Extending occupancy grid mapping for dynamic environments

In this paper, the commonly used filtering technique occupancy grid mapping for static environments is extended for dynamic environments. The proposed method is able to estimate velocities indirectly. We apply a distribution model of the respective state variable to estimate the cell dynamics by means of prediction and update cycle, as known by standard tracking filters. Therefore, we present a straight forward derivation of the prediction and update rule. Furthermore, we validate our approach by simple one dimensional simulations, and show how it can be extended into a two dimensional world, including the resulting consequences, e.g. in terms of memory requirements.