Evolving network architectures with custom computers for multi-spectral feature identification

R. Porter; M. Gokhale; N. Harvey; S. Perkins; C. Young

doi:10.1109/EH.2001.937970

Evolving network architectures with custom computers for multi-spectral feature identification

R. Porter, M. Gokhale, N. Harvey, S. Perkins, C. Young

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Scopus citations

Abstract

This paper investigates the design of evolvable FPGA circuits where the design space is severely constrained to an interconnected network of meaningful high-level operators. The specific design domain is image processing, especially pattern recognition in remotely sensed images. Preliminary experiments are reported that compare neural networks with a recently introduced variant known as morphological networks. A novel network node is then presented that is particularly suited to the problem of pattern recognition in multi-spectral data sets. More specifically, the node can exploit both spectral and spatial information, and implements both feature extraction and classification components of a typical image processing pipeline. Once trained, the network can be applied to large image data sets, for at the sensor to extract features of interest with two orders of magnitude speed-up compared to software implementations.

Original language	English (US)
Title of host publication	Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001
Editors	Ricardo Salem Zebulum, Jason Lohn, Adrian Stoica, Didier Keymeulen
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	261-270
Number of pages	10
ISBN (Electronic)	0769511805
DOIs	https://doi.org/10.1109/EH.2001.937970
State	Published - 2001
Externally published	Yes
Event	3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001 - Long Beach, United States Duration: Jul 12 2001 → Jul 14 2001

Publication series

Name	Proceedings - NASA/DoD Conference on Evolvable Hardware, EH
Volume	2001-January
ISSN (Print)	1550-6029

Other

Other	3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001
Country/Territory	United States
City	Long Beach
Period	7/12/01 → 7/14/01

ASJC Scopus subject areas

General Engineering

Access to Document

10.1109/EH.2001.937970

Cite this

Porter, R., Gokhale, M., Harvey, N., Perkins, S., & Young, C. (2001). Evolving network architectures with custom computers for multi-spectral feature identification. In R. S. Zebulum, J. Lohn, A. Stoica, & D. Keymeulen (Eds.), Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001 (pp. 261-270). Article 937970 (Proceedings - NASA/DoD Conference on Evolvable Hardware, EH; Vol. 2001-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EH.2001.937970

Evolving network architectures with custom computers for multi-spectral feature identification. / Porter, R.; Gokhale, M.; Harvey, N. et al.
Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001. ed. / Ricardo Salem Zebulum; Jason Lohn; Adrian Stoica; Didier Keymeulen. Institute of Electrical and Electronics Engineers Inc., 2001. p. 261-270 937970 (Proceedings - NASA/DoD Conference on Evolvable Hardware, EH; Vol. 2001-January).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Porter, R, Gokhale, M, Harvey, N, Perkins, S & Young, C 2001, Evolving network architectures with custom computers for multi-spectral feature identification. in RS Zebulum, J Lohn, A Stoica & D Keymeulen (eds), Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001., 937970, Proceedings - NASA/DoD Conference on Evolvable Hardware, EH, vol. 2001-January, Institute of Electrical and Electronics Engineers Inc., pp. 261-270, 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001, Long Beach, United States, 7/12/01. https://doi.org/10.1109/EH.2001.937970

Porter R, Gokhale M, Harvey N, Perkins S, Young C. Evolving network architectures with custom computers for multi-spectral feature identification. In Zebulum RS, Lohn J, Stoica A, Keymeulen D, editors, Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001. Institute of Electrical and Electronics Engineers Inc. 2001. p. 261-270. 937970. (Proceedings - NASA/DoD Conference on Evolvable Hardware, EH). doi: 10.1109/EH.2001.937970

Porter, R. ; Gokhale, M. ; Harvey, N. et al. / Evolving network architectures with custom computers for multi-spectral feature identification. Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001. editor / Ricardo Salem Zebulum ; Jason Lohn ; Adrian Stoica ; Didier Keymeulen. Institute of Electrical and Electronics Engineers Inc., 2001. pp. 261-270 (Proceedings - NASA/DoD Conference on Evolvable Hardware, EH).

