Weed image classification using Gabor wavelet and gradient field distribution

Asnor Juraiza Ishak, Aini Hussain, Mohd. Marzuki Mustafa

Research output: Contribution to journalArticle

49 Citations (Scopus)

Abstract

This paper presents an image analysis technique that utilizes a combination of a Gabor wavelet (GW) and gradient field distribution (GFD) techniques to extract a new set of feature vectors based on their directional texture properties for the classification of weed types. The feature extraction process involves the use of GW to enhance the directional feature of the images, followed by GFD implementation to produce histogram gradient orientation angles and additional steps to generate the histogram envelope. Next, curve fitting technique is used to estimate the envelope function to determine its quadratic polynomial equation, y = ax2 + bx + c and by taking the second derivative, the curvature value, a, is determined and used as a single input feature vector. The proposed technique was compared with another technique that also uses a single input feature obtained via GW algorithm implementation only. The overall classification accuracy utilizing the proposed technique is 94%, whereas using a GW only feature obtained 84% accuracy. The results obtained showed that this proposed technique is effective in performing weed classification.

Original languageEnglish
Pages (from-to)53-61
Number of pages9
JournalComputers and Electronics in Agriculture
Volume66
Issue number1
DOIs
Publication statusPublished - Apr 2009

Fingerprint

Image classification
image classification
wavelet
weed
weeds
histogram
Curve fitting
Image analysis
Feature extraction
Textures
methodology
Polynomials
image analysis
curvature
Derivatives
texture
distribution
chemical derivatives
extracts

Keywords

  • Feature extractions
  • Image processing
  • Single layer perceptron
  • Weed classification

ASJC Scopus subject areas

  • Agronomy and Crop Science
  • Horticulture
  • Forestry
  • Computer Science Applications
  • Animal Science and Zoology

Cite this

Weed image classification using Gabor wavelet and gradient field distribution. / Ishak, Asnor Juraiza; Hussain, Aini; Mustafa, Mohd. Marzuki.

In: Computers and Electronics in Agriculture, Vol. 66, No. 1, 04.2009, p. 53-61.

Research output: Contribution to journalArticle

@article{006986b52ccf497d8547d51478068e40,
title = "Weed image classification using Gabor wavelet and gradient field distribution",
abstract = "This paper presents an image analysis technique that utilizes a combination of a Gabor wavelet (GW) and gradient field distribution (GFD) techniques to extract a new set of feature vectors based on their directional texture properties for the classification of weed types. The feature extraction process involves the use of GW to enhance the directional feature of the images, followed by GFD implementation to produce histogram gradient orientation angles and additional steps to generate the histogram envelope. Next, curve fitting technique is used to estimate the envelope function to determine its quadratic polynomial equation, y = ax2 + bx + c and by taking the second derivative, the curvature value, a, is determined and used as a single input feature vector. The proposed technique was compared with another technique that also uses a single input feature obtained via GW algorithm implementation only. The overall classification accuracy utilizing the proposed technique is 94{\%}, whereas using a GW only feature obtained 84{\%} accuracy. The results obtained showed that this proposed technique is effective in performing weed classification.",
keywords = "Feature extractions, Image processing, Single layer perceptron, Weed classification",
author = "Ishak, {Asnor Juraiza} and Aini Hussain and Mustafa, {Mohd. Marzuki}",
year = "2009",
month = "4",
doi = "10.1016/j.compag.2008.12.003",
language = "English",
volume = "66",
pages = "53--61",
journal = "Computers and Electronics in Agriculture",
issn = "0168-1699",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Weed image classification using Gabor wavelet and gradient field distribution

AU - Ishak, Asnor Juraiza

AU - Hussain, Aini

AU - Mustafa, Mohd. Marzuki

PY - 2009/4

Y1 - 2009/4

N2 - This paper presents an image analysis technique that utilizes a combination of a Gabor wavelet (GW) and gradient field distribution (GFD) techniques to extract a new set of feature vectors based on their directional texture properties for the classification of weed types. The feature extraction process involves the use of GW to enhance the directional feature of the images, followed by GFD implementation to produce histogram gradient orientation angles and additional steps to generate the histogram envelope. Next, curve fitting technique is used to estimate the envelope function to determine its quadratic polynomial equation, y = ax2 + bx + c and by taking the second derivative, the curvature value, a, is determined and used as a single input feature vector. The proposed technique was compared with another technique that also uses a single input feature obtained via GW algorithm implementation only. The overall classification accuracy utilizing the proposed technique is 94%, whereas using a GW only feature obtained 84% accuracy. The results obtained showed that this proposed technique is effective in performing weed classification.

AB - This paper presents an image analysis technique that utilizes a combination of a Gabor wavelet (GW) and gradient field distribution (GFD) techniques to extract a new set of feature vectors based on their directional texture properties for the classification of weed types. The feature extraction process involves the use of GW to enhance the directional feature of the images, followed by GFD implementation to produce histogram gradient orientation angles and additional steps to generate the histogram envelope. Next, curve fitting technique is used to estimate the envelope function to determine its quadratic polynomial equation, y = ax2 + bx + c and by taking the second derivative, the curvature value, a, is determined and used as a single input feature vector. The proposed technique was compared with another technique that also uses a single input feature obtained via GW algorithm implementation only. The overall classification accuracy utilizing the proposed technique is 94%, whereas using a GW only feature obtained 84% accuracy. The results obtained showed that this proposed technique is effective in performing weed classification.

KW - Feature extractions

KW - Image processing

KW - Single layer perceptron

KW - Weed classification

UR - http://www.scopus.com/inward/record.url?scp=61349112689&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=61349112689&partnerID=8YFLogxK

U2 - 10.1016/j.compag.2008.12.003

DO - 10.1016/j.compag.2008.12.003

M3 - Article

AN - SCOPUS:61349112689

VL - 66

SP - 53

EP - 61

JO - Computers and Electronics in Agriculture

JF - Computers and Electronics in Agriculture

SN - 0168-1699

IS - 1

ER -