respectively. Datasets 1 and 2 both have higher
Cm
,
Cr
, and
Ql
. It implies that the extraction of the buildings is successful
in a whole. Over 0.94 of
Cr
means that most of the buildings
were extracted, and over 0.91 of
Cm
means that only very
few false detections occurred. Of course that it is due to the
orderly ranged and similarly sized housing pattern. It also re-
flects the good capability of the proposed method in separat-
ing out the close houses. The performance on Dataset 3 is not
as good as those on Datasets 1 and 2, because the sizes of the
buildings in Dataset 3 are variable and some markers were not
extracted correctly. Nevertheless,
Ql
for Dataset 3 still reaches
to 0.76. The much higher
Cr
than
Cm
for Dataset 3 implies
fewer false alarms but more missing.
Conclusions
The objective of this work was to precisely extract houses
from
DSM
of high-density residential areas. The proposed
method aims to solve the problem of under segmentation for
touching houses caused by low quality
DSM
. To this end, we
proposed a method of marker labeling as the indicator for
house extraction. Considering that the houses are usually of
simple shapes and roofs in high-density housing areas, we
reconstruct the roof of each house as a dome through morpho-
logical operations and extract the local maximal region as the
marker. A dome is reconstructed using a collection of slices
that are derived using iterative differential openings of the
DSM
under a series of scales corresponding to the house. We
proposed a modified granulometry and associated technique
to detect these scales.
The experiments on real datasets demonstrate that the
scale detection method is robust to different data, because it
yields a scale range instead of a single value. This scale range
guarantees that at least one slice can be extracted for most of
the smaller or lower houses thus be labeled as the marker. The
evaluations for the performance shows that the quality of the
proposed extraction technique is satisfied with higher correct-
ness and completeness and fewer mistakes. Compared with
other morphology-based segmentation techniques such as the
one that Pesaresi and Benediktsson (2001) and other research-
ers proposed, our method reduces under segmentation greatly
in dense residential areas. The disadvantage is that the sparse
tiny houses may be missed because there may be no scales
corresponding to detecting them. The proposed technique can
be able to work on other data sets, including the cases of more
complex building distribution patterns.
References
Ameri, B., and D. Fritsch, 2000. Automatic 3D building recon-
struction using plane-roof structures,
Proceedings of the ASPRS
Annual Conference
, 22-25 May, Washington, D.C., pp. 22–26.
Awrangleb, M., C. Zhang, and C.S. Fraser, 2012. Building
detection in complex scenes through effective separation of
buildings from trees,
Photogrammetric Engineering & Remote
Sensing
,78(7):729–745.
Awrangjeb, M., and C.S. Fraser, 2014. Automatic segmentation of raw
LIDAR data for extraction of building roofs,
Remote Sensing
,
6(5):3716–3751.
Awrangjeb, M., C. Zhang, and C.S. Fraser, 2013. Automatic extraction
of building roofs using LIDAR data and multispectral imagery,
ISPRS Journal of Photogrammetry and Remote Sens
ing, 83:1–18.
Awrangjeb, M., M. Ravanbakhsh, and C.S. Fraser, 2010. Automatic
detection of residential buildings using LIDAR data and
multispectral imagery,
ISPRS Journal of Photogrammetry and
Remote Sens
ing, 65(5):457–467.
Brunn, A., and U. Weidner, 1997. Building extraction from digital
surface models,
IAPRS 3D Reconstruction and Modeling of
Topographic Objects
, Stuttgart, Germany, 17-19 September, Vol.
32, Part 3-4W2.
Cheng, L., J. Gong, M. Li, and Y. Liu, 2011. 3D building model
reconstruction from multi-view aerial imagery and lidar data,
Photogrammetric Engineering & Remote Sensing
, 77(2):125–139.
Fazan, A.J., and A.P.D. Poz, 2010. Building roof contours extraction
from aerial imagery based on snakes, and dynamic programming,
Proceedings of the FIG Congress 2010 - Facing the Challenges -
Building the Capacity,
11-16 April 2010, Sydney, Australia, pp.
