A stereo image registration platform that automates spatial alignment using feature-based computer vision algorithms built as a B.Sc. research project in Surveying & Geoinformatics at the University of Lagos.
The Image Registration Tool project focuses on building an intelligent image registration system that aligns stereo image pairs using algorithms such as SIFT, ORB, AKAZE, and FAST. The system provides accurate geometric transformations and visualization for mapping, photogrammetry, and 3D reconstruction workflows.
Implements robust algorithms SIFT, ORB, AKAZE, FAST to extract reliable keypoints for stereo correspondence.
Applies RANSAC to estimate projective transformations and reject outliers during alignment.
Measures registration accuracy using RMSE, MAE, and pixel displacement consistency.
Integrates FastAPI backend with responsive Nuxt frontend for smooth visualization and user interaction.
Academic Supervisor
Department of Surveying & Geoinformatics, University of Lagos
Student Developer – B.Sc. Surveying & Geoinformatics
Responsible for algorithm implementation, web development, and performance evaluation under academic supervision.
This B.Sc. research contributes to advancing geospatial technology in Nigeria through the automation of stereo image registration processes, improving mapping accuracy and operational efficiency.
Accuracy Rate
Algorithms Tested
Experience automated stereo image registration with Image Registration Tool advanced algorithms and intuitive interface.