Temporal ground reversion (TGR) uses the temporary ground feature to solve the partial under-segmentation issue. To segment the ground plane accurately even when the ground is elevated with several layers, region-wise vertical plane fitting (R-VPF) is also introduced.
<h3>
What is meant by Ground segmentation?</h3>
Ground segmentation is a crucial issue in the realm of 3D perception employing 3D LiDAR sensors for several reasons, including the detection of traversable areas and the recognition of objects. Several ground segmentation techniques have been put forth in this situation. However, there are still certain restrictions present.
First off, certain approaches for ground segmentation call for adjusting parameters based on the environment, which is overly difficult and time-consuming. Furthermore, even with properly set parameters, a partial under-segmentation issue could still arise, implying ground segmentation failures in some areas.
Finally, when the ground is above another structure, such as a retaining wall, ground segmentation methods frequently fail to estimate an acceptable ground plane. We suggest Patchwork++, an expansion of Patchwork, a reliable ground segmentation technique, to solve these issues. Utilizing adaptive ground likelihood estimation (A-GLE), Patchwork++ uses prior ground segmentation results to determine the relevant parameters.
Additionally, temporal ground reversion (TGR) uses the temporary ground feature to solve the partial under-segmentation issue. To segment the ground plane accurately even when the ground is elevated with several layers, region-wise vertical plane fitting (R-VPF) is also introduced.
To learn more about 3d point cloud refer to:
brainly.com/question/28152656
#SPJ4
