Remote estimation of target height from unmanned aerial vehicle (UAV) images

Tonini, Andrea (author), Redweik, Paula (author), Painho, Marco (author), Castelli, Mauro (author)

This paper focuses on how the height of a target can be swiftly estimated using images acquired by a digital camera installed into moving platforms, such as unmanned aerial vehicles (UAVs). A pinhole camera model after distortion compensation was considered for this purpose since it does not need extensive processing nor vanishing lines. The pinhole model has been extensively employed for similar purposes in past studies but mainly focusing on fixed camera installations. This study analyzes how to tailor the pinhole model for gimballed cameras mounted into UAVs, considering camera parameters and flight parameters. Moreover, it indicates a solution that foresees correcting only a few needed pixels to limit the processing overload. Finally, an extensive analysis was conducted to define the uncertainty associated with the height estimation. The results of this analysis highlighted interesting relationships between UAV-to-target relative distance, camera pose, and height uncertainty that allow practical exploitations of the proposed approach. The model was tested with real data in both controlled and uncontrolled environments, the results confirmed the suitability of the proposed method and outcomes of the uncertainty analysis. Finally, this research can open consumer UAVs to innovative applications for urban surveillance

• Language: English
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: EF - School of Economics and Business
• Publication version: Version of Record
• Year: 2020
• Number of pages: Str. 1-24
• Numbering: Vol. 12, št. 21 (art. 3602)
• PID: 20.500.12556/RUL-124230
• UDC: 659.2:004
• ISSN on article: 2072-4292
• DOI: 10.3390/rs12213602
• COBISS.SI-ID: 38407939
• Publication date in RUL: 11.01.2021

Record is a part of a journal

• Title: Remote sensing
• Shortened title: Remote sens.
• Publisher: MDPI
• ISSN: 2072-4292
• COBISS.SI-ID: 32345133

Licences

• License: CC BY 4.0, Creative Commons Attribution 4.0 International
• Link: http://creativecommons.org/licenses/by/4.0/
• Description: This is the standard Creative Commons license that gives others maximum freedom to do what they want with the work as long as they credit the author.
• Licensing start date: 11.01.2021