Nondestructive Banana Ripeness Classification using Neural Network

Banana contributes to the total fruit production in Indonesia. To ensure the quality of a banana, Indonesian government with its national standard agency (BSN) has issued the national standard of fruit. This study proposed the classification method of a banana using the neural network based on color and texture. Data was collected before model creation with numbers of banana of ripeness level 2, 3, and 4 were 30, 30, and 45 respectively. Three neural network model with three different variables, i.e. color image and texture. Testing result show a high accuracy result of 95.24%.