Resolving the Broccoli Problem: Identifying Optimal Computational Algorithms for the Accuracy Assessment of Tree Delineations from Remotely-sensed Data

Murray, Jonathan and Gullick, David Stephen and Blackburn, George Alan and Whyatt, James Duncan and Edwards, Christopher James (2018) Resolving the Broccoli Problem: Identifying Optimal Computational Algorithms for the Accuracy Assessment of Tree Delineations from Remotely-sensed Data. In: GISRUK 2018, 2018-04-17 - 2018-04-20, Leicester University.

Text (GISRUK2018_Contribution_117)
GISRUK2018_Contribution_117.pdf - Accepted Version
Download (443kB)

Abstract

For many different investigative purposes, trees and forests are aerially scanned using light detection and ranging (LiDAR). Often, this also requires the manual measurement of ground reference (GR) plots within the LiDAR scan. Upon analysis, there is regularly a mismatch between the alignment of GR and LiDAR tree locations, crown areas and tree heights. This anomaly is frequently overlooked and under-reported in the current literature. This study investigates the suitability of match pairing algorithms for the quantification of misalignment errors between two datasets representing GR and LiDAR data, and recommends an algorithm for accurately quantifying match-pairing differences.