sim2radar_artifacts / README.md
AnonymousAuthors0018's picture
Add files using upload-large-folder tool
bd8c181 verified
|
Raw
History Blame Contribute Delete
2.99 kB
---
license: cc-by-4.0
tags:
- radar
- mmwave
- fmcw
- point-cloud
- robotics
pretty_name: Sim2Radar Artifacts
---
# Sim2Radar — Artifacts
Artifacts for the Sim2Radar simulation framework (anonymous submission under
review): real FMCW mmWave radar recordings used for evaluation. All radar
data is raw ADC / pre-FFT IQ samples at 77 GHz (TI IWR1843BOOST + DCA1000EVM).
## Contents
```
data/
rover/
lecture_hall/ # indoor environment 1 (recordings tagged "env")
depth/ # sensor moves toward/away from a static object
lateral/ # sensor moves laterally past a static object
squash/ # indoor environment 2 (recordings tagged "squash"/"court")
depth/
lateral/
microbench/ # controlled single-object motion primitives
```
### data/rover/ — moving sensor, static object
A mobile robot carries the radar plus an RGB-D camera past a static target
object placed on a stand. Sweep: 0.275 m/s, 1 m travel, 4 passes per
recording. Start distances of 0.5 m, 1 m, and 2 m.
Recording folder naming:
```
<timestamp>_<site>_<object>_<start_dist>_<sweep_len>_<Depth|Lateral>_x<passes>
```
Note: recordings whose timestamps begin with `1969_12_31` come from a capture
machine without a synchronized real-time clock; the relative timing within
each recording is still consistent.
Each recording folder contains:
- `radar_data/raw/*.bin` — raw ADC captures, one file per radar frame
(sequence-numbered)
- `rosbag/` — ROS 2 bag (sqlite3) with topics:
- `/camera/color/image_raw`, `/camera/color/camera_info`
- `/camera/depth/image_rect_raw`, `/camera/depth/camera_info`
- `/radar/frame_ref` — radar frame reference messages
- `/synced/pair` — JSON strings pairing RGB/depth frames with radar
frames by timestamp and sequence number (the embedded `radar_raw_bin`
paths reflect the original capture machine layout; match on the
`seq######` suffix to locate the file under `radar_data/raw/`)
Objects are household items: cans, bottles, cardboard box, olive oil,
fire extinguisher, and a trihedral corner reflector.
### data/microbench/ — controlled motion primitives
Single objects (corner reflector, soda can) undergoing simple controlled
transformations: turntable rotation (both directions), forward/back,
right/left, diagonal translation, and a static reference. Per recording:
- `rawADC_*.bin` — raw DCA1000 ADC capture
- `rawIQ_*.mat` — the same capture parsed to an IQ array (MATLAB v7.3 / HDF5)
File naming: `rawADC_<motion>__<range>__<duration>__lab__<timestamp>.bin`
`setup_photos/` contains photographs of the linear-rail and turntable capture
setups and live micro-Doppler processing during collection.
## Radar configuration
TI IWR1843BOOST, 77 GHz FMCW, captured through a DCA1000EVM. Chirp/profile
configuration files are included where available; raw `.bin` frames are
131,072 bytes each in the rover recordings.
## License
CC BY 4.0.