A hybrid SIMD/MIMD architecture for image understanding

Kai Hwang, V. K. Prasanna Kumar, Dongseung Kim, Hussein M. Alnuweiri

Research output: Contribution to journalConference article

Abstract

This paper presents an integrated parallel architecture for image analysis and computer vision. The architecture consists of an SIMD array processor with multiple broadcast busses, an MIMD multiprocessor with orthogonal access busses, and a two-dimensional memory array shared by both processor systems. Low-level image processing is performed on the SIMD array, while intermediate and high-level image analysis is performed on the multiprocessor system. The interaction between the two levels is supported by a common shared memory. To illustrate the power of such a two-level system we show that a wide class of image and vision analysis tasks can be performed efficiently on the architecture. Representative problems include transitive closure, histogramming image component labeling, pyramid computation, Hough transform, pattern clustering, and scene labeling.

Original languageEnglish
Pages (from-to)15-26
Number of pages12
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume1058
DOIs
Publication statusPublished - 17 May 1989
EventHigh Speed Computing II 1989 - Los Angeles, United States
Duration: 15 Jan 198920 Jan 1989

Fingerprint

MIMD (computers)
SIMD (computers)
Image Understanding
Image understanding
Labeling
Image analysis
image analysis
Container closures
Data storage equipment
Image Analysis
Hough transforms
Parallel architectures
Parallel processing systems
Computer vision
marking
central processing units
Transitive Closure
Image processing
Hough Transform
Parallel Architectures

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Cite this

A hybrid SIMD/MIMD architecture for image understanding. / Hwang, Kai; Prasanna Kumar, V. K.; Kim, Dongseung; Alnuweiri, Hussein M.

In: Proceedings of SPIE - The International Society for Optical Engineering, Vol. 1058, 17.05.1989, p. 15-26.

Research output: Contribution to journalConference article

@article{0a0595652ade410eb515eca063f72435,
title = "A hybrid SIMD/MIMD architecture for image understanding",
abstract = "This paper presents an integrated parallel architecture for image analysis and computer vision. The architecture consists of an SIMD array processor with multiple broadcast busses, an MIMD multiprocessor with orthogonal access busses, and a two-dimensional memory array shared by both processor systems. Low-level image processing is performed on the SIMD array, while intermediate and high-level image analysis is performed on the multiprocessor system. The interaction between the two levels is supported by a common shared memory. To illustrate the power of such a two-level system we show that a wide class of image and vision analysis tasks can be performed efficiently on the architecture. Representative problems include transitive closure, histogramming image component labeling, pyramid computation, Hough transform, pattern clustering, and scene labeling.",
author = "Kai Hwang and {Prasanna Kumar}, {V. K.} and Dongseung Kim and Alnuweiri, {Hussein M.}",
year = "1989",
month = "5",
day = "17",
doi = "10.1117/12.951662",
language = "English",
volume = "1058",
pages = "15--26",
journal = "Proceedings of SPIE - The International Society for Optical Engineering",
issn = "0277-786X",
publisher = "SPIE",

}

TY - JOUR

T1 - A hybrid SIMD/MIMD architecture for image understanding

AU - Hwang, Kai

AU - Prasanna Kumar, V. K.

AU - Kim, Dongseung

AU - Alnuweiri, Hussein M.

PY - 1989/5/17

Y1 - 1989/5/17

N2 - This paper presents an integrated parallel architecture for image analysis and computer vision. The architecture consists of an SIMD array processor with multiple broadcast busses, an MIMD multiprocessor with orthogonal access busses, and a two-dimensional memory array shared by both processor systems. Low-level image processing is performed on the SIMD array, while intermediate and high-level image analysis is performed on the multiprocessor system. The interaction between the two levels is supported by a common shared memory. To illustrate the power of such a two-level system we show that a wide class of image and vision analysis tasks can be performed efficiently on the architecture. Representative problems include transitive closure, histogramming image component labeling, pyramid computation, Hough transform, pattern clustering, and scene labeling.

AB - This paper presents an integrated parallel architecture for image analysis and computer vision. The architecture consists of an SIMD array processor with multiple broadcast busses, an MIMD multiprocessor with orthogonal access busses, and a two-dimensional memory array shared by both processor systems. Low-level image processing is performed on the SIMD array, while intermediate and high-level image analysis is performed on the multiprocessor system. The interaction between the two levels is supported by a common shared memory. To illustrate the power of such a two-level system we show that a wide class of image and vision analysis tasks can be performed efficiently on the architecture. Representative problems include transitive closure, histogramming image component labeling, pyramid computation, Hough transform, pattern clustering, and scene labeling.

UR - http://www.scopus.com/inward/record.url?scp=85075300295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075300295&partnerID=8YFLogxK

U2 - 10.1117/12.951662

DO - 10.1117/12.951662

M3 - Conference article

AN - SCOPUS:85075300295

VL - 1058

SP - 15

EP - 26

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

SN - 0277-786X

ER -