| |
ISPRS Test Project on Urban Classification and 3D
Building Reconstruction
The automated extraction of urban objects from data acquired by
airborne sensors has been an important topic of research in
photogrammetry for at least two decades. One observation that can be
made is that most of the papers dealing with urban object extraction
focus on a single object class, e.g. buildings, roads or, less
frequently, trees. Typically, results are published for a few test
sites available to the authors.
The aim of this project is to provide state-of-the-art data sets which
can be used by interested researchers in order to test own methods and
algorithms on urban object classification and building reconstruction.
The results can be delivered to our working group
and will be tested against ground truth. The final (assessment) results
will be published on
the ISPRS congress in Melbourne 2012, and on this
website if desired. Further, we plan to publish in an
international photogrammetric journal.
Detailled information is available from a pdf file which can be downloaded here.
In
the
following
a
brief
overview
is
provided.
If you are interested in participation and want to download the data,
please fill in the questionnaire.
Once
we
have
reviewed
your
input,
you
will
be
provided
with
instructions
on
how
to
access the datasets.
Object classes
Two test
sites, each containing several test areas for which reference
data are available, are provided for the participants in this project
in order to evaluate techniques for the extraction of various urban
object classes:
- Urban object detection: The participants may choose
to detect single object classes, or they can try to extract several
object classes simultaneously, for instance to benefit from context
information, i.e. the information contained in the mutual arrangement
of objects in complex urban scenes such as those distributed in this
project. Reference data are available for a variety of object classes.
The object classes typically discerned in urban classification are
building, road, tree, low vegetation/grass, and artificial ground.
This
list
is
sometimes
augmented
by
car
if very high-resolution data (such as
those in this project) are used. Since such a definition is not useful
for evaluating the results of road extraction, in this project, the
class artificial ground was split into road and artificial ground other
than road.
- 3D Building Reconstruction: The participants shall
reconstruct detailed
3D roof structures in the test areas. Detailed 3D models of
roofs are
available as reference data. They will be used to evaluate the quality
of the roof plane segmentation process as well as the geometrical
accuracy of the outline polygons of the roof planes.
Dataset 1: Vaihingen/Enz, Germany
This test data set was captured over Vaihingen in
Germany. The data set is a subset of the data used for the test
of
digital
aerial
cameras carried out by the German Association of
Photogrammetry and Remote Sensing (DGPF). It consists of three test
areas for which reference data for various object classes are available
(yellow areas in the figure below) and a larger test site “Roads” for
road extraction (blue area in the same figure):
- Area 1: “Inner City”: This test area is situated in
the centre of the
city of Vaihingen. It is characterized by dense development consisting
of historic buildings having rather complex shapes, but also has some
trees.
- Area 2: “High Riser”: This area is characterized by a
few high-rising residential buildings that are surrounded by
trees.
- Area 3: “Residential Area”: This is a purely
residential area with small detached houses.
- “Roads”: This area encloses all the other test sites
and can be used for testing urban road extraction techniques.
For each of these test areas, the following data are distributed:
- Digital Aerial Images and Orientation Parameters: The
images are a part of the Intergraph/ZI DMC block with 8 cm ground
resolution. Each area is visible in multiple images from several
strips. The orientation parameters are distributed together with the
images.
- Airborne Laserscanner Data: The test area was covered
by altogether 10
strips captured with a Leica ALS50 system. Inside an individual strip
the average point density is 4 points/m2. Apart from the original point
clouds, digital surface model (DSM) is also made available.
Dataset 2: Toronto, Canada
This data set covers an area of about
1.45 km2 in the central area of the City of Toronto in Canada which was
captured by the Microsoft Vexcel’s UltraCam-D (UCD) camera and the
Optech’s airborne laserscanner (ALTM-ORION M). The images have a ground
resolution of 15 cm, while the laserscanner provided 6 points/m2. The
“Downtown Toronto”
data contains representative scene characteristics of modern mega city
in the North America including a mixture of low- and high-story
buildings with a wide variety of rooftop structures and street and road
features. The test data can be divided into three test sites, 'Area 4’,
‘Area 5’, and ‘Entire Area’. The areas ‘Area 4’ and ‘Area 5’ should be
used for testing object extraction and building reconstruction
techniques, while the entire test area can be used for road detection.
- Area 4: This area contains a mixture of low and high
story buildings, showing various degrees of shape complexity in rooftop
structure and rooftop furniture. The scene also contains different
urban objects including cars, trees, street furniture, roads and
parking lots.
- Area 5: This area represents a typical example of a
cluster of high-rise buildings in a modern mega city in North America.
The scene contains shadows casted by high buildings, under which
various types of urban objects (e.g., cars, street furniture, and
roads) can be found.
Conditions of use
Both datasets are subject to different conditions of use which are
detailled in the document. You
will need to agree to those
conditions. Especially for the Toronto
dataset two non-data disclosure agreement forms need to be filled-in
and a short description
of the intended research need to be provided and sent to the address as
indicated in the pdf document. The
documents can be downloaded here. Upon
receiption of the filled-in forms you will be provided with
instructions on how to access the data.
Important dates
- 30 September 2011: Deadline for submitting results by
the participants (if to be considered for the ISPRS congress in
Melbourne)
- 30 October 2011: Participants are informed about the
evaluation of their results
- 28 November 2011: Deadline for submitting full papers
for the
peer-reviewed track of ISPRS Commission III at the ISPRS Congress in
Melbourne
- 31 May 2012: Final deadline for submitting results by
the participants (for journal paper)
- 30 June 2012: Participants are informed about the
evaluation of their results
- 24 August–3 September 2012: ISPRS Congress in
Melbourne
- 30 September 2012: Submission of journal papers about
the test results
- 2013: Special issue of a photogrammetric journal on
the test project
Submission of results
The results of automatic object
extraction can be uploaded to the ftp server at ftp.itc.nl/incomming. The
following steps should be undertaken:
- ask M. Gerke for the credentials, in order
to be able to logon to the upload site
- connect to
ftp.itc.nl/incomming using the credentials
- create a sub folder with a suitable name
- copy the result
files in the format as indicated in the pdf document (sections 3.1 and
3.2), including an
accompanying letter with a brief description of the method (and/or
reference
to a paper) into the new folder
- provide M. Gerke with the folder name for
further processing
Help us to spread it
out and circulate the flyer!
|
wm) |
© ISPRS |