Desarrollo de un Algoritmo de Visión Artificial para la Clasificación del Mango Edward por su Estado de Madurez

This study presents the development of a computer vision algorithm for the automated classification of Edward mangoes according to their ripeness stage. A total of 2,301 images were labeled using the Roboflow platform into three categories: ripe, unripe, and defective. To enhance dataset variability, preprocessing and data augmentation techniques were applied. The YOLOv11 model was trained on Google Colab using PyTorch and subsequently deployed in Visual Studio Code for real-time validation through camera-based testing. The system achieved an accuracy of 98.1%, demonstrating its potential to reduce economic losses and optimize key processes in the agro-industrial sector.