Comparison of spruce and beech tree attributes from field data, airborne and terrestrial laser scanning using manual and automatic methods

0544858 - ÚVGZ 2022 RIV NL eng J - Článek v odborném periodiku
Novotný, Jan - Navrátilová, Barbora - Albert, J. - Cienciala, E. - Fajmon, Lukáš - Brovkina, Olga
Comparison of spruce and beech tree attributes from field data, airborne and terrestrial laser scanning using manual and automatic methods.
Remote Sensing Applications. Roč. 23, AUG (2021), č. článku 100574. E-ISSN 2352-9385
Grant CEP: GA MZe(CZ) QK1910150
Výzkumná infrastruktura: CzeCOS III - 90123
Institucionální podpora: RVO:86652079
Klíčová slova: Airborne LiDAR * Beech * Field-map * Spruce * Terrestrial laser scanning * Time efficiency * Tree-level forest inventory
Obor OECD: Environmental sciences (social aspects to be 5.7)
Způsob publikování: Open access
https://www.sciencedirect.com/science/article/pii/S2352938521001105?via%3Dihub

This study's aim was to make a comparative analysis of airborne (ALS) and terrestrial (TLS) laser scanning and field data collection methods in estimating the tree attributes height (H), diameter at breast height (DBH), crown base (CB), and crown diameter (CD). Manual and automatic methods of extracting tree attributes from ALS and TLS data were evaluated for spruce and beech at a study site within the Czech part of the Silesian Beskids. Strong correlation (r) was observed among scanning H and field-measured H for spruce (r = 0.91 for ALS automatic vs. field, r = 0.94 for ALS manual vs. field, r = 0.91 for TLS manual vs. field), and with standard error of the estimates (SEE) from 1.4 m to 2.2 m. DBH estimates from TLS manual method showed r = 0.96 for spruce and beech with average SEE of 2.9 cm. DBH estimates for beech from TLS automatic method were poorer (r = 0.73, SEE = 9.1 cm) than for spruce (r = 0.96, SEE = 2.5 cm). TLS estimates of CB were better both for manual and automatic methods (r = 0.85, SEE = 1.4 m for manual, r = 0.73, SEE = 2.4 for automatic) than ALS estimates (r = 0.67, SEE = 4.0 m for manual, r = 0.61, SEE = 4.2 m for automatic). CD estimates from ALS showed better agreement with field measurements (r = 0.71, SEE = 1.2 m) than did CD estimates from TLS (r = 0.52, SEE = 1.4 m). TLS technology was evaluated as being most time demanding for both manual and automatic modes in comparison with ALS and field technologies. Field technology for in situ measurement was shown to be time effective compared with ALS and/or TLS technologies that used manual methods to extract tree attributes. ALS technology with automatic methods of tree attributes extraction was found to be most efficient, showing that automation is a key process and has strong potential for applications to inventory trees within forests over large areas in Central Europe.
Trvalý link: http://hdl.handle.net/11104/0322538