3D reconstruction course project
About this project
- Python.
- Pytorch for optimization of positions of cameras and positions of triangulated 3D points.
- 3D reconstruction, structure from motion, multi-view geometry, bundle adjustment.
- Duration: 6 weeks.
This project was done in the course TSBB15 Computer Vision at Linköping university.
The goal was to construct a system able to perform 3D reconstruction of a static scene given images of this scene takenfrom different angles. This task is thoroughly studied in the field of computer vision and known as Structure from Motion (SfM).
Links
- Here’s the link to the Github repo.
- And here’s the project report.