In: Wagner W„ Szekely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B
647
the detection mask. Varying water level can result in detected
changes at the coast.
Obviously, every object or parameter which caused a
significant backscatter change will result in a detection.
However, customers may only be interested in specific changes,
like e.g. in new constructions.
To analyse the scenery to gain additional knowledge for the
qualification of detected changes, different measures have been
used for segmentation and classification: Coefficient of
variation, Coherence and statistics of the intensity.
Five broad classes have been evaluated: water, forest, open
land, urban and coastal areas. The result of this classification is
shown in Figure 5.
Assuming a certain application and that a user is only interested
in changes on land and not in variations on the sea e.g. due to
waves, varying water level and ships, the following reduction
has been applied. Changes on water and coastal areas have been
eliminated as unrequested changes.
The information about the change direction of the backscatter
(reduced or increased) has been combined with the
classification results of Figure 5. The result is shown in Figure
6 and a detail of it in Figure 7.
Figure 7: Detail of detected and assessed changes.
The clearing on the left hand side has been classified as a
change type in forest areas with a decreased backscatter, which
is typically the case for clearings. At the new forest edges an
increased backscatter is caused by layover.
The new construction at the area around the airport were
classified as open land, which is true since in the first
acquisition there was bare soil and the increased backscatter
gives the indication of a new object. The temporarily parked
wagons at the runway are classified as a change in urban areas,
where usually many changes are occurring independent on the
change direction. Domain specific interpretation knowledge is
needed to verify the plausibility of such changes and qualify it
as temporary change.
The detected and assessed changes are plausible. However, the
assessment depends on the quality of the classification. Even
though no apriori information and only a small amount of SAR
data sets have been available the customer can be provided
with, a rough idea about the type of occurring changes.
5. CONCLUSION
In this paper an approach for a combined detection and
qualification of changes has been presented. The integration of
additional information derived directly from the images is
helpful to restrict search areas and to give more information to
judge the plausibility of changes, if other a-priori information is
not available.
The assessment depends on the accuracy of the classification of
the imagery. However, an indication about the change direction
can be given and can be helpful for further interpretation.
In future tasks, the improvement of the assessment can be
conducted by using e.g. dual-polarised data and a comparison
of classifications of pre and post event images.
REFERENCES
Hanssen, R. F., 2001. Radar Interferometry, Data
Interpretation and error analysis. Kluwer Academic Publishers,
Dordrecht.
Preiss M., Gray D., Stacy N., 2003. A change detection
technique for repeat pass interferometric SAR. IEEE
International Geoscience and Remote Sensing Symposium
(IGARSS), Vol. 2, pp. 938-940.
Preiss M., and Stacy N.J.S., 2006. Coherent Change Detection:
Theoretical Description and Experimental Results. STO-TR-
1851
Scheuchl B., Ullmann T., Koudogbo F., 2009. Change
Detection using high resolution TerraSAR-X data preliminary
results. The International Archives of the Photogrammetry,
Remote Sensing and Spatial Information Sciences. XXXVIII-1-
4-7/W5.
Schmitt A., Wessel B., Roth A., 2009. Curvelet approach for
SAR Image denoising structure enhancement and change
detection. City Models, Roads and Traffic (CMRT), 03,-
04.09.2009 , Paris (France).
Touzi, R. Lopes, A. Bousquet, P., 1988. A statistical and
geometrical edge detector for SAR images. IEEE Transactions
on Geoscience and Remote Sensing, Vol 26 (6), pp. 764-773.
Wegmiiller, U., Strozzi, T., Wiesmann, A., Werner, C., 2003.
ENVISAT ASAR for land cover information. In Proceedings of
IGARSS’03, Toulouse.
Wright P., Macklin T., Willis C., Rye T., 2005. Coherent
Change Detection with SAR. 2nd EMRS DTC Technical
Conference, Edinburgh