SISTEMA PROTOTIPO PARA CLASIFICACIÓN DE Eugenia stipitata POR GRADO DE MADUREZ MEDIANTE REDES NEURONALES ARTIFICIALES

This paper presents the implementation of a prototype system for the classification of Arazá (Eugenia stipitata
Mc Vaugh) by Degree of maturity by using artificial neural networks. The degree of maturity is an important
physical index seen in the classification process of Arazá. This physical factor can be measured in two ways. One
is given by the producer experience when classifying the fruit and the other of technical type, using a measuring
device called a colorimeter in a laboratory. The proposed prototype system consists of an imaging system and a
graphical user interface called SISCA. The neural network is a multilayer network with Backpropagation
learning rule and classifies six states of ripeness of Arazá in RGB color space. The system was tested with a
significant population of fruit. The classification of fruit maturity and the response time of SICAwere evaluated
and the results were compared with other existing methods. System reliability was established at 90.32% of
matches and qualifying time of fruit fell by 90%, which increases the number of fruits classified per unit of time.