Clustering Persistent Scatterer Points Based on a Hybrid Distance Metric
P. Schneider, and U. Soergel. ICPR International Workshops and Challenges. ICPR 2021. Lecture Notes in Computer Science,, page 621--632. Cham, Springer International Publishing, (2021)
Abstract
Persistent Scatterer Interferometry (PSI) is a powerful radar-based remote sensing technique, able to monitor small displacements by analyzing a temporal stack of coherent synthetic aperture radar images. In an urban environment it is desirable to link the resulting PS points to single buildings and their substructures to allow an integration into building information and monitoring systems. We propose a distance metric that, combined with a dimension reduction, allows a clustering of PS points into local structures which follow a similar deformation behavior over time. Our experiments show that we can extract plausible substructures and their deformation histories on medium sized and large buildings. We present the results of this workflow on a relatively small residential house. Additionally we demonstrate a much larger building with several hundred PS points and dozens of resulting clusters in a web-base platform that allows the investigation of the results in three dimensions.
%0 Conference Paper
%1 10.1007/978-3-030-92659-5_40
%A Schneider, Philipp J.
%A Soergel, Uwe
%C Cham
%D 2021
%E Bauckhage, Christian
%E Gall, Juergen
%E Schwing, Alexander
%I Springer International Publishing
%J Pattern Recognition
%K conference review
%N vol 12667
%P 621--632
%T Clustering Persistent Scatterer Points Based on a Hybrid Distance Metric
%V ICPR International Workshops and Challenges. ICPR 2021. Lecture Notes in Computer Science,
%X Persistent Scatterer Interferometry (PSI) is a powerful radar-based remote sensing technique, able to monitor small displacements by analyzing a temporal stack of coherent synthetic aperture radar images. In an urban environment it is desirable to link the resulting PS points to single buildings and their substructures to allow an integration into building information and monitoring systems. We propose a distance metric that, combined with a dimension reduction, allows a clustering of PS points into local structures which follow a similar deformation behavior over time. Our experiments show that we can extract plausible substructures and their deformation histories on medium sized and large buildings. We present the results of this workflow on a relatively small residential house. Additionally we demonstrate a much larger building with several hundred PS points and dozens of resulting clusters in a web-base platform that allows the investigation of the results in three dimensions.
%@ 978-3-030-92659-5
@inproceedings{10.1007/978-3-030-92659-5_40,
abstract = {Persistent Scatterer Interferometry (PSI) is a powerful radar-based remote sensing technique, able to monitor small displacements by analyzing a temporal stack of coherent synthetic aperture radar images. In an urban environment it is desirable to link the resulting PS points to single buildings and their substructures to allow an integration into building information and monitoring systems. We propose a distance metric that, combined with a dimension reduction, allows a clustering of PS points into local structures which follow a similar deformation behavior over time. Our experiments show that we can extract plausible substructures and their deformation histories on medium sized and large buildings. We present the results of this workflow on a relatively small residential house. Additionally we demonstrate a much larger building with several hundred PS points and dozens of resulting clusters in a web-base platform that allows the investigation of the results in three dimensions.},
added-at = {2022-01-28T09:02:33.000+0100},
address = {Cham},
author = {Schneider, Philipp J. and Soergel, Uwe},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2f58a5b942ef76643f98998adc8c9936d/markusenglich},
editor = {Bauckhage, Christian and Gall, Juergen and Schwing, Alexander},
interhash = {b7af683c00a1217c42f0702580f11710},
intrahash = {f58a5b942ef76643f98998adc8c9936d},
isbn = {978-3-030-92659-5},
journal = {Pattern Recognition},
keywords = {conference review},
number = {vol 12667},
pages = {621--632},
publisher = {Springer International Publishing},
timestamp = {2022-03-08T07:08:38.000+0100},
title = {Clustering Persistent Scatterer Points Based on a Hybrid Distance Metric},
volume = {ICPR International Workshops and Challenges. ICPR 2021. Lecture Notes in Computer Science, },
year = 2021
}