N.N. Vasin – Dr.Sc.(Eng.), Professor, Department «Networks and Communication Systems», 

Volga State University of Telecommunications and Informatics (Samara)

E-mail: vasin.psuti@mail.ru, vasin-nn@psuti.ru

R.R. Diyazitdinov – Ph.D.(Eng.), Associate Professor, Department «Networks and Communication Systems», Volga State University of Telecommunications and Informatics (Samara) E-mail: rinat.diyazitdinov@gmail.com

The article was described the algorithm of 1D and 2D digital video signal superposition with additive and multiplicative noise. This task also is named estimation optical flow in the world researches. The applying filed is video compression, solid object deformation estimation, aim object’s tracking, velocity measurement, metrological characteristic estimation for measurement system. The problem of increasing accuracy of superposition is very important at some task. The accuracy of superposition defines a measurement’s repeatability. The repeatability is the serious parameters of measurement system. This one defines the cost of system and possibility of real applying at manufacturing process. The example of applying is superposition digital video signal for velocity measurement of aim object which is sensitive with additive and multiplicative noise of video surveillance. Thus the problem of increasing accuracy of superposition digital video signal with extra noise is important for measurement system development. Describing algorithm was developed by math model which is defined two signals. One of these signals contains additive and multiplicative variables. The Taylor series in the surrounding area of finding variable allows taking high processing procedure of estimation. This one allows defining linear equation system. The unknown variables of this system are parameter optical flow. This analytical way requires less process time than full scan algorithms as correlation or residual methods. Also the article was shown the superposition examples for 1-dimensional and 2-dimensional case. The accuracy characteristics were estimated by computer modeling. It consists to adjunction Gaussian noise to aim signals, estimate superposition and compare estimation with real value. The result of compassion was shown as plots of root mean square depend on noise power. These plots were shown that accuracy of the developing algorithm is nearly as full scan methods but time processing is nearly as methods with Taylor series (it is much less than full scan methods). Thus developing algorithm has advantages by main properties. It can be used for modernization measurement process for accuracy increasing without additional constraint to processor power.

1. Vasin N.N., Kurinskii V.Yu. Obrabotka videosignalov dlya izmereniya skorosti dvizheniya zheleznodorozhnykh vagonov na sortirovochnoi gorke. Kompyuternaya optika. 2005. № 27. S. 185−188.
2. Brox T., Bruhn A., Papenberg N., Weickert J. High Accuracy Optical Flow Estimation Based on a Theory for Wrapping. ECCV 2004. Lecture Notes in Computer Science. 2004. V. 3024. P. 25−36.
3. Lucas B.D., Kanade T. An iterative image registration technique with an application to stereo vision. Proc. Seventh International Joint Conference on Artificial Intelligence. 1981. P. 674−679.
4. Lucas B.D. Generalized Image Matching by the Method of Differences. Pittsburgh. Carnegie-MelionUniversitypublisher. 1984.
5. Tashlinskii A.G. Otsenivanie parametrov prostranstvennykh deformatsii posledovatelnostei izobrazhenii. Ulyanovsk: Izd-vo UlGTU. 2000.
6. Horn B.K., Schunck B.G. Determining optical flow. Artificial Intell. 1981. V. 17. P. 185−204.
7. Androsov V.A., Boiko Yu.I., Bochkarev A.M., Odnorog A.P. Sovmeshchenie izobrazhenii v usloviyakh neopredelennosti. Zarubezhnaya radioelektronika. 1985. № 4. S. 32−41.
8. Gubanov A.V., Efimov V.M., Kirichuk V.S., Pustovskikh A.I., Reznik A.L. Metody otsenivaniya vzaimnogo smeshcheniya fragmentov izobrazhenii. Avtometriya. 1988. № 3. S. 70−73.
9. Pankova T.L., Reznik A.L. Effektivnost algoritmov pretsizionnogo sovmeshcheniya tsifrovykh izobrazhenii. Avtometriya. 1991. № 5. S. 39−43.
10. Popov P.G. Sovmeshchenie izobrazhenii televizionnogo i teplovizionnogo kanalov. Avtometriya. 1993. № 1. S. 35−39.
11. Croon G., Ho H., Wagter C., Kampen E., Remes B., Chu Q. Optic-flow based slope estimation for autonomous landing. International Journal of Micro Air Vehicles. 2013. V. 5. P. 287−297.
12. Hordijk B., Scheper K., Croon G. Vertical landing for micro air vehicles using event-based optical flow. Field Robotics. 2017. V. 35. P. 69−90.
13. Kueng B., Mueggler E., Gallego G., Scaramuzza D. Low latency visual odometry using event-based feature tracks. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 2016. P. 16−23.
14. Liu M., Delbruck T. Block-matching optical flow for dynamic vision sensors: Algorithm and FPGA implementation. IEEE International Symposium on Circuits and Systems (ISCAS). 2017. P. 1−4.
15. Barranco F., Fermuller C., Aloimonos Y. Contour motion estimation for asynchronous event-driven cameras. Proceedings of the IEEE. 2014. V. 102. P. 1537−1556.