0 2 4 6 8 10
= P
EEE
p
Figure 5. Change trends of brightness, contrast and average gradients when p value from 1-10 in color infrared aerial image shadow
removal
6. CONCLUSION
In this paper, the Shades of Grey (SoG) algorithm based on
color constancy is explored to remove shadows in true color and
color infrared urban aerial images. The shadow removal effect is
analyzed when p (the parameter in Minkowski norm) takes
different values in each experiment. Three quantitative
evaluations are utilized to assess quality of the shadow removal
result. The main results can be described as follows:
(1)For aerial images, whether true color or infrared, based on
the shadow detection result, the Shades of Grey algorithm can
achieve better shadow removal results when p takes a proper
value. In this paper, this method obtained the best overall
shadow removal effect as p=2. It is different from the normal
nature scene images which get good shadow removal result
when p=6. (2)When p =1, SoG turns into Gray-World one. But
the result is not as good when p=1. It follows that aerial
images can not be regarded simply as grey world images, and
the algorithm based on Gray-World assumptions is not suitable
for aerial images.
Since the method used here does not need a priori statistics
models, and is easier in calculating shadow removal, it is more
suitable for large area aerial image shadow removal.
REFERENCE
Arbel, E. and Hel-Or, H. 2007. Texture-Preserving Shadow
Removal in Color Images Containing Curved Surfaces. In:
Proceedings of IEEE Computer Society Conference on
Computer Vision and Pattern Recognition. Minneapolis,
Minnesota, U.S.A., pp.1-8
Buchsbaum, G., 1980. A Spatial Processor Model for Object
Colour Perception. Journal of Franklin Institute, 310(1), pp. 1-
26.
Finlayson, G.D., Hordley, S.D., and Hubel, P., 2001. Color by
correlation:A simple, unifying framework for color constancy.
IEEE Transaction Pattern Analysis Machine Intelligence, 23(11)
pp. 1209-1221.
Finlayson, G.D., Hordley, S.D., and Drew, M.S., 2002a.
Removing Shadows from Images. In: Proceedings of the 7th
European Conference on Computer Vision, Copenhagen,
Denmark , Part IV, pp. 823-850
Finlayson, G.D., Hordley, S.D., and Drew, M.S., 2002b.
Removing Shadows from Images Using Retinex. In:
Proceedings of IS&T/ SID Tenth Color Imaging Conference:
Color Science and Engineering Systems, Technologies, and
Applications, Scottsdale, Arizona, USA, pp.73-79.
Finlayson, G.D., and Trezzi, E., 2004. Shades of gray and
colour constancy. In: Proceedings of IS&T/ SID Twelfth Color
Imaging Conference, Scottsdale, Arizona, USA, pp. 37-41.
Fredembach, C., et al.., 2005. Hamiltonian Path Based Shadow
Removal. In: BMVC, Oxford,U.K., Vol. 2, pp. 502-511.
Land, E. and McCann, J., 1971. Lightness and Retinex Theory.
Journal of the Optical Society of America, 61(1),pp. 1-11
Shor, Y. and Lischinski, D., 2008 The Shadow Meets the Mask:
Pyramid-Based Shadow Removal. Computer Graphics Forum,
27(2), pp. 577-586
Susuki, A., Shio, A., Arai, H., and Ohtsuka, S., 2000. Dynamic
shadow compensation of aerial images based on color and
spatial analysis. In: Proceedings of the 15th International
Conference on Patten Recognition, Barcelona, Catalonia, Spain,
Vol. I , pp.317-320.
Weijer, J. van de, Gevers, Th., Gijsenij, A., 2007. Edge-Based
Color Constancy. IEEE Transaction Image Processing, 16(9),
pp. 2207-2214.
Joint ISO/CIE Standard ISO 11664-2:2007(EYCIE S 014-
2/E:2006
http://www.cie.co.at/index.php/index.php?i ca_id=484
ACKNOWLEDGEMENTS
The aerial images dates was supported by Shanghai Municipal
Institute of Surveying and Mapping in this research. The author
would like to thank for the help.
KEY WC
ABSTRA
The recer
three-dim
back it is
provide tl
should be
performar
NUI sens
Human |
measureir
Applicatic
applicatio
systems a
measure
solutions
white lig
overview
2005). Lit
from $3:
widesprea
Natural U
years as
Interfaces
is to free f
mouse or
touch scre
and point
Different
flight cam
However
as a NUI «
NUI sens
impact oi
Sensors w
Corp., 20
units were
This easil
units sold
below $9
only for
Sensor so
since the
pose, app
evident.
