Accurate lane detection using commercial GNSS devices

Commercial GNSS devices are very common nowadays. The rather accurate position and velocity estimation made navigation very efficient. Lane detection, however, requires a level of accuracy which is hardly reached by such devices. In this work, we present a novel framework for lane detection using GNSS devices. We suggest a framework for analyzing rapid changes in the satellite´s signal-strength, caused by a momentary blockage due to utility and light poles. A method for detecting such momentary changes between LOS/NLOS is presented, followed by a geometric algorithm that improves location-accuracy of commercial GNSS devices to less than 1 meter error-range. The algorithm was verified using both simulation and field experiments.