Topographic and non-topographic neural network based computational platform for UAV applications

Cs Rekeczky, G. Tímár, D. Bálya, I. Szatmári, A. Zarándy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this work, we present an architecture and algorithmic framework where topographic and nontopographic computation is combined on the basis of several artificial neural network models. The algorithm cores utilize an analogic (analog and logical) architecture consisting of a high resolution optical sensor, a low resolution cellular sensorprocessor (cellular nonlinear network - CNN - based chip) and a digital signal processor. The proposed framework makes the acquisition of a spatially and temporally consistent image flow possible even in case of extreme variations in the environment. It ideally supports the handling of difficult problems on a moving platform such as terrain identification, navigation parameter estimation and multi-target tracking. The proposed spatio-temporal adaptation relies on a feature based optical flow estimation that can be efficiently calculated on available CNN chips. We will illustrate how multi-channel visual flow analysis and classifier (ART, KN) driven visual attention-selection mechanisms can be efficiently supported by an analogic architecture. The experiments performed on an analogic CNN hardware prototype will highlight some of the application potentials for unmanned air vehicle (UAV) applications.

Original languageEnglish
Title of host publicationIEEE International Conference on Neural Networks - Conference Proceedings
Pages1763-1768
Number of pages6
Volume3
Publication statusPublished - 2004
Event2004 IEEE International Joint Conference on Neural Networks - Proceedings - Budapest, Hungary
Duration: Jul 25 2004Jul 29 2004

Other

Other2004 IEEE International Joint Conference on Neural Networks - Proceedings
CountryHungary
CityBudapest
Period7/25/047/29/04

Fingerprint

Nonlinear networks
Neural networks
Optical flows
Optical sensors
Digital signal processors
Air
Target tracking
Parameter estimation
Identification (control systems)
Navigation
Classifiers
Hardware
Experiments

ASJC Scopus subject areas

  • Software

Cite this

Rekeczky, C., Tímár, G., Bálya, D., Szatmári, I., & Zarándy, A. (2004). Topographic and non-topographic neural network based computational platform for UAV applications. In IEEE International Conference on Neural Networks - Conference Proceedings (Vol. 3, pp. 1763-1768)

Topographic and non-topographic neural network based computational platform for UAV applications. / Rekeczky, Cs; Tímár, G.; Bálya, D.; Szatmári, I.; Zarándy, A.

IEEE International Conference on Neural Networks - Conference Proceedings. Vol. 3 2004. p. 1763-1768.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rekeczky, C, Tímár, G, Bálya, D, Szatmári, I & Zarándy, A 2004, Topographic and non-topographic neural network based computational platform for UAV applications. in IEEE International Conference on Neural Networks - Conference Proceedings. vol. 3, pp. 1763-1768, 2004 IEEE International Joint Conference on Neural Networks - Proceedings, Budapest, Hungary, 7/25/04.
Rekeczky C, Tímár G, Bálya D, Szatmári I, Zarándy A. Topographic and non-topographic neural network based computational platform for UAV applications. In IEEE International Conference on Neural Networks - Conference Proceedings. Vol. 3. 2004. p. 1763-1768
Rekeczky, Cs ; Tímár, G. ; Bálya, D. ; Szatmári, I. ; Zarándy, A. / Topographic and non-topographic neural network based computational platform for UAV applications. IEEE International Conference on Neural Networks - Conference Proceedings. Vol. 3 2004. pp. 1763-1768
@inproceedings{50c7f45014e74e9b8dc63d8d522505ed,
title = "Topographic and non-topographic neural network based computational platform for UAV applications",
abstract = "In this work, we present an architecture and algorithmic framework where topographic and nontopographic computation is combined on the basis of several artificial neural network models. The algorithm cores utilize an analogic (analog and logical) architecture consisting of a high resolution optical sensor, a low resolution cellular sensorprocessor (cellular nonlinear network - CNN - based chip) and a digital signal processor. The proposed framework makes the acquisition of a spatially and temporally consistent image flow possible even in case of extreme variations in the environment. It ideally supports the handling of difficult problems on a moving platform such as terrain identification, navigation parameter estimation and multi-target tracking. The proposed spatio-temporal adaptation relies on a feature based optical flow estimation that can be efficiently calculated on available CNN chips. We will illustrate how multi-channel visual flow analysis and classifier (ART, KN) driven visual attention-selection mechanisms can be efficiently supported by an analogic architecture. The experiments performed on an analogic CNN hardware prototype will highlight some of the application potentials for unmanned air vehicle (UAV) applications.",
author = "Cs Rekeczky and G. T{\'i}m{\'a}r and D. B{\'a}lya and I. Szatm{\'a}ri and A. Zar{\'a}ndy",
year = "2004",
language = "English",
volume = "3",
pages = "1763--1768",
booktitle = "IEEE International Conference on Neural Networks - Conference Proceedings",

}

TY - GEN

T1 - Topographic and non-topographic neural network based computational platform for UAV applications

AU - Rekeczky, Cs

AU - Tímár, G.

AU - Bálya, D.

AU - Szatmári, I.

AU - Zarándy, A.

PY - 2004

Y1 - 2004

N2 - In this work, we present an architecture and algorithmic framework where topographic and nontopographic computation is combined on the basis of several artificial neural network models. The algorithm cores utilize an analogic (analog and logical) architecture consisting of a high resolution optical sensor, a low resolution cellular sensorprocessor (cellular nonlinear network - CNN - based chip) and a digital signal processor. The proposed framework makes the acquisition of a spatially and temporally consistent image flow possible even in case of extreme variations in the environment. It ideally supports the handling of difficult problems on a moving platform such as terrain identification, navigation parameter estimation and multi-target tracking. The proposed spatio-temporal adaptation relies on a feature based optical flow estimation that can be efficiently calculated on available CNN chips. We will illustrate how multi-channel visual flow analysis and classifier (ART, KN) driven visual attention-selection mechanisms can be efficiently supported by an analogic architecture. The experiments performed on an analogic CNN hardware prototype will highlight some of the application potentials for unmanned air vehicle (UAV) applications.

AB - In this work, we present an architecture and algorithmic framework where topographic and nontopographic computation is combined on the basis of several artificial neural network models. The algorithm cores utilize an analogic (analog and logical) architecture consisting of a high resolution optical sensor, a low resolution cellular sensorprocessor (cellular nonlinear network - CNN - based chip) and a digital signal processor. The proposed framework makes the acquisition of a spatially and temporally consistent image flow possible even in case of extreme variations in the environment. It ideally supports the handling of difficult problems on a moving platform such as terrain identification, navigation parameter estimation and multi-target tracking. The proposed spatio-temporal adaptation relies on a feature based optical flow estimation that can be efficiently calculated on available CNN chips. We will illustrate how multi-channel visual flow analysis and classifier (ART, KN) driven visual attention-selection mechanisms can be efficiently supported by an analogic architecture. The experiments performed on an analogic CNN hardware prototype will highlight some of the application potentials for unmanned air vehicle (UAV) applications.

UR - http://www.scopus.com/inward/record.url?scp=10844292590&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10844292590&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:10844292590

VL - 3

SP - 1763

EP - 1768

BT - IEEE International Conference on Neural Networks - Conference Proceedings

ER -