Share Email Print

Proceedings Paper

Study on the tie point selection for DEM extraction from stereo PRISM images
Author(s): Yoshiyuki Kawata; Yukihiro Funatsu; Satoshi Yoshii; Kazuya Takemata
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The accuracy of DEM extraction was analyzed from the view of tie point selection in the stereo ALOS/PRISM images, using PCI Geomatica software. In the analysis we considered three different parameters in the automatic tie point selection, namely, 1) the number of tie points, 2) the image correlation coefficient of tie points, and 3) the spatial resolution of DEM extraction. We found that a better DEM extraction accuracy was possible when we adopted a single tie point with large image correlation coefficient (around 0.8) and the spatial resolution of 2.5 (m) in the automatic tie point selection from the stereo PRISM images. In addition, we examined the dependence of the DEM extraction accuracy on the tie point’s elevation in the manual tie point selection. However, no clear dependence on the tie point’s elevation was found because of large DEM noises at tie points in the mountain area. Finally, some preliminary analysis results of DEM extraction accuracy were presented from the stereo QuickBird images.

Paper Details

Date Published: 21 November 2012
PDF: 10 pages
Proc. SPIE 8524, Land Surface Remote Sensing, 85242A (21 November 2012); doi: 10.1117/12.977183
Show Author Affiliations
Yoshiyuki Kawata, Kanazawa Institute of Technology (Japan)
Yukihiro Funatsu, Kanazawa Institute of Technology (Japan)
Satoshi Yoshii, Kanazawa Institute of Technology (Japan)
Kazuya Takemata, Kanazawa Institute of Technology (Japan)

Published in SPIE Proceedings Vol. 8524:
Land Surface Remote Sensing
Dara Entekhabi; Yoshiaki Honda; Haruo Sawada; Jiancheng Shi; Taikan Oki, Editor(s)

© SPIE. Terms of Use
Back to Top