- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Faculty Research and Publications /
- Determining Optimal Video Length for the Estimation...
Open Collections
UBC Faculty Research and Publications
Determining Optimal Video Length for the Estimation of Building Height through Radial Displacement Measurement from Space Plowright, Andrew; Tortini, Riccardo; Coops, Nicholas C.
Abstract
We presented a methodology for estimating building heights in downtown Vancouver, British Columbia, Canada, using a high definition video (HDV) recorded from the International Space Station. We developed an iterative routine based on multiresolution image segmentation to track the radial displacement of building roofs over the course of the HDV, and to predict the building heights using an ordinary least-squares regression model. The linear relationship between the length of the tracking vector and the height of the buildings was excellent (r² ≤ 0.89, RMSE ≤ 8.85 m, p < 0.01). Notably, the accuracy of the height estimates was not improved considerably beyond 10 s of outline tracking, revealing an optimal video length for estimating the height or elevation of terrestrial features. HDVs are demonstrated to be a viable and effective data source for target tracking and building height prediction when high resolution imagery, spectral information, and/or topographic data from other sources are not available.
Item Metadata
Title |
Determining Optimal Video Length for the Estimation of Building Height through Radial Displacement Measurement from Space
|
Creator | |
Publisher |
Multidisciplinary Digital Publishing Institute
|
Date Issued |
2018-09-18
|
Description |
We presented a methodology for estimating building heights in downtown Vancouver, British Columbia, Canada, using a high definition video (HDV) recorded from the International Space Station. We developed an iterative routine based on multiresolution image segmentation to track the radial displacement of building roofs over the course of the HDV, and to predict the building heights using an ordinary least-squares regression model. The linear relationship between the length of the tracking vector and the height of the buildings was excellent (r² ≤ 0.89, RMSE ≤ 8.85 m, p < 0.01). Notably, the accuracy of the height estimates was not improved considerably beyond 10 s of outline tracking, revealing an optimal video length for estimating the height or elevation of terrestrial features. HDVs are demonstrated to be a viable and effective data source for target tracking and building height prediction when high resolution imagery, spectral information, and/or topographic data from other sources are not available.
|
Subject | |
Genre | |
Type | |
Language |
eng
|
Date Available |
2019-06-18
|
Provider |
Vancouver : University of British Columbia Library
|
Rights |
CC BY 4.0
|
DOI |
10.14288/1.0379487
|
URI | |
Affiliation | |
Citation |
ISPRS International Journal of Geo-Information 7 (9): 380 (2018)
|
Publisher DOI |
10.3390/ijgi7090380
|
Peer Review Status |
Reviewed
|
Scholarly Level |
Faculty
|
Rights URI | |
Aggregated Source Repository |
DSpace
|
Item Media
Item Citations and Data
Rights
CC BY 4.0