Teknologi Artificial Intelligence (AI) Vision Swift dalam Sistem Pemantauan Latihan Bulu Tangkis dengan Algoritma Optical Flow

Abstract

Badminton is one of the most popular sports in Indonesia. In fact, Indonesia often wins various badminton competitions at the international level. Many people enjoy playing badminton, but many of them do not know whether their shots are good or whether they understand the basic techniques of badminton correctly. Additionally, many of them want to improve their skills but do not have enough time to train with a professional coach. This research aims to develop a badminton training monitoring system based on AI Vision technology using the Swift programming language. This system is expected to help badminton players evaluate and improve the quality of their shots independently. The main focus of this research is to optimize computational accuracy by using the Optical Flow algorithm to track the movement of the shuttlecock during training. In developing this system, the Optical Flow algorithm is used to analyze the shuttlecock's trajectory and its drop points. The results of this research show that testing 20 shots using shuttlecock trajectory can be accurately detected by the system with an accuracy of 97.22%. Meanwhile, the system's accuracy in tracking the placement of the shuttlecock in the opponent's service area is 94.50%.