Development of an Intelligent System for the Detection and Classification of Polymetallic Stones by Size in Crushing Processes

León León R.A.; Cáceres Cedrón J.D.; Castro C.E. · Memorias de la Conferencia Iberoamericana de Complejidad, Informatica y Cibernetica, CICIC (2025) · DOI: 10.54808/cicic2025.01.350

Resumen

The abstract should summarize the content of the paper. Try to This project aims to develop an intelligent system based on artificial intelligence for the detection and classification of polymetallic stones by size in crushing processes, thus optimizing operational efficiency. The methodology included the meticulous collection and labeling of a dataset of 4,154 images of polymetallic stones, captured under real conditions of lighting and angle. Kaggle was used for data processing and cleaning, and the YOLOv8 model was selected for real-time detection due to its accuracy and speed. During training in Google Colab, hyperparameters were adjusted and constant evaluations were conducted to maximize accuracy. The results show that the model achieved an accuracy rate of 84% in the classification of the "stone" category and reached nearly 100% accuracy at high confidence levels, indicating solid performance in detecting polymetallic stones. Copyright 2025. © por el International Institute of Informatics and Systemics.

Instituciones: Universidad Privada del Norte