Development of an artificial vision algorithm for the classification of papaya in its three states of maturity

León León R.A.; Atalay Serrano J.P.; Yengle Horna M.; Rojas Valverde A.; Romero Luis M.; Meregildo Aliaga B. · Memorias de la Conferencia Iberoamericana de Complejidad, Informatica y Cibernetica, CICIC (2025) · DOI: 10.54808/cicic2025.01.364

Resumen

This study presents the development of an automated computer vision system to classify papayas into three maturity stages: immature, mature, and overripe, using the YOLOv8 model. The methodology involved collecting and labeling 3,172 images of papayas in various maturity stages and capture conditions, utilizing the OpenImages platform for efficient data labeling. The YOLOv8 model was developed in the PyCharm development environment and trained on Google Colab with a Tesla T4 GPU, taking advantage of PyTorch compatibility, which is essential for the processing and training of convolutional neural networks. The results showed an average accuracy of 98% in classifying the different maturity stages of the papaya. The use of YOLOv8 for papaya classification by maturity is an innovative solution that addresses the need for automation in the agro-food industry. Copyright 2025. © por el International Institute of Informatics and Systemics.

Instituciones: Universidad Privada del Norte