Using 3D imaging and machine learning to predict liveweight and carcass characteristics of live finishing beef cattle

G A Miller*, JJ Hyslop, David Barclay, Andrew Edwards, William Thompson, Carol-Anne Duthie

*Corresponding author for this work

Research output: Contribution to journalArticle

231 Downloads (Pure)

Abstract

Selection of finishing beef cattle for slaughter and evaluation of performance is currently achieved through visual assessment and/or by weighing through a crush. ThusConsequently, large numbers of cattle are not meeting target specification at the abattoir. Video imaging analysis (VIA) is increasingly used in abattoirs to grade carcasses with high accuracy. There is potential for three-dimensional (3D) imaging to be used on farm to predict carcass characteristics of live animals and to optimise slaughter selections. The objectives of this study were to predict liveweight (LW) and carcass characteristics of live animals using 3D imaging technology and machine learning algorithms (artificial neural networks). Three dimensional images and LW’s were passively collected from finishing steer and heifer beef cattle of a variety of breeds and sexes pre slaughter (either on farm or after entry to the abattoir lairage) using an automated camera system. Sixty potential predictor variables were automatically extracted from the live animal 3D images using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios and were used to develop predictive models for liveweight and carcass characteristics. Cold carcass weights (CCW) for each animal were provided by the abattoir. Saleable meat yield (SMY) and EUROP fat and conformation grades were also determined for each individual by VIA of half of the carcass. From the 3D images, 60 potential predictor variables were automatically extracted using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios. Performance of prediction models was assessed using R2 and RMSE parameters following regression of predicted and actual variables for LW (R2 = 0.7, RMSE = 42), CCW (R2 = 0.88, RMSE = 14) and SMY (R2 = 0.72, RMSE = 14). The models predicted EUROP fat and conformation grades with 54% and 55% accuracy (R2), respectively. This study demonstrated that 3D imaging coupled with machine learning analytics can be used to predict LW, SMY and traditional carcass characteristics of live animals. This system presents an opportunity to reduce a considerable inefficiency in beef production enterprises through autonomous monitoring of finishing cattle on the farm and marketing of animals at the optimal time.
Original languageEnglish
Article number30
JournalFrontiers in Sustainable Food Systems
Volume3
Early online date1 May 2019
DOIs
Publication statusFirst published - 1 May 2019

Fingerprint

artificial intelligence
carcass characteristics
beef cattle
finishing
image analysis
slaughterhouses
body weight
animals
slaughter
meat
carcass weight
farms
lairage
cattle
lipids
cameras
neural networks
marketing
heifers
beef

Keywords

  • 3D imaging
  • carcass characteristics
  • finishing beef cattle
  • machine learning
  • precision livestock farming

Cite this

@article{63bda737e28547939df039fac195ff13,
title = "Using 3D imaging and machine learning to predict liveweight and carcass characteristics of live finishing beef cattle",
abstract = "Selection of finishing beef cattle for slaughter and evaluation of performance is currently achieved through visual assessment and/or by weighing through a crush. ThusConsequently, large numbers of cattle are not meeting target specification at the abattoir. Video imaging analysis (VIA) is increasingly used in abattoirs to grade carcasses with high accuracy. There is potential for three-dimensional (3D) imaging to be used on farm to predict carcass characteristics of live animals and to optimise slaughter selections. The objectives of this study were to predict liveweight (LW) and carcass characteristics of live animals using 3D imaging technology and machine learning algorithms (artificial neural networks). Three dimensional images and LW’s were passively collected from finishing steer and heifer beef cattle of a variety of breeds and sexes pre slaughter (either on farm or after entry to the abattoir lairage) using an automated camera system. Sixty potential predictor variables were automatically extracted from the live animal 3D images using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios and were used to develop predictive models for liveweight and carcass characteristics. Cold carcass weights (CCW) for each animal were provided by the abattoir. Saleable meat yield (SMY) and EUROP fat and conformation grades were also determined for each individual by VIA of half of the carcass. From the 3D images, 60 potential predictor variables were automatically extracted using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios. Performance of prediction models was assessed using R2 and RMSE parameters following regression of predicted and actual variables for LW (R2 = 0.7, RMSE = 42), CCW (R2 = 0.88, RMSE = 14) and SMY (R2 = 0.72, RMSE = 14). The models predicted EUROP fat and conformation grades with 54{\%} and 55{\%} accuracy (R2), respectively. This study demonstrated that 3D imaging coupled with machine learning analytics can be used to predict LW, SMY and traditional carcass characteristics of live animals. This system presents an opportunity to reduce a considerable inefficiency in beef production enterprises through autonomous monitoring of finishing cattle on the farm and marketing of animals at the optimal time.",
keywords = "3D imaging, carcass characteristics, finishing beef cattle, machine learning, precision livestock farming",
author = "Miller, {G A} and JJ Hyslop and David Barclay and Andrew Edwards and William Thompson and Carol-Anne Duthie",
year = "2019",
month = "5",
day = "1",
doi = "10.3389/fsufs.2019.00030",
language = "English",
volume = "3",
journal = "Frontiers in Sustainable Food Systems",
issn = "2571-581X",

}

Using 3D imaging and machine learning to predict liveweight and carcass characteristics of live finishing beef cattle. / Miller, G A; Hyslop, JJ; Barclay, David; Edwards, Andrew; Thompson, William; Duthie, Carol-Anne.

