ANALYSIS OF THE DEGREE OF URBAN IMPERVIOUS SURFACE BASED ON OBJECT-
ORIENTED METHOD
Qulin Tan 1, Zhengjun Liu 2, Xiaofang Li 1
1.Photogrammetric Engineering & Remote Sensing, School of Civil Engineering, Beijing Jiaotong University, Beijing,
100044, China. E-mail: qltanbjtu@163.com; 2. Institute of Photogrammetry and Remote Sensing, Chinese Academy of
Surveying and Mapping, Beijing, China; E-mail: zjliu@casm.ac.cn
ABSTRACT
Analysis of the degree of urban impervious surface can
provide scientific information for urban-related issues. In
this paper, we performed a classification of the degree of
imperviousness in Miyun urban areas of China’s Beijing
region based on object-oriented method with Landsat TM
data in 2006. The approach mainly involved establishing an
image object hierarchy consisting of three levels of
different resolution. The three levels were: (1) a classified
super-scale image object level by fusing a pre-classified
thematic image objects, (2) a classified sub-scale image
object level by multi-resolution segmentation in a very high
resolution, and (3) a middle level of image objects by multi-
resolution segmentation in a relative middle scale. We
performed a final class-related classification in the middle
level using the above super-scale and sub-scale information.
After the classification completed, a thematic map of
showing the degree of imperviousness in urban areas has
been created.
Index Terms—Remote sensing, Multi-resolution
segmentation, Fuzzy classification, Object-oriented
1. INTRODUCTION
It is necessary to be aware of the development status in
urban areas in time for the planner and the manager of city.
In recent years, impervious surface has emerged not only as
an indicator of the degree of urbanization, but also as a
major indicator of environmental quality [1]. Therefore,
estimating and mapping (detecting, monitoring, and
analyzing) impervious surface is valuable not only for
environmental management but also for urban planning.
Driven by societal needs and technological advances, many
municipal government agencies have started to collect and
map impervious surface data for civic and environmental
uses [2].
Thanks to NSFC (40401037) & BJTU (2005SM036) for funding.
978-1-4244-2808-3/08/$25.00 ©2008 IEEE
Authorized licensed use limited to: GOVERNMENT COLLEGE OF TECHNOLOGY. Downloaded on June 26,2010 at 22:00:32 UTC from IEEE Xplore. Restrictions apply.
In spite of its significance, the methods for estimating
and mapping impervious surfaces and applications of
impervious surface data have not been sufficiently explored
[2]. Among the various digital remote sensing approaches
[2~6], image classification based on pixel analysis is one of
the most widely used methods in the extraction of
impervious surfaces, but results are often not satisfactory
because of the heterogeneity of urban landscapes and the
high spectral variation within the same land-cover class.
Object Oriented (OO) approach to image analysis has
advantages over pixel-based methods in many aspects [7].
The basic difference to pixel-based procedures is that OO
method does not classify single pixels, but rather image
objects (image regions). In comparison to a single pixel, an
image object offers substantially more information. Beyond
purely spectral information, image objects contain a lot of
additional attributes which can be used for classification:
shape, texture and a whole set of relational/contextual
information (semantic information). Utilizing not only
image object attributes, but also the relationship between
networked image objects, results in sophisticated
classification incorporating local context. Moreover, the
object-oriented method which first extracts homogeneous
regions and then classifies them avoids the annoying salt-
and-pepper effect of the more or less spatially finely
distributed classification results that are typical of pixel-
based analysis. Thus, with the integration of a broad
spectrum of different object features and semantic
information, the OO classification is more robust and more
accurate. The object-oriented image analysis significantly
extends the range of image analysis applications and turns
remote sensing data into more accurately classified
geographic information for various purposes [7, 8].
In this paper, we performed a classification of the degree
of imperviousness in the urban areas of county of Miyun,
China’s Beijing Region based on object-oriented method,
with the support of a subset of Landsat TM data in 2006
and eCognition software. Fig.1 shows the composite image
of the study area with TM band 1, 2 and 3. The basic
workflow of object-oriented image analysis can be found in
V - 164 IGARSS 2008
 the user guide of eCognition software. Grassland and Rural Woodland. Structure Group 3 just
contained only Water body. In the following classification-
2. METHODOLOGY based segmentation (fusion), all neighboring objects that
belong to the same structure group were combined to larger
The approach mainly involves establishing an image object objects. In this study, lengthy objects can be expected for
hierarchy consisting of three levels of different resolution. water bodies, very large objects for rural areas, and medium
In our study, different classifications have been applied to sized objects for the urban areas.
different image object levels. Information from different
image object levels has to be used in the process of
mapping the degree of imperviousness.
Firstly, we created a classified super-scale image object
level by fusing pre-classified thematic image objects. On
the highest object level only three classes would be
distinguished: possibly impervious (urban areas), which are
the areas of interest, rural areas that are not impervious, and
water. To achieve this, three abstract classes were
constructed for the highest object level. They are Water
body, Not Impervious, and Possibly Impervious. Our main
interest is the urban areas, which is where urban
imperviousness has to be expected.
