Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle

Livestock welfare and management could be greatly enhanced by the replacement of branding or ear tagging with less invasive visual biometric identification methods. Biometric identification of cattle from muzzle patterns has previously indicated promising results. Significant barriers exist in the t...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ali Shojaeipour, Greg Falzon, Paul Kwan, Nooshin Hadavi, Frances C. Cowley, David Paul
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	biometric identification cattle deep learning few-shot learning livestock welfare muzzle detection Agriculture S
Acceso en línea:	https://doaj.org/article/684cef23a2854930b488e49da43bfc5c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:684cef23a2854930b488e49da43bfc5c
record_format	dspace
spelling	oai:doaj.org-article:684cef23a2854930b488e49da43bfc5c2021-11-25T16:12:41ZAutomated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle10.3390/agronomy111123652073-4395https://doaj.org/article/684cef23a2854930b488e49da43bfc5c2021-11-01T00:00:00Zhttps://www.mdpi.com/2073-4395/11/11/2365https://doaj.org/toc/2073-4395Livestock welfare and management could be greatly enhanced by the replacement of branding or ear tagging with less invasive visual biometric identification methods. Biometric identification of cattle from muzzle patterns has previously indicated promising results. Significant barriers exist in the translation of these initial findings into a practical precision livestock monitoring system, which can be deployed at scale for large herds. The objective of this study was to investigate and address key limitations to the autonomous biometric identification of cattle. The contributions of this work are fourfold: (1) provision of a large publicly-available dataset of cattle face images (300 individual cattle) to facilitate further research in this field, (2) development of a two-stage YOLOv3-ResNet50 algorithm that first detects and extracts the cattle muzzle region in images and then applies deep transfer learning for biometric identification, (3) evaluation of model performance across a range of cattle breeds, and (4) utilizing few-shot learning (five images per individual) to greatly reduce both the data collection requirements and duration of model training. Results indicated excellent model performance. Muzzle detection accuracy was 99.13% (1024 × 1024 image resolution) and biometric identification achieved 99.11% testing accuracy. Overall, the two-stage YOLOv3-ResNet50 algorithm proposed has substantial potential to form the foundation of a highly accurate automated cattle biometric identification system, which is applicable in livestock farming systems. The obtained results indicate that utilizing livestock biometric monitoring in an advanced manner for resource management at multiple scales of production is possible for future agriculture decision support systems, including providing useful information to forecast acceptable stocking rates of pastures.Ali ShojaeipourGreg FalzonPaul KwanNooshin HadaviFrances C. CowleyDavid PaulMDPI AGarticlebiometric identificationcattledeep learningfew-shot learninglivestock welfaremuzzle detectionAgricultureSENAgronomy, Vol 11, Iss 2365, p 2365 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	biometric identification cattle deep learning few-shot learning livestock welfare muzzle detection Agriculture S
spellingShingle	biometric identification cattle deep learning few-shot learning livestock welfare muzzle detection Agriculture S Ali Shojaeipour Greg Falzon Paul Kwan Nooshin Hadavi Frances C. Cowley David Paul Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
description	Livestock welfare and management could be greatly enhanced by the replacement of branding or ear tagging with less invasive visual biometric identification methods. Biometric identification of cattle from muzzle patterns has previously indicated promising results. Significant barriers exist in the translation of these initial findings into a practical precision livestock monitoring system, which can be deployed at scale for large herds. The objective of this study was to investigate and address key limitations to the autonomous biometric identification of cattle. The contributions of this work are fourfold: (1) provision of a large publicly-available dataset of cattle face images (300 individual cattle) to facilitate further research in this field, (2) development of a two-stage YOLOv3-ResNet50 algorithm that first detects and extracts the cattle muzzle region in images and then applies deep transfer learning for biometric identification, (3) evaluation of model performance across a range of cattle breeds, and (4) utilizing few-shot learning (five images per individual) to greatly reduce both the data collection requirements and duration of model training. Results indicated excellent model performance. Muzzle detection accuracy was 99.13% (1024 × 1024 image resolution) and biometric identification achieved 99.11% testing accuracy. Overall, the two-stage YOLOv3-ResNet50 algorithm proposed has substantial potential to form the foundation of a highly accurate automated cattle biometric identification system, which is applicable in livestock farming systems. The obtained results indicate that utilizing livestock biometric monitoring in an advanced manner for resource management at multiple scales of production is possible for future agriculture decision support systems, including providing useful information to forecast acceptable stocking rates of pastures.
format	article
author	Ali Shojaeipour Greg Falzon Paul Kwan Nooshin Hadavi Frances C. Cowley David Paul
author_facet	Ali Shojaeipour Greg Falzon Paul Kwan Nooshin Hadavi Frances C. Cowley David Paul
author_sort	Ali Shojaeipour
title	Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
title_short	Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
title_full	Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
title_fullStr	Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
title_full_unstemmed	Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle
title_sort	automated muzzle detection and biometric identification via few-shot deep transfer learning of mixed breed cattle
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/684cef23a2854930b488e49da43bfc5c
work_keys_str_mv	AT alishojaeipour automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle AT gregfalzon automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle AT paulkwan automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle AT nooshinhadavi automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle AT francesccowley automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle AT davidpaul automatedmuzzledetectionandbiometricidentificationviafewshotdeeptransferlearningofmixedbreedcattle
_version_	1718413260633407488

Automated Muzzle Detection and Biometric Identification via Few-Shot Deep Transfer Learning of Mixed Breed Cattle

Ejemplares similares