Muladi Muladi, Dyah Lestari, Dedy Tri Prasetyo, Aji Prasetya Wibawa, Triyanna Widiyaningtiyas, Utomo Pujianto
Fruit is in the highest ranking together with protein as dietary food that containing high energy but low sugar. The risk of some chronic disease may be reduced by consuming fruit. One of the fruits chosen as a diet is apple because it has high phytochemical and anti-oxidant content. As one of the fruits most sought after by the public to maintain diet and competition with imported fruit, handling post-harvest apples is very important. Color is one of the indicators considered when consumers choose fruit. The classification method in sorting fruit automatically is based largely on the color of the fruit because in this method there is no physical contact with the fruit which can cause fruit damage. In this research, the classification system is developed for the Malang's local grown apple Manalagi based on the extraction of the average RGB color feature using the Backpropagation Artificial Neural Network algorithm. The purpose of the developed system is to distinguish the Manalagi, the famous Malang grown apple of worth consumption and not worth the consumption. The fruit classification is based on the average RGB color composition of the fruit skin. The developed system can classify Manalagi in worth consumption and not worth the consumption with an accuracy rate of 90%. © 2019 IEEE.
Universitas Negeri Malang, Department of Electrical Engineering, Malang, Indonesia