The pre-classified thematic image objects were acquired
by segmentation using default scale parameter 10, and then
performing classifications using Nearest Neighbor classifier
and Membership Function. Nearest neighbor is a classifier
used to classify image objects based on given sample
objects within a defined feature space. After a number of Fig.1 Landsat TM image (R: band3;G:band2;B:band1)
sample objects have been declared as initial information for
a nearest neighbor classification, we can classify the image
objects in the whole scene using the NN classifier. After the
first NN classification, four classes: Woodland General,
Grassland General, Impervious General and Water-body
were classified. With the class-related features derived from
the first classification, we performed further classification
for the above general classes. For example, using the class-
related feature of the relative area of objects of the class
Impervious General within a perimeter of 10 pixels around
the object in question, the child classes (Urban Grassland
and Rural Grassland) of Grassland General can be further
classified. Using the same class-related feature, the child
classes (Urban Woodland and Rural Woodland) of
Woodland General can be also further classified. With the
Impervious General, the child classes of it (Urban
Impervious and Rural Impervious) were classified by
combining the features of area and density with the class- Fig.2 Classified super-scale image object level
related feature of relative border to neighbor objects
classified as Urban Impervious. The latest class-related Now this highest image object level already existed due
classification provided the knowledge base for the fusion of to the classification-based fusion. The objects in this level
the image objects to larger, meaningful objects. will become super-objects of all the objects in the two
After the above two classifications, three structure following generated lower levels. We had to further classify
groups of classes were constructed. Structure Group 1 the highest image object level, because super-object’s
contained all urban classes: Urban Impervious, Urban classification information was needed to use for the
Grassland and Urban Woodland. Structure Group 2 classification of the sub-objects by determining whether
contained all rural classes: Rural Impervious, Rural

V - 165

Authorized licensed use limited to: GOVERNMENT COLLEGE OF TECHNOLOGY. Downloaded on June 26,2010 at 22:00:32 UTC from IEEE Xplore. Restrictions apply.
 they belong to a possibly impervious or not impervious
super object.
On this image object level, water objects can easily be
discerned using a spectral feature (the near infrared mean
value). To fully assign an object to Water body class, its
near infrared mean value must be 50 or smaller. All objects
that urban imperviousness can occur in were assigned to
Possibly Impervious. It was known from the former
classification and fusion that these are objects that have
been created out of the former urban objects. With that
knowledge, the following class rules can be built. Objects
of the class Possibly Impervious: (1) must have a near
infrared mean value larger than 90 and less than 95, (2)
must be of 25,000 pixels or smaller in size, and (3) are not
objects of the class Water body. Objects of the class Not
Impervious are neither objects of the class Possibly
Impervious nor objects of the class Water body. Fig. 2
shows the final classified super-scale image object level.
This classification provides super-scale information for
the classification of the lower levels. That is, for each sub-
object in a lower level, it will know if it belongs to a
possibly impervious urban area, a non-impervious rural
area or water body. In this way, the super-scale
classification information can aid us in further
classifications.
Fig.3 Classified small-scale image object level
Secondly, we created a classified sub-scale image object
level by multi-resolution segmentation on the near infrared
layer of the data subset. In order to create image objects
that represent single structures like rows of houses or
streets in urban environments, a very high resolution (scale
parameter: 3) had been chosen for the segmentation. The
segmentation yielded some image objects that should be
small enough to represent urban structures and can be
V - 166
Authorized licensed use limited to: GOVERNMENT COLLEGE OF TECHNOLOGY. Downloaded on June 26,2010 at 22:00:32 UTC from IEEE Xplore. Restrictions apply.
classified as Completely Impervious.
Now the objective is to determine the objects in the
lowest level that represent completely impervious or dense
spots. It has already been ascertained that the near infrared
layer is best for analyzing urban structures. It is advisable
to extract features from it for the classification of dense
areas as well. So, the following class rules can be
formulated: objects of the class Completely Impervious
have a near infrared mean value less than or equal to 59,
and also have a mean value in blue band layer of 70 or
higher. After classification on this level completion, urban
structures are well represented. Fig. 3 shows the classified
small-scale image object level.
Fig.4 Classification of the degree of imperviousness in
urban areas on middle-scale image object level
Thirdly, we created middle level of image objects by
multi-resolution segmentation in a relative middle scale
(scale parameter: 23) using only the near infrared layer
again. This segmentation yielded objects that were
relatively homogeneous in their degree of urban
imperviousness. In essence, we classify the image on this
level to create an imperviousness thematic map of the area.
Together with the classification-based-fusion super scale
level and the sub-scale level, we have now created three
image object levels of different resolution. Each of the
image objects created knows its neighbors as well as its
super- and sub-objects along with all their attributes and
class assignments.
From the lowest level, we can acquire information on the
actual degree of urban impervious surface by calculating
the relative area of sub-objects classified as Completely
Impervious. By using super-scale information, we can
determine where urban imperviousness can occur for
objects of middle image object level. An object will be
 assigned to Urban Area, if it has a super-object classified as
Possibly Impervious. By analogy, an object is assigned to
the class Pervious Rural Area, if it is not Urban Area and
not Water body.
