February 2010 Volume 76, Number 2people.ucalgary.ca/~gjhay/PDFs/Peer_reviewed/2010_Hay...February...
Transcript of February 2010 Volume 76, Number 2people.ucalgary.ca/~gjhay/PDFs/Peer_reviewed/2010_Hay...February...
&PE RSPPH
OTO
GRA
MM
ETRI
C
HO
TOG
RAM
MET
RIC
EN
GIN
EERI
NG
&
NG
INEE
RIN
G &
REM
OTE
EM
OTE
SEN
SIN
GEN
SIN
G
Th
e o
ffici
al j
ou
rna
l fo
r im
ag
ing
an
d g
eo
spa
tial i
nfo
rma
tion
sci
en
ce a
nd
tech
no
log
y
February 2010 Volume 76, Number 2
Special Issue: Geographic Object-Based Image Analysis (GEOBIA)
Cover.indd 1 1/15/2010 2:04:31 PM
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING Februar y 2010 121
From global climate change to natural disaster response and national
defense, remote sensing has provided critical information on vast areas of
the Earth’s surface for over 30 years, and continues to do so today. Daily,
terabytes of data are acquired from space- and air-borne platforms, result-
ing in massive archives with incredible information potential; however it
is only recently that we have begun to mine the spatial wealth of these
archives. In essence, we are data rich, but geospatial information poor. In
most cases, data/image access is constrained by technological, national,
and security barriers, and tools for analyzing, visualizing, comparing, and
sharing these data and their extracted information are still in their infancy.
Furthermore, policy, legal, and remuneration issues related to who owns
(and are responsible for) value-added products resulting from the original
data sources, or from products that represent the culmination of many
different users input (i.e., citizen sensors) are not well understood and still
developing. Thus, myriad opportunities exist for improved geospatial in-
formation generation and exploitation.
Over the last decade a quiet paradigm shift in remote sensing image
processing has been taking place that promises to change the way we
think about, analyze and use remote sensing imagery. With it we will have
moved from more than 20 years of a predominantly pixel-spectra based
model to a dynamic multiscale object-based contextual model that at-
tempts to emulate the way humans interpret images (Hay and Castilla,
2008). However, along this new path from pixels, to objects, to (geo-)in-
telligence and the consolidation of this new paradigm, there are numerous
challenges still to be addressed (Hay and Castilla, 2006). In an effort to
better identify these challenges and their potential solutions the interna-
tional conference GEOBIA, 2008 – Pixels, Objects, Intelligence: Geo-
graphic Object Based Image Analysis for the 21st Century was held at the
University of Calgary, Alberta, Canada, 05–08 August, 5–8 in partnership
with the Canadian Space Agency, the American Society for Photogram-
metry and Remote Sensing (ASPRS), and the International Society for
Photogrammetry and Remote Sensing (ISPRS). In total, 137 participants
from 19 different countries attended this conference that included eight
industry led workshops, three keynote addresses, and 65 regular oral pre-
sentations. A special joint session titled “GEOBIA in Support of Govern-
ment of Canada Needs” was also held, as were poster sessions and a stu-
dent award for best paper/presentation. A key objective of the conference
was to facilitate a forum for this growing international community, and to
share in the latest developments of GEOBIA theory, methods, and appli-
cations. Our theme — “Pixels, Objects, Intelligence: GEOgraphic Ob-
ject-Based Image Analysis for the 21st Century” was intended to highlight
this goal, as well as the evolution of this new discipline. GEOBIA (pro-
nounced ge-o-be-ah) is a sub-discipline of GIScience devoted to develop-
ing automated methods to partition remote sensing imagery into mean-
ingful image-objects, and assessing their characteristics through scale
(Hay and Castilla, 2008). Its primary objective is the generation of geo-
graphic information (in GIS-ready format) from which new geo-intelli-
gence can be obtained (Hay, 2008). Here, geo-intelligence is defi ned as
geospatial content in context.
