Sector 2
Sector for Digital Optics
Head of Sector
– Dr. Nikolay Merzlyakov
Tel.: (095)
209-28-83; E-mail: nick@iitp.ru;
Nick_Merzlyakov@hotmail.com
The leading researchers of the laboratory include:
|
Dr.Sc. (Math.) |
L. Yaroslavsky |
Dr. |
V. Kober |
|
Dr. |
T. Belikova |
Dr. |
V. Lashin |
|
Dr. |
I. Bockstein |
Dr. |
M. Mozerov |
|
Dr. |
V. Karnaukhov |
|
|
Directions of activity:
·
development of relative databases and DBMS;
·
digital image enhancement;
·
optical-digital methods of image processing and pattern recognition;
·
synthesis of two-dimensional digital filters;
·
multimedia;
·
three-dimensional scene reconstruction;
·
medical image analysis and classification;
·
digital holography.
The method of digital restoration of different kinds of
watermark images (watermarks, laid, and chain lines that were met in
manuscripts, incunabula, and other historical documents) has been developed.
The developed algorithms and software tools are focused on restoration of
grayscale watermark images, acquired by X-ray, electron or beta radiography
methods. The experimental results fulfilled for real watermark images confirmed
the applicability of the developed method for solving the tasks of
chronological identification of manuscripts, incunabula, and other historical
documents.
The method of data indexing in text-graphical databases of
watermarks has been developed. On its basis fast search algorithms for the
multidimensional data in the specialized database were realized, using both
data indexing algorithms and methods of direct search. This method and
algorithms are developed in the frame of a prototype of the integrated system
for the database management.
Researches for developing a test-graphical database on the
history of Russian science were started in 2000 and continued in 2001. A
relational database and a bank of images have been developed to present some
personal funds of the Archive of RAS (Russian Academy of Sciences) in digital form. In particular, we processed
the following funds:
1) Fund No. 1916 of the academician A. P. Aleksandrov, RAS
President in 1975-1986. The fund comprised 322 storage units referred to
1932-1986 years. These images were included into the Archive in 1987, and 87
storage units with 517 photographs were digitized, restored, and input into the
text- graphic database.
2) Fund No. 1729 of the academician M. V. Keldysh, RAS
President in 1961-1975. The fund comprises 272 storage units referred to
1937-1986 years. Above 200 storage units were digitized, their processing and
restoration is in progress.
3) Portrait gallery of Russian scientists of past time
selected from Musin-Pushkin collection. The collection is a part of personal
fund of the academician N. A. Morozov. The collection consists of 2651 storage
units which are almost unknown to a world-wide scientific community. More than
260 images of scientists were selected from this collection and digitized.
Their processing and restoration is in progress.
Maintenance of the State registered database of the RAS
history and membership since 1724 (RAS2000) has been performed. The site with
this database, http://hp.iitp.ru , works since
1998 and is being intensively visited.
The developed generic structure of the text- graphics
database was applied to store an archival collection of medals (RAS Archive,
Category XIII). Now-a-days processing and restoration of all archival units is
finished and the database is completely populated. It is logically linked to
RAS2000 (database of the RAS history, created in 2000), thus enriching the
content of these databases.
An effective method of suppression of the image distortions
caused by the Gibbs phenomena has been developed and realized on the basis of
Fourier transform technique typical for image restoration approach. The
efficiency of the proposed method has been confirmed in the task of free-size
image restoration and implemented using
fast Fourier transform.
An original algorithm of iterative restoration of
multispectral images has been developed. The algorithm is implemented to the
restoration of color images and allows the use of a priori information about color balance of images under
processing. The developed algorithm was approved experimentally on a set of
test images and real color images.
A new technique for local contrast enhancement using
rank-order filters with spatially adaptive neighborhoods has been proposed. The
technique is based on the unsharp masking operation. However, instead of linear
lowpass filtering we used various rank-order smoothing operations. The
smoothing is performed over the pixels of spatially adaptive neighborhoods of
details to be enhanced and their surrounding backgrounds. Various rank-order
filters for local enhancement of small and middle-size details are implemented.
Extensive testing using test and real images has shown that the proposed
algorithms outperformed the conventional algorithms in terms of a subjective
visual criterion.
