   • pts_in_hull.npy

2. Add the path to your black and white image in the script.

3. Run the script:

   python main.py

Project Structure

Contributing

Contributions are welcome! Please open an issue or submit a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Richard Zhang for the colorization model and resources.
• OpenCV and Caffe for providing the framework.