Diseño e implementación de un sistema de Visión Artificial para determinar condiciones de fatiga en una persona mediante el índice Perclos, utilizando Open CV y una tarjeta Raspberry Pi 3

Abstract

Vision is in many respects the most complex sense of human physiology, overcome in complexity only by organs such as the brain. The Artificial Vision has developed a series of tools, to emulate the functioning of the view, solutions and practical application alternatives. One of these options has been implemented to determine fatigue or tiredness in people, for vehicle assistance systems, supervision modules in nuclear power plants, etc. In this work, the design of a low cost system is developed using a Raspberry PI 3 card, where an algorithm designed in Open CV is housed. This system determines the fatigue using the PERCLOS index. The system can be used indistinctly, both at night with light in the near infrared spectrum and also in the day without additional light. The operation of the system was verified in two light moving vehicles and in a concierge cabin both day and night.