Optimized feature fusion of lidar and hyperspectral data for tree species mapping in closed forest canopies

Frieke Vancoillie, Wenzhi Liao, P Kempeneers, K Vandekerkhove, Sidharta Gautama, Wilfried Philips, Robert De Wulf, Akira Iwasaki (Editor), Naoto Yokoya, Jocelyn Chanussot

Research output: Contribution to conferencePaperpeer-review

6 Citations (Scopus)

Abstract

This study deals with data fusion of hyperspectral and LiDAR sensors for tree species mapping in complex, closed forest canopies in Belgium. In particular, seven tree species were mapped: Beech, Ash, Larch, Poplar, Copper beech, Chestnut and Oak. The added value of LiDAR height profile data on tree species mapping was assessed. Sensor data were fused in the PCA domain, while optimal feature combination was derived from the best classification performance (in terms of Kappa and producer's accuracy) based on 5-fold cross-validation. Besides, varying training set sizes were tested (resp. 10%, 30% and 50% number of samples per tree species class). Feature fusion of PCA-transformed HS and LiDAR data was most effective for small sample set sizes reaching a Kappa accuracy improvement of 10.51%.
Original languageEnglish
Pages1-4
Number of pages4
DOIs
Publication statusPublished - 2015
Event7th workshop on hyperspectral image and signal processing : evolution in remote sensing 2015 - Tokyo, Japan
Duration: 2 Jun 20155 Jun 2015

Workshop

Workshop7th workshop on hyperspectral image and signal processing : evolution in remote sensing 2015
Country/TerritoryJapan
CityTokyo
Period2/06/155/06/15

Keywords

  • LiDAR
  • hyperspectral
  • feature fusion
  • closed forest canopy
  • tree species mapping
  • atmospheric modeling
  • geophysical image processing
  • image classification
  • principal component analysis

Fingerprint

Dive into the research topics of 'Optimized feature fusion of lidar and hyperspectral data for tree species mapping in closed forest canopies'. Together they form a unique fingerprint.

Cite this