jkk-research / urban_road_filter

Real-time LIDAR-based Urban Road and Sidewalk detection for Autonomous Vehicles 🚗
BSD 3-Clause "New" or "Revised" License
289 stars 78 forks source link

I would like to know if this algorithm can still work for single frame point clouds synthesized by three mechanical radars #20

Closed FengYuQ closed 1 year ago

FengYuQ commented 1 year ago

I now want to perform ground removal on the synthesized point cloud, and I am not sure if this algorithm can work in the synthesized point cloud. Mechanical radars are all mounted on cars, with one radar overhead and two radars on the side

horverno commented 1 year ago

I don't think it will work on radars. Even on solid-state LIDAR there are limitations. You can read about the limitations in Conclusions and Future Work section of the paper.

FengYuQ commented 1 year ago

I don't think it will work on radars. Even on solid-state LIDAR there are limitations. You can read about the limitations in Conclusions and Future Work section of the paper.

Sorry, this is my pen mistake. What I mean is a point cloud composed of three rotating mechanical Lidar

horverno commented 1 year ago

That could work theoretically. Try only enabling the star_shaped_method. x_zero_method and x zero method use channel information of a single LIDAR, which is not relevant if the pointcloud is merged from 3 LIDARS.

FengYuQ commented 1 year ago

So I still have a question is that if I use the point cloud of merged and what the value of channel should I have set?

horverno commented 1 year ago

As mentioned star_shaped_method does not rely on channel info, so it doesn't matter. The only problem with this approach is that it uses one method instead of three, so it won't be as accurate.