Implementation of feature extraction technique from gray level co-occurrence matrix to classify narrow and broad weed in oil palm plantation

Kamarul Hawari Ghazali, Mohd. Marzuki Mustafa, Aini Hussain

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Fundamental step in digital image processing started with image acquisition, pre-processing, processing, representation and classification. In this paper, image processing technique namely Gray Level Co-occurrence Matrix (GLCM) has been used for texture analysis. Texture of an image is an important feature to both human and machine in making decision for a classification task. It has very unique concept and normally cannot be interpreted by a mathematical formulation. To a computer, texture is classified as rough when the difference between the high and low points is large. On the contrary, silky or smooth texture would have little difference between the high and low points. Due to this unique characteristic, feature extraction based on texture context is chosen for the two categorical weed pattern classification purposes. Hence, two feature vectors from the weed texture context namely, contrast and regularity of images are extracted distinguishing between broad and narrow weed. The performance of the proposed feature extraction was evaluated in terms of classification accuracies and the result shows that the proposed technique gives correct classification rate is above 80%.

Original languageEnglish
Pages (from-to)68-75
Number of pages8
JournalEuropean Journal of Scientific Research
Volume20
Issue number1
Publication statusPublished - 2008

Fingerprint

Gray Level Co-occurrence Matrix
Palm oil
Elaeis guineensis
Feature Extraction
weed
Feature extraction
Texture
Oils
plantation
Textures
plantations
texture
weeds
Classify
matrix
oil
image processing
Image processing
methodology
Digital Image Processing

Keywords

  • Contrast
  • GLCM
  • Regularity
  • Texture
  • Weed

ASJC Scopus subject areas

  • General

Cite this

@article{0c474c9784394320b7ee3ee518d8202c,
title = "Implementation of feature extraction technique from gray level co-occurrence matrix to classify narrow and broad weed in oil palm plantation",
abstract = "Fundamental step in digital image processing started with image acquisition, pre-processing, processing, representation and classification. In this paper, image processing technique namely Gray Level Co-occurrence Matrix (GLCM) has been used for texture analysis. Texture of an image is an important feature to both human and machine in making decision for a classification task. It has very unique concept and normally cannot be interpreted by a mathematical formulation. To a computer, texture is classified as rough when the difference between the high and low points is large. On the contrary, silky or smooth texture would have little difference between the high and low points. Due to this unique characteristic, feature extraction based on texture context is chosen for the two categorical weed pattern classification purposes. Hence, two feature vectors from the weed texture context namely, contrast and regularity of images are extracted distinguishing between broad and narrow weed. The performance of the proposed feature extraction was evaluated in terms of classification accuracies and the result shows that the proposed technique gives correct classification rate is above 80{\%}.",
keywords = "Contrast, GLCM, Regularity, Texture, Weed",
author = "Ghazali, {Kamarul Hawari} and Mustafa, {Mohd. Marzuki} and Aini Hussain",
year = "2008",
language = "English",
volume = "20",
pages = "68--75",
journal = "European Journal of Scientific Research",
issn = "1450-202X",
publisher = "European Journals Inc.",
number = "1",

}

TY - JOUR

T1 - Implementation of feature extraction technique from gray level co-occurrence matrix to classify narrow and broad weed in oil palm plantation

AU - Ghazali, Kamarul Hawari

AU - Mustafa, Mohd. Marzuki

AU - Hussain, Aini

PY - 2008

Y1 - 2008

N2 - Fundamental step in digital image processing started with image acquisition, pre-processing, processing, representation and classification. In this paper, image processing technique namely Gray Level Co-occurrence Matrix (GLCM) has been used for texture analysis. Texture of an image is an important feature to both human and machine in making decision for a classification task. It has very unique concept and normally cannot be interpreted by a mathematical formulation. To a computer, texture is classified as rough when the difference between the high and low points is large. On the contrary, silky or smooth texture would have little difference between the high and low points. Due to this unique characteristic, feature extraction based on texture context is chosen for the two categorical weed pattern classification purposes. Hence, two feature vectors from the weed texture context namely, contrast and regularity of images are extracted distinguishing between broad and narrow weed. The performance of the proposed feature extraction was evaluated in terms of classification accuracies and the result shows that the proposed technique gives correct classification rate is above 80%.

AB - Fundamental step in digital image processing started with image acquisition, pre-processing, processing, representation and classification. In this paper, image processing technique namely Gray Level Co-occurrence Matrix (GLCM) has been used for texture analysis. Texture of an image is an important feature to both human and machine in making decision for a classification task. It has very unique concept and normally cannot be interpreted by a mathematical formulation. To a computer, texture is classified as rough when the difference between the high and low points is large. On the contrary, silky or smooth texture would have little difference between the high and low points. Due to this unique characteristic, feature extraction based on texture context is chosen for the two categorical weed pattern classification purposes. Hence, two feature vectors from the weed texture context namely, contrast and regularity of images are extracted distinguishing between broad and narrow weed. The performance of the proposed feature extraction was evaluated in terms of classification accuracies and the result shows that the proposed technique gives correct classification rate is above 80%.

KW - Contrast

KW - GLCM

KW - Regularity

KW - Texture

KW - Weed

UR - http://www.scopus.com/inward/record.url?scp=53749103202&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=53749103202&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:53749103202

VL - 20

SP - 68

EP - 75

JO - European Journal of Scientific Research

JF - European Journal of Scientific Research

SN - 1450-202X

IS - 1

ER -