Marine underwater
MC-lab Underwater Dataset
Underwater stereo visual-inertial odometry and refractive camera self-calibration
Twenty-four underwater pool recordings combine five grayscale views, inertial sensing, pressure and actuator telemetry under varying motion and illumination.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Marine underwater
- Environment
- Indoor
- Data origin
- Real-world
- Sensor count
- 9
- Ground truth
- Not reported
- Calibration
- Parameters Only
- Annotations
- Not reported
Sensor overview
High-level sensor availability before the detailed sensor records below.
Ground truth
Reference scope, method, rate, coverage, and provenance.
Calibration and synchronization
Reported calibration level, reproducibility signals, and supporting notes.
- Level
- Parameters Only
- Processed parameters
- Available
- Raw calibration data
- Not reported
- Calibration targets
- Not reported
Parameters only
Processed calibration parameters are reported, but raw calibration data is not reported.
Notes
Released air/water camchain YAML, target definitions and calibration reports provide intrinsics/extrinsics. Complete raw calibration recordings are not established.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
This entry counts the original twenty-four motion/illumination recordings. Separate air-water recordings and the six motion-capture recordings added in 2024 are outside that count and do not establish motion-capture truth for the original recordings.
The maintained lossless migration inventory documents five image topics for the original motion subset, while the selected paper evaluates the front stereo pair. The varying-media subset has only the front pair and Alphasense IMU.
AprilTag revisits are intermittent drift checks, with tag maps derived from odometry. Recorded ROVIO estimates are not independent ground truth.
The October 2026 host move preserves original messages but merges split bags. Calibration parameters and reports are public; complete raw calibration sessions and measured timing residuals remain unverified.
Cameras
Camera 1
- Model
- Sony IMX-287 / Alphasense Core
- Setup
- Stereo
- Count
- 1
- Effective cameras
- 2
- Modality
- Grayscale
- Resolution
- 720 x 540
- Rate
- 20 Hz
- Shutter
- Global
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Front stereo pair cam0/cam1, used by the paper. The lossless migration inventory lists 20 Hz mono8 images, and released camchain calibration confirms 720x540.
Camera 2
- Model
- Sony IMX-287 / Alphasense Core
- Setup
- Mono
- Count
- 3
- Effective cameras
- 3
- Modality
- Grayscale
- Resolution
- 720 x 540
- Rate
- 20 Hz
- Shutter
- Global
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Top and two side views cam2/cam3/cam4 are documented in the original motion-subset bag inventory. They are additional physical views, not copies of the stereo pair.
IMUs
Alphasense integrated IMU
- Model
- Bosch BMI085
- Count
- 1
- Accel / gyro
- Not reported
Pixhawk flight-controller inertial output
- Model
- Not reported
- Count
- 1
- Accel / gyro
- Not reported
GNSS
LiDAR
Radar
Additional sensors
Barometer
- Name
- ROV pressure readings
- Count
- 1
- Rate
- 50 Hz
- Notes
- Original motion-subset inventory lists /mavros/imu/static_pressure at 50 Hz; physical component model/count are unreported.
Odometry
- Name
- Recorded online ROVIO estimate
- Count
- 1
- Notes
- Pose/odometry topics are onboard estimator outputs, not external reference truth.
Other sensor
- Name
- ROV actuator PWM commands
- Count
- 1
- Rate
- 50 Hz
- Notes
- Channels 1–8 command thrusters and 9–10 command onboard LEDs. These are control channels, not independently measured thrust or a sensor count.
Annotations
Reported annotation availability and task support.
- Availability
- Not reported
- Format
- Not reported
- Class count
- Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{singh2023mclabunderwater,
author = {Mohit Singh and Mihir Dharmadhikari and Kostas Alexis},
title = {An Online Self-calibrating Refractive Camera Model with Application to Underwater Odometry},
year = {2023},
url = {https://arxiv.org/abs/2310.16658v1}
}
Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy