YOLOv8 para Controle de Produção Pós-colheita e Beneficiamento de Frutos

Abstract

In this research, the performance of the YOLOv8 network in identifying fruit containers through images and videos was evaluated. The results showed that the network can achieve high accuracy in both images and videos, even under adverse conditions. Operational research (OR) plays a fundamental role in this work, as it is used to model the business problem identified as the lack of real-time control over the quantity of fruits being processed. Monitoring containers with harvested fruits can provide metrics for essential harvest control estimates in fruit processing.