A new method of constructing the rank filters for
suppression mixed additive and impulsive noise with a given variance and
distortion probability has been proposed. Computer simulation on test and real
images has confirmed the superiority of the filters constructed by the proposed
method comparing with the conventional algorithms.
A new method for 3D data segmentation with the help of rank
algorithms in the local-connected areas has been proposed. Algorithms of
segmentation data conversion to Open Graphic Library has been developed and
realized. Fast geometrical transform in 3D homographic space was proposed and
implemented for real data processing in computer experiments.
A method of the multi-spectral images matching for one
display visualization has been developed. The developed
algorithm is realized for the night vision system: high sensitive CCD camera
plus IR camera sensor.
A series of methods for detecting and segmenting informative
objects, situated on a complex background, has been developed. Model-based
detection was applied for automatic detection and segmentation of the objects
of interest on initial images and images after optimal filtering. The optimal
linear filter was used to improve imaging of the object (its informative
features) on the observed image. Filtering of small-size details to improve
false alarm and misdetection rates then followed the segmentation procedure.
Developed series of methods were tested on test images and real medical images
(lung tomograms) with small solitary nodules. A comparison of segmentation
results obtained before and after optimal filtering showed that optimal
filtering allowed better outlining the object regions. The developed series of
methods can be useful for computer-assisted detection, segmentation, and
analysis of low contrast flaws (lesions) on a complex image background. They
can help to identify more precisely diagnostically important features (the
object margin, shape, and area) that is important for decision of numerous
medical tasks and in technical tasks of material inspection.
Methods of image processing, estimation of informative features, and following image classification to automate diagnosis of idiopathic pulmonary fibrosis have been developed. The methods include preprocessing images for enhancement diagnostically important details, segmentation of the lesion and measurements of the lesion area to estimate the defeat degree. The carried out experiments with use of survey roentgenography, spiral computer tomography (SCT), and x-ray–morphological comparisons have shown the developed methods gave an objective and effective estimate of the activity of idiopathic pulmonary fibrosis at different stages of disease.
GRANTS FROM:
·
INTAS (00–00081): "A Distributed Database and
Processing System for Watermarks".
·
Austrian Science Fund FWF
(FWF-13289-ARS):
"Watermarks in Middle Age Manuscripts".
·
Russian
Foundation for Basic Researches (No. 99-07-90017): "Development of both watermark
image database and system for watermark image processing and analysis".
·
Russian Foundation for Basic Researches
(No. 99-01-00265): "Development
of algorithmic and physical model of 3D view device".
·
Russian Foundation for Basic Researches
(No. 99-01-00269):
"Development of optical-digital methods for real-time pattern
recognition".
·
Russian Foundation for Basic Researches
(No. 00-07-90032):
"Development and creation of text-graphical database on Russian Basic
Science History based on RAS archive resources".
·
Russian Foundation for Basic Researches
(No. 01-07-90354):
"Distributed database for chronological identification of manuscripts and
incunabula".
·
The project of the State scientific and
technical subprogram "Information of Russia" of priority direction of
"Information technologies and electronics" (No. 037.03.319.17/1-99): "Development of integrated system
for storage and processing multidimensional signals for information maintenance
of scientific researches, culture and education", subpart
"Development of technology and software for construction of multilevel
text-graphic digital archives on the basis of conventional database management
system".
1. Wenger E.,
Karnaukhov V., Haidinger A., Stieglecker M.. A Digital Image and Database
System for Watermarks in Medieval Manuscripts. – In: David Bearman and Franca
Garzotto, editors, ICHIM '01 Proceedings, Cultural Heritage and Technologies in
the Third Millennium, Milano, Italy, 2001, pp. 259-264.
2. Karnaukhov A.,
Merzlyakov N., Milyukova O., Karnaukhov V., Wenger E., Aizenberg I., Karnaukhov
V. Digital Restoration of Watermark Images. – In: Proceedings of EVA’01,
Moscow, Centre PIC of Ministry Culture of Russia, STG, Moscow, 2001, pp.
196-199.
