학술논문

MAPPING NATURAL FOREST STANDS WITH LOW-COST DRONES.
Document Type
Article
Source
Mathematical & Computational Forestry & Natural Resource Sciences. Apr2022, Vol. 14 Issue 1, p33-43. 11p.
Subject
*AERIAL photographs
*LONGLEAF pine
*DIGITAL elevation models
*PLANT phenology
*FOREST mapping
*GROUND controlled approach
Language
ISSN
1946-7664
Abstract
We used a low-cost hobby drone to produce high resolution aerial photographs of a 12 ha mature longleaf pine (Pinus palustris) stand. The photos were combined into orthophoto mosaics and digital surface models to produce repeatable crown maps. Repeated flights allowed the use of tree phenology to separate longleaf from loblolly (Pinus taeda) and pond (Pinus serotina) pines, as well as some hardwood species. Careful ground control was necessary to produce aerial crown maps that matched field measured stems. However, average crown area/stem basal area ratio of 15 m radius plots produced correlation coefficients comparable to open single tree measures, ground control improved the relationship especially for loblolly and pond pine. With ground control height measurement was comparable to SfM (Surface from Motion) research results but had a positive bias greater than 1m. The most difficult problem was determining individual trees associated with a mapped crown area. Height measures were hampered by our inability to determine a correction for the true ellipsoid height of the camera. [ABSTRACT FROM AUTHOR]