A.V. Bogoslovsky - Dr. Sc. (Eng.), Honored Scientist of RF, Professor, Department of Mathematics, MESC «Zhukovsky-Gagarin Air Force Academy» (Voronezh) E-mail: p-digim@mail.ru M.A. Pantyukhin - Post-graduate Student, MESC «Zhukovsky-Gagarin Air Force Academy» (Voronezh) E-mail: ol-max@mail.ru I.V. Zhigulina - Ph. D. (Eng.), Associate Professor, Professor, Department of Mathematics, MESC «Zhukovsky-Gagarin Air Force Academy» (Voronezh) E-mail: ira_zhigulina@mail.ru

Currently, the various recognizing optoelectronic systems widely use the correlation-extreme methods, fractal processing methods and various heuristics. The first methods - efficiency depends strongly due to degree of patterns - real objects matching. Even small changes on the brightness of the observed image field, as well as the sub-pixel image distortion reduce their effectiveness significantly. The disadvantage of the second group is the lack of computing techniques for all fractal characteristics. It has an adversely affects to processing effects. The third group is not adapted to process low-sized samples sufficiently. In general, the final probability rates of mentioned methods do not allow controlling the processing parameters. The mechanism of matched to image-s properties filtering allow to estimate processing performance due to variation the aperture size of the synthesized filter and calculating the output image energy. However, it is necessary to adapt these methods to real images processing. Therefore, the aim of this paper is creating the algorithm to find the filter-s size aperture on high spatial frequencies filtering matched with an input image-s properties. It is shown that in order to select an object of interest can be used a binary form instead of full, for example, grayscale image of the searched object. This form is described schematically by its outer contour combined with the inner filling. It is also shown that the nonlinear output image processing transformation improves quality by reducing the occurrence of false contour elements and correct processing of the true image contours. In addition it is shown that to achieve the final results is possible to use aperture smaller than the maximum possible from the point of view of output image energy. The results of real images processing are presented.

 

1. Sovmeshhenie izobrazhenijj v korreljacionno-ehkstremalnykh navigacionnykh sistemakh / Pod red. L.N. Kostjashkina, M.B. Nikiforova. M.: Radiotekhnika. 2015. 208 s.
2. Avtomaticheskoe obnaruzhenie i soprovozhdenie dinamicheskikh obektov na izobrazhenijakh, formiruemykh optiko-ehlektronnymi priborami v uslovijakh apriornojj neopredelennosti. Metody i algoritmy / Pod red. A.A. KHramicheva. M.: Radiotekhnika. 2015. 280 s.
3. Objectdetectionandrecognitionindigitalimages: digitalandpractice / B. Cyganek. AJ. Wiley & Sons, Ltd, Publication. 2013. 556 p.
4. Gruzman I.S., Kirichuk V.S., Kosykh V.P., Peretjagin G.I., Spektor A.A. Cifrovaja obrabotka izobrazhenijj v informacionnykh sistemakh: Ucheb. posobie. Novosibirsk: Izd-vo NGTU. 2002. 352 s.
5. Obrabotka mnogomernykh signalov. V 2-kh knigakh. Kn.1. Linejjnaja mnogomernaja diskretnaja obrabotka signalov. Metody analiza i sinteza / Pod red. A.V. Bogoslovskogo. M.: Radiotekhnika. 2013. 168 s.
6. Bogoslovskijj A.V., ZHigulina I.V., Popov V.A., JAkovlev V.A. Vybor apertury dvumernogo filtra pri filtracii tipa Vinera-KHopfa // Radiotekhnika. 2010. № 5. S. 51−54.
7. Sledy vojjny na sputnikovykh snimkakh. URL = http://www.topwar.ru/uploads/posts/2013-10/1381821341_irak.irakel-habbaniya.jpg (10.09.2016).
8. Bogoslovskijj A.V.ZHigulina I.V., Pantjukhin M.A. Soglasovannaja so svojjstvami izobrazhenijj verkhnikh prostranstvennykh chastot izobrazhenijj / Sb. tezisov dokladov MNK «Televidenie: peredacha i obrabotka izobrazhenijj». SPb.: LEHTI. 2016. S. 105−108.
9. Llojjd D. Sistemy teplovidenija. M.: Mir. 1978. 416 s.
10. Izobrazhenie aehrodroma s samoletami. URL = http://armyman.info/uploads/posts/2013-10/thumbs/1381325541_rossiya-uglovoevladivostok.jpg (10.09.2016).