Robust environment perception based on occupancy grid maps for autonomous vehicle

For a purposes of reduction of driving workload, traffic accidents, and so on, autonomous vehicle systems, which can drive even though no human driver rides on, have been developed all over the world. Recently, it is considered that a part of such techniques play an important roll to energy saving. To navigate the autonomous vehicle in a complex environment, it is necessary to extract static objects precisely and robustly. Moreover, about dynamic objects, it is not enough to consider its existing position, and it is considered that estimating its motion and predicting its future position are strongly demanded. In this paper, for a solution of such problems, we propose an environment perception method using a LIDAR based on Occupancy Grid Maps.