Edge representation and recognition using neural networks

O. Kwon, C. Lee

Research output: Contribution to conferencePaper

2 Citations (Scopus)

Abstract

In this paper, we propose a new approach to represent and recognize edges of objects using multilayer feedforward neural networks. First, we will show how an edge of an object can be represented by neural networks. This is accomplished by generating two classes consisting of samples that lie on each side of the edge and then by training a neural network to classify the two classes. If the training is successfully accomplished, the resulting neural network will have a decision boundary that matches the edge we want to represent. Second, we will propose a matching algorithm that identifies an arbitrarily rotated and shifted edge. The matching algorithm uses a gradient descent algorithm. The proposed algorithm can be used in the area of object representation and recognition. In addition, we will investigate the relationship between the number of hidden neurons and complexity of edges.

Original languageEnglish
Pages110-113
Number of pages4
Publication statusPublished - 2001 Jan 1
Event2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001) - Pusan, Korea, Republic of
Duration: 2001 Jun 122001 Jun 16

Other

Other2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001)
CountryKorea, Republic of
CityPusan
Period01/6/1201/6/16

Fingerprint

Neural networks
Feedforward neural networks
Multilayer neural networks
Neurons

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

Kwon, O., & Lee, C. (2001). Edge representation and recognition using neural networks. 110-113. Paper presented at 2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001), Pusan, Korea, Republic of.
Kwon, O. ; Lee, C. / Edge representation and recognition using neural networks. Paper presented at 2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001), Pusan, Korea, Republic of.4 p.
@conference{aaf194f19b02432a912e823c9623fed3,
title = "Edge representation and recognition using neural networks",
abstract = "In this paper, we propose a new approach to represent and recognize edges of objects using multilayer feedforward neural networks. First, we will show how an edge of an object can be represented by neural networks. This is accomplished by generating two classes consisting of samples that lie on each side of the edge and then by training a neural network to classify the two classes. If the training is successfully accomplished, the resulting neural network will have a decision boundary that matches the edge we want to represent. Second, we will propose a matching algorithm that identifies an arbitrarily rotated and shifted edge. The matching algorithm uses a gradient descent algorithm. The proposed algorithm can be used in the area of object representation and recognition. In addition, we will investigate the relationship between the number of hidden neurons and complexity of edges.",
author = "O. Kwon and C. Lee",
year = "2001",
month = "1",
day = "1",
language = "English",
pages = "110--113",
note = "2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001) ; Conference date: 12-06-2001 Through 16-06-2001",

}

Kwon, O & Lee, C 2001, 'Edge representation and recognition using neural networks', Paper presented at 2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001), Pusan, Korea, Republic of, 01/6/12 - 01/6/16 pp. 110-113.

Edge representation and recognition using neural networks. / Kwon, O.; Lee, C.

2001. 110-113 Paper presented at 2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001), Pusan, Korea, Republic of.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Edge representation and recognition using neural networks

AU - Kwon, O.

AU - Lee, C.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - In this paper, we propose a new approach to represent and recognize edges of objects using multilayer feedforward neural networks. First, we will show how an edge of an object can be represented by neural networks. This is accomplished by generating two classes consisting of samples that lie on each side of the edge and then by training a neural network to classify the two classes. If the training is successfully accomplished, the resulting neural network will have a decision boundary that matches the edge we want to represent. Second, we will propose a matching algorithm that identifies an arbitrarily rotated and shifted edge. The matching algorithm uses a gradient descent algorithm. The proposed algorithm can be used in the area of object representation and recognition. In addition, we will investigate the relationship between the number of hidden neurons and complexity of edges.

AB - In this paper, we propose a new approach to represent and recognize edges of objects using multilayer feedforward neural networks. First, we will show how an edge of an object can be represented by neural networks. This is accomplished by generating two classes consisting of samples that lie on each side of the edge and then by training a neural network to classify the two classes. If the training is successfully accomplished, the resulting neural network will have a decision boundary that matches the edge we want to represent. Second, we will propose a matching algorithm that identifies an arbitrarily rotated and shifted edge. The matching algorithm uses a gradient descent algorithm. The proposed algorithm can be used in the area of object representation and recognition. In addition, we will investigate the relationship between the number of hidden neurons and complexity of edges.

UR - http://www.scopus.com/inward/record.url?scp=0034844098&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034844098&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0034844098

SP - 110

EP - 113

ER -

Kwon O, Lee C. Edge representation and recognition using neural networks. 2001. Paper presented at 2001 IEEE International Symposium on Industrial Electronics Proceedings (ISIE 2001), Pusan, Korea, Republic of.