Scientific and Technical Journal


ISSN Print 2221-3937
ISSN Online 2221-3805
In automated systems of visual information processing - in industry, in medicine, in security systems method of classification largely determines the accuracy of decisions. Education in the classification of such systems is often carried out by small sets of parameters for the complex shape of the clusters and their intersection in the feature space. When classifying into operation when it is necessary to recognize objects invariant to transformations of scale, rotate, shear in images with noise and distorted form, it can lead to reduced reliability. Increased reliability by increasing the data set increases resource use and increases the time for learning these systems in the case of introduction of new products in the industry, improvements in methods of medical drugs, etc. When training and debugging system in such cases it is necessary to select the parameters of the classifier - from a range of values of coefficients that determine the shape of the surfaces separating the classes in the feature space. The authors have developed a method of classification using multi-start optimization with wavelet transform. In this paper, we propose a method of classification, which can help to determine a set of sub-ranges of factors separating surfaces to allow for the selection of the classifier reliability required from the perspective of pragmatic adequacy, based on the known properties of the wavelet transform to carry out spatial processing with adjustable detail.

1. Abakumov V., Rybin A., Svatosh I., and Sinecop U. Sistemy otobrageniya v medicine [The Systems of Reflection in the Medicine], (2001), Kiev, Ukraine, Un. Publ., 336 p. (In Russian).

2. Katys P., and Katys G. Sistemy mashinnogo videniya s intellectualnymi videodatchikami [The Systems of Machine Vision with Intellectual Video Sensors], (2001), Informacionnye Tehnologii, Ukraine, No. 10, pp. 28 – 33 (In Russian).

3. Antoshchuk S., Krylov V., and Shcherbakova G., (2007), The Integrated Circuits Photo Masks Images Alignment for Automated Optical Inspection System, DAAAM International Scientific Book, B. Katalinic (Ed). Published by DAAAM International, Chapter 26, pp. 287 – 294 (In English).

4. Shcherbakova G., Krylov V., Abakumov V., Brovkov V., and Kozina I., (2011), Sub Gradient Iterative Method for Neural Networks Training, The IEEE International Conference on Intelligent Data Acquisition and Advanced Computing System, 15-17 September 2011, Prague, Czech Republic, pp. 361 – 364 (In English).

5. Moganti M., Ercal F., Dagli C., and Tsunekawa S., (1996), Automatic PCB Inspection Algorithms: a Survey, Computer Vision and Image Understanding, Vol. 63, No. 2, pp. 287 – 313 (In English), available at:

6. Abakumov V., Antoshchuk S., and Krylov V., Raspoznavanie visualnoi informacii v avtomatizirovannih sistemah [Recognition of Visual Information in Automation Systems], (2003), Electronika i Sviaz, Kiev, Ukraine,No. 19, pp. 46 – 48 (In Russian).

7. Shcherbakova G. Subgradientniy metod klassificacii v prostranstve weivlet preobrazovaniya dlia technicheskoy diagnostiki [Sub-gradient Classification Method in the Wavelet Domain for Technical Diagnostics], (2010), Electrotehnicheskie i Kompiuternie Sistemy, Kiev, Ukraine, No.01(77), pp. 136 – 142 (InRussian).

8. KrylovV., and ShcherbakovaG. Ierarhicheskiysubgradientnyimetodoptimizaciivoblastiveivletpreobrazovaniya [TheHierarchicalSub-gradientIterativeOptimizationMethodintheWaveletTransformingDomain], (2008), ElectronikaiSviaz, Kiev, Ukraine, No. 6, pp. 28 – 31 (InRussian).

9. Krylov V., and PolyakovaM. Konturnaya segmentacia izobrageniy v prostranstve veivlet preobrazovaniya s ispolzovaniem liftinga [ContourImagesSegmentationinSpaceofWaveletTransformwiththeUseofLifting], (2007), Optical-electronic Informatively-power Technologies,No. 2 (12), pp. 48 – 58 (In Russian).

10. Shcherbakova G., and Krylov V. Subgradientnyi metod optimizazii v prostranstve veivlet-preobrasovaniya s ograniceniyami vtorogo roda pri avtomatisirovannom tehnicheskom diagnostirovanii [Sub-gradient Optimization Method in Wavelet Domain with Inequality Constrain in Automation Technical Diagnostics], (2012), Electrotehnicheskie i Kompiuternie Sistemy,Kiev, Ukraine, No.07 (83), pp. 127 – 131 (In Russian).
11. Shcherbakova G., Krylov V., and Dilevskiy A. Multistartoviy metod optimizazii v prostranstve veivlet preobrasovaniya s iterativnoi ocenkoi ograniceniy vtorogo roda [Multi-start Optimization Method in Wavelet Domain with Iterative Inequality Constrain], (2013), Electrotehnicheskie i Kompiuternie Sistemy, Kiev, Ukraine, No.09 (85), pp. 180 – 185(In Russian).

12. Polak E., (1971), Computational Methods in Optimization. A Unified Approach, University of California, Berkeley, New York, London: Academic Press (In English).

Last download:
27 Nov 2020

[ © KarelWintersky ] [ All articles ] [ All authors ]
[ © Odessa National Polytechnic University, 2014-2018. Any use of information from the site is possible only under the condition that the source link! ]