350 rub
Journal Neurocomputers №3 for 2014 г.
Article in number:
Determination of object redetection system based on neural network
Keywords: neural networks tracking descriptors Bhattachary-s distance SOFTMAX-normalization
Authors:
A. A. Noskov - Post-graduate Student, Yaroslavl Demidov State University. E-mail: noskoff.andrey@gmail.com
E. A. Aminova - Post-graduate Student, Yaroslavl Demidov State University. E-mail: lena@piclab.ru
A. L. Priorov - Dr.Sc. (Eng.), Associate Professor, Yaroslavl Demidov State University. E-mail: andcat@yandex.ru
I. N. Trapeznikov - Post-graduate Student, Yaroslavl Demidov State University. E-mail: trapeznikoff@list.ru
Abstract:
The case when the object of interest is lost from the sight of the camera for various reasons and then reappeared is actual for detection and tracking issues. The main reasons of the tracking object moving termination are overlapping fixed obstacle between moving object and camera or object moving from the first camera field of view in the field of view of another. The mentioned above scenario is typical for the different systems of detection and tracking, in which the object of interest is human. The various mobile and stationary obstacles can survey in the observation scene. Important information which is difficult to obtain by conventional methods is counting visitor statistics of a particular territory with the exception of each new detection as a new person. The goal of this article is to propose the determination of object redetection method based on a neural network. Analyses of system quality decency on the essential parameters of the offered neural network are presented. Topology and structure of the neural network are described. Proposed system used to achieve the challenges of correct recovery of the object tracking on the video if there are obstacles in the path of motion and to solve the issues which are connected with the automation of pretreatment processes of incoming information. System independency on noises increases with the usage of modern preprocessing algorithms of the input image. Applying the neural network topology and the proposed structure gives the system effectiveness over 96 - 98%.The obtained results can be integrated into automated security, monitoring and quality control of the company systems.
Pages: 36-43
References

  1. Bradski G.R. Computer Vision Face Tracking For Use in a Perceptual User Interface // Proc. IEEE Workshop on Applications of Comp. Vision. Princeton, 1998. P. 214-219.
  2. Chellappa R., Wilson C., Sirohey S. Human and Machine Recognition of Faces: A Survey // Proc. IEEE. 1995. V. 83. P. 705-740.
  3. Moghaddam B., Pentland A. Probabilistic Visual Learning for Object Detection // Int-l Conf. Computer Vision, 1995. P. 786-793.
  4. Viola P., Jones M. Rapid Object Detection Using a Boosted Cascade of Simple Features // Proc. Int. Conf. on Computer Vision and Pattern Recognition. 2001. № 1. P. 511-518.
  5. Fukunaga K. Introduction to Statistical Pattern Recognition. Academic Press. Boston, 1990.
  6. Bryuxanov Ju.A., Priorov A.L., Dzhigan V.I., Xryashhev V.V. Osnovy' czifrovoj obrabotki signalov: ucheb. posobie. Jaroslavl': Jarosl. gos. universitet. 2013. 344 s.
  7. Priorov A.L., Apal'kov A.V., Xryashhev V.V. Cifrovaya obrabotka izobrazhenij: ucheb. posobie. Jaroslavl' Jarosl. gos. universitet. 2007. 235 s.
  8. Dvorkovich V.P., Dvorkovich A.V. Cifrovy'e videoinformaczionny'e sistemy' (teoriya i praktika). M.: Texnosfera. 2012. 1008 s.
  9. Priorov A.L., Voloxov V.A., Sergeev E.V., Mochalov I.S. Parallel'naya proczedura fil'traczii izobrazhenij na osnove analiza glavny'x komponent nelokal'noj obrabotki // E'lektromagnitny'e volny' i e'lektronny'e sistemy'. 2012. № 11. S. 64‑70.
  10. Priorov A., Tumanov K., Volokhov V., Sergeev E., Mochalov I. Applications of Image Filtration Based on Principal Component Analysis and Nonlocal Image Processing // IAENG International Journal of Computer Science. 40:2. R. 62-80.
  11. Priorov A., Volokhov V., Sergeev E., Mochalov I., Tumanov K. Parallel Filtration Based on Principle Component Analysis and Nonlocal Image Processing // Proc. International MultiConference of Engineers and Computer Scientists 2013. Hong Kong, 2013. V. 1. P. 430-435.
  12. Priorov A.L., Xryashhev V.V., Pavlov E.A., Gerasimov N.B., Shemyakov A.M. Oczenka kachestva czvetny'x izobrazhenij pri podavlenii impul'snogo shuma // Radiotexnika. 2013. № 5. S. 41-49.
  13. Xryashhev V.V., Priorov A.L., Solov'ev V.E., Shemyakov A.M. Opredelenie tipa iskazheniya izobrazheniya v zadache nee'talonnoj oczenki kachestva // Nelinejny'j mir. 2013. T. 11. № 1. S. 32-35.
  14. Sergeev E.V., Mochalov I.S., Voloxov V.A., Priorov A.L. Nelokal'ny'j algoritm fil'traczii izobrazhenij na osnove metoda glavny'x komponent // Uspexi sovremennoj radioe'lektroniki. 2012. № 3. S. 80-88.
  15. Priorov A.L., Voloxov V.A., Sergeev E.V., Mochalov I.S. Parallel'naya proczedura fil'traczii izobrazhenij na osnove analiza glavny'x komponent nelokal'noj obrabotki // E'lektromagnitny'e volny' i e'lektronny'e sistemy'. 2012. № 11. S. 64-70.
  16. Diane Seku Abdel' Kader Raspoznavanie i generacziya obrazov v nejronnoj seti s ierarxicheskoj svyaznost'yu // Nejrokomp'yutery': razrabotka, primenenie. 2014. № 1. S. 47-57.
  17. Kalinovskij I.A., Spiczy'n V.G. Algoritm obnaruzheniya licz na osnove svertochnoj nejronnoj seti // Nejrokomp'yutery': razrabotka, primenenie. 2013. № 10. S. 48-53.
  18. Timofeev A.V., Derin O.A. Selekcziya ob''ektov v mul'tiizobrazheniyax // Nejrokomp'yutery': razrabotka, primenenie. 2010. № 7. S. 45-56.
  19. Sirota A.A., Voronova E.V. Modelirovanie izobrazhenij prostranstvenno-raspredelenny'x ob''ektov na odnorodnom fone s ispol'zovaniem iskusstvenny'x nejronny'x setej // Nejrokomp'yutery': razrabotka, primenenie 2010. № 10. S. 27-35.
  20. Zagorujko S.N., Skribczov P.V. Intellektual'ny'j algoritm umen'sheniya zashumlennosti izobrazhenij // Nejrokomp'yutery': razrabotka, primenenie 2013. № 8. S. 38-41.