Sistem Klasifikasi Penumpang Bus Trans Padang Berdasarkan Pakaian Menggunakan Metode Image Processing

Abstract

Public transportation is one of the things that is often used by everyone. One type of public transportation that is often used in Indonesia is the Bus Rapid Transit (BRT) or commonly known as the busway. The city of Padang, West Sumatra has a BRT to serve the community called the Trans Padang Bus. However, in this bus service there are still several shortcomings that make the bus service ineffective. The gate system which is still manual on this bus makes service slow and inefficient. By using a flap barrier gate that has an image processing system for suit passengers, the service on this bus will be more effective. Utilizing the Raspberry Pi and You Only Look Once (YOLO), this system can be made as efficient as possible for the place and use. Passengers who will enter will be immediately detected by the PIR sensor and the system will start with a camera that will take pictures of passenger clothes. The image will be processed using YOLO and the results of the process in the form of passenger categories and prices will be displayed on the LCD. The author uses YOLO and the Fast R-CNN algorithm to determine passenger categories. With this method, the test results show an average success percentage of 98.3%.

Keywords

• Machine Learning
• Flap Barrier
• Image Processing
• Classification