Implementación de un sistema de reconocimiento de colores RGB con un sistema neuronal artificial análogo

Abstract

This research internship presents the simulation and implementation of an analog artificial neural network designed for RGB color pattern classification. The initiative arises from the need to explore alternatives to digital processing, taking advantage of the lower energy consumption and real-time operation provided by analog systems. The project implemented three analog neurons, each responsible for classifying one of the RGB colors. These neurons were developed using operational amplifiers and LDR sensors with optical filters, which captured light reflections and converted them into voltage signals used as inputs for the neural network.