3. Wenger E.,
Karnaukhov V., Haidinger A., Merzlyakov N., van Thienen G., Oukhanova E.,
Erastov D. A Distributed Database and Processing System for Watermarks: an
INTAS Project. – In: Proceedings of EVA’01, Moscow, Centre PIC of Ministry
Culture of Russia, STG, Moscow, 2001, pp. 200-206.
4. Kober V.,
Mozerov M., Alvarez-Borrego J. Noise suppression using nonlinear filters with
spatially connected neighborhoods // Electronic Imaging Newsletters. 2001, vol.
12, no. 1, p. 5.
5. Kober V.,
Mozerov M., Alvarez-Borrego J. Nonlinear filters with spatially-connected
neighborhoods // Optical Engineering. 2001, vol. 40, No. 6, pp. 971-983.
6. Kober V.,
Mozerov M., Alvarez-Borrego J., Ovseyevich I. Rank image processing using
spatially adaptive neighborhoods // Pattern Recognition and Image Analysis.
2001, vol. 11, no. 3, pp. 542-552.
7. Kober V.,
Mozerov M., Alvarez-Borrego J., Ovseyevich I. Fast algorithms of rank-order
filters with spatially adaptive neighborhoods // Pattern Recognition and Image
Analysis. 2001, vol. 11, no. 4,
pp. 690-698.
8. Mozerov M.,
Kober V., Choi Tae S. Improved Motion Stereo Matching Based on a Modified
Dynamic Programming // Optical Engineering. 2001, vol. 40, no. 10, pð. 2234-2239.
9. Mozerov M.,
Kober V., Ovseyevich I. Real-time processing of night vision // Pattern
Recognition and Image Analysis. 2001, vol. 11, no. 2, pp. 347-349.
10. Kober V., Mozerov M.,
Alvarez-Borrego J., Hidalgo Silva H. Image enhancement using nonlinear filters
with spatially-adaptive neighborhoods // Annual meeting, Applications of
Digital Image Processing XXIV, San-Diego, July 2001, pp. 508-517.
11. Osipenko V.I., Belikova T.P.,
Shehter A.I., Kogan E.A., Popova E.N. A posteriori linear filtering of
roentgenogramms in differential diagnosis and estimation of idiopathic
pulmonary fibrosis (ipf) activity // Proc. of the Conference Medical
Informatics Technologies: Medicine, Health, Computer. Moscow, VVC, 2001, p.
11-13.
12. Perner P., Belikova T. A hybrid Tool for Data Mining in Picture
Archiving System. – In: Machine Learning and Data Mining in Pattern
Recognition. Springer Verlag 2001, vol. 2123, p. 141-156.
13. Aizenberg I.,
Butakov C., Karnaukhov V., Merzlyakov N., and Milyukova O. Blurred Image
Restoration Using the Type of Blur and Blur Parameters Identification on the
Neural Network // SPIE Proceedings. 2002,
vol. 4667 (accepted).
14. Karnaukhov V.N., Merzlyakov
N.S., Osipova N.M., Rubanov L.I. Experience of electronic databases creation in
the Archive of the Russian Academy of Sciences // Rossiiskie Arhivy. 2001, no.
6 (in print).
15. Rubanov L., Merzlyakov N.,
Karnaukhov V., Osipova N. Strategy of creation of digital archives accessible
through the Internet // Proc. of IS&T/SPIE’s 14th Annual Symposium
«Electronic Imaging 2002: Science and Technology», 20-25 January 2002, San
Jose, California, USA, SPIE Proceedings, 2001á vol. 4667 (accepted).
16. Kober V., Mozerov M.,
Alvarez-Borrego J., Ovseyevich I. Unsharp masking using rank-order filters with
spatially adaptive neighborhoods // Pattern Recognition and Image Analysis.
2002, vol. 12, no. 1 (in print).
17. Belikova Ò., Palenichka R., Ivasenko I. Ñomputer-aided
detection and segmentation of objects on medical images // The 10-th
International Conference in Central Europe on Computer Graphics, Visualization
and Computer Vision'2002 (accepted).
18. Belikova Ò., Palenichka R., Ivasenko I. Improved detection and segmentation of
objects on medical images // J. Computer Vision and Image Understanding (in
print).