In: Frontiers in Sustainable Food Systems, Vol. 3, 30, 01.05.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Using 3D imaging and machine learning to predict liveweight and carcass characteristics of live finishing beef cattle

AU - Miller, G A

AU - Hyslop, JJ

AU - Barclay, David

AU - Edwards, Andrew

AU - Thompson, William

AU - Duthie, Carol-Anne

PY - 2019/5/1

Y1 - 2019/5/1

N2 - Selection of finishing beef cattle for slaughter and evaluation of performance is currently achieved through visual assessment and/or by weighing through a crush. ThusConsequently, large numbers of cattle are not meeting target specification at the abattoir. Video imaging analysis (VIA) is increasingly used in abattoirs to grade carcasses with high accuracy. There is potential for three-dimensional (3D) imaging to be used on farm to predict carcass characteristics of live animals and to optimise slaughter selections. The objectives of this study were to predict liveweight (LW) and carcass characteristics of live animals using 3D imaging technology and machine learning algorithms (artificial neural networks). Three dimensional images and LW’s were passively collected from finishing steer and heifer beef cattle of a variety of breeds and sexes pre slaughter (either on farm or after entry to the abattoir lairage) using an automated camera system. Sixty potential predictor variables were automatically extracted from the live animal 3D images using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios and were used to develop predictive models for liveweight and carcass characteristics. Cold carcass weights (CCW) for each animal were provided by the abattoir. Saleable meat yield (SMY) and EUROP fat and conformation grades were also determined for each individual by VIA of half of the carcass. From the 3D images, 60 potential predictor variables were automatically extracted using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios. Performance of prediction models was assessed using R2 and RMSE parameters following regression of predicted and actual variables for LW (R2 = 0.7, RMSE = 42), CCW (R2 = 0.88, RMSE = 14) and SMY (R2 = 0.72, RMSE = 14). The models predicted EUROP fat and conformation grades with 54% and 55% accuracy (R2), respectively. This study demonstrated that 3D imaging coupled with machine learning analytics can be used to predict LW, SMY and traditional carcass characteristics of live animals. This system presents an opportunity to reduce a considerable inefficiency in beef production enterprises through autonomous monitoring of finishing cattle on the farm and marketing of animals at the optimal time.

AB - Selection of finishing beef cattle for slaughter and evaluation of performance is currently achieved through visual assessment and/or by weighing through a crush. ThusConsequently, large numbers of cattle are not meeting target specification at the abattoir. Video imaging analysis (VIA) is increasingly used in abattoirs to grade carcasses with high accuracy. There is potential for three-dimensional (3D) imaging to be used on farm to predict carcass characteristics of live animals and to optimise slaughter selections. The objectives of this study were to predict liveweight (LW) and carcass characteristics of live animals using 3D imaging technology and machine learning algorithms (artificial neural networks). Three dimensional images and LW’s were passively collected from finishing steer and heifer beef cattle of a variety of breeds and sexes pre slaughter (either on farm or after entry to the abattoir lairage) using an automated camera system. Sixty potential predictor variables were automatically extracted from the live animal 3D images using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios and were used to develop predictive models for liveweight and carcass characteristics. Cold carcass weights (CCW) for each animal were provided by the abattoir. Saleable meat yield (SMY) and EUROP fat and conformation grades were also determined for each individual by VIA of half of the carcass. From the 3D images, 60 potential predictor variables were automatically extracted using bespoke algorithms; these variables included lengths, heights, widths, areas, volumes and ratios. Performance of prediction models was assessed using R2 and RMSE parameters following regression of predicted and actual variables for LW (R2 = 0.7, RMSE = 42), CCW (R2 = 0.88, RMSE = 14) and SMY (R2 = 0.72, RMSE = 14). The models predicted EUROP fat and conformation grades with 54% and 55% accuracy (R2), respectively. This study demonstrated that 3D imaging coupled with machine learning analytics can be used to predict LW, SMY and traditional carcass characteristics of live animals. This system presents an opportunity to reduce a considerable inefficiency in beef production enterprises through autonomous monitoring of finishing cattle on the farm and marketing of animals at the optimal time.

KW - 3D imaging

KW - carcass characteristics

KW - finishing beef cattle

KW - machine learning

KW - precision livestock farming

UR - http://www.scopus.com/inward/record.url?scp=85077613011&partnerID=8YFLogxK

U2 - 10.3389/fsufs.2019.00030

DO - 10.3389/fsufs.2019.00030

M3 - Article

VL - 3

JO - Frontiers in Sustainable Food Systems

JF - Frontiers in Sustainable Food Systems

SN - 2571-581X

M1 - 30

ER -