11–16.
Foody, G., 2002. Status of land cover classification accuracy
assessment,
Remote Sensing of Environment
, 80(1):185–201.
Forlani, G., and C. Nardinocchi, 2001. Building detection and roof
extraction in laser scanning data,
International Archives of
Photogrammetry and Remote Sensing
, 34.
Förstner, W., and L. Plüumer (editors), 1997.
Semantic Modeling for
the Acquisition of Topographic Information from Images and
Maps
, Birkhüauser Verlag, Bonn, Germany.
Gerke, M., 2009. Dense matching in high resolution oblique airborne
images.
International Archives of Photogrammetry and Remote
Sensing
,
2009, XXXVIII:77–82.
Habib, A.F., R. Zhai, and K. Changjae, 2010. Generation of complex
polyhedral building models by integrating stereo-aerial imagery
and lidar data,
Photogrammetric Engineering & Remote Sensing
,
76(5):609–623.
Hao, T., J. Yang, and Y. Wang, 2010. Towards automatic building
extraction: Variational level set model using prior shape
knowledge,
Acta Automatica Sinica,
36(11):1502–1511.
Huang, X., and L. Zhang, 2011. A multidirectional and multiscale
morphological index for automatic building extraction from
multispectral Geoeye-1 imagery,
Photogrammetric Engineering &
Remote Sensing
, 77(7):721–732.
Huang, X., and L. Zhang, 2012. Morphological building/shadow
index for building extraction from high-resolution imagery over
urban areas,
IEEE Journal of Selected Topics in Applied Earth
Observations and Remote Sensing
,
5(1):161–172.
Hug, C., and A. Wehr, 1997. Detecting and identifying topographic
objects in imaging laser altimetry data,
International Archives of
Photogrammetry and Remote Sensing
, 32/3-4W2.
Katartzis, A., and H. Sahli, 2008. A stochastic framework for the
identification of building rooftops using a single remote sensing
image,
IEEE Transactions on Geoscience and Remote Sensing
,
46:259–271.
Lari, Z., and H. Ebadi, 2007. Automated building extraction from
high-resolution satellite imagery using spectral and structural
information based on artificial neural networks,
Proceedings of
the ISPRS Hannover Workshop 2007 - High-Resolution Earth
Imaging for Geospatial Information
, 29 May - 01 June, Hannover,
Germany.
Lee, D.S., J. Shan, and J.S. Bethel, 2003. Class-guided building
extraction from Ikonos imagery,
Photogrammetric Engineering &
Remote Sensing
, 69(2):143–150.
Lefèvre, S., J. Weber, and D. Sheeren, 2007. Automatic building
extraction in VHR images using advanced morphological
operators,
Proceedings of the IEEE/ISPRS Joint Workshop on
Remote Sensing
, URBAN, 11-13 April, Paris, France, pp. 1–5.
Li, Y., and L. Zhu, H. Shimamura, and K. Tachibana, 2012. A refining
method for building object aggregation and footprint modelling
using multi-source data,
ISPRS Archives for Photogrammetry
and Remote Sensing
, Volume XXXIX-B3, Melbourne, Australia,
pp.41–46.
Li, Y., B. Yong, H. Wu, R. An, H. Xu, J. Xu, and Q. He, 2014. Filtering
airborne lidar data by modified white top-hat transform with
directional edge constraints,
Photogrammetric Engineering &
Remote Sensing
, 80(2):133–141.
Matikainen, L., J. Hyyppä, and H. Hyyppä, 2001. Automatic
detection of buildings from laser scanner data for map updating,
International Archives of Photogrammetry and Remote Seneing
,
34/3-W13.
Meng, X., N. Currit, L. Wang, and X. Yang, 2012. Detect residential
buildings from lidar and aerial photographs through object-
oriented classification,
Photogrammetric Engineering & Remote
Sensing
, 78(1):35–44.
28
January 2016
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING