Exploring the temporal lag between the structure and function of urban areas - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 150

housing
high density

low density

Belfast

Rep. of
Ireland

United Kingdom

FIGURE 11.1 Structural and functional representations of the Belfast study area: IKONOS image (top left), postal addresses
(top right), census housing surface (bottom left, for whole of NE Belfast; bottom right, the same spatial dimensions as
IKONOS image and postal addresses).

urban neighborhoods where increasing irregularity is reflected
in less space-filling of urban structure and summarized by a
greater fractal dimension ( D ). A useful complement to D is the
contagion index, which measures the degree of fragmentation
within the neighborhood, and a reciprocal to D is the lacunarity
measurement of the spatial distribution of gaps and holes in the
overall urban structural fabric (for full explanation see Mesev,
2005; 2007, Myint, Mesev and Lam, 2006).
Another example is provided by Fig. 11.2 which illustrates the
coupling of structural representation from unclassified IKONOS
imagery with the COMPAS postal data representing buildings
(solid boxes). The goal of structural - functional links is to use the
location and the 'area' attributes from COMPAS to help classify
built land cover from the image by, for example in Bayesian mod-
ifications of maximum likelihood, constraining apriori spectral

space. The dashed boxes represent the desired classification of
individual residential buildings that have similar areal dimen-
sions to those of the measured building areas as determined from
the postal data. Incidentally, the areal dimensions of buildings are
derived from total floorspace area and represented as graduated
squares. Encouraging results are documented in Mesev (2007)
from preliminary empirical testing on IKONOS imagery using
aerial photography at 15 cm spatial resolution. An iterative com-
putational procedure is being operationalized which links the
spatial delineation of buildings from high resolution sensor data
with the functional characteristics from postal records. Using the
software Definiens, a spectra-spatial classification using IKONOS
and COMPAS address points and implementing a nearest neigh-
bor contextual rule reported accuracies of 92.8% compared to
86.6% from a multispectral-only classification.