ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
1

Lidar Snowpole localization

asked 2021-03-16 03:