GPS-Based Location Estimation System in a NLOS Environment Using a Screen Coordinate System

Abstract

GPS-based location estimation systems are the optimum way to track where an object is in an outdoor environment. However, location errors occur when signals transmitted by GPS satellites are interfered with by buildings or other obstacles, since the location tracking system uses these signals. To modify the positional errors incurred by GPS signal interference, this study proposes a measure to minimize the errors of GPS-based location estimation systems in NLOS environments. In this system, location errors are reduced by organizing the screen coordinates with collected image data, establishing an environment favorable for the application of visual information to the screen coordinate system. This data is then applied using perspective projection formulas, substituting the actual GPS data to the divided screen. When tracking a person located at the base of a building, the error range was reduced by around5m when using the proposed system as compared to GPS signals alone.