Treffer: Jurnal Pengembangan Aplikasi Pelacakan Objek dengan Kalman Filter untuk Objek Bergerak Cepat

This article presents the development of an object tracking application using the Kalman Filter method for fast-moving objects in video, such as balls, vehicles, or runners. The project was developed in Python using the OpenCV library, focusing on real-time tracking based on color detection and predictive estimation. The Kalman Filter is applied to predict object positions in frames where detection might be unstable or temporarily missing. The system is tested with three types of video: a thrown ping-pong ball, a moving vehicle, and a running person. The tracking performance achieved up to 95% accuracy in ideal conditions, with an average processing time of 11 ms per frame. This work can be reproduced using the provided Python code and can serve as a practical reference for those learning about computer vision, object tracking, or real-time video processing.