Please use this identifier to cite or link to this item:
https://hdl.handle.net/20.500.14279/9545
Title: | Change detection from very high resolution satellite time series with variable off-nadir angle | Authors: | Barazzetti, Luigi Brumana, Raffaella Cuca, Branka Previtali, Mattia |
Major Field of Science: | Engineering and Technology | Field Category: | Civil Engineering;Civil Engineering | Keywords: | Automation;Change Detection;Monitoring;Natural Disaster;Satellite Images | Issue Date: | 1-Jan-2015 | Source: | 3rd International Conference on Remote Sensing and Geoinformation of the Environment, RSCy 2015; Paphos; Cyprus; 16 March 2015 through 19 March 2015 | DOI: | 10.1117/12.2192429 | Abstract: | Very high resolution (VHR) satellite images have the potential for revealing changes occurred overtime with a superior level of detail. However, their use for metric purposes requires accurate geo-localization with ancillary DEMs and GCPs to achieve sub-pixel terrain correction, in order to obtain images useful for mapping applications. Change detection with a time series of VHS images is not a simple task because images acquired with different off-nadir angles have a lack of pixel-to-pixel image correspondence, even after accurate geo-correction. This paper presents a procedure for automatic change detection able to deal with variable off-nadir angles. The case study concerns the identification of damaged buildings from pre- and post-event images acquired on the historic center of L'Aquila (Italy), which was struck by an earthquake in April 2009. The developed procedure is a multi-step approach where (i) classes are assigned to both images via object-based classification, (ii) an initial alignment is provided with an automated tile-based rubber sheeting interpolation on the extracted layers, and (iii) change detection is carried out removing residual mis-registration issues resulting in elongated features close to building edges. The method is fully automated except for some thresholds that can be interactively set to improve the visualization of the damaged buildings. The experimental results proved that damages can be automatically found without additional information, such as digital surface models, SAR data, or thematic vector layers. | URI: | https://hdl.handle.net/20.500.14279/9545 | ISBN: | 978-162841700-5 | Rights: | © 2015 Copyright SPIE. | Type: | Conference Papers | Affiliation : | Cyprus University of Technology Politecnico di Milano |
Publication Type: | Peer Reviewed |
Appears in Collections: | Δημοσιεύσεις σε συνέδρια /Conference papers or poster or presentation |
CORE Recommender
Page view(s) 20
459
Last Week
1
1
Last month
14
14
checked on Aug 28, 2024
Google ScholarTM
Check
Altmetric
Items in KTISIS are protected by copyright, with all rights reserved, unless otherwise indicated.