ImageColorizer

ImageColorizer is a tool for automatically colorizing black and white images using deep learning. This project uses a pre-trained model to add natural colors to grayscale photos, making them look more vivid and realistic.

Features

• Automatic colorization of black and white images
• Utilizes state-of-the-art deep learning models
• Easy-to-use command line interface
• Supports common image formats (JPEG, PNG, etc.)

Installation

1. Clone the repository

   git clone https://github.com/ayanbag/ImageColorizer.git
   cd ImageColorizer

2. Create and activate a virtual environment (recommended)

   python3 -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install the required dependencies

   pip install -r requirements.txt

Usage

To colorize a black-and-white image:

python colorize.py --input path/to/your/image.jpg --output path/to/save/colorized.jpg

Parameters:

• --input: Path to the input grayscale image.
• --output: Path where the colorized image will be saved.

Example:

python colorize.py --input images/grayscale_photo.jpg --output results/colorized_photo.jpg

Demo

You can try the tool using the sample images provided in the images directory:

python colorize.py --input images/example_bw.jpg --output results/example_color.jpg

Requirements

• Python 3.6+
• See requirements.txt for the full list of dependencies.

Project Structure

ImageColorizer/
├── colorize.py
├── requirements.txt
├── README.md
├── images/
└── results/

Contribution

Contributions are welcome! Please open issues or submit pull requests for improvements or bug fixes.

License

This project is licensed under the MIT License. See LICENSE for details.

Acknowledgements

• Pre-trained model and research inspiration from Zhang et al., 2016
• Thanks to the open-source community for supporting this project.