作者: A. Antonello , S. Franceschi , V. Floreancig , F. Comiti , G. Tonon
DOI: 10.5194/ISPRS-ARCHIVES-XLII-4-W2-27-2017
关键词:
摘要: Abstract. In the present study, we applied Particle Swarming Optimization (PSO) procedure to parametrize two Local Maxima (LM) algorithms and a pattern recognition model based on raster point-cloud datasets in order extract treetops of coniferous forests from high resolution LiDAR-data different forest structures (monoplane, biplane multi-layer) Alps region. The approach uses geomorphon algorithm DSM detect treetops. gave good results terms matching rates (Rmat: 0.8) with intermediate values commission omission (Rcom: 0.22, Rom: 0.2). Therefore, it could be valid alternative LM-algorithms when only products (DSM – CHM) are available. has been proved powerful method properly stands complex structures. This allows obtain detection estimation accuracy volume, also comparison most recent available literature data. models developed Java under Free Open Source license integrated JGrassTools library, which is now as SpatialToolbox GIS gvSIG.