This repository contains all my practical work completed for Computer Vision coursework, progressively updated as I advance through the learning curriculum.
This repository explores state-of-the-art techniques used in advanced Computer Vision and 3D Vision.
References :
- Computer Vision: Algorithms and Applications 2nd Edition - Richard Szeliski
- Multiple View Geometry - Hartley & Zisserman (2004)
Architectures & Techniques:
- Interactive Point Selection
- Homography Estimation
- Panorama Construction
Technical Stack:
- C++
- Imagine++
Architectures & Techniques:
- SIFT feature detection and matching
- Normalized 8-points estimation algorithm (with Hartley Normalization)
- RANSAC Algorithm
- Epipolar lines
Technical Stack:
- C++
- Imagine++
Architectures & Techniques:
- Depth Estimation in Stereo View
- 3D Reconstruction
- Disparity Map
- Seeds Propagation
Technical Stack:
- C++
- Imagine++
Architectures & Techniques:
- Depth Estimation in Stereo View
- 3D Reconstruction
- Disparity Map
- Graph Cuts
- Min-Cut & Maxflow algorithms for graphs
Technical Stack:
- C++
- OpenCV
- Imagine++
- Programming Language: C++
- Computer Vision Libraries: OpenCV - Imagine++
Each lab folder contains:
README.md- Detailed lab description and findings- script of the lab:
- Commented functions
- Visualization outputs
This project is based on material from the course 3D Vision taught at MVA master (M2 AI Master Program of Ecole Normale Supérieure Paris-Saclay), with contributions by:
- Pascal Monasse, Ecole des Ponts ParisTech See personal webpage
- Loïc Landrieu, Ecole des Ponts ParisTech See personal webpage
- Arslan Artykov, Ecole des Ponts ParisTech See personal webpage
The implementation in all labs leverages the Imagine++ library. Please keep this attribution if you reuse, adapt, or share this project.