Interest in GEOBIA is worldwide and rapidly evolving. GEOBIA 2008
built upon the success of OBIA 2006 (Lang et al., 2006) — the 1st Inter-
national Conference on Object Based Image Analysis — held in Salzburg
Austria, which was attended by over 120 participants from 24 different
countries. An edited book (Blaschke et al., 2008) was published from
extended peer-reviewed OBIA 2006 conference papers, and OBIA and
GEOBIA Wikis have been developed to facilitate community interaction
with of over 20,000 combined views (Wiki, 2009). More recently, Bl-
aschke (2009) conducted a comprehensive literature review, analyzing
more than 820 OBIA/GEOBIA related articles (comprising 145 journal
papers, 84 book chapters, and nearly 600 conference papers). From this
review, it is evident that the early developmental years of OBIA/GEOBIA
were characterized by a dominance of grey literature; however, over the
last four to fi ve years the number of peer-reviewed journal articles has
increased sharply. This suggests that an image-processing paradigm shift
is indeed taking place within the remote-sensing community. Similarly,
GEOBIA 2008 website statistics (from 12 April 2007–05 August 2008)
revealed 58,623 conference page views from all over the world (Figure
1). Specifi cally, these views represent 17,209 visits from 5,865 individu-
als in 111 different countries/territories spread over 1,647 unique cities.
Figure 1. GEOBIA 2008 web statistics, showing examples of major page-view locations from around the world (source: Google Statistics )
In order to provide greater dissemination of the information shared
during GEOBIA 2008, conference proceedings (Hay et al., 2008) are
freely available from two online sources1. In addition, three new peer-re-
viewed GEOBIA related special journal issues are either underway (Al-
pin and Smith, 2010) or in preparation (Johansen and Bartolo, 2010;
Addink, 2011). We also note that GEOBIA 2010 will be held (29 June–02
July, 2010) in Ghent, Belgium (http://geobia.ugent.be/), with planning
already in-progress for GEOBIA 2012.
In support of the GEOBIA 2008 conference theme, this special issue is
composed of three main areas: We begin with two papers representing the
pixel theme. Johansen et al., present a comparison of geo-object- and pix-
el-based change detection applied to a high-resolution multispectral forest
scene, followed by Kim et al., describing two studies that illustrate the
importance of incorporating both spectral and non-spectral ancillary data
for GEOBIA vegetation classifi cations from very high resolution (VHR)
imagery. Within the object theme, Lizarazo and Barros present a new fuzzy
Foreword
Special Issue: Geographic Object-Based Image Analysis (GEOBIA) by Geoffrey J. Hay and Thomas Blaschke
1. Two online archives of GEOBIA 2008 proceedings may be found at (http://www.ucalgary.ca/geobia/Publishing) and (http://www.isprs.org/publications/archives.aspx) (last date accessed 02 December 2009).
February Layout 2.indd 121 1/15/2010 1:05:17 PM
122 Februar y 2010 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
image segmentation method for urban land-cover classifi cation, followed
by Smith, who argues for incorporating and exploiting existing digital car-
tography within the GEOBIA framework. This leads to a related study by
Radoux and Defourny who describe an automated GEOBIA method to
detect discrepancies between an existing (vector) forest map and a VHR
image. The fi nal theme is intelligence — referring to geo-intelligence —
which denotes the “right (geographically referenced) information” (i.e., the
content) in the “right situation” so as to satisfy a specifi c query or queries
within user specifi ed constraints (i.e., the context). The fi rst paper in this
section by Moreno et al., describes a novel geographic object-based vector
approach for cellular automata modeling to simulate land-use change that
incorporates the concept of a dynamic neighborhood. This represents a
very different approach for partitioning a scene, compared to the commonly
used GEOBIA segmentation techniques, while producing a form of tempo-
ral geospatial information with a unique heritage and attributes. The fi nal
paper by Tiede et al., presents a fully operational workfl ow for the model-
ing of 31,698 biotope complexes at the regional level with geo-objects and
a-priori knowledge. It represents one of the few published (to date) meth-
odologically sound, yet operational and transferable, approaches to semi-
automatically delineate biotope complexes.
Due to publication limitations, we regret that a number of very worthy
manuscripts were unable to be included in this special issue. Initially 21
papers were submitted, only seven have been published. Our objective in
selecting these papers is to provide a broad and relatively comprehensive
sample of the many different kinds of research topics that are being ad-
dressed with Geographic Object-Based Image Analysis. We also wish to
thank the 54 reviewers involved in the double- (and sometimes triple-)
blind review process, whose comments have enhanced the high quality
contributions found in this special issue. For those seeking additional re-
sources, we invite you to further peruse the OBIA 2006 and GEOBIA
2008 proceedings, to sample the 43 chapters of the recent book by Bl-
aschke et al., (2008) and to join us in Ghent, Belgium for GEOBIA 2010.
