Marine surface

SeePerSea

Maritime object detection and classification

Ego-centric RGB and LiDAR observations of in-water ships, buoys and other hazards, recorded from surface vessels in sea and freshwater environments.

Metadata: Complete Metadata needs review

Quick facts

Vehicle type
Marine surface
Environment
Outdoor
Data origin
Real-world
Sensor count
4
Ground truth
Available
Calibration
Parameters Only
Annotations
Available

Sensor overview

High-level sensor availability before the detailed sensor records below.

Ground truth

Reference scope, method, rate, coverage, and provenance.

Ground-truth references for SeePerSea
Scope Method Reference system Count Rate Coverage Reference provenance
External
Measured using equipment independent of the platform’s normal onboard sensor pipeline.
Derived
Computed primarily from the dataset’s own recorded sensors.
Hybrid
Combines independent reference equipment with onboard measurements.
Other Manual annotation Not reported Not reported Not reported Unknown Reference provenance
External
Measured using equipment independent of the platform’s normal onboard sensor pipeline.
Derived
Computed primarily from the dataset’s own recorded sensors.
Hybrid
Combines independent reference equipment with onboard measurements.
Derived
Reference notes
Object annotations comprise 2D boxes and 3D boxes with position, dimensions, yaw and class. Roll/pitch are not supplied. Image and LiDAR annotation processes have three expert review rounds; KITTI format and benchmark evaluation use the shared FoV. No platform trajectory ground truth is established.

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

Intrinsic calibration and camera–LiDAR extrinsic calibration are described, with per-sequence parameters and an overlay checking tool. Dynamic on-water misalignment remains.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Ship class predominates; dusk/night samples are limited (14.9%/5.2%, versus 79.9% day).

Small and distant objects may have very sparse LiDAR returns.

Object boxes supply yaw while assuming approximately zero pitch and roll.

Noisy extrinsics and motion cause spatial and temporal cross-modal misalignment.

Cameras

rgb

Model
Not reported
Setup
Mono
Count
1
Effective cameras
1
Modality
Rgb
Resolution
640 x 480
Rate
Not reported
Shutter
Not Reported
Lens type
Not Reported
HFOV
91.8 deg
VFOV
75.5 deg
Notes
CMOS OV2710 image sensor; complete camera model unreported. Full-HD 1080P and nighttime IR support describe the device; reported image resolution is 640 by 480. Acquisition is 30 Hz; delivered synchronized-frame cadence is unverified. IR does not establish a thermal camera.

IMUs

Pixhawk 4 flight-controller IMU

Model
Not reported
Count
1
Accel / gyro
Not reported

GNSS

u-blox M8N

Model or name
u-blox M8N
Count
1
Rate
Not reported
GNSS type
NOT REPORTED
Position output
Not reported
Velocity output
Not reported
Heading output
Not reported

LiDAR

Ouster OS1-64 gen2

Model or name
Ouster OS1-64 gen2
Count
1
Dimensions
3
Rate
Not reported
Channels
64
Range
120 m

Radar

Not reported

Pitot tube

Not reported

Additional sensors

Not reported

Annotations

Reported annotation availability and task support.

Availability
Available
Format
JSON; converters for YOLO and KITTI
Class count
[ship buoy other]
Semantic segmentation Not reported
Instance segmentation Not reported
Object detection Available
Optical flow Not reported
Depth ground truth Not reported

Citation

Citation key and BibTeX kept at the end of the page for reference.

Citation key
Not reported

BibTeX

Not reported

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy