Robust edge detection based on canny algorithm for noisy images

Haider O. Lawend, Anuar Mikdad Muad, Aini Hussain

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

The aim of many edge detection techniques is to highlight edges in an image. However, due to nature of the edge detection that is based on the derivative operation, this process often amplifies noises too. Therefore, there is always a trade-off in the edge detection technique between extracting information and suppressing noise. There is variety of edge detectors or operators with different sizes of kernel. This paper proposes an edge detection technique based on traditional Canny edge detector. Unlike many established edge detection techniques that focus on the gradient in grayscale image, the proposed technique includes two more features: the length and the directional change of the edges. The inclusion of the two features helps to increase the robustness of the proposed technique towards noise. The proposed technique is tested with synthetic and natural images. Results are compared with other established edge detection techniques and demonstrate that the proposed technique is able to detect low contrast edges and highly resistance to different types of noise. As a result, the trade-off between the information and noise in image edge detection is reduced.

Original languageEnglish
Pages (from-to)5104-5114
Number of pages11
JournalJournal of Theoretical and Applied Information Technology
Volume95
Issue number19
Publication statusPublished - 15 Oct 2017

Fingerprint

Edge Detection
Edge detection
Detectors
Trade-offs
Detector
Derivatives
Inclusion
kernel
Gradient
Robustness
Derivative
Operator

Keywords

  • Canny edge detection
  • Directional change
  • Edge gradient
  • Edge length
  • Noise suppression

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Robust edge detection based on canny algorithm for noisy images. / Lawend, Haider O.; Muad, Anuar Mikdad; Hussain, Aini.

In: Journal of Theoretical and Applied Information Technology, Vol. 95, No. 19, 15.10.2017, p. 5104-5114.

Research output: Contribution to journalArticle

@article{6e9d86ab8ef94f15b05c5e514b10ae14,
title = "Robust edge detection based on canny algorithm for noisy images",
abstract = "The aim of many edge detection techniques is to highlight edges in an image. However, due to nature of the edge detection that is based on the derivative operation, this process often amplifies noises too. Therefore, there is always a trade-off in the edge detection technique between extracting information and suppressing noise. There is variety of edge detectors or operators with different sizes of kernel. This paper proposes an edge detection technique based on traditional Canny edge detector. Unlike many established edge detection techniques that focus on the gradient in grayscale image, the proposed technique includes two more features: the length and the directional change of the edges. The inclusion of the two features helps to increase the robustness of the proposed technique towards noise. The proposed technique is tested with synthetic and natural images. Results are compared with other established edge detection techniques and demonstrate that the proposed technique is able to detect low contrast edges and highly resistance to different types of noise. As a result, the trade-off between the information and noise in image edge detection is reduced.",
keywords = "Canny edge detection, Directional change, Edge gradient, Edge length, Noise suppression",
author = "Lawend, {Haider O.} and Muad, {Anuar Mikdad} and Aini Hussain",
year = "2017",
month = "10",
day = "15",
language = "English",
volume = "95",
pages = "5104--5114",
journal = "Journal of Theoretical and Applied Information Technology",
issn = "1992-8645",
publisher = "Asian Research Publishing Network (ARPN)",
number = "19",

}

TY - JOUR

T1 - Robust edge detection based on canny algorithm for noisy images

AU - Lawend, Haider O.

AU - Muad, Anuar Mikdad

AU - Hussain, Aini

PY - 2017/10/15

Y1 - 2017/10/15

N2 - The aim of many edge detection techniques is to highlight edges in an image. However, due to nature of the edge detection that is based on the derivative operation, this process often amplifies noises too. Therefore, there is always a trade-off in the edge detection technique between extracting information and suppressing noise. There is variety of edge detectors or operators with different sizes of kernel. This paper proposes an edge detection technique based on traditional Canny edge detector. Unlike many established edge detection techniques that focus on the gradient in grayscale image, the proposed technique includes two more features: the length and the directional change of the edges. The inclusion of the two features helps to increase the robustness of the proposed technique towards noise. The proposed technique is tested with synthetic and natural images. Results are compared with other established edge detection techniques and demonstrate that the proposed technique is able to detect low contrast edges and highly resistance to different types of noise. As a result, the trade-off between the information and noise in image edge detection is reduced.

AB - The aim of many edge detection techniques is to highlight edges in an image. However, due to nature of the edge detection that is based on the derivative operation, this process often amplifies noises too. Therefore, there is always a trade-off in the edge detection technique between extracting information and suppressing noise. There is variety of edge detectors or operators with different sizes of kernel. This paper proposes an edge detection technique based on traditional Canny edge detector. Unlike many established edge detection techniques that focus on the gradient in grayscale image, the proposed technique includes two more features: the length and the directional change of the edges. The inclusion of the two features helps to increase the robustness of the proposed technique towards noise. The proposed technique is tested with synthetic and natural images. Results are compared with other established edge detection techniques and demonstrate that the proposed technique is able to detect low contrast edges and highly resistance to different types of noise. As a result, the trade-off between the information and noise in image edge detection is reduced.

KW - Canny edge detection

KW - Directional change

KW - Edge gradient

KW - Edge length

KW - Noise suppression

UR - http://www.scopus.com/inward/record.url?scp=85031751859&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85031751859&partnerID=8YFLogxK

M3 - Article

VL - 95

SP - 5104

EP - 5114

JO - Journal of Theoretical and Applied Information Technology

JF - Journal of Theoretical and Applied Information Technology

SN - 1992-8645

IS - 19

ER -