Panorama Stitching with ORB and Homography in OpenCV

This project implements a classical panorama stitching pipeline using feature detection, matching, and homography estimation in Python with OpenCV.

What this project does

• Loads 3 input images
• Detects keypoints and descriptors using ORB
• Matches descriptors using Brute-Force Matcher (Hamming)
• Filters matches and estimates a homography with RANSAC
• Warps images into a common frame and stitches them into a panorama
• Removes black borders by cropping the valid region

Project structure

• feature-matching-panorama-stitching.py - main script
• assets/ - input images (24.png, 11.png, 42.png)
• report/ - PDF report

Requirements

• Python 3.7 (tested)
• Dependencies in requirements.txt

Setup (Windows)

python -m venv .venv
.\.venv\Scripts\activate
pip install -r requirements.txt