Object Tracking Using Lidar Data Filtered by Minimized Kalman Filter on Turtlebot3 Mobile Robot

The development of autonomous vehicles requires high accuracy and precision in sensor data for effective interaction with the environment and execution of functions. Processing this data with efficient algorithms positively influences vehicle decision-making. In this study, the TurtleBot3 platform, an ideal simulation model for autonomous vehicles, is used to detect and track nearby objects in the sub-system Robotic Operating System (ROS) Noetic environment. The lidar sensor data from this platform is refined using interpolation and a minimized Kalman filter to remove noise and irregularities. This approach provides clearer and more reliable measurement data, resulting in more stable and fine-tuned responses in the vehicle's motion planning. Compared to the general Kalman filter theory, this method offers faster implementation without relying on the exact error tolerance of the sensor to provide acceptable results.