Eventos Anais de eventos
COBEM 2021
26th International Congress of Mechanical Engineering
Machine vision algorithm for continuous counting of poultry in slaughterhouses
Submission Author:
Guilherme de Santana Weizenmann , SC
Co-Authors:
GLEISON Stopassola, Guilherme de Santana Weizenmann, Marcos Virgilio da Costa
Presenter: Guilherme de Santana Weizenmann
doi://10.26678/ABCM.COBEM2021.COB2021-1707
Abstract
With the high growth of the refrigeration industries and the need to improve their processes, the use of computer vision techniques combined with the application of image processing has increased. The of these two techniques is able to realize significant production improvements. The correct implementation can result in an improvement in the quality of products, an increase in production, or even provide the identification of losses. At poultry slaughterhouses, in the initial stages of slaughter, the animals are counted, seeking to identify losses occurred at each stage of the process. As this counting must be as correct as possible, and the current solutions do not present a high accuracy, the possibility of implementing a system using computer vision and image processing was proposed. This article consists on the development of an algorithm that is capable of counting the chickens in a continuous flow, aiming the implementation in the slaughterhouses in the future. For the development of this work, we sought to obtain the best possible performance, reaching as a final result 99.94% accuracy.
Keywords
Machine vision, POULTRY SLAUGHTERHOUSE, Industrial Application
