https://github.com/kekeblom/strayvisualizer
Visualize Data From Stray Scanner https://keke.dev/blog/2021/03/10/Stray-Scanner.html
https://github.com/kekeblom/strayvisualizer
3d-reconstruction camera computer-vision dataset depth-maps mapping open3d pointclouds python reconstruction rgb-d slam
Last synced: about 2 months ago
JSON representation
Visualize Data From Stray Scanner https://keke.dev/blog/2021/03/10/Stray-Scanner.html
- Host: GitHub
- URL: https://github.com/kekeblom/strayvisualizer
- Owner: kekeblom
- License: mit
- Created: 2021-03-13T17:02:48.000Z (about 4 years ago)
- Default Branch: main
- Last Pushed: 2023-01-04T09:52:13.000Z (over 2 years ago)
- Last Synced: 2023-04-06T10:26:09.286Z (about 2 years ago)
- Topics: 3d-reconstruction, camera, computer-vision, dataset, depth-maps, mapping, open3d, pointclouds, python, reconstruction, rgb-d, slam
- Language: Python
- Homepage:
- Size: 31.3 KB
- Stars: 49
- Watchers: 7
- Forks: 11
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# StrayVisualizer
A set of scripts to work with data collected using [Stray Scanner](https://apps.apple.com/us/app/stray-scanner/id1557051662).
## Usage
To access the data from your device, see [this wiki entry](https://github.com/kekeblom/StrayVisualizer/wiki/Accessing-Data).
### Installing Dependencies
Install dependencies with `pip -r requirements.txt`.
### Example Datasets
If you don't have your own dataset, you can download one of these example datasets:
```
wget https://stray-data.nyc3.digitaloceanspaces.com/datasets/dinosaur.tar.gz
tar -xvf dinosaur.tar.gz
python stray_visualize.py dinosaur/
```
## Visualizing the data
Run `python stray_visualize.py `.
Available command line options are:
- `--point-clouds` shows pointclouds.
- `--confidence=` there are three levels of confidence for the depth outputs: 0, 1 and 2. Higher is more confident. Only points having confidence higher or equal to the given value are shown.
- `--frames` shows the camera pose coordinate frames.
- `--every=` determines how often the coordinate frame is drawn. Default is to draw every 60th frame.
- `--trajectory` shows a black line for the trajectory of the camera.
- `--integrate` will run the data through the Open3D RGB-D integration pipeline and visualize the resulting mesh.
- `--voxel-size=` sets the voxel size in meters for RGB-D integration.
- `--mesh-filename` save the mesh from RGB-D integration into the given file. Defaults to no mesh saved.
## Creating a Video From the Depth Maps
`python make_video.py ` will combine depth maps into a video. It will create a file `depth_video.mp4` inside the dataset folder.
## Converting to a ROS bag
`python create_rosbag.py --out ros.bag` will create a rosbag from your scene. This has to be done with your ROS distribution sourced, as it requires some ros packages, such as `sensor_msgs`, `cv_bridge` and `rosbag`.
## Reporting Issues
If you find any issues with this project or bugs in the Stray Scanner app, you can open an issue on this repository.
## Related Projects
[PyojinKim](https://github.com/PyojinKim) created a set of [helpful scripts](https://github.com/PyojinKim/StrayScannerVisualizer) to work with data collected with the App in Matlab.