ReferencesAddink, E., (editor), 2011. GEOBIA 2011: Special Issue, Journal of Ap-
plied Earth Observation and Geoinformation (in progress).
Alpine, P., and G. Smith (editors), 2010. Special Issue on Object-Based
Landscape Analysis, International Journal of Geographical Informa-
tion Science (in progress).
Blaschke,T., 2009. Object Based Image Analysis for Remote Sensing, IS-
PRS Journal of Photogrammetry and Remote Sensing, in press, 42 p.
Blaschke,T., S. Lang, and G.J. Hay (editors), 2008. Object-Based Image
Analysis, Spatial concepts for knowledge-drives remote sensing appli-
cations, Series: XVII, Lecture Notes in Geoinformation and Cartogra-
phy, Springer-Verlag, p.818, 304 illustrations with CD-ROM, ISBN:
978-3-540-77057-2, URL: http://www.springer.com/978-3-540-77057-
2 (last date accessed: 02 December 2009).
Hay, G.J., 2009. GEOgraphic Object-Based Image Analysis (GEOBIA),
Developing a new sub-disclipline in GIScience, Oral presentation and
abstract, 20-22 Feb, Spatial Knowledge and Information – Canada, Fer-
nie, B.C., URL: http://rose.geog.mcgill.ca/ski/ (last date accessed: 02
December 2009).
Hay, G.J., T. Blaschke, and D. Marceau (editors), 2008. Proceedings of
GEOBIA 2008 – Pixels, Objects, Intelligence. GEOgraphic Object
Based Image Analysis for the 21st Century, University of Calgary, Cal-
gary, Alberta, Canada, 05–08 August, ISPRS Vol. XXXVIII-4/C1, Ar-
chives ISSN No. 1682-1777, 373 p., URL: http://www.ucalgary.ca/geo-
bia/Publications (last date accessed: 02 December 2009).
Hay, G.J., and G. Castilla, 2008. Geographic Object-Based Image Analy-
sis (GEOBIA): A new name for a new discipline?, in Object-Based Im-
age Analysis – Spatial Concepts for Knowledge-driven Remote Sensing
Applications, T. Blaschke, S. Lang, and G.J. Hay (editors), Springer-
Verlag, Chapter 1.4, pp. 81- 92.
Hay, G.J., and G.C. Castilla, 2006. Object-Based Image Analysis:
Strengths, weaknesses, opportunities and threats (SWOT), Proceedings
from Bridging Remote Sensing and GIS: International Symposium on
Object-based Image Analysis, 04–05 July, Salzburg, Center for Geoin-
formatics, URL:(http://www.commission4.isprs.org/obia06/ (last date
accessed: 02 December 2009).
Johansen, K., and R. Bartolo (editors), 2010. Geographic Object Based Im-
age Analysis - Special Issue, Journal of Spatial Science (in progress).
Lang. S., T. Blaschke, and E. Schöpfer (editors), 2006. Proceedings of the
1st International Conference on Object-Based Image Analysis (OBIA
2006), Salzburg University, Austria, 04–05 July, ISPRS Archieves, Vol.
No. XXXVI–4/C42ISSN 1682–1777, URL: http://www.commission4.
isprs.org/obia06/ (last date accessed: 02 December 2009).
Wiki, 2009. GEOBIA WIKI: University of Calgary, Alberta, Canada,
URL: http://wiki.ucalgary.ca/page/GEOBIA (last date accessed: 02 De-
cember 2009).
AuthorsGeoffrey J. Hay
University of Calgary
Department of Geography
2500 University Dr. NW
Calgary, AB, Canada, T2N 1N4
Phone: +1 (403) 220-8761
Fax: +1 (403) 2200-4768
Thomas Blaschke
Z_GIS Centre for Geoinformatics and
Department for Geography and Geology
University of Salzburg
Hellbrunner Str. 34, A-5020
Salzburg, Austria
February Layout 2.indd 122 1/15/2010 1:05:18 PM