Develop an Algorithm for Peach Ripeness Classification Using Artificial Vision

Minchola Quispe R.C.; Rodriguez Aguilar E.F.; León León R.A. · Memorias de la Conferencia Iberoamericana de Complejidad, Informatica y Cibernetica, CICIC (2025) · DOI: 10.54808/cicic2025.01.373

Resumen

The objective of this research was to develop a peach maturity classification algorithm based on computer vision, utilizing the YOLOv8 architecture. To achieve this, 1,114 images of peaches at various stages of maturity were collected and meticulously labeled on the Roboflow (plataforma de anotación) platform. The YOLOv8 model was trained on Google Colab (entorno de desarrollo), using a Tesla T4 GPU and configuring specific parameters such as batch size and learning rate, across 350 epochs to ensure precise convergence. The trained model achieved 99% accuracy in classifying peach maturity, demonstrating its effectiveness and adaptability across diverse conditions. This YOLOv8-based approach provides a fast and accurate solution for automated peach classification, optimizing quality control in production and contributing to the modernization of agricultural processes. Copyright 2025. © por el International Institute of Informatics and Systemics.

Instituciones: Universidad Privada del Norte