Object-wise comparison of LiDAR occupancy grid scan rendering methods

Computing Occupancy grids with LiDAR data, is a popular strategy for environment representation. In the last two decades, several authors have proposed different methods to render sensed information into grids, seeking obtain computational efficiency or accurate modeling. However, no comparison regarding their performance under object detection in autonomous driving applications has been found literature. As result, this work compares six representative scan rendering strategies quantitative manner. To that end, novel evaluation framework occupancy proposed. It addresses main steps of detection: segmentation and features estimation, proposing meaningful procedure, repeatable other OG approaches. The code available https://git-autopia.car.upm-csic.es/open_source/occupancy_grid_object_detection_evaluation.git.