Improvement of Canny algorithm and implementation in hardware

Hai Ping Wei; Bao Jun Zhao; Lin Bo Tang; Pei Kun He

Improvement of Canny algorithm and implementation in hardware

Hai Ping Wei^*, Bao Jun Zhao, Lin Bo Tang, Pei Kun He

^*Corresponding author for this work

School of Information and Electronics

Beijing Institute of Technology

Research output: Contribution to journal › Article › peer-review

15 Citations (Scopus)

Abstract

In the traditional Canny algorithm, the image is smoothed at first, and then the gradient is calculated by the planar filter. Based on the gradient histogram, the double thresholds can be found. At last, non-maximum suppression based on double thresholds is used to detect the edge. It is improved properly so that it can be implemented in hardware. Template in the improved algorithm is used in stead of convolution in the traditional algorithm, in order that it can be implemented in field programmable gate array (FPGA) by logic diagram totally.

Original language	English
Pages (from-to)	263-266
Number of pages	4
Journal	Guangxue Jishu/Optical Technique
Volume	32
Issue number	2
Publication status	Published - Mar 2006

Keywords

Canny operator
Double threshold
Edge detection
FPGA
Non-maximum suppression

Cite this

@article{0273ca01511b4c9c8c4571806822c097,

title = "Improvement of Canny algorithm and implementation in hardware",

abstract = "In the traditional Canny algorithm, the image is smoothed at first, and then the gradient is calculated by the planar filter. Based on the gradient histogram, the double thresholds can be found. At last, non-maximum suppression based on double thresholds is used to detect the edge. It is improved properly so that it can be implemented in hardware. Template in the improved algorithm is used in stead of convolution in the traditional algorithm, in order that it can be implemented in field programmable gate array (FPGA) by logic diagram totally.",

keywords = "Canny operator, Double threshold, Edge detection, FPGA, Non-maximum suppression",

author = "Wei, \{Hai Ping\} and Zhao, \{Bao Jun\} and Tang, \{Lin Bo\} and He, \{Pei Kun\}",

year = "2006",

month = mar,

language = "English",

volume = "32",

pages = "263--266",

journal = "Guangxue Jishu/Optical Technique",

issn = "1002-1582",

publisher = "Optical Technique",

number = "2",

}

TY - JOUR

T1 - Improvement of Canny algorithm and implementation in hardware

AU - Wei, Hai Ping

AU - Zhao, Bao Jun

AU - Tang, Lin Bo

AU - He, Pei Kun

PY - 2006/3

Y1 - 2006/3

N2 - In the traditional Canny algorithm, the image is smoothed at first, and then the gradient is calculated by the planar filter. Based on the gradient histogram, the double thresholds can be found. At last, non-maximum suppression based on double thresholds is used to detect the edge. It is improved properly so that it can be implemented in hardware. Template in the improved algorithm is used in stead of convolution in the traditional algorithm, in order that it can be implemented in field programmable gate array (FPGA) by logic diagram totally.

AB - In the traditional Canny algorithm, the image is smoothed at first, and then the gradient is calculated by the planar filter. Based on the gradient histogram, the double thresholds can be found. At last, non-maximum suppression based on double thresholds is used to detect the edge. It is improved properly so that it can be implemented in hardware. Template in the improved algorithm is used in stead of convolution in the traditional algorithm, in order that it can be implemented in field programmable gate array (FPGA) by logic diagram totally.

KW - Canny operator

KW - Double threshold

KW - Edge detection

KW - FPGA

KW - Non-maximum suppression

UR - http://www.scopus.com/inward/record.url?scp=33646354896&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:33646354896

SN - 1002-1582

VL - 32

SP - 263

EP - 266

JO - Guangxue Jishu/Optical Technique

JF - Guangxue Jishu/Optical Technique

IS - 2

ER -

Improvement of Canny algorithm and implementation in hardware

Abstract

Keywords

Other files and links

Fingerprint

Cite this