Identification of Red Dragon Fruit Using Backpropagation Method Based on Android

Damar Prasetyo (Department of Information Systems, Faculty of Information Technology and Electrical Engineering University of Technology Yogyakarta, Indonesia)
Renggo Danu Murti Bimantaka (Department of Information Systems, Faculty of Information Technology and Electrical Engineering University of Technology Yogyakarta, Indonesia)

International Journal of Applied Business and Information Systems
Vol. 2 No. 2 (2018)

Ripeness identification of red dragon fruit using conventional methods has a lack of ripeness accuracy, due to the subjective nature of the election or lack of understanding of science in choosing a ripe red dragon fruit. This research was conducted to create a system to identify the ripeness level of red dragon fruit using artificial neural networks backpropagation method with image processing. The stages of the research are 4 steps process, namely preprocessing, training, testing carried out in Matlab and predictions made on the Android system. The data used are 30 images of red dragon fruit which have different levels of ripeness, 10 raw categories, 10 ripe categories, and 10 categories too ripe. the results of the identification of each of the 20 raw dragon fruit images, ripe, and too ripe, can recognize 100% in raw category, 100% in ripe category, and 85% in too ripe category