Klasifikasi Buah Jeruk secara Realtime Menggunakan Algoritma You Only Look Once V5 Berbasis Android

Abstract

The utilization of computer vision technology applied in Android smartphones can assist humans in many ways. This technology enables computer to see and analyze object around its surroundings using image processing approaches. The You Only Look Once (YOLO) algorithm version 5 based on Android platform has been implemented in this research, which able to classify the different types of citrus fruit (makrut lime, kaffir lime, and lime) based on image processing of the citrus fruit's skin texture. The image used in this study consist of testing and training data in .jpg and .jpeg formats, captured using a 12MP resolution smartphone camera. The designed system follows the stages of image acquisition, image pre-processing, and image classification, organized in the form of code in the YOLO algorithm version 5. The accuracy rate of the system's real-time testing on an Android smartphone device reached 97.9%.