Enhancing Indoor Position Estimation Accuracy: Integration of Accelerometer, Raw Distance Data, and Extended Kalman Filter in Comparison to Vicon Motion Capture Data

Abstract

Indoor positioning systems are a significant area of research and development, helping people navigate within buildings where GPS signals are unavailable. These systems have diverse applications, including aiding navigation in places like shopping malls, airports, and hospitals and improving emergency evacuation processes. The purpose of this study is to evaluate various technologies and algorithms used in indoor positioning. This study focuses on using raw distance data and Kalman filters to enhance indoor position accuracy. It employs a trilateration algorithm based on Recursive Least Squares (RLS) for initial position estimation and combines the results with accelerometer data. The designed algorithm using real sensor data collected in an ROS(Robot Operating System) environment was tested, and the results obtained were compared with data obtained from the Vicon Indoor Positioning System. In this comparison, the Root Mean Square Error metric was used. As a result of the comparison, it was observed that the error obtained from the designed algorithm is less than that of the Vicon system.