Finally, we performed a final class-related classification
in the middle level, using the above sub-scale and super-
scale information. Urban imperviousness only occurs in
areas represented by objects classified as Urban Area.
Other than Water and Not Impervious, for each object
classified as Urban Area the part of its sub-objects’ area
classified as Completely Impervious will be calculated.
Thus, image objects in urban areas have been divided into
six classes (0%, 1-25%, 26-50%, 51-75%, 76-99%, and
100% impervious) according to the actual degree of
imperviousness they represent. After the classification
completed, a thematic map of showing the degree of
imperviousness in urban areas has been created. Fig.4
shows the final classification of the degree of
imperviousness in urban areas on middle-scale image object
level.
From the final classification of imperviousness degree,
we can examine what share of the urban area the single
impervious degrees cover by dividing the total areas of the
impervious degree classes by the total areas (urban area).
Tab.1 shows the statistical result. Also, analysis of their
spatial distributions can give us important information
about the economic and environmental implications relating
with the city development status.
Tab. 1 Statistical Result of the Degree of Urban
Imperviousness
Share of urban area
Class
(%)
0% impervious 4.1
1-25% impervious 11.0
26-50% impervious 24.6
51-75% impervious 27.7
76-99% impervious 31.8
100% impervious 0.8
3. DISCUSSIONS
Most previous researches for extraction of impervious
surfaces in urban areas adopted pixel-based methods on the
medium spatial resolution images such as Landsat
TM/ETM+ and Terra’s ASTER images. However, their
spatial resolutions are regarded as too coarse, due to the
heterogeneity of urban landscapes and the complexity of
impervious surface materials. Urban landscapes are
typically composed of features that are smaller than the
spatial resolution of such sensors and are a complex
V - 167
Authorized licensed use limited to: GOVERNMENT COLLEGE OF TECHNOLOGY. Downloaded on June 26,2010 at 22:00:32 UTC from IEEE Xplore. Restrictions apply.
combination of buildings, roads, grass, trees, soil, water,
and so on. The object-based approach is still fairly new in
various remote sensing application fields, but it is attracting
more and more attention on the part of this community. In
our study, although we also just tested the Landsat TM
dataset, but we have proved that for estimating and
mapping of the degree of urban impervious surface, the
object-oriented remote sensing image analysis method is an
effective, simple and rapid way to acquire the urban
impervious surface information. It is not only suitable for
the medium spatial resolution images, but also especially
suitable for high-resolution images for its huge data volume,
because segmentation drastically reduces the sheer number
of units to be handled for classification. Even if a lot of
intelligence is applied to the analysis of each single image
object, the classification works relatively fast. Furthermore,
homogeneous image objects provide a significantly
increased signal-to-noise ratio compared to single pixels as
to the attributes to be used for classification. Thus,
independent of the multitude of additional information, the
classification is more robust and more accurate comparing
with pixel-based methods when using the same resolution
remote sensing data.
4. REFERENCES
[1] Arnold, C.L. Jr., and Gibbons, C.J., “Impervious surface
coverage: the emergence of a key environmental indicator”.
Journal of the American Planning Association, 1996, 62, 243–258.
[2] Qihao Weng, Remote sensing of impervious surfaces (edited
by Qihao Weng), Taylor & Francis series in remote sensing
applications, Boca Raton, CRC Press, 2008.
[3] S. Phinn, M. Stanford, P. Scarth, A. T. Murray, and P. T. Shyy,
“Monitoring the composition of urban environments based on the
vegetation impervious surface-soil (VIS) model by subpixel
analysis techniques,” Int. J. Remote Sens., vol. 23, no. 20, pp.
4131-4153, 1999.
[4] M. Ji and J. Jensen, “Effectiveness of subpixel analysis in
detecting and quantifying urban imperviousness from Landsat
Thematic Mapper imagery,” Geocarto Int., vol. 14, no. 4, pp. 33-
41, 1999.
[5] C. Wu and A. T. Murray, “Estimating impervious surface
distribution by spectral mixture analysis,” Remote Sens. Environ.,
vol. 84, pp. 493-505, 2003.
[6] Xiaojun Yang, “Estimating Landscape Imperviousness Index
From Satellite Imagery”, IEEE Geoscience and Remote Sensing
Letters, vol. 3, no. 1, January 2006, pp. 6-9.
[7] Baatz, M., Heyen, M., and Hofman, P. (2004). eCognition ̢
Object oriented image analysis. User Guide. Definiens AG.
Abrufbar unter: http://www.definiens-imaging.com.
[8] Baatz, M. and Sch¨ape, A. (2000). Multiresolution
segmentation – an optimization approach for high quality multi-
scale image segmentation. In: J. Strobl, and T. Blaschke (Eds.):
Angewandte Geographische Informationsverarbeitung XII.
Beitr¨age zum AGIT-Symposium Salzburg. Heidelberg, New
York.