Abstract:
In this study, a simple machine vision system was developed for sorting three maturity classes
of tomatoes grown in Ethiopia. For the sorting analysis, RGB color features were extracted
from each class of tomato images. Six different color features were calculated from RGB
color space. An artificial neural network classifier with Back propagation method was tested.
The input layer consists of six color features, the hidden layer consists of 40 nodes and the
output layer consists of three nodes representing three tomato classes (green, pink and red).
The best sorting accuracies in testing data set was 76% for all the three classes (green, pink
and red) of tomato images. That means the overall sorting accuracy was 76%. Finally, based
on the obtained results, a tomato sorting machine can be designed to categorize 3 colors of
tomatoes decreasing human labor and to reducing sorting time.
