Software de monitoreo y control de flujo de personas en tiempo real con IA y visión computacional.

Abstract

This document presents the development structure of a "real-time monitoring and control system for people flow using artificial intelligence and computer vision." Designed for a furniture company, this system enhances the efficiency, safety, and optimization of commercial spaces, surpassing manual counting methods. The project employs the FURPS+ model to thoroughly evaluate the requirements (Araujo Sandoval, 2020; Yungan Gualli et al., 2019) and follows the Unified Process for iterative development (Sumano, 2012; Hernández González, 2017). Technologies such as Python, OpenCV, PyTorch, and Laravel are essential for its implementation (Itseez, 2018; Paszke et al., 2019; Otwell, 2020). The project results show a minimal difference compared to manual counting, with an error margin between 0.97% and 22.39%. It supports each detection with images and segments by age and gender, providing valuable data for strategic decision-making.