*****************************************************DATA CREATORS*****************************************************

CREATORS:		Cedric Le Gentil, Jack Naylor, Nuwan Munasinghe, Jasprabhijit Mehami, Benny Dai, 			Mikhail Asavkin, Donald G. Dansereau, Teresa Vidal-Calleja
CONTACT:		Donald G. Dansereau, donald.dansereau@sydney.edu.au

********************************************************DATA FILES*********************************************************

FILE NAMES:		ROS Bags are named according to the following convention -
			T[X]_RA_[Y]_ES_[Z]_[lighting condtions]
			X = Trajectory Number (1-3)
			Y = Sunlight Source %
			Z = Earthshine Source Included (1=True, 0=False)
			lighting conditions = Description of Y and Z combinations with local co-located lighting.


FOLDER STRUCTURE:	 dataset_0307 contains the augmented parts of the main dataset from the paper. Subfolders contain calibration and training data for the no port, and two texture (data augmentation cases). This also contains the ground truth calibration as a yml file and raw calibration ROS bag in the calibration folder. Note that the calibration is provided for the DAVIS 346 Camera only, the Prophesee data is provided without extrinsics.

dataset_0307_notexture contains the unaugmented data for training and testing.

hard_cases_benchmark contains the benchmark hard cases from the paper. hard_cases_additional provides additional difficult cases not reported in the paper.

test_datasets_aditional contains additional test datasets without calibration data. These are not benchmarked in the paper.

ROSbag Topics:
/dvs/image_raw - RGB Image from DAVIS 346 Event Camera
/dvs/events - Event Stream from DAVIS 346 Event Camera
/tf - TF Tree for UR5e Robotic Arm
/joint_states - Joint states for UR5e Robotic Arm
/prophesee/camera/cd_events_buffer - Events buffer from Prophesee EVK3+ Event Camera
/prophesee/camera/camera_info - Camera Info from Prophesee EVK3+ Event Camera
/rosout - Logging info

*****************************************************DATA SOURCES*****************************************************

WHERE:		Robotic Imaging Laboratory, Australian Centre for Robotics, The University of Sydney
WHEN:			June – September 2024
HOW:			Captured analogue satellite docking/proximity operations environment, under controlled illumination conditions.
****************************************************DATA DEFINITIONS****************************************************

UNITS:			Time in seconds, distances in meters.

**********************************************TECHNICAL REQUIREMENTS**********************************************

HARDWARE:		DAVIS 240C, Prophesee EVK3, Universal Robots UR5e robotic arm
SOFTWARE:		ROS Noetic+
FILE FORMATS:	Calibration in .yml file, datasets in ROSbag form.

**********************************************HOW TO RUN**********************************************

1. Download all .tar files from the online repository (there are a few) and the manifest.csv and restore_tars.sh file.
2. Edit restore_tars.sh, setting the TAR_DIR variable to the location containing all tars, and the RESTORE_DIR variable to the directory where you would like the folder structure to populate.
3. Run restore_tars.sh to restore the tar-ed dataset back to the original structure.
4. Play back the ROS bags.