Multiprocessor-based video motion detection using adaptive neural systems

J. C. Lee; B. J. Sheu; C. F. Chang; R. Chellappa

doi:10.1109/VTSA.1991.246708

Multiprocessor-based video motion detection using adaptive neural systems

J. C. Lee
, B. J. Sheu
, C. F. Chang
, R. Chellappa

University of Southern California

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

Abstract

Fast motion detection is an important step in the high-speed video/vision processing systems. System-level design of a 2-dimensional mesh-concerned competitive neural network for video motion detection is presented. The motion information from a sequence of images can be obtained through mixed-analog/digital signal processing. Massively parallel neurocomputing is performed by multiple copies of compact and efficient neuroprocessors. Inter-processor data transfer is carried out by dedicated point-to-point analog interconnections. Global data communication between the host computer and neuroprocessors is through the digital common bus to maintain strong signal strength. An extendable analog winner-take-all circuit is used to implement the competition function with a minimal delay time. A 1.5×2.8-cm² chip in a 1.2- μ m CMOS technology can accommodate 64 velocity-selective neuroprocessors. This chip can achieve 83.2 Giga-connections per second. By using 128 VLSI neural chips, the speed-up factor over the Sun-4/60 SPARC station-1 is estimated to by 54545.

Original language	English
Title of host publication	1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	74-78
Number of pages	5
ISBN (Electronic)	078030036X, 9780780300361
DOIs	https://doi.org/10.1109/VTSA.1991.246708
State	Published - 1991
Externally published	Yes
Event	1991 International Symposium on VLSI Technology, Systems, and Applications, VTSA 1991 - Taipei, Taiwan Duration: 22 05 1991 → 24 05 1991

Publication series

Name	International Symposium on VLSI Technology, Systems, and Applications, Proceedings
ISSN (Print)	1930-8868

Conference

Conference	1991 International Symposium on VLSI Technology, Systems, and Applications, VTSA 1991
Country/Territory	Taiwan
City	Taipei
Period	22/05/91 → 24/05/91

Bibliographical note

Publisher Copyright:
© 1991 IEEE.

Access to Document

10.1109/VTSA.1991.246708

Cite this

Lee, J. C., Sheu, B. J., Chang, C. F., & Chellappa, R. (1991). Multiprocessor-based video motion detection using adaptive neural systems. In 1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991 (pp. 74-78). Article 246708 (International Symposium on VLSI Technology, Systems, and Applications, Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/VTSA.1991.246708

Lee, J. C. ; Sheu, B. J. ; Chang, C. F. et al. / Multiprocessor-based video motion detection using adaptive neural systems. 1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991. Institute of Electrical and Electronics Engineers Inc., 1991. pp. 74-78 (International Symposium on VLSI Technology, Systems, and Applications, Proceedings).

@inproceedings{bb8cfed3f02842ab991444532c5654be,

title = "Multiprocessor-based video motion detection using adaptive neural systems",

abstract = "Fast motion detection is an important step in the high-speed video/vision processing systems. System-level design of a 2-dimensional mesh-concerned competitive neural network for video motion detection is presented. The motion information from a sequence of images can be obtained through mixed-analog/digital signal processing. Massively parallel neurocomputing is performed by multiple copies of compact and efficient neuroprocessors. Inter-processor data transfer is carried out by dedicated point-to-point analog interconnections. Global data communication between the host computer and neuroprocessors is through the digital common bus to maintain strong signal strength. An extendable analog winner-take-all circuit is used to implement the competition function with a minimal delay time. A 1.5×2.8-cm2 chip in a 1.2- μ m CMOS technology can accommodate 64 velocity-selective neuroprocessors. This chip can achieve 83.2 Giga-connections per second. By using 128 VLSI neural chips, the speed-up factor over the Sun-4/60 SPARC station-1 is estimated to by 54545.",

author = "Lee, \{J. C.\} and Sheu, \{B. J.\} and Chang, \{C. F.\} and R. Chellappa",

note = "Publisher Copyright: {\textcopyright} 1991 IEEE.; 1991 International Symposium on VLSI Technology, Systems, and Applications, VTSA 1991 ; Conference date: 22-05-1991 Through 24-05-1991",

year = "1991",

doi = "10.1109/VTSA.1991.246708",

language = "英语",

series = "International Symposium on VLSI Technology, Systems, and Applications, Proceedings",

