Quantifying cereal crop movement through hemispherical video analysis of agricultural plots

Alexander Q. Susko, Peter M Marchetto, D. Jo Heuschele, Kevin P Smith

Research output: Contribution to journalArticle

Abstract

Background: Violent movement of crop stems can lead to failure under high winds. Known as lodging, this phenomenon is particularly detrimental to cool-season cereals such as oat, barley, and wheat; contributing to yield and economic losses. Phenotyping the movement of cereal crops in real-time could aid in the breeding and selecting of lodging resistant cereals. Since no methods exist to quantify dynamic, real time plant responses in an agricultural setting, we devised a video analysis protocol to quantify mean frequency and amplitude of plant movement for a 360° field of view camera system. Results: We present both the image analysis method for identifying predefined regions of a 2D field design as they appear on 360° field of view video, as well as a signal processing pipeline to quantify movement from time varying color signals from plot canopies within these predefined field regions. We detected significant differences in the natural frequency and amplitude of plant movement from video of 16 cereal cultivars planted in a randomized complete block design on five different windy days. Natural frequencies quantified by this method averaged 1.37 Hz, while over 2.5-fold differences in amplitude within similar frequency ranges were detected across the 16 cereal cultivars. Conclusions: This method is sensitive enough to systematically differentiate small frequency and amplitude differences in cultivar movement, and shows promise for investigating the physiological basis for differences in cereal movement and lodging resistance. The relative accuracy of the plot demarcation protocol suggests it could be used for other high-throughput phenotyping applications that require both high image resolution and a large field of view.

Original languageEnglish (US)
Article number55
JournalPlant Methods
Volume15
Issue number1
DOIs
StatePublished - May 24 2019

Fingerprint

grain crops
lodging
cultivars
lodging resistance
phenotype
methodology
cameras
plant response
Hordeum
oats
Triticum
Breeding
barley
Edible Grain
image analysis
canopy
Color
economics
Economics
wheat

Keywords

  • 360 Camera
  • Barley
  • High throughput phenotyping
  • Image analysis
  • Lodging
  • Oat
  • Wheat

PubMed: MeSH publication types

  • Journal Article

Cite this

Quantifying cereal crop movement through hemispherical video analysis of agricultural plots. / Susko, Alexander Q.; Marchetto, Peter M; Jo Heuschele, D.; Smith, Kevin P.

In: Plant Methods, Vol. 15, No. 1, 55, 24.05.2019.

Research output: Contribution to journalArticle

@article{99fabbc790b8415c80e68fe924377da8,
title = "Quantifying cereal crop movement through hemispherical video analysis of agricultural plots",
abstract = "Background: Violent movement of crop stems can lead to failure under high winds. Known as lodging, this phenomenon is particularly detrimental to cool-season cereals such as oat, barley, and wheat; contributing to yield and economic losses. Phenotyping the movement of cereal crops in real-time could aid in the breeding and selecting of lodging resistant cereals. Since no methods exist to quantify dynamic, real time plant responses in an agricultural setting, we devised a video analysis protocol to quantify mean frequency and amplitude of plant movement for a 360° field of view camera system. Results: We present both the image analysis method for identifying predefined regions of a 2D field design as they appear on 360° field of view video, as well as a signal processing pipeline to quantify movement from time varying color signals from plot canopies within these predefined field regions. We detected significant differences in the natural frequency and amplitude of plant movement from video of 16 cereal cultivars planted in a randomized complete block design on five different windy days. Natural frequencies quantified by this method averaged 1.37 Hz, while over 2.5-fold differences in amplitude within similar frequency ranges were detected across the 16 cereal cultivars. Conclusions: This method is sensitive enough to systematically differentiate small frequency and amplitude differences in cultivar movement, and shows promise for investigating the physiological basis for differences in cereal movement and lodging resistance. The relative accuracy of the plot demarcation protocol suggests it could be used for other high-throughput phenotyping applications that require both high image resolution and a large field of view.",
keywords = "360 Camera, Barley, High throughput phenotyping, Image analysis, Lodging, Oat, Wheat",
author = "Susko, {Alexander Q.} and Marchetto, {Peter M} and {Jo Heuschele}, D. and Smith, {Kevin P}",
year = "2019",
month = "5",
day = "24",
doi = "10.1186/s13007-019-0437-5",
language = "English (US)",
volume = "15",
journal = "Plant Methods",
issn = "1746-4811",
publisher = "BioMed Central",
number = "1",

}

TY - JOUR

T1 - Quantifying cereal crop movement through hemispherical video analysis of agricultural plots

AU - Susko, Alexander Q.

AU - Marchetto, Peter M

AU - Jo Heuschele, D.

AU - Smith, Kevin P

PY - 2019/5/24

Y1 - 2019/5/24

N2 - Background: Violent movement of crop stems can lead to failure under high winds. Known as lodging, this phenomenon is particularly detrimental to cool-season cereals such as oat, barley, and wheat; contributing to yield and economic losses. Phenotyping the movement of cereal crops in real-time could aid in the breeding and selecting of lodging resistant cereals. Since no methods exist to quantify dynamic, real time plant responses in an agricultural setting, we devised a video analysis protocol to quantify mean frequency and amplitude of plant movement for a 360° field of view camera system. Results: We present both the image analysis method for identifying predefined regions of a 2D field design as they appear on 360° field of view video, as well as a signal processing pipeline to quantify movement from time varying color signals from plot canopies within these predefined field regions. We detected significant differences in the natural frequency and amplitude of plant movement from video of 16 cereal cultivars planted in a randomized complete block design on five different windy days. Natural frequencies quantified by this method averaged 1.37 Hz, while over 2.5-fold differences in amplitude within similar frequency ranges were detected across the 16 cereal cultivars. Conclusions: This method is sensitive enough to systematically differentiate small frequency and amplitude differences in cultivar movement, and shows promise for investigating the physiological basis for differences in cereal movement and lodging resistance. The relative accuracy of the plot demarcation protocol suggests it could be used for other high-throughput phenotyping applications that require both high image resolution and a large field of view.

AB - Background: Violent movement of crop stems can lead to failure under high winds. Known as lodging, this phenomenon is particularly detrimental to cool-season cereals such as oat, barley, and wheat; contributing to yield and economic losses. Phenotyping the movement of cereal crops in real-time could aid in the breeding and selecting of lodging resistant cereals. Since no methods exist to quantify dynamic, real time plant responses in an agricultural setting, we devised a video analysis protocol to quantify mean frequency and amplitude of plant movement for a 360° field of view camera system. Results: We present both the image analysis method for identifying predefined regions of a 2D field design as they appear on 360° field of view video, as well as a signal processing pipeline to quantify movement from time varying color signals from plot canopies within these predefined field regions. We detected significant differences in the natural frequency and amplitude of plant movement from video of 16 cereal cultivars planted in a randomized complete block design on five different windy days. Natural frequencies quantified by this method averaged 1.37 Hz, while over 2.5-fold differences in amplitude within similar frequency ranges were detected across the 16 cereal cultivars. Conclusions: This method is sensitive enough to systematically differentiate small frequency and amplitude differences in cultivar movement, and shows promise for investigating the physiological basis for differences in cereal movement and lodging resistance. The relative accuracy of the plot demarcation protocol suggests it could be used for other high-throughput phenotyping applications that require both high image resolution and a large field of view.

KW - 360 Camera

KW - Barley

KW - High throughput phenotyping

KW - Image analysis

KW - Lodging

KW - Oat

KW - Wheat

UR - http://www.scopus.com/inward/record.url?scp=85066431179&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066431179&partnerID=8YFLogxK

U2 - 10.1186/s13007-019-0437-5

DO - 10.1186/s13007-019-0437-5

M3 - Article

C2 - 31139244

AN - SCOPUS:85066431179

VL - 15

JO - Plant Methods

JF - Plant Methods

SN - 1746-4811

IS - 1

M1 - 55

ER -