Development of an Artificial Vision System for Supervising the Use of Safety Harness in Work at Heights

León León R.A.; Alvarez Ocas Y.E.; Avila Torres E.G.; Florian Camacho D.D.; Gonzales Gonzales G.R.; Contreras Campos A.A. · Memorias de la Conferencia Iberoamericana de Complejidad, Informatica y Cibernetica, CICIC (2025) · DOI: 10.54808/cicic2025.01.315

Resumen

This study developed a convolutional neural network model designed for image classification, specifically applied to object recognition in industrial contexts. The MobileNetV2 architecture was utilised for feature extraction, while dense layers and a Softmax activation function were integrated to enhance classification performance. Experimental results demonstrated 100% accuracy across all evaluation metrics, highlighting the model’s effectiveness in terms of precision, recall, and F1-Score. This project not only evidenced the technical feasibility of deep learning models in targeted applications but also laid the groundwork for their deployment in real-world industrial environments. Copyright 2025. © por el International Institute of Informatics and Systemics.

Instituciones: Universidad Privada del Norte