SISTEM KLASIFIKASI SAMPAH BERBASIS CONVOLUTIONAL NEURAL NETWORK
Daftar Isi:
- Unmanned Surface Vehicle (USV) becomes a new solution in overcoming waste pollution in rivers. Use of USV replaces the garbage management system that is still done manually. Therefore the USV should be able to work as a garbage collector tool in the water by autonomous. To be able to detect the presence of garbage in the river, USV needs a garbage classification system capable of distinguishing garbage with plants. Many classification systems object to difficulty in identifying two objects that have a form that has been deformed and has the same color so that the approach is done using Deep Learning The implementation of this study used a Convutional Neural Network-based Garbage classification system (CNN). CNN-based Deep Learning is able to identify multiple objects at the same time in realtime and few of research availabe is CNN-based garbage classification systems specifically for river garbage. This research makes and tests CNN-based garbage classification system in identifying garbage in realtime in rivers. The type of CNN used in this study is You Only Look Once (YOLO). After its application, YOLO is able to classify garbage (Milo box) and water hyacinth plants with a percentage success of 65%. Keywords: USV, CNN, Waste Classification System, YOLO.