Portal de Periódicos da CAPES

Automatic Mapping of Forest Stands Based on Three-Dimensional Point Clouds Derived from Terrestrial Laser-Scanning

2017; Multidisciplinary Digital Publishing Institute; Volume: 8; Issue: 8 Linguagem: Inglês

10.3390/f8080265

1999-4907

Tim Ritter, Marcel Schwarz, Andreas Tockner, Friedrich Leisch, Arne Nothdurft,

Forest Ecology and Biodiversity Studies

Mapping of exact tree positions can be regarded as a crucial task of field work associated with forest monitoring, especially on intensive research plots. We propose a two-stage density clustering approach for the automatic mapping of tree positions, and an algorithm for automatic tree diameter estimates based on terrestrial laser-scanning (TLS) point cloud data sampled under limited sighting conditions. We show that our novel approach is able to detect tree positions in a mixed and vertically structured stand with an overall accuracy of 91.6%, and with omission- and commission error of only 5.7% and 2.7% respectively. Moreover, we were able to reproduce the stand’s diameter in breast height (DBH) distribution, and to estimate single trees DBH with a mean average deviation of ±2.90 cm compared with tape measurements as reference.