Land-cover identification from airborne waveform LiDAR using a waveform stacking method

Hsien Ming Wu, Tee-Ann Teo

Research output: Contribution to conferencePaperpeer-review

Abstract

Airborne waveform LiDAR acquires high precision 3-D points and records a succession of the entire backscattered energy. Traditionally, the received waveform was analyzed individually to obtain physical features by waveform decomposition technique. The relationship between neighboring waveforms is usually ignored in waveform analysis. However, the integration of adjacency waveforms may produce 3-D backscattered energy rather than 1-D backscattered energy. The objective of this study is to explore the feasibility of waveform stacking technique from airborne waveform LiDAR data. This study employs several waveform features in land cover identification, for example, echo width, amplitude, intensity, normalized digital surface model, echo ratio, backscattering coefficient and the features analyzed from waveform stacking. The benefit of waveform stacking in land cover identification is also discussed. The proposed scheme contains three major works. First, all the adjacency 1-D waveforms are stacked into 3-D grids based on pulse coordinates (X, Y, Z). The results of waveform stacking represent the return energy of objects for each cell. The principal components analysis (PCA) is applied to 3-D grids to reduce data volume, and the entropy is used to describe the space texture. Second, we use cross-correlation to validate the relevance of different features. Finally, we combine the results of waveform stacking and other waveform features in object-oriented classification. After image segmentation, a nearest neighbor configuration (NN) classifier is selected in classification. The test data was located at Miao-li City, Taiwan. These multi-strip lidars were acquired by Riegl Q680i. The experimental results indicated that the waveform stacking technique is able to combine multiple waveforms into 3-D voxels for object identification.

Original languageEnglish
StatePublished - 1 Jan 2015
Event36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015 - Quezon City, Metro Manila, Philippines
Duration: 24 Oct 201528 Oct 2015

Conference

Conference36th Asian Conference on Remote Sensing: Fostering Resilient Growth in Asia, ACRS 2015
CountryPhilippines
CityQuezon City, Metro Manila
Period24/10/1528/10/15

Keywords

  • Identification
  • Land cover
  • Lidar
  • Waveform
  • Waveform stacking

Fingerprint Dive into the research topics of 'Land-cover identification from airborne waveform LiDAR using a waveform stacking method'. Together they form a unique fingerprint.

Cite this