Transhir of retinal technology to applications beyond biological vision

Frank S. Werblin, T. Roska, Leon Chua, Adam Jacobs, Tibor Kozek, A. Zarándy

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

It has been possible to emulate the image processing functions of îhc vertebrate retina in great detail using a new massively parallel computational technology called "Cellular Nonlinear Network." CNN is in many ways the silicon counterpart of a vertebrate retina, comprised of scries of planar arrays ol locally connected analog processing elements. The strength of interactions between the neighboring elements determines the image processing operation, and a sequence of such operations can be stored, then sequentially executed on-chip. 'fhcse features endow the chip with a computational speed greater than 10': operations per second, thousands of times faster then conventional digital technology. The chips have been configured to perform a variety of retinal functions and could serve as the "smart camera" for a prosthetic visual system connected to the optic nerve or visual cortex that requires flexible, programmable miniaturized high speed computational analysis of the visual scene. In addition, algorithms, borrowed from biological vision can he implemented on the chip lo solve a variety of industrial and medical image processing challenges where extremely high speed, small size and low power are required. Examples include the fusion of infrared and visible images to improve night vision performance using single or double opponent color coding schemes, implementation of edge-detecting algorithms for the rapid detection of lesions such as spiculated masses or calcifications in mammograms. image segmentation for the separation of text and graphics for high speed copl or image analysis systems. The chip can also serve as conduit for reverse technology tlow whereby principles of image processing borrowed from machine vision, implemented on the chip, can be used to analyze biological visual function.

Original languageEnglish
JournalInvestigative Ophthalmology and Visual Science
Volume38
Issue number4
Publication statusPublished - 1997

Fingerprint

Technology
Vertebrates
Retina
Night Vision
Silicon
Visual Cortex
Optic Nerve
Color
Power (Psychology)

ASJC Scopus subject areas

  • Ophthalmology

Cite this

Transhir of retinal technology to applications beyond biological vision. / Werblin, Frank S.; Roska, T.; Chua, Leon; Jacobs, Adam; Kozek, Tibor; Zarándy, A.

In: Investigative Ophthalmology and Visual Science, Vol. 38, No. 4, 1997.

Research output: Contribution to journalArticle

@article{de074c4d66794b5e900666b755537015,
title = "Transhir of retinal technology to applications beyond biological vision",
abstract = "It has been possible to emulate the image processing functions of {\^i}hc vertebrate retina in great detail using a new massively parallel computational technology called {"}Cellular Nonlinear Network.{"} CNN is in many ways the silicon counterpart of a vertebrate retina, comprised of scries of planar arrays ol locally connected analog processing elements. The strength of interactions between the neighboring elements determines the image processing operation, and a sequence of such operations can be stored, then sequentially executed on-chip. 'fhcse features endow the chip with a computational speed greater than 10': operations per second, thousands of times faster then conventional digital technology. The chips have been configured to perform a variety of retinal functions and could serve as the {"}smart camera{"} for a prosthetic visual system connected to the optic nerve or visual cortex that requires flexible, programmable miniaturized high speed computational analysis of the visual scene. In addition, algorithms, borrowed from biological vision can he implemented on the chip lo solve a variety of industrial and medical image processing challenges where extremely high speed, small size and low power are required. Examples include the fusion of infrared and visible images to improve night vision performance using single or double opponent color coding schemes, implementation of edge-detecting algorithms for the rapid detection of lesions such as spiculated masses or calcifications in mammograms. image segmentation for the separation of text and graphics for high speed copl or image analysis systems. The chip can also serve as conduit for reverse technology tlow whereby principles of image processing borrowed from machine vision, implemented on the chip, can be used to analyze biological visual function.",
author = "Werblin, {Frank S.} and T. Roska and Leon Chua and Adam Jacobs and Tibor Kozek and A. Zar{\'a}ndy",
year = "1997",
language = "English",
volume = "38",
journal = "Investigative Ophthalmology and Visual Science",
issn = "0146-0404",
publisher = "Association for Research in Vision and Ophthalmology Inc.",
number = "4",

}

TY - JOUR

T1 - Transhir of retinal technology to applications beyond biological vision

AU - Werblin, Frank S.

AU - Roska, T.

AU - Chua, Leon

AU - Jacobs, Adam

AU - Kozek, Tibor

AU - Zarándy, A.

PY - 1997

Y1 - 1997

N2 - It has been possible to emulate the image processing functions of îhc vertebrate retina in great detail using a new massively parallel computational technology called "Cellular Nonlinear Network." CNN is in many ways the silicon counterpart of a vertebrate retina, comprised of scries of planar arrays ol locally connected analog processing elements. The strength of interactions between the neighboring elements determines the image processing operation, and a sequence of such operations can be stored, then sequentially executed on-chip. 'fhcse features endow the chip with a computational speed greater than 10': operations per second, thousands of times faster then conventional digital technology. The chips have been configured to perform a variety of retinal functions and could serve as the "smart camera" for a prosthetic visual system connected to the optic nerve or visual cortex that requires flexible, programmable miniaturized high speed computational analysis of the visual scene. In addition, algorithms, borrowed from biological vision can he implemented on the chip lo solve a variety of industrial and medical image processing challenges where extremely high speed, small size and low power are required. Examples include the fusion of infrared and visible images to improve night vision performance using single or double opponent color coding schemes, implementation of edge-detecting algorithms for the rapid detection of lesions such as spiculated masses or calcifications in mammograms. image segmentation for the separation of text and graphics for high speed copl or image analysis systems. The chip can also serve as conduit for reverse technology tlow whereby principles of image processing borrowed from machine vision, implemented on the chip, can be used to analyze biological visual function.

AB - It has been possible to emulate the image processing functions of îhc vertebrate retina in great detail using a new massively parallel computational technology called "Cellular Nonlinear Network." CNN is in many ways the silicon counterpart of a vertebrate retina, comprised of scries of planar arrays ol locally connected analog processing elements. The strength of interactions between the neighboring elements determines the image processing operation, and a sequence of such operations can be stored, then sequentially executed on-chip. 'fhcse features endow the chip with a computational speed greater than 10': operations per second, thousands of times faster then conventional digital technology. The chips have been configured to perform a variety of retinal functions and could serve as the "smart camera" for a prosthetic visual system connected to the optic nerve or visual cortex that requires flexible, programmable miniaturized high speed computational analysis of the visual scene. In addition, algorithms, borrowed from biological vision can he implemented on the chip lo solve a variety of industrial and medical image processing challenges where extremely high speed, small size and low power are required. Examples include the fusion of infrared and visible images to improve night vision performance using single or double opponent color coding schemes, implementation of edge-detecting algorithms for the rapid detection of lesions such as spiculated masses or calcifications in mammograms. image segmentation for the separation of text and graphics for high speed copl or image analysis systems. The chip can also serve as conduit for reverse technology tlow whereby principles of image processing borrowed from machine vision, implemented on the chip, can be used to analyze biological visual function.

UR - http://www.scopus.com/inward/record.url?scp=0342307533&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0342307533&partnerID=8YFLogxK

M3 - Article

VL - 38

JO - Investigative Ophthalmology and Visual Science

JF - Investigative Ophthalmology and Visual Science

SN - 0146-0404

IS - 4

ER -