Faster R-CNN Implementation Method for Multi-Fruit Detection Using Tensorflow Platform

摘要

Fruit is a commodity highly potential crop in Indonesia. While harvest, fruit production is very abundant. however, the slow harvest process makes the quality decreased. Consequently, the selling price is cheap. In our research, we propose a Deep learning method using faster R-CNN to detect classification a multi-fruits. The input used mango and pitaya fruits. The dataset is a real data taken from a farmer at harvest time and then we into 2 classes, the classification are mango and pitaya for the purpose of training object detection. We used the MobileNet model on TensorFlow platform. In this study, we achieved the accuracy score of about 99%. This method is very appropriate for developed the process of sorting multi-fruits in real-time so as to maintain the quality of the fruit.