Canny Edge Detector Implementation 🔍

A complete from-scratch implementation of the Canny edge detection algorithm for image processing applications.

Features ✨

• Complete Canny Algorithm: Full implementation of all five stages (Gaussian smoothing, gradient calculation, non-maximum suppression, double thresholding, hysteresis tracking)
• Configurable Parameters: Adjustable sigma, kernel size, and threshold values for different use cases
• Class-Based Design: Clean, reusable object-oriented implementation
• Image Utilities: Built-in functions for loading, converting RGB to grayscale, and visualization
• Educational Examples: Comprehensive Jupyter notebooks demonstrating the algorithm pipeline
• Image Alignment Applications: Extended use cases for practical image processing tasks

Tech Stack / Built With 🛠️

• Python 3.7+: Core programming language
• NumPy: Numerical computations and array operations
• SciPy: Scientific computing and image processing functions
• Matplotlib: Image visualization and plotting
• Jupyter: Interactive development and demonstration

Prerequisites

• Python 3.7 or higher
• pip package manager

Installation 📥

# Clone the repository
git clone https://gitlab.com/user4302_Projects/coding/python/university/iit/canny-edge-detector.git
cd canny-edge-detector

# Install required packages
pip install numpy scipy matplotlib jupyter

Usage / Quick Start ⚡

Basic Edge Detection

from cannyedgedetector import CannyEdgeDetector
from utils import utils

# Load and preprocess images
images = utils.load_data('path/to/images')

# Initialize detector with custom parameters
detector = CannyEdgeDetector(
    images, 
    sigma=1.4, 
    kernel_size=5, 
    low_threshold=0.09, 
    high_threshold=0.17
)

# Detect edges
edge_images = detector.detect()

# Visualize results
utils.visualize(edge_images, 'gray')

Advanced Configuration

# Custom detector configuration
detector = CannyEdgeDetector(
    images,
    sigma=1.0,                    # Gaussian smoothing parameter
    kernel_size=5,                # Size of Gaussian kernel
    weak_pixel=75,               # Intensity for weak edges
    strong_pixel=255,            # Intensity for strong edges
    low_threshold=0.05,          # Lower threshold ratio
    high_threshold=0.15          # Higher threshold ratio
)

Project Structure 📂

canny-edge-detector/
├── cannyedgedetector.py          # Main CannyEdgeDetector class
├── utils/
│   └── utils.py                  # Utility functions (load, visualize, rgb2gray)
├── Canny_Edge_Detector.ipynb     # Main demonstration notebook
├── image_alignment_task.ipynb     # Extended image processing examples
├── face_images/                  # Sample images directory
├── README.md                     # This file
└── .git/                         # Git repository

Configuration 🔧

The CannyEdgeDetector class accepts the following parameters:

• sigma (float, default=1): Standard deviation for Gaussian kernel
• kernel_size (int, default=5): Size of the Gaussian kernel (must be odd)
• weak_pixel (int, default=75): Intensity value for weak edges
• strong_pixel (int, default=255): Intensity value for strong edges
• low_threshold (float, default=0.05): Lower threshold ratio (0-1)
• high_threshold (float, default=0.15): Higher threshold ratio (0-1)

Development / Running Locally 🏗️

# Start Jupyter notebook for interactive development
jupyter notebook

# Run specific notebooks
jupyter nbconvert --to notebook --execute Canny_Edge_Detector.ipynb
jupyter nbconvert --to notebook --execute image_alignment_task.ipynb

Testing 🧪

The project includes demonstration notebooks that serve as functional tests:

# Run the main demonstration
jupyter nbconvert --to notebook --execute Canny_Edge_Detector.ipynb --inplace

# Run image alignment examples
jupyter nbconvert --to notebook --execute image_alignment_task.ipynb --inplace

Building for Production 🏭

This is a pure Python library with no build steps required. Simply import the modules:

import cannyedgedetector
from utils import utils

Deployment 🚀

The library can be deployed as:

1. Python Package: Install via pip from local directory
2. Docker Container: Containerize with Python base image
3. Cloud Function: Deploy as serverless function for image processing
4. Web Service: Wrap in Flask/FastAPI for HTTP API

Contributing 🤝

We welcome contributions! Please follow these guidelines:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Merge Request

Important: For any questions, bug reports, feature requests, or security concerns, please open an issue on GitLab: https://gitlab.com/user4302_Projects/coding/python/university/iit/canny-edge-detector/-/issues

License 📄

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

Support & Contact 👋

For any questions, bug reports, feature requests, or security concerns, please open an issue on GitLab: https://gitlab.com/user4302_Projects/coding/python/university/iit/canny-edge-detector/-/issues

No email or direct messaging support is available.

Acknowledgments 🙏

• John Canny for the original edge detection algorithm (1986)
• SciPy and NumPy communities for excellent numerical computing tools
• Computer vision researchers and practitioners who continue to advance edge detection techniques