L.Sh. Biktimirov – Assistant, Department «Radio Engineering», Ulyanovsk State Technical University E-mail: linarbiktimirov@rambler.ru

A.G. Tashlinskii – Dr. Sc. (Eng.), Professor, Head of Department «Radio Engineering», Ulyanovsk State Technical University E-mail: tag@ulstu.ru

The computational costs of the search algorithm based on the fragment pattern on the image, realizing the management of the set of search procedures, the working areas of which cover the examined image, are analyzed. Management is based on the analysis of the penalty function of the procedures and the granting of priority to the execution of the next iteration of the procedure with a minimum penalty. It is believed that the desired fragment is in the region in which the search procedure first reached the specified threshold number of iterations. Discrete probability distributions are obtained for the number of iterations performed by procedures for situations of presence and absence of the desired fragment on the image, with the use of which expressions were found that determine computational costs. It is shown that the use of the algorithm leads to a significant reduction. In this case, the cost advantage increases with the number of managed procedures. Examples of discrete distributions of the number of iterations for simulated images with a correlation function close to Gaussian are given.

1. Ipatov Yu.A., Kreveczkij A.V. Modelirovanie metodov obnaruzheniya i prostranstvennoj lokalizaczii gruppovy'x tochechny'x ob''ektov // Nauka. Texnologii. Proizvodstvo. 2014. № 2. S. 7−11.
2. Gerasimova N.I., Verxoturova A.E'. Poisk fragmenta izobrazheniya s ispol'zovaniem nejronnoj seti Koxonena // Sb. nauchny'x trudov «Informaczionny'e texnologii v nauke, upravlenii, soczial'noj sfere i mediczine». Tomsk: TPU. 2014. Ch. 1. S. 68−70.
3. Chambon S., Crouzil A. Dense matching using correlation: new measures that are robust near occlusions // British Machine Vision Conference, Norwich, Great Britain. 2003. V. 1. P. 143−152.
4. Czy'pkin Ya.Z. Informaczionnaya teoriya identifikaczii. M.: Nauka. Fizmatlit. 1995. 336 s.
5. Zitova B., Flusser J. Image registration methods: a survey // Image and vision computing. 2003. V. 21. № 11. P. 977−1000.
6. Tashlinskij A.G. Oczenivanie parametrov prostranstvenny'x deformaczij posledovatel'nostej. Ul'yanovsk: UlGTU. 2000. 131 s.
7. Szeliski R. Image alignment and stitching: A tutorial // Technical Report MSR-TR-2004-92. Microsoft Research. 2004. 93 p.
8. Tashlinskii A.G. Pseudogradient Estimation of Digital Images Interframe Geometrical De-formations / Vision Systems: Segmentation & Pattern Recognition. 2007. Vienna (Austria): I Tech Education and Publishing. P. 465−494.
9. Zhukova A.V., Voronov S.V. Analiz e'ffektivnosti primeneniya korrelyaczionny'x mer v kachestve czelevy'x funkczij v rekurrentny'x proczedurax privyazki izobrazhenij // Radioe'lektronnaya texnika. 2016. № 9. S. 11−16.
10. Voronov S.V., Tashlinskii A.G. Efficiency analysis of information theoretic measures in image registration // Pattern recognition and image analysis. 2016. V. 26. № 3. P. 502−505.
11. Tashlinskii A.G., Muratkhanov D.S. Structural Optimization of Pseudogradient Algorithms for Measuring Interframe Image Deformations // Pattern Recognition and Image Analysis. 2003. V. 13. № 1. P. 177−178.
12. Biktimirov L.Sh., Tashlinskij A.G. Oczenka veroyatnosti otsutstviya iskomogo fragmenta na izobrazhenii dlya algoritma s upravleniem mnozhestvom proczedur poiska // Radiotexnika. 2016. № 9. S. 6−10.