Occupancy Grid Fusion Prototyping Using Automotive Virtual Validation Environment

This paper walks through the occupancy grid fusion algorithm prototyping process. The implementation consists of core fusion algorithms using probability derived from inverse sensor models. Preliminary results are obtained using an automotive virtual validation tool and phenomenological sensor models of radars, lidars and selected functions of vision sensors. The purpose of the developed framework is to perform a relative performance assessment between certain grid computation and fusion methods. Assessment is carried out by comparing computed results with reference data. Virtual validation is used to enable quick and cost effective reference data generation in comparison to real world testing.