Please use this identifier to cite or link to this item: https://ktisis.cut.ac.cy/handle/10488/14307
Title: Open source software DASOS: efficient accumulation, analysis, and visualisation of full-waveform lidar
Authors: Miltiadou, Milto 
Grant, Michael G. 
Campbell, Nell D.F 
Warren, Mark 
Crewley, Daniel 
Hadjimitsis, Diofantos G. 
Keywords: Software Engineering;Remote Sensing;Data analysis;Full-waveform LiDAR;Forestry
Category: Agriculture Forestry and Fisheries
Field: Engineering and Technology
Issue Date: Jun-2019
Source: https://www.spiedigitallibrary.org/conference-proceedings-of-spie/11174/2537915/Open-source-software-DASOS--efficient-accumulation-analysis-and-visualisation/10.1117/12.2537915.short
Project: Advancement of Tree Structure Observation Algorithms for FOREST Monitoring 
Journal: SPIE Library 
Conference: Seventh International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2019) 
Abstract: Full-waveform (FW) LiDAR have been available for 20 years, but compared to discrete LiDAR, there are very few researchers exploiting these data due to the increased complexity. DASOS is an open source command-line software developed for improving the adoption of FW LiDAR in Earth Observation related applications. It uses voxelisation for interpreting the data, which is fundamentally different from the state-of-art tools interpreting FW LiDAR. There are four key features of DASOS: (1) Generation of polygonal meshes by extracting an iso-surface from the voxelised data. (2) the 2D FW LiDAR metrics exported in standard GIS format; each pixel corresponds to a column from the voxelised space and contains information about the spread of the non-open voxels, (3) efficient alignment with hyperspectral imagery using a hashed table with buckets of geolocated hyperspectral pixels. The outputs of the alignment are coloured polygonal meshes, and aligned metrics. (4) The extraction of 3D raw or composite features into vectors using 3D-windows; these feature vectors can be used in machine learning for describing objects, such as trees. Machine learning approaches (e.g. random forest) could be used for classifying trees in the 3D-voxelised space.
URI: https://ktisis.cut.ac.cy/handle/10488/14307
DOI: 10.1117/12.2537915
Type: Conference Papers
Appears in Collections:Άρθρα/Articles

Show full item record

Page view(s)

15
Last Week
2
Last month
0
checked on Oct 18, 2019

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.