https://github.com/gunjanmimo/spherical-3d-reconstruction-pipeline
https://github.com/gunjanmimo/spherical-3d-reconstruction-pipeline
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/gunjanmimo/spherical-3d-reconstruction-pipeline
- Owner: gunjanmimo
- Created: 2024-03-26T17:53:46.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-04-03T10:53:18.000Z (about 1 year ago)
- Last Synced: 2025-01-01T06:42:31.066Z (5 months ago)
- Language: Python
- Size: 22.5 KB
- Stars: 0
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Spherical 3D Reconstruction Pipeline
### PROPOSED METHOD OVERVIEW
1. **METHOD 1**
**Method 1** used OpenMVG to do camera pose calculation and then it uses OpenMVS for Densify the generated pointcloud.
2. **METHOD 2**
**Method 2** uses AliceVision to preprocesse 360 images and then it uses Colmap sfm tool to generate sparse and dense pointcloud file.
### SETUP
#### 1. DATA PREPARATION
**FOLDER STRUCTURE**
```
.
└── data/
├── config.yaml
└── images/
├── image_1.jpeg
├── image_2.jpeg
└── image_N.jpeg
```
`config.yaml` contains different configuration to run the proposed methods. For details check **README.md** of individual method.
**RUN PIPELINE**
Individual method folder has `Dockerfile` to run the 3D reconstruction pipeline.
To build the Docker image, run
```
cd
docker build -t pipeline .
```
To run the 3D pipeline and generate `.ply` file, use the following command
```
docker run --gpus all -v :/app/data -t pipeline
```
It will save the final dense pointcloud `.ply` file.