Evaluation of different quality functions for road course estimation using imaging radar

This work presents three different quality functions for road course estimation using an imaging radar sensor. The quality functions work on gridmap data. A gridmap integrates each measurement in chronological order. Range estimation is found out to be necessary on country roads and a solution is introduced. All quality functions are evaluated using a huge set of data consisting of highways and country roads. The driven trajectory is taken as ground truth for the evaluation. The results show that on highways the quality functions perform nearly equal. However on country roads there are differences. The huge error reduction achieved by the range estimation is pointed out. At the end the quality function performing best is determined.