3D vegetation mapping using small-footprint full-waveform airborne laser scanners

  • Authors:

  • W. Wagner;M. Hollaus;C. Briese;V. Ducic

  • Affiliations:

  • Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing, Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, 1040 Wien, Austria;Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing, Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, 1040 Wien, Austria;Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing, Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, 1040 Wien, Austria;Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing, Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, 1040 Wien, Austria

  • Venue:
  • International Journal of Remote Sensing - 3D Remote Sensing in Forestry
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Small-footprint full-waveform airborne laser scanning (ALS) is a remote sensing technique capable of mapping vegetation in three dimensions with a spatial sampling of about 0.5-2 m in all directions. This is achieved by scanning the laser beam across the Earth's surface and by emitting nanosecond-long infrared pulses with a high frequency of typically 50-150 kHz. The echo signals are digitized during data acquisition for subsequent off-line waveform analysis. In addition to delivering the three-dimensional (3D) coordinates of scattering objects such as leaves or branches, full-waveform laser scanners can be calibrated for measuring the scattering properties of vegetation and terrain surfaces in a quantitative way. As a result, a number of physical observables are obtained, such as the width of the echo pulse and the backscatter cross-section, which is a measure of the electromagnetic energy intercepted and re-radiated by objects. The main aim of this study was to build up an understanding of the scattering characteristics of vegetation and the underlying terrain. It was found that vegetation typically causes a broadening of the backscattered pulse, while the backscatter cross-section is usually smaller for canopy echoes than for terrain echoes. These scattering properties allowed classification of the 3D point cloud into vegetation and non-vegetation echoes with an overall accuracy of 89.9% for a dense natural forest and 93.7% for a baroque garden area. In addition, by removing the vegetation echoes before the filtering process, the quality of the digital terrain model could be improved.