🎨 Image Colorisation

A deep learning–based computer vision application that automatically converts grayscale images into colorized images. The project provides an interactive web interface where users can upload a black-and-white image and obtain its colorized version.

🚀 Features

• 🖼️ Convert grayscale images into colorized images
• 🤖 Deep learning–based color prediction
• ⚙️ Automated image preprocessing and model inference
• 🌐 Flask-based web interface
• 📤 Upload images directly through the browser
• 📥 View the generated colorized output
• 🔄 End-to-end pipeline from image upload to colorized result

🛠️ Tech Stack

• Programming Language: Python
• Machine Learning: Deep Learning / Neural Networks
• Computer Vision: Image Processing
• Backend: Flask
• Frontend: HTML, CSS
• Model: Pre-trained image colorization model

📂 Project Structure

Image-Colorisation/
│
├── app.py                 # Flask application
├── model.py               # Model architecture and inference logic
├── model/                 # Trained model files
│── index.html             # Image upload page
│── result.html            # Colorized image output page
├── static/                # Static assets
├── requirements.txt       # Python dependencies
└── README.md              # Project documentation

🔄 How It Works

The application follows the following pipeline:

Input Grayscale Image
        ↓
Image Upload
        ↓
Image Preprocessing
        ↓
Deep Learning Model
        ↓
Color Prediction
        ↓
Post-processing
        ↓
Colorized Image
        ↓
Web Interface

⚙️ Installation

1. Clone the repository

git clone https://github.com/vj2304/Image-Colorisation.git
cd Image-Colorisation

2. Create a virtual environment

python -m venv venv

Activate it:

Windows

venv\Scripts\activate

Linux / macOS

source venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

▶️ Running the Application

Start the Flask application:

python app.py

Open your browser and navigate to:

http://127.0.0.1:5000

Upload a grayscale image and the application will generate the corresponding colorized output.

📸 Example

Input

A grayscale/black-and-white image is uploaded through the web interface.

Output

The deep learning model predicts suitable colors and generates a colorized version of the input image.

Note: Model output quality depends on the input image and the training/model architecture. Colorization is an inherently predictive task, so generated colors may not always represent the original colors.

🧠 Key Concepts

This project demonstrates practical applications of:

• Deep Learning
• Computer Vision
• Image Processing
• Image-to-Image Transformation
• Neural Network Inference
• Model Deployment
• Flask Web Development

🔮 Future Improvements

• Improve colorization quality using more advanced architectures
• Train on larger and more diverse datasets
• Add GPU acceleration for faster inference
• Add batch image processing
• Provide drag-and-drop image upload
• Add before/after image comparison
• Deploy the application using a cloud platform
• Add quantitative image-quality evaluation metrics

👨‍💻 Author

VJ2304

GitHub: https://github.com/vj2304