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

pages = "74--78",

booktitle = "1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991",

address = "美国",

}

Lee, JC, Sheu, BJ, Chang, CF & Chellappa, R 1991, Multiprocessor-based video motion detection using adaptive neural systems. in 1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991., 246708, International Symposium on VLSI Technology, Systems, and Applications, Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 74-78, 1991 International Symposium on VLSI Technology, Systems, and Applications, VTSA 1991, Taipei, Taiwan, 22/05/91. https://doi.org/10.1109/VTSA.1991.246708

Multiprocessor-based video motion detection using adaptive neural systems. / Lee, J. C.; Sheu, B. J.; Chang, C. F. et al.
1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991. Institute of Electrical and Electronics Engineers Inc., 1991. p. 74-78 246708 (International Symposium on VLSI Technology, Systems, and Applications, Proceedings).

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

TY - GEN

T1 - Multiprocessor-based video motion detection using adaptive neural systems

AU - Lee, J. C.

AU - Sheu, B. J.

AU - Chang, C. F.

AU - Chellappa, R.

PY - 1991

Y1 - 1991

N2 - Fast motion detection is an important step in the high-speed video/vision processing systems. System-level design of a 2-dimensional mesh-concerned competitive neural network for video motion detection is presented. The motion information from a sequence of images can be obtained through mixed-analog/digital signal processing. Massively parallel neurocomputing is performed by multiple copies of compact and efficient neuroprocessors. Inter-processor data transfer is carried out by dedicated point-to-point analog interconnections. Global data communication between the host computer and neuroprocessors is through the digital common bus to maintain strong signal strength. An extendable analog winner-take-all circuit is used to implement the competition function with a minimal delay time. A 1.5×2.8-cm2 chip in a 1.2- μ m CMOS technology can accommodate 64 velocity-selective neuroprocessors. This chip can achieve 83.2 Giga-connections per second. By using 128 VLSI neural chips, the speed-up factor over the Sun-4/60 SPARC station-1 is estimated to by 54545.

AB - Fast motion detection is an important step in the high-speed video/vision processing systems. System-level design of a 2-dimensional mesh-concerned competitive neural network for video motion detection is presented. The motion information from a sequence of images can be obtained through mixed-analog/digital signal processing. Massively parallel neurocomputing is performed by multiple copies of compact and efficient neuroprocessors. Inter-processor data transfer is carried out by dedicated point-to-point analog interconnections. Global data communication between the host computer and neuroprocessors is through the digital common bus to maintain strong signal strength. An extendable analog winner-take-all circuit is used to implement the competition function with a minimal delay time. A 1.5×2.8-cm2 chip in a 1.2- μ m CMOS technology can accommodate 64 velocity-selective neuroprocessors. This chip can achieve 83.2 Giga-connections per second. By using 128 VLSI neural chips, the speed-up factor over the Sun-4/60 SPARC station-1 is estimated to by 54545.

UR - https://www.scopus.com/pages/publications/85069681465

U2 - 10.1109/VTSA.1991.246708

DO - 10.1109/VTSA.1991.246708

M3 - 会议稿件

AN - SCOPUS:85069681465

T3 - International Symposium on VLSI Technology, Systems, and Applications, Proceedings

SP - 74

EP - 78

BT - 1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 1991 International Symposium on VLSI Technology, Systems, and Applications, VTSA 1991

Y2 - 22 May 1991 through 24 May 1991

ER -

Lee JC, Sheu BJ, Chang CF, Chellappa R. Multiprocessor-based video motion detection using adaptive neural systems. In 1991 International Symposium on VLSI Technology, Systems, and Applications - Proceedings of Technical Papers, VTSA 1991. Institute of Electrical and Electronics Engineers Inc. 1991. p. 74-78. 246708. (International Symposium on VLSI Technology, Systems, and Applications, Proceedings). doi: 10.1109/VTSA.1991.246708

Multiprocessor-based video motion detection using adaptive neural systems

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this