Dataset: sar-dag_raycast.zip, 2.8 GB Access Condition: Open access Description: The "3D_terrain_data" folder contains data required to rebuild 3D terrain. Subfolder "20220206 downloaded data" contains NASA'S DEM and OSM map, both downloaded on 06/02/2022. Subfolder "geo" contains generated 3D terrain (30 m and 100 m resolution) in Houdini's .bgeo.sc format. The folder "drone_data" contains the "CSVs" subfolder, with 9 CVSs prepared as the raycaster's input. Subfolder "flights" contains data for the 7 flights. There are 3 subfolders ("DET", "GT" and "REC") in each flight's folder. "DET" contains images with our detections marked. "GT" contains ground truth stored in text files. "REC" contains raw recorded images (with metadata). Experiment details are available in the "Application of Raycast Method for Person Geolocalization and Distance Determination Using UAV Images in Real-World Land Search and Rescue Scenarios" paper. (English)
Documentation: readme.txt, 1.48 KB Access Condition: Open access
Ivašić Kos, M., Paulin, G. & Sambolek, S. (2023). SAR-DAG_raycast DATASET [Data set]. https://urn.nsk.hr/urn:nbn:hr:195:686105.
Ivašić Kos, Marina, et al. SAR-DAG_raycast DATASET. Fakultet informatike i digitalnih tehnologija, 2023. 25 Aug 2024. https://urn.nsk.hr/urn:nbn:hr:195:686105.
Ivašić Kos, Marina, Goran Paulin, and Saša Sambolek. 2023. SAR-DAG_raycast DATASET. Fakultet informatike i digitalnih tehnologija. https://urn.nsk.hr/urn:nbn:hr:195:686105.
Ivašić Kos, M., Paulin, G. and Sambolek, S. 2023. SAR-DAG_raycast DATASET. Fakultet informatike i digitalnih tehnologija. [Online]. [Accessed 25 August 2024]. Available from: https://urn.nsk.hr/urn:nbn:hr:195:686105.
Ivašić Kos M, Paulin G, Sambolek S. SAR-DAG_raycast DATASET. [Internet]. Fakultet informatike i digitalnih tehnologija: Kutina, HR; 2023, [cited 2024 August 25] Available from: https://urn.nsk.hr/urn:nbn:hr:195:686105.
M. Ivašić Kos, G. Paulin and S. Sambolek, SAR-DAG_raycast DATASET, Fakultet informatike i digitalnih tehnologija, 2023. Accessed on: Aug 25, 2024. Available: https://urn.nsk.hr/urn:nbn:hr:195:686105.
Title (english)
SAR-DAG_raycast DATASET
Author
Marina Ivašić Kos Faculty of Informatics and Digital Technologies, University of Rijeka; Centre for Artificial Intelligence, University of Rijeka
Author
Goran Paulin Kreativni odjel d.o.o.
Author
Saša Sambolek High school Tina Ujevića, Kutina
Scientific / art field, discipline and subdiscipline
SOCIAL SCIENCES Information and Communication Sciences Information Systems and Information Science
Scientific / art field, discipline and subdiscipline
TECHNICAL SCIENCES Computing Data Processing
Abstract (english)
A dataset is used to train and define a system that can precisely geolocate persons automatically detected in offline processed images recorded during the SAR mission. The dataset contains data 3D simulations for a few real-world terrains of different configurations and complexity using a custom-made 3D terrain generator and raycaster, along with a person detections with YOLO deep neural network on the real terrain. The collected data is used to define a method for geolocating detected persons based on raycasting, which allows using low-cost commercial drones with a monocular camera in SAR missions.
Number: 101087348 Title (english): HORIZON EUROPE Widening participation and spreading excellence INNO2MARE project Acronym: INNO2MARE Leader: UNIRI Jurisdiction: eu
Publisher
Fakultet informatike i digitalnih tehnologija Faculty of Informatics and Digital Technologies
Access conditions
Open access
Terms of use
Public note (english)
Experiment details are available in the "Application of Raycast Method for Person Geolocalization and Distance Determination Using UAV Images in Real-World Land Search and Rescue Scenarios" paper.