Classification of Starfruit Ripeness using Neural Network Technique

Starfruit is one of tropical food has been exported by Malaysia country to Europe, Middle East and Canada. The demand for this fruit is high. The exported fruits have been graded by FAMA. The graded system for starfruit ripeness is manually done by a human. The human cannot cope with high demand graded the ripeness of starfruit. This paper proposed a classification of starfruit ripeness system using artificial neural network. The methodology of image processing has successfully demonstrated. The segmentation technique using a Euclidean distance metric has been demonstrated. The classification using a sigmoid activation function in ANN improved the recognition system. The classification system has an accuracy of 97.33%. The system can recognize the unripe, ripe and overripe of starfruit.