Key points are not available for this paper at this time.
This paper proposes a road detection approach based solely on dense 3D-LIDAR data. The approach is built up of four stages: (1) 3D-LIDAR points are projected to a 2D reference plane; then, (2) dense height maps are computed using an upsampling method; (3) applying a sliding-window technique in the upsampled maps, probability distributions of neighbouring regions are compared according to a similarity measure; finally, (4) morphological operations are used to enhance performance against disturbances. Our detection approach does not depend on road marks, thus it is suitable for applications on rural areas and inner-city with unmarked roads. Experiments have been carried out in a wide variety of scenarios using the recent KITTI-ROAD benchmark, obtaining promising results when compared to other state-of-art approaches.
Fernandes et al. (Wed,) studied this question.