List of Datasets

UGV

CitrusFarm

Multimodal farm localization, mapping and crop monitoring

Seven Jackal sequences in three citrus fields include RGB, grayscale, thermal and red-green-NIR imagery, LiDAR, IMU and wheel odometry with RTK position references.

Metadata: Complete Metadata source documented Released 2023

Primary places to inspect, download, read about, or reproduce the dataset.

Quick facts

Vehicle type
UGV
Environment
Outdoor
Data origin
Real-world
Sensor count
13
Ground truth
Available
Calibration
Full Raw Calibration Data
Annotations
Not reported

Sensor overview

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

Cameras

5 reported

IMUs

2 reported

GNSS

1 reported

LiDAR

1 reported

Radar

Not reported

Magnetometers

2 reported

Barometers

Not reported

UWB

Not reported

Ground truth

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

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.
3D position RTK GNSS SwiftNav Duro RTK · SwiftNav Duro 1 rover GNSS receiver 10 Hz Full 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
Processed WGS84 position reference reuses the independently released rover RTK fixes and a local base station. Fixed mode is reported throughout all seven runs. Position only: not a 6-DoF attitude reference or a numeric one-centimeter error guarantee.

Calibration and synchronization

Reported calibration level, reproducibility signals, and supporting notes.

Level
Full Raw Calibration Data
Processed parameters
Available
Raw calibration data
Available
Calibration targets
Not reported

Reproducible calibration data

Raw calibration data is reported, so users can reproduce or inspect the calibration process.

Notes

Raw calibration bags, configuration, scripts and results are documented. Kalibr, Allan variance and LiDAR-camera calibration are complemented by CAD/mechanical wheel/GNSS extrinsics; these are not independent empirical accuracy guarantees.

Known or intentional limitations

Reported constraints and characteristics to check before using the dataset.

Seven outdoor farm runs cover three citrus fields; seasonal, weather and nighttime diversity is limited. Rounded 1.7-hour/7.5-km homepage figures differ from the summed paper sequence values.

Zed2i depth and pose are computed products; R-G-NIR channels share one Mapir camera. RTK supplies position only and its reference reuses the released fixes.

Camera clocks can drift and final alignment/reference errors are unknown. Wheel/GNSS extrinsics use mechanical geometry; canopy and repeated vegetation challenge perception.

Cameras

Camera 1

Model
FLIR Blackfly BFS-U3-16S2M-CS
Setup
Mono
Count
1
Effective cameras
1
Modality
Grayscale
Resolution
1440 x 1080
Rate
10 Hz
Shutter
Global
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported

Camera 2

Model
FLIR ADK
Setup
Mono
Count
1
Effective cameras
1
Modality
Thermal
Resolution
640 x 512
Rate
10 Hz
Shutter
Global
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Thermal images; see the report for spectral-band context.

Camera 3

Model
Stereolabs Zed2i
Setup
Stereo
Count
1
Effective cameras
2
Modality
Rgbd
Resolution
1280 x 720
Rate
10 Hz
Shutter
Rolling
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Stereo RGB plus GPU-computed depth, represented once as a physical camera system.

Camera 4

Model
Mapir Survey3
Setup
Mono
Count
1
Effective cameras
1
Modality
Nir
Resolution
1280 x 720
Rate
10 Hz
Shutter
Rolling
Lens type
Not Reported
HFOV
Not reported
VFOV
Not reported
Notes
Three channels are red, green and near-infrared; nir is a coarse modality mapping, not a claim of NIR-only output.

IMUs

MicroStrain navigation IMU

Model
MicroStrain 3DM-GX5-AHRS
Count
1
Accel / gyro
200 / 200 Hz

Zed2i internal IMU

Model
Not reported
Count
1
Accel / gyro
Not reported

GNSS

SwiftNav Duro

Model or name
SwiftNav Duro
Count
1
Rate
10 Hz
GNSS type
RTK
Position output
Available
Velocity output
Not reported
Heading output
Not available

LiDAR

Velodyne VLP-16

Model or name
Velodyne VLP-16
Count
1
Dimensions
3
Rate
10 Hz
Channels
16
Range
Not reported

Radar

Not reported

Additional sensors

Magnetometer

Name
MicroStrain integrated magnetometer
Count
1
Notes
Released /microstrain/mag topic in the existing IMU package; native cadence and magnetic chip are unknown.

Magnetometer

Name
Zed2i integrated magnetometer
Count
1
Notes
Released /zed2i/zed_node/imu/mag topic in the existing camera package; native cadence and magnetic chip are unknown.

Odometry

Name
Jackal wheel odometry
Count
1
Rate
50 Hz
Notes
Encoder-derived odometry, not independent truth.

Odometry

Name
Zed2i driver pose
Count
1
Notes
Computed pose stream; rate is not established.

Annotations

Reported annotation availability and task support.

Availability
Not reported
Format
Not reported
Class count
Not reported
Semantic segmentation Not reported
Instance segmentation Not reported
Object detection Not reported
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
teng2023citrusfarm

BibTeX

@inproceedings{teng2023citrusfarm,
  author = {{Hanzhe Teng} and {Yipeng Wang} and {Xiaoao Song} and {Konstantinos Karydis}},
  title = {Multimodal Dataset for Localization, Mapping and Crop Monitoring in Citrus Tree Farms},
  year = {2023},
  booktitle = {International Symposium on Visual Computing},
  eprint = {2309.15332},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2309.15332v2}
}

Dataset corrections

Notice a specific issue after reviewing the page?

Report an inaccuracy