![](bg6.png)
by a white line in Figure 3a and 3b. End result after
shadow artifact removing is represented in Figure 3c
and 3d.
In this paper we present only two shadowed
images, but our system was tested by prepress
specialists on about 200 images and sufficient
quality was obtained in 75% cases. In other cases
additional correction was required. Automation of
shadow correction process reduces time of
processing one shadowed image from about one
hour to 2-3 minutes.
8 CONCLUSIONS
The developed information technology for color
correction of shadowed images admits a high
automation degree. Its software implementation has
resulted in essential reducing the time expenditures
for prepress of colorful images. Apart from
preparation of painting reproductions, the
technology may be employed to process the areas of
images obtained by aerospace monitoring systems
and intended for print, and to provide the services of
improving the quality of digital images to a wide
range of users.
There is a difference in internal parallelism and
computational complexity degree between the steps
of the technology developed. Accounting for these
factors when elaborating a program complex, in
particular, implementation of image processing
algorithms in the GPU, allows its productivity to be
increased.
ACKNOWLEDGEMENTS
We render our thanks to the specialists of Agni
Publishing House for their qualified assistance in
computer-aided color correction and testing of
software on real images.
This work was supported by the Russian
Foundation for Basic Research (Project No. 09-07-
00269-a).
REFERENCES
Cheng L. Removing shadows from colorful images, PhD
Thesis – Simon Fraser University, 2006. – 155 p.
Weiss Y. Deriving intrinsic images from image
sequences, ICCV01 – IEEE, 2001. – V.II – P. 68-75.
McCamy M., Davidson J. G. A color-rendition chart, J.
App. Photog. In Eng. – 1976 – V.2 – P. 95-99.
Gevers T., Smeulders A. W. M. Color-based object
recognition / Patt.Rec. – 1999. – V.32 P. 453-464.
Fursov V. A., Gavrilov A. V. Parallel algorithm of data
selection using relative conforming estimate criterion /
Proceedings of The 12th ISPE International
Conference on Concurrent Engineering: Research and
Applications, Ft. Worth/Dallas, U.S.A., 25 - 29 July,
2005, p. 375-380.
Fursov V. A., Nikonorov A. V. Conformity estimation in
color lookup tables preprocessing problem. 7th
International Conference on Pattern Recognition and
Image Analysis: New Information Technologies,
St.Peterburg, 2004, Russian Federation, Volume I,
p. 213-216.
Nikonorov A. V. Constructing the conforming estimates
of non linear parameters, 4th European Congress on
Computational Methods in Applied Sciences and
Engineering, 2004, Finland. – PP. 404, 429.
Judd D., Wyszecki G., Color in business, science, and
industry – New York: Wiley, 1975. – 553 с.
Murashev D. M. Automated cytological specimen image
segmentation technique based on the active contour
model, Proceedings of Moscow Institute of Physics
and Technology. – 2009. – V.1 – N.1. P. 80-89. – (in
Russian).
Kass M., Witkin A., Terzopoulos D., Snakes: Active
contour models, Int. J. Computer vision. – N.1. –
1987. – PP. 321-331.
Soifer V. A. et al. Computer Image Processing
/ – Lightning Source Inc, 2009 – 296 pp.
Bibikov S. A., Fursov V. A., Color correction based on
models identification using test image patches,
Computer optics. – Samara-Moscow, 2008 – Т.32 №
3. – С 302-307 – (in Russian).
Smart J. Cross-platform GUI Programming with
wxWidgets – Prentice Hall, 2000. -714 p.
Munshi A. Kronos OpenCL Working Group, The OpenCL
Specification. – 2009. – 314p.
Nikonorov A. V., Fursov V. A. Distributed computational
system of color images correction, XV All-Russian
conference Telematica. – 2008. – Vol.1. – p. 88-89 –
(in Russian).
DESKTOP SUPERCOMPUTING TECHNOLOGY FOR SHADOW CORRECTION OF COLOR IMAGES
129