@inproceedings{bb6f68b040fc43fe86b8c60254826f29,

title = "Evolving network architectures with custom computers for multi-spectral feature identification",

abstract = "This paper investigates the design of evolvable FPGA circuits where the design space is severely constrained to an interconnected network of meaningful high-level operators. The specific design domain is image processing, especially pattern recognition in remotely sensed images. Preliminary experiments are reported that compare neural networks with a recently introduced variant known as morphological networks. A novel network node is then presented that is particularly suited to the problem of pattern recognition in multi-spectral data sets. More specifically, the node can exploit both spectral and spatial information, and implements both feature extraction and classification components of a typical image processing pipeline. Once trained, the network can be applied to large image data sets, for at the sensor to extract features of interest with two orders of magnitude speed-up compared to software implementations.",

author = "R. Porter and M. Gokhale and N. Harvey and S. Perkins and C. Young",

note = "Publisher Copyright: {\textcopyright} 2001 IEEE.; 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001 ; Conference date: 12-07-2001 Through 14-07-2001",

year = "2001",

doi = "10.1109/EH.2001.937970",

language = "English (US)",

series = "Proceedings - NASA/DoD Conference on Evolvable Hardware, EH",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "261--270",

editor = "Zebulum, {Ricardo Salem} and Jason Lohn and Adrian Stoica and Didier Keymeulen",

booktitle = "Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001",

}

TY - GEN

T1 - Evolving network architectures with custom computers for multi-spectral feature identification

AU - Porter, R.

AU - Gokhale, M.

AU - Harvey, N.

AU - Perkins, S.

AU - Young, C.

PY - 2001

Y1 - 2001

N2 - This paper investigates the design of evolvable FPGA circuits where the design space is severely constrained to an interconnected network of meaningful high-level operators. The specific design domain is image processing, especially pattern recognition in remotely sensed images. Preliminary experiments are reported that compare neural networks with a recently introduced variant known as morphological networks. A novel network node is then presented that is particularly suited to the problem of pattern recognition in multi-spectral data sets. More specifically, the node can exploit both spectral and spatial information, and implements both feature extraction and classification components of a typical image processing pipeline. Once trained, the network can be applied to large image data sets, for at the sensor to extract features of interest with two orders of magnitude speed-up compared to software implementations.

AB - This paper investigates the design of evolvable FPGA circuits where the design space is severely constrained to an interconnected network of meaningful high-level operators. The specific design domain is image processing, especially pattern recognition in remotely sensed images. Preliminary experiments are reported that compare neural networks with a recently introduced variant known as morphological networks. A novel network node is then presented that is particularly suited to the problem of pattern recognition in multi-spectral data sets. More specifically, the node can exploit both spectral and spatial information, and implements both feature extraction and classification components of a typical image processing pipeline. Once trained, the network can be applied to large image data sets, for at the sensor to extract features of interest with two orders of magnitude speed-up compared to software implementations.

UR - http://www.scopus.com/inward/record.url?scp=84952933473&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84952933473&partnerID=8YFLogxK

U2 - 10.1109/EH.2001.937970

DO - 10.1109/EH.2001.937970

M3 - Conference contribution

AN - SCOPUS:84952933473

T3 - Proceedings - NASA/DoD Conference on Evolvable Hardware, EH

SP - 261

EP - 270

BT - Proceedings - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001

A2 - Zebulum, Ricardo Salem

A2 - Lohn, Jason

A2 - Stoica, Adrian

A2 - Keymeulen, Didier

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 3rd NASA/DoD Workshop on Evolvable Hardware, EH 2001

Y2 - 12 July 2001 through 14 July 2001

ER -

Evolving network architectures with custom computers for multi-spectral feature identification

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this