Pengembangan Sistem Pendeteksian dan Penghitungan Manusia pada Video Secara Real Time Menggunakan Metode You Only Look Once (Yolo)

Abstract

This study aims to develop a system to detect and calculate the number of people in a video real time. Research and development for the system is carried out using the YOLO V3 method whic is used because it can quickly detect object. In the detection and counting people in the research also make notifications that are sent in the form of messages that will be sent to the user via the Telegram app. The notification is made with the intention to informing the user if there are detected people entering or leaving the line that created in real time. The researchers used 100 human images used as a data set. The results of the training conducted demonstrated that he was able to discover humans in the video. The software development used in this research is a progressive method. The end result of this study is the application of the graphical user interface to detect and calculate the number of people entering and leaving the meter line that only has 12,6% of avg loss and can give early warning notice to the user from telegram message.