Methodological approach for fast high-resolution image selection: FAHRIS algorithm.

Recent research highlights advancements in collecting Artificial Light at Night (ALAN) data using radiosondes on stratospheric balloons, revealing a need for enhanced in-flight image stabilization. This paper proposes a twofold approach: Firstly, it introduces a design concept for a high-resolution image acquisition and stabilization system for aerial instruments (e.g., drones, balloons). Secondly, it presents a novel Fast Algorithm for High-Resolution Image Selection (FAHRIS) for rapid image selection, grouping and stitching of acquired imagery. FAHRIS' effectiveness is validated using datasets from three flights over Italy: a stratospheric balloon flight reaching 34 kms over Florence, and drone flights using a DJI Mavic 2 up to 253 m over Trevisoand 330 m over Padua. Limitations and challenges encountered during the validation of FAHRIS, such as computational constraints affecting dataset processing, are addressed. Additionally, the results of the image stitching process highlight potential distortions and stretching issues, particularly evident in images with significant relative angles.•Design proposition of stabilization system for aerial instruments.•Development of a novel and fast image selection, grouping and stitching algorithm (FAHRIS).•Validation of algorithm against data sets from three flights.