Textural imaging and discriminant analysis for distinguishing weeds for spot spraying

G. E. Meyer, T. Mehta, M. F. Kocher, D. A. Mortensen, Ashok K Samal

Research output: Contribution to journalArticle

140 Citations (Scopus)

Abstract

Advanced computer vision and statistical methods were employed for identifying living plants from soil/residue background for two species of grasses (Shattercane, Green Foxtail) and two broadleaf species (Velvetleaf, Red Root Pigweed) weeds. The excess green index method was used as a contrast enhancement for specifically identifying plant from soil regions. Excess green classified plant and soil regions correctly over the entire three-week observation period with high accuracies (99% plus). Plant and soil binary images were derived from excess green images and provided edge boundaries. These boundaries were used with corresponding gray scale images to extract four classical textural features for plants and soil: angular second moment, inertia, entropy, and local homogeneity. These features were derived from the co-occurrence matrix. Stepwise and canonical discriminant analyses were used to test the classification performance of the texture and excess green features. Discrimination models of local homogeneity, inertia, and angular second moment were found to classify grass and broadleaf categories of plants, with classification accuracies of 93 and 85%, respectively. Classification accuracies of individual species only ranged from 30 to 77%. Soil classification accuracies were also high for textural feature algorithms (97%). The time required to produce tokensets ranged from 15 to 20 s on a UNIX computer system. Additional time required for the system to reach a plant/soil classification ranged from 5 to 10 s. This translated into an overall system response time of 20 to 30 s, with the preprocessing step constituting the major part of the system response time.

Original languageEnglish (US)
Pages (from-to)1189-1197
Number of pages9
JournalTransactions of the American Society of Agricultural Engineers
Volume41
Issue number4
StatePublished - Jul 1 1998

Fingerprint

Discriminant Analysis
Discriminant analysis
Spraying
discriminant analysis
spraying
weed
Soil
weeds
image analysis
Soils
Imaging techniques
soil classification
soil
Poaceae
taxonomy
inertia
homogeneity
Reaction Time
grass
Viridiplantae

Keywords

  • Color index
  • Discriminant analysis
  • Image processing
  • Texture

ASJC Scopus subject areas

  • Agricultural and Biological Sciences (miscellaneous)

Cite this

Textural imaging and discriminant analysis for distinguishing weeds for spot spraying. / Meyer, G. E.; Mehta, T.; Kocher, M. F.; Mortensen, D. A.; Samal, Ashok K.

In: Transactions of the American Society of Agricultural Engineers, Vol. 41, No. 4, 01.07.1998, p. 1189-1197.

Research output: Contribution to journalArticle

@article{62ec5f5017d44935a2157e38d17fa513,
title = "Textural imaging and discriminant analysis for distinguishing weeds for spot spraying",
abstract = "Advanced computer vision and statistical methods were employed for identifying living plants from soil/residue background for two species of grasses (Shattercane, Green Foxtail) and two broadleaf species (Velvetleaf, Red Root Pigweed) weeds. The excess green index method was used as a contrast enhancement for specifically identifying plant from soil regions. Excess green classified plant and soil regions correctly over the entire three-week observation period with high accuracies (99{\%} plus). Plant and soil binary images were derived from excess green images and provided edge boundaries. These boundaries were used with corresponding gray scale images to extract four classical textural features for plants and soil: angular second moment, inertia, entropy, and local homogeneity. These features were derived from the co-occurrence matrix. Stepwise and canonical discriminant analyses were used to test the classification performance of the texture and excess green features. Discrimination models of local homogeneity, inertia, and angular second moment were found to classify grass and broadleaf categories of plants, with classification accuracies of 93 and 85{\%}, respectively. Classification accuracies of individual species only ranged from 30 to 77{\%}. Soil classification accuracies were also high for textural feature algorithms (97{\%}). The time required to produce tokensets ranged from 15 to 20 s on a UNIX computer system. Additional time required for the system to reach a plant/soil classification ranged from 5 to 10 s. This translated into an overall system response time of 20 to 30 s, with the preprocessing step constituting the major part of the system response time.",
keywords = "Color index, Discriminant analysis, Image processing, Texture",
author = "Meyer, {G. E.} and T. Mehta and Kocher, {M. F.} and Mortensen, {D. A.} and Samal, {Ashok K}",
year = "1998",
month = "7",
day = "1",
language = "English (US)",
volume = "41",
pages = "1189--1197",
journal = "Transactions of the ASABE",
issn = "2151-0032",
publisher = "American Society of Agricultural and Biological Engineers",
number = "4",

}

TY - JOUR

T1 - Textural imaging and discriminant analysis for distinguishing weeds for spot spraying

AU - Meyer, G. E.

AU - Mehta, T.

AU - Kocher, M. F.

AU - Mortensen, D. A.

AU - Samal, Ashok K

PY - 1998/7/1

Y1 - 1998/7/1

N2 - Advanced computer vision and statistical methods were employed for identifying living plants from soil/residue background for two species of grasses (Shattercane, Green Foxtail) and two broadleaf species (Velvetleaf, Red Root Pigweed) weeds. The excess green index method was used as a contrast enhancement for specifically identifying plant from soil regions. Excess green classified plant and soil regions correctly over the entire three-week observation period with high accuracies (99% plus). Plant and soil binary images were derived from excess green images and provided edge boundaries. These boundaries were used with corresponding gray scale images to extract four classical textural features for plants and soil: angular second moment, inertia, entropy, and local homogeneity. These features were derived from the co-occurrence matrix. Stepwise and canonical discriminant analyses were used to test the classification performance of the texture and excess green features. Discrimination models of local homogeneity, inertia, and angular second moment were found to classify grass and broadleaf categories of plants, with classification accuracies of 93 and 85%, respectively. Classification accuracies of individual species only ranged from 30 to 77%. Soil classification accuracies were also high for textural feature algorithms (97%). The time required to produce tokensets ranged from 15 to 20 s on a UNIX computer system. Additional time required for the system to reach a plant/soil classification ranged from 5 to 10 s. This translated into an overall system response time of 20 to 30 s, with the preprocessing step constituting the major part of the system response time.

AB - Advanced computer vision and statistical methods were employed for identifying living plants from soil/residue background for two species of grasses (Shattercane, Green Foxtail) and two broadleaf species (Velvetleaf, Red Root Pigweed) weeds. The excess green index method was used as a contrast enhancement for specifically identifying plant from soil regions. Excess green classified plant and soil regions correctly over the entire three-week observation period with high accuracies (99% plus). Plant and soil binary images were derived from excess green images and provided edge boundaries. These boundaries were used with corresponding gray scale images to extract four classical textural features for plants and soil: angular second moment, inertia, entropy, and local homogeneity. These features were derived from the co-occurrence matrix. Stepwise and canonical discriminant analyses were used to test the classification performance of the texture and excess green features. Discrimination models of local homogeneity, inertia, and angular second moment were found to classify grass and broadleaf categories of plants, with classification accuracies of 93 and 85%, respectively. Classification accuracies of individual species only ranged from 30 to 77%. Soil classification accuracies were also high for textural feature algorithms (97%). The time required to produce tokensets ranged from 15 to 20 s on a UNIX computer system. Additional time required for the system to reach a plant/soil classification ranged from 5 to 10 s. This translated into an overall system response time of 20 to 30 s, with the preprocessing step constituting the major part of the system response time.

KW - Color index

KW - Discriminant analysis

KW - Image processing

KW - Texture

UR - http://www.scopus.com/inward/record.url?scp=0032122198&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032122198&partnerID=8YFLogxK

M3 - Article

VL - 41

SP - 1189

EP - 1197

JO - Transactions of the ASABE

JF - Transactions of the ASABE

SN - 2151-0032

IS - 4

ER -