| |
|
|
| Registros recuperados : 361 | |
| Registros recuperados : 361 | |
|
|
|
Registro Completo
|
Biblioteca(s): |
Embrapa Agrossilvipastoril. |
|
Data corrente: |
10/02/2024 |
|
Data da última atualização: |
10/02/2024 |
|
Tipo da produção científica: |
Resumo em Anais de Congresso |
|
Autoria: |
BENICIO, L. M.; XAVIER, D. B.; LIMA, I. B. G. de; PEREIRA, D. H.; CONDOTTA, I. C. F. da S.; LOPES, L. B. |
|
Afiliação: |
LUANA MARIA BENICIO, UNIVERSIDADE DE ILLINOIS; DIEGO BATISTA XAVIER, CPAMT; ITALO BRAZ GONÇALVES DE LIMA, UNIVERSIDADE DE ILLINOIS; DALTON HENRIQUE PEREIRA, UNIVERSIDADE FEDERAL DE MATO GROSSO; ISABELLA CARDOSO FERREIRA DA SILVA CONDOTTA, UNIVERSIDADE DE ILLINOIS; LUCIANO BASTOS LOPES, CPAMT. |
|
Título: |
Individual cattle face recognition through computer vision. |
|
Ano de publicação: |
2023 |
|
Fonte/Imprenta: |
In: JORNADA CIENTÍFICA DA EMBRAPA AGROSSILVIPASTORIL, 12., 2023. Sinop. Resumos... Brasília, DF: Embrapa, 2023. p. 24. |
|
Série: |
(Embrapa Agrossilvipastoril. Eventos Técnicos & Científicos, 1) |
|
Idioma: |
Inglês |
|
Conteúdo: |
Abstract: Identifying a single animal in a herd allows for individual management and association, and traceability over time of each animal. Over the years, methods have been adopted to identify animals individually, such as using ear tags, visible ear tags, tattoos, and radio frequency identification devices (RFID). However, a disadvantage of this system is that the tool requires much work to configure (identification and tracking system), and the cost to install and replace the RFID tags must be considered. Systems based on image processing can be considered alternatives for individual identification of farm animals, as it is a non-invasive tool, is easy to apply, and can be adopted in real-time. Furthermore, it is possible to monitor more than one animal simultaneously individually. Several authors have used images to identify animals by face images, using the individual pattern as an ID. This work aims to analyze to develop an automatic individual identification of Nellore cattle through facial recognition using computer vision techniques. Facial images were collected from 30 Nellore males during weighing using the Intel RealSense camera. Then, only the face region of the animals was selected automatically, and an image classification model, using Yolo architecture version 8, was developed. Preliminary results showed an accuracy of 95% when correlating the animal's face with the identification number without the need to use identification ear tags. Thus, the proposed model presents potential for field applicability in individual animal recognition. The following steps of this research will include a more significant number of animals in the model and field testing to validate the methodology. MenosAbstract: Identifying a single animal in a herd allows for individual management and association, and traceability over time of each animal. Over the years, methods have been adopted to identify animals individually, such as using ear tags, visible ear tags, tattoos, and radio frequency identification devices (RFID). However, a disadvantage of this system is that the tool requires much work to configure (identification and tracking system), and the cost to install and replace the RFID tags must be considered. Systems based on image processing can be considered alternatives for individual identification of farm animals, as it is a non-invasive tool, is easy to apply, and can be adopted in real-time. Furthermore, it is possible to monitor more than one animal simultaneously individually. Several authors have used images to identify animals by face images, using the individual pattern as an ID. This work aims to analyze to develop an automatic individual identification of Nellore cattle through facial recognition using computer vision techniques. Facial images were collected from 30 Nellore males during weighing using the Intel RealSense camera. Then, only the face region of the animals was selected automatically, and an image classification model, using Yolo architecture version 8, was developed. Preliminary results showed an accuracy of 95% when correlating the animal's face with the identification number without the need to use identification ear tags. Thus, the proposed mod... Mostrar Tudo |
|
Thesaurus NAL: |
Animal identification; Computer vision; Face. |
|
Categoria do assunto: |
X Pesquisa, Tecnologia e Engenharia |
|
URL: |
https://ainfo.cnptia.embrapa.br/digital/bitstream/doc/1161943/1/2023-cpamt-lbl-individual-cattle-face-recognition-through-computer-vision-p24.pdf
|
|
Marc: |
LEADER 02475nam a2200217 a 4500
001 2161943
005 2024-02-10
008 2023 bl uuuu u00u1 u #d
100 1 $aBENICIO, L. M.
245 $aIndividual cattle face recognition through computer vision.$h[electronic resource]
260 $aIn: JORNADA CIENTÍFICA DA EMBRAPA AGROSSILVIPASTORIL, 12., 2023. Sinop. Resumos... Brasília, DF: Embrapa, 2023. p. 24.$c2023
490 $a(Embrapa Agrossilvipastoril. Eventos Técnicos & Científicos, 1)
520 $aAbstract: Identifying a single animal in a herd allows for individual management and association, and traceability over time of each animal. Over the years, methods have been adopted to identify animals individually, such as using ear tags, visible ear tags, tattoos, and radio frequency identification devices (RFID). However, a disadvantage of this system is that the tool requires much work to configure (identification and tracking system), and the cost to install and replace the RFID tags must be considered. Systems based on image processing can be considered alternatives for individual identification of farm animals, as it is a non-invasive tool, is easy to apply, and can be adopted in real-time. Furthermore, it is possible to monitor more than one animal simultaneously individually. Several authors have used images to identify animals by face images, using the individual pattern as an ID. This work aims to analyze to develop an automatic individual identification of Nellore cattle through facial recognition using computer vision techniques. Facial images were collected from 30 Nellore males during weighing using the Intel RealSense camera. Then, only the face region of the animals was selected automatically, and an image classification model, using Yolo architecture version 8, was developed. Preliminary results showed an accuracy of 95% when correlating the animal's face with the identification number without the need to use identification ear tags. Thus, the proposed model presents potential for field applicability in individual animal recognition. The following steps of this research will include a more significant number of animals in the model and field testing to validate the methodology.
650 $aAnimal identification
650 $aComputer vision
650 $aFace
700 1 $aXAVIER, D. B.
700 1 $aLIMA, I. B. G. de
700 1 $aPEREIRA, D. H.
700 1 $aCONDOTTA, I. C. F. da S.
700 1 $aLOPES, L. B.
Download
Esconder MarcMostrar Marc Completo |
|
Registro original: |
Embrapa Agrossilvipastoril (CPAMT) |
|
|
Biblioteca |
ID |
Origem |
Tipo/Formato |
Classificação |
Cutter |
Registro |
Volume |
Status |
Fechar
|
| Nenhum registro encontrado para a expressão de busca informada. |
|
|
