🎨 False Coloration of Black & White Images

📌 Overview

This project presents a hybrid image colorization system that transforms grayscale images into colored images using two approaches:

• 🎯 False Coloration using OpenCV colormaps
• 🤖 Realistic Colorization using a deep learning model (Caffe)
• 🖥️ Interactive GUI built with PyQt5

The system is designed for educational and practical applications in signal processing, computer vision, and image enhancement.

---

🚀 Features

• Apply multiple colormaps (Jet, Parula, Hot, HSV, etc.)
• Adjust color intensity using a slider
• Perform realistic image colorization using deep learning
• User-friendly GUI (no coding required)
• Load and save images easily
• Real-time preview of results

---

🛠️ Technologies Used

• Python 3.9
• OpenCV
• NumPy
• PyQt5
• Caffe Model (via OpenCV DNN module)

---

📂 Project Structure

false-colorization-project/
│
├── src/
│   └── main.py
│
├── models/
│   ├── colorization_deploy_v2.prototxt
│   ├── colorization_release_v2.caffemodel
│   └── pts_in_hull.npy
│
├── assets/
│   ├── GUI.png
│   ├── HSV_colormap.png
│   ├── bone_and_pink_colormap.png
│   └── realistic_1.png
│
├── docs/
│   └── report.pdf
│
├── requirements.txt
└── README.md

---

⚙️ Installation

git clone https://github.com/fida-121/false-colorization-project.git
cd false-colorization-project
pip install -r requirements.txt

---

📥 Model Download

Due to GitHub file size limits, the trained model is not included in this repository.

Download it from the link below:

👉 https://drive.google.com/file/d/1KKetd41ADCXG4TLaJCbajIQCajrgrws8/view?usp=drive_link

After downloading, place the file in the following directory:

models/colorization_release_v2.caffemodel

---

▶️ Run the Project

python src/main.py

---

📸 Screenshots

Below are some results generated by the application, including GUI interface, colormap outputs, and deep learning-based realistic colorization.

🖥️ GUI Interface

---

🎨 Colormap Results

HSV Colormap

Bone & Pink Colormap

---

🤖 Realistic Colorization

---

📊 Results and Discussion

• Colormaps provide fast and visually appealing enhancements
• Deep learning produces semantically meaningful colors
• GUI allows easy comparison and interaction
• Effective for educational and visualization purposes

---

📘 Report

The complete project report is available in:

📄 docs/report.pdf

---

👨‍💻 Authors

• Fida Hussain
• Basil Iftikhar
• Syed Irtaza Abbas

---

🎓 Academic Info

Department of Mechatronics and Control Engineering University of Engineering and Technology, Lahore

Course: Signal Processing (MCT-244L)

---

🔮 Future Work

• Real-time video colorization
• Web deployment using Flask/Django
• User-guided color hints
• Image enhancement techniques (histogram equalization, contrast adjustment)

---

📄 License

This project is developed for educational purposes only.