Нейросетевой поиск особых точек для системы технического зрения…
11
Neuronet search of special points for computer vision
system when determining mobile platform movements
© M.O. Korlyakova, E.Yu. Prokopov, P.K. Novikov
1
Kaluga Branch of Bauman Moscow State Technical University, Kaluga, 248000, Russia
The main purpose of the article is to present a computer vision system model for solving
the problem of mobile robot navigation in natural environment. We consider the possibil-
ities of using different ways of searching certain points and set the requirements to
onboard systems. As the main approach we offer to use neuronet models of information
processing, and show advantages and disadvantages of this approach. The article exam-
ines Hopfild networks abilities and networks on the basis of radial neurons. We carried
out modeling and determined the most effective networks models in terms of accuracy
and speed. The received model is checked on real examples. We studied possible methods
of certain points search acceleration and defined the ways of implementing the proposed
approach within onboard systems of computer vision system.
Keywords:
neuron networks, object recognition, computer vision system
REFERENCES
Devyaterikov E.A., Mikhailov B.B. Sistema tekhnicheskogo zreniya dlya izme-
reniia puti mobil'nogo robota [The vision system for measuring path of the mo-
bile robot].
Sbornik trudov nauch.-tekhn. konf. «Tekhnicheskoe zrenie v siste-
makh upravleniya–2011»
[Proc. of the sci. and eng. conf. "Technical Vision in
Control Systems’2011"]. Moscow, SRI RAS, 2012, pp. 219–224.
Ivashina E.A., Korlyakova M.O., Pilipenko A.Yu. Formirovanie priznakov dlya
neirosetevogo poiska napravleniya peremeshcheniya ob"ekta v stsene [For-
mation of signs for the neural network search of the direction of the object
movement in the scene].
XV Vserossiyskaya nauchno-tekhnicheskaya konfer-
entsiya «Neiroinformatika–2013»: Sbornik nauchnykh trudov
[XV All-Russian
Scientific and Technical Conference "Neuroinformatics’2013": Collection of
scientific papers]. In 3 parts. Part 2. Moscow, NRNU MEPhI, 2013, pp. 40–49.
Kokareva E.A., Korlyakova M.O. Pilipenko A.Yu. Reshenie zadachi stereore-
konstruktsii v neirosetevom bazise [Solution of stereo reconstruction problem in
the neural network basis].
XIV Vserossiyskaya nauchno-tekhnicheskaya konfer-
entsiya «Neiroinformatika–2012»: Sbornik nauchnykh trudov.
[XIV All-Russian
Scientific Conference "Neuroinformatics’2012": Collection of scientific papers].
In 3 parts. Part 1. Moscow, NRNU MEPhI, 2012, pp. 160–169.
Shapiro L., Stokman J.
Kompyuternoe zrenie
[Computer vision]. Transl. from
Engl., Moscow, BINOM, Laboratoriia znanii Publ., 2006, 752 p.
Korlyakova M.O.
(b. 1969) graduated from Kaluga Branch of Bauman Moscow State Tech-
nical University in 1992. Ph.D., Assoc. Professor of the Automatic Control Systems Depart-
ment at Kaluga Branch of Bauman Moscow State Technical University. Scientific interests
include neural networks, intelligence control systems. e-mail:
Novikov P.K
. (b. 1991) is a student of the Automatic Control Systems Department at
Kaluga Branch of Bauman Moscow State Technical University. Scientific interests in-
clude neural networks, pattern recognition. e-mail:
Prokopov E.Yu.
(b. 1992) is a student of the Automatic Control Systems Department at
Kaluga Branch of Bauman Moscow State Technical University. Scientific interests in-
clude neural networks, pattern recognition. e-mail:
