IEEE Geoscience and Remote Sensing Magazine - March 2015 - 30

data. A standardization work has also been done in order
to ensure compatibility among the different libraries and
their products: all the functions can be combined together,
thus reducing integration issues. Every function included in
the source code is completed with a brief description of its
role along with the necessary inputs and outputs. The entire
documentation is also available online [119].
Starting from the basic functions, more elaborated
workflows have been developed to handle different situations and indicators:
◗ Stack satellite is meant to process a stack of Landsat TM
images, taking care of different and necessary aspects.
The downloaded images are processed using one of
them as reference to correct a displacement, if any. Indicators like built-up areas and age of built-up areas are
extracted according to different implemented techniques, both pixel-based and object-based. The main
advantage of this workflow is its capability to process
the entire set in one click, reducing the user input to the
bare minimum.
◗ Segmentation optimizer has been specifically developed
to help the final user in defining the parameters for a
chosen segmentation by simply providing a reference of
the desired feature.
◗ Features from segments is a workflow designed to compute user-determined features from each segment and
assign the value directly to the considered segment.
These features can be used to increase the distance
among classes for a better performing classification
or to use segments as aggregators. In order to increase
the processing speed, a multi-processing version is also
made available.
◗ Building footprints are extracted from a homonymous
algorithm using a combination of very high-resolution
imagery with supervised classification and specific
morphological filters.
◗ Building height is computed using date and time of acquisition in combination with the extracted shadow
length. This software is also capable of assigning -in an
automatic way- each computed value to the corresponding building.
◗ Building density is determined by evaluating the number and areas of buildings found in a region of interest
around each polygon.
◗ Building regularity is calculated by simply finding and
measuring the two main axes of a polygon. The result is
the assignment of a "regular"/"non-regular" flag.
6. GitHub distribution
All the described workflows and functions have been
thought and designed to be free-to-use and free-to-modify and are released with the GNU General Public License
(GNU GPL) License v3. All the produced code is available
through the public GitHub repositories [117,118]. Users are
invited to test, adjust and use the code according to their
necessity, and naturally to redistribute the modified code
30

under the terms of the cited license. An installation guide is
also provided for source code and plugins with instructions
on how to setup the working environment. We also encourage readers and users to provide us feedback on the algorithms and share results in order to improve and increase
the capabilities of the produced algorithms.
7. ConClusions
In this paper we discuss, through a concrete example of algorithms definition and standard-compliant, open-source
implementation, how Earth Observation data can be useful in the context of multi-risk vulnerability assessment.
Whereas it is still frequent today that the proposed use of
"innovative" EO data in a risk estimation context sparks
hopes and then boils down to mostly visual interpretation, we tried and focus on operational set up of automated
procedures that have been recognized as suitable for riskrelated information extraction. Starting from this point,
and especially from the open-source sharing of the generated software, we would like to grow a community around
the developed tools and have more programmers to contribute. We believe in the importance of sharing code and
avoid duplicating efforts at least where the primary goal is
not commercial but rather scientific and societal at large.
ACknowledGments
Funding of this work by the EU through FP 7 projects SENSUM (call: FP7-SPACE-2012-1, project ID: 312972), RASOR
(call: FP7-SPACE-2013-1, project ID: 606888) and MARSITE (call: FP7-ENV-2012-two-stage, project ID 308417) is
gratefully acknowledged.
referenCes
[1] (2014, Aug.). Framework to integrate space-based and in-situ sensing for dynamic vulnerability and recovery monitoring (SENSUM
Project). [Online]. Available: http://www.sensum-project.eu/
[2] J. Birkmann, Measuring Vulnerability to Natural Hazards: Towards
Disaster Resilient Societies. Tokyo, Japan: United Nations Univ.
Press, 2006.
[3] United Nations Office for Disaster Risk Reduction (UNISDR),
Living with Risk: A Global Review Of Disaster Reduction Initiatives.
Geneva, Switzerland: United Nations publication, 2004.
[4] C. J. van Westen, "Multi-hazard risk assessment: RiskCity: Distance education," Enschede Netherlands, ITC, 2010.
[5] A. R. As-syakur, W. S. Adnyana, W. Arthana, and W. Nuarsa, "Enhanced built-up and bareness index (EBBI) for mapping built-up
and bare land in an urban area," Remote Sens., vol. 4, no. 12, pp.
2957-2970, 2012.
[6] J. Chen, M. Li, Y. Liu, C. Shen, and W. Hu, "Extract residential areas automatically by new built-up index," in Proc. 18th Int. Conf.
Geoinformatics, Beijing, June 2010, pp. 1-5.
[7] D. Lu and Q. Weng, "Spectral mixture analysis of the urban
landscapes in Indianapolis with Landsat ETM+ imagery," Photogramm. Eng. Remote Sens., vol. 70, no. 4, pp. 1053-1062, Sept.
2004.
[8] C. J. van der Sandea, S. M. de Jongb, and A. P. J. de Roo, "A segmentation and classification approach of IKONOS-2 imagery for
land cover mapping to assist flood risk and flood damage assessment," Int. J. Appl. Earth Observ. Geoinform., vol. 4, no. 3, pp.
217-229, June 2003.
ieee Geoscience and remote sensing magazine

march 2015

http://www.sensum-project.eu/

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2015