Three dimensional chemometric analyses of hyperspectral images for beef tenderness forecasting

Govindarajan Konda Naganathan, Kim Cluff, Ashok K Samal, Chris R. Calkins, David D. Jones, George E. Meyer, Jeyamkondan Subbiah

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

A prototype on-line hyperspectral imaging system (λ = 400-1000 nm) was developed and used to acquire images of exposed ribeye muscle on hanging beef carcasses (n = 274) at 2-day postmortem in a commercial beef packing plant. After image acquisition, a strip steak was cut from each carcass and vacuum packaged. After aging for 14 days, the steaks were cooked and Warner-Bratzler shear force values were collected as a measure of tenderness. Four different principal component analysis-based dimensionality reduction methods were implemented to reduce information redundancy in beef hyperspectral images. Textural features extracted from the 2-day hyperspectral images were modeled using Fisher's linear discriminant (FLD), support vector machines (SVM), and decision tree (DT) models to predict 14-day aged, cooked beef tenderness. Based on a true validation procedure using 101 samples, the FLD model yielded a tender certification accuracy of 86.7%. In addition, wavelengths corresponding to myoglobin and its derivatives (541, 577, and 635 nm), beef aging (541, 577, 635, 756, and 980 nm), protein (910 nm), fat (928 nm), and water (739, 756, and 988 nm) were identified.

Original languageEnglish (US)
Pages (from-to)309-320
Number of pages12
JournalJournal of Food Engineering
Volume169
DOIs
StatePublished - Jan 1 2016

Fingerprint

hyperspectral imagery
chemometrics
beef
steaks
packing houses
beef carcasses
myoglobin
certification
Decision Trees
prototypes
Myoglobin
shear stress
Certification
wavelengths
Vacuum
Principal Component Analysis
principal component analysis
chemical derivatives
image analysis
Linear Models

Keywords

  • Decision tree
  • Fisher's linear discriminant model
  • Instrument grading
  • Partial least squares analysis
  • Principal component analysis
  • Support vector machines

ASJC Scopus subject areas

  • Food Science

Cite this

Three dimensional chemometric analyses of hyperspectral images for beef tenderness forecasting. / Konda Naganathan, Govindarajan; Cluff, Kim; Samal, Ashok K; Calkins, Chris R.; Jones, David D.; Meyer, George E.; Subbiah, Jeyamkondan.

In: Journal of Food Engineering, Vol. 169, 01.01.2016, p. 309-320.

Research output: Contribution to journalArticle

Konda Naganathan, Govindarajan ; Cluff, Kim ; Samal, Ashok K ; Calkins, Chris R. ; Jones, David D. ; Meyer, George E. ; Subbiah, Jeyamkondan. / Three dimensional chemometric analyses of hyperspectral images for beef tenderness forecasting. In: Journal of Food Engineering. 2016 ; Vol. 169. pp. 309-320.
@article{6d8bea0b989341fcaae0dc532b657a62,
title = "Three dimensional chemometric analyses of hyperspectral images for beef tenderness forecasting",
abstract = "A prototype on-line hyperspectral imaging system (λ = 400-1000 nm) was developed and used to acquire images of exposed ribeye muscle on hanging beef carcasses (n = 274) at 2-day postmortem in a commercial beef packing plant. After image acquisition, a strip steak was cut from each carcass and vacuum packaged. After aging for 14 days, the steaks were cooked and Warner-Bratzler shear force values were collected as a measure of tenderness. Four different principal component analysis-based dimensionality reduction methods were implemented to reduce information redundancy in beef hyperspectral images. Textural features extracted from the 2-day hyperspectral images were modeled using Fisher's linear discriminant (FLD), support vector machines (SVM), and decision tree (DT) models to predict 14-day aged, cooked beef tenderness. Based on a true validation procedure using 101 samples, the FLD model yielded a tender certification accuracy of 86.7{\%}. In addition, wavelengths corresponding to myoglobin and its derivatives (541, 577, and 635 nm), beef aging (541, 577, 635, 756, and 980 nm), protein (910 nm), fat (928 nm), and water (739, 756, and 988 nm) were identified.",
keywords = "Decision tree, Fisher's linear discriminant model, Instrument grading, Partial least squares analysis, Principal component analysis, Support vector machines",
author = "{Konda Naganathan}, Govindarajan and Kim Cluff and Samal, {Ashok K} and Calkins, {Chris R.} and Jones, {David D.} and Meyer, {George E.} and Jeyamkondan Subbiah",
year = "2016",
month = "1",
day = "1",
doi = "10.1016/j.jfoodeng.2015.09.001",
language = "English (US)",
volume = "169",
pages = "309--320",
journal = "Journal of Food Engineering",
issn = "0260-8774",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Three dimensional chemometric analyses of hyperspectral images for beef tenderness forecasting

AU - Konda Naganathan, Govindarajan

AU - Cluff, Kim

AU - Samal, Ashok K

AU - Calkins, Chris R.

AU - Jones, David D.

AU - Meyer, George E.

AU - Subbiah, Jeyamkondan

PY - 2016/1/1

Y1 - 2016/1/1

N2 - A prototype on-line hyperspectral imaging system (λ = 400-1000 nm) was developed and used to acquire images of exposed ribeye muscle on hanging beef carcasses (n = 274) at 2-day postmortem in a commercial beef packing plant. After image acquisition, a strip steak was cut from each carcass and vacuum packaged. After aging for 14 days, the steaks were cooked and Warner-Bratzler shear force values were collected as a measure of tenderness. Four different principal component analysis-based dimensionality reduction methods were implemented to reduce information redundancy in beef hyperspectral images. Textural features extracted from the 2-day hyperspectral images were modeled using Fisher's linear discriminant (FLD), support vector machines (SVM), and decision tree (DT) models to predict 14-day aged, cooked beef tenderness. Based on a true validation procedure using 101 samples, the FLD model yielded a tender certification accuracy of 86.7%. In addition, wavelengths corresponding to myoglobin and its derivatives (541, 577, and 635 nm), beef aging (541, 577, 635, 756, and 980 nm), protein (910 nm), fat (928 nm), and water (739, 756, and 988 nm) were identified.

AB - A prototype on-line hyperspectral imaging system (λ = 400-1000 nm) was developed and used to acquire images of exposed ribeye muscle on hanging beef carcasses (n = 274) at 2-day postmortem in a commercial beef packing plant. After image acquisition, a strip steak was cut from each carcass and vacuum packaged. After aging for 14 days, the steaks were cooked and Warner-Bratzler shear force values were collected as a measure of tenderness. Four different principal component analysis-based dimensionality reduction methods were implemented to reduce information redundancy in beef hyperspectral images. Textural features extracted from the 2-day hyperspectral images were modeled using Fisher's linear discriminant (FLD), support vector machines (SVM), and decision tree (DT) models to predict 14-day aged, cooked beef tenderness. Based on a true validation procedure using 101 samples, the FLD model yielded a tender certification accuracy of 86.7%. In addition, wavelengths corresponding to myoglobin and its derivatives (541, 577, and 635 nm), beef aging (541, 577, 635, 756, and 980 nm), protein (910 nm), fat (928 nm), and water (739, 756, and 988 nm) were identified.

KW - Decision tree

KW - Fisher's linear discriminant model

KW - Instrument grading

KW - Partial least squares analysis

KW - Principal component analysis

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=84941985405&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84941985405&partnerID=8YFLogxK

U2 - 10.1016/j.jfoodeng.2015.09.001

DO - 10.1016/j.jfoodeng.2015.09.001

M3 - Article

VL - 169

SP - 309

EP - 320

JO - Journal of Food Engineering

JF - Journal of Food Engineering

SN - 0260-8774

ER -