This repository contains Python code for image matching, feature detection, and 3D reconstruction using COLMAP, a popular computer vision toolkit.

Small script for easier image adding in colmap

Toolkit for a collaborative and shardable repository of the 2d & 3d likeness of objects.

Utilities for interacting with SFM Data in golang

This repository contains the code developed for my MSc project titled "Exploring Photogrammetry for High-Fidelity 3D Model Generation: A Power Grid Inspection Study", as well as some of the results. The thesis was written in collaboration with SINTEF, and the degree issued by DTU. The full thesis can be downloaded from the below link.

Image to Mesh Alignment experiments with PixSfM

The aim of my project was to develop a user-friendly method for generating realistic 3D models and displaying them on smartphones using augmented reality. The only input required from the user is a short video of the desired object, which can be uploaded to the app. This project uses a modified version of Nvidia's Instant-NeRF.

3D Surface and Mesh Reconstruction from RGB Videos using COLMAP and Open3D

Blender plugin which generates a dataset for colmap by exporting blender camera poses and rendering scene.

3D Gaussian Splatting SAM

Senior project with Austin Theriault and Ryan Hudson

Transforming 2D images into 3D semantically segmented scenes using innovative CNN architecture and COLMAP reconstruction.

Improved depth from mono camera images fusing Colmap and LiDAR data

A repo for my Graduate Computer Vision I semester project

Render COLMAP reconstructed Model with Pytorch3D

Provides a function to delete image ids from points3D.txt file for COLMAP

The missing tool kit for MVS researchers

Automatic Scale Factor Estimation of 3D Reconstruction in COLMAP Utilizing Aruco Marker

visual odometry c++ wrapered by pybind11 that containes the feature extraction(traditional sift, orb, etc opencv supported, pytorch supported learning feature), feature matching(flann, supperglue), essential estimator(five, eight, p3p etc), triangulation

Easily visualize, registrate and work with COLMAP reconstructions.