University of North Texas
High Resolution Satellite Images and LiDAR Data for Small-Area Building Extraction and Population Estimation
Abstract
dc:descriptionPopulation estimation in inter-censual years has many important applications. In this research, high-resolution pan-sharpened IKONOS image, LiDAR data, and parcel data are used to estimate small-area population in the eastern part of the city of Denton, Texas. Residential buildings are extracted through object-based classification techniques supported by shape indices and spectral signatures. Three population indicators -building count, building volume and building area at block level are derived using spatial joining and zonal statistics in GIS. Linear regression and geographically weighted regression (GWR) models generated using the three variables and the census data are used to estimate population at the census block level. The maximum total estimation accuracy that can be attained by the models is 94.21%. Accuracy assessments suggest that the GWR models outperformed linear regression models due to their better handling of spatial heterogeneity. Models generated from building volume and area gave better results. The models have lower accuracy in both densely populated census blocks and sparsely populated census blocks, which could be partly attributed to the lower accuracy of the LiDAR data used.
Degree
thesis:*- Grantor dc:publisher
- University of North Texas
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ramesh, Sathya
- Contributors dc:contributor
-
- Dong, Pinliang
- Yuan, Xiaohui (Computer science researcher)
- Tiwari, Chetan
Subjects
dc:subject × 9Rights
dc:rights- Statement dc:rights
-
- Public
- Copyright
- Ramesh, Sathya
- Copyright is held by the author, unless otherwise noted. All rights reserved.
- Language dc:language
- English
Identifiers
dc:identifier.*- Identifier
-
oclc: 605876293
untcat: b3825938
https://digital.library.unt.edu/ark:/67531/metadc12188/
ark: ark:/67531/metadc12188