A Hopfield neural network for adaptive image segmentation: An active surface paradigm

Dinggang Shen, Horace H S Ip

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

This paper presents an adaptive thresholding technique for separating objects from noisy and non-uniformly illuminated images. The construction of the threshold surface is formulated as an active surface optimization problem, which is then solved by a Hopfield neural network. We proposed four constraints which ensure the active threshold surface to conform with the underlying image topography. Compared with Yanowitz and Bruckstein's method, this method produces superior segmentations particularly when the edge segments are sparsely distributed in the image and under non-uniform illuminations. Using three types of artificial and real images, we show that this method converges faster and produces better segmentations compared with previous interpolation-based adaptive thresholding techniques.

Original languageEnglish
Pages (from-to)37-48
Number of pages12
JournalPattern Recognition Letters
Volume18
Issue number1
Publication statusPublished - 1997 Jan 1
Externally publishedYes

Fingerprint

Hopfield neural networks
Image segmentation
Topography
Interpolation
Lighting

Keywords

  • Active threshold surface
  • Adaptive thresholding
  • Document segmentation
  • Hopfield neural network
  • Image segmentation
  • Interpolation
  • Optimization problem

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Electrical and Electronic Engineering

Cite this

A Hopfield neural network for adaptive image segmentation : An active surface paradigm. / Shen, Dinggang; Ip, Horace H S.

In: Pattern Recognition Letters, Vol. 18, No. 1, 01.01.1997, p. 37-48.

Research output: Contribution to journalArticle

@article{051e463c027144018145d8b1b44ebb27,
title = "A Hopfield neural network for adaptive image segmentation: An active surface paradigm",
abstract = "This paper presents an adaptive thresholding technique for separating objects from noisy and non-uniformly illuminated images. The construction of the threshold surface is formulated as an active surface optimization problem, which is then solved by a Hopfield neural network. We proposed four constraints which ensure the active threshold surface to conform with the underlying image topography. Compared with Yanowitz and Bruckstein's method, this method produces superior segmentations particularly when the edge segments are sparsely distributed in the image and under non-uniform illuminations. Using three types of artificial and real images, we show that this method converges faster and produces better segmentations compared with previous interpolation-based adaptive thresholding techniques.",
keywords = "Active threshold surface, Adaptive thresholding, Document segmentation, Hopfield neural network, Image segmentation, Interpolation, Optimization problem",
author = "Dinggang Shen and Ip, {Horace H S}",
year = "1997",
month = "1",
day = "1",
language = "English",
volume = "18",
pages = "37--48",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - A Hopfield neural network for adaptive image segmentation

T2 - An active surface paradigm

AU - Shen, Dinggang

AU - Ip, Horace H S

PY - 1997/1/1

Y1 - 1997/1/1

N2 - This paper presents an adaptive thresholding technique for separating objects from noisy and non-uniformly illuminated images. The construction of the threshold surface is formulated as an active surface optimization problem, which is then solved by a Hopfield neural network. We proposed four constraints which ensure the active threshold surface to conform with the underlying image topography. Compared with Yanowitz and Bruckstein's method, this method produces superior segmentations particularly when the edge segments are sparsely distributed in the image and under non-uniform illuminations. Using three types of artificial and real images, we show that this method converges faster and produces better segmentations compared with previous interpolation-based adaptive thresholding techniques.

AB - This paper presents an adaptive thresholding technique for separating objects from noisy and non-uniformly illuminated images. The construction of the threshold surface is formulated as an active surface optimization problem, which is then solved by a Hopfield neural network. We proposed four constraints which ensure the active threshold surface to conform with the underlying image topography. Compared with Yanowitz and Bruckstein's method, this method produces superior segmentations particularly when the edge segments are sparsely distributed in the image and under non-uniform illuminations. Using three types of artificial and real images, we show that this method converges faster and produces better segmentations compared with previous interpolation-based adaptive thresholding techniques.

KW - Active threshold surface

KW - Adaptive thresholding

KW - Document segmentation

KW - Hopfield neural network

KW - Image segmentation

KW - Interpolation

KW - Optimization problem

UR - http://www.scopus.com/inward/record.url?scp=0030644855&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030644855&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0030644855

VL - 18

SP - 37

EP - 48

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 1

ER -