blobboards-examples / README.md
prittjam's picture
Boards artifact: definitions only — renders regenerate from recorded blobs
5a667fc verified
|
Raw
History Blame Contribute Delete
1.43 kB
---
license: mit
pretty_name: BlobBoards Examples
tags:
- computer-vision
- camera-calibration
- fiducial-markers
- julia
---
# BlobBoards example data
The data [BlobBoards.jl](https://github.com/prittjam/BlobBoards.jl)'s
pipeline examples run on, fetched automatically through the package's
`Artifacts.toml` catalog — no dataset setup required.
| file | contents | serves |
|---|---|---|
| `bb_example_boards.tar.gz` | 50 board definitions (JSON only, ~0.7 MB). Renders regenerate on demand from the recorded blob lists via BlobBoards' authoring extension (`import CairoMakie`), cached in a local scratch space — the byte-frozen print-run renders live in the paper's Zenodo `assets` tier | everything |
| `bb_example_still.tar.gz` | one 4240×2832 still + its camera store (`cameras.json`, `vgio/cameras/v1`) | `find_boards`, calibrated + uncalibrated |
| `bb_example_video.tar.gz` | one 1920×1080 handheld clip + its camera store | `track_boards` + trace replay, calibrated + uncalibrated |
Each tarball is a Julia Pkg artifact (content-addressed, sha256-pinned in
BlobBoards' `Artifacts.toml`); the raw trees are also browsable here.
Cameras are stored in VisualGeometryIo's canonical `vgio/cameras/v1`
schema and read with `VisualGeometryIo.load(CameraModel, "cameras.json")`.
Consume via the examples:
```
julia --project=. examples/pipeline/find_boards.jl
julia --project=. examples/pipeline/track_boards.jl
```