Supervised / unsupervised change detection

de Alwis Pitts, Dilkushi and De Vecchi, Daniele and Harb, Mostapha and So, Emily and Dell'Acqua, Fabio (2014) Supervised / unsupervised change detection. University of Cambridge. ISBN UNSPECIFIED

Documents
2.6_final_dec23_1.pdf
[img]
[Download]
[img]
Preview
PDF
2.6_final_dec23_1.pdf - Table of Contents

1MB
Item Type:Book or Monograph
Item Status:Live Archive

Abstract

The aim of this deliverable is to provide an overview of the state of the art in change detection techniques and a critique of what could be programmed to derive SENSUM products. It is the product of the collaboration between UCAM and EUCENTRE. The document includes as a necessary requirement a discussion about a proposed technique for co-registration. Since change detection techniques require an assessment of a series of images and the basic process involves comparing and contrasting the similarities and differences to essentially spot changes, co-registration is the first step. This ensures that the user is comparing like for like. The developed programs would then be used on remotely sensed images for applications in vulnerability assessment and post-disaster recovery assessment and monitoring. One key criterion is to develop semi-automated and automated techniques.
A series of available techniques are presented along with the advantages and disadvantages of each method. The descriptions of the implemented methods are included in the deliverable D2.7 ”Software Package SW2.3”.
In reviewing the available change detection techniques, the focus was on ways to exploit medium resolution imagery such as Landsat due to its free-to-use license and since there is a rich historical coverage arising from this satellite series.
Regarding the change detection techniques with high resolution images, this was also examined and a recovery specific change detection index is discussed in the report.

Keywords:Change detection, Image Processing, satellite data
Subjects:G Mathematical and Computer Sciences > G400 Computer Science
F Physical Sciences > F640 Earth Science
Divisions:College of Science > School of Geography
ID Code:30024
Deposited On:09 Mar 2018 15:52

Repository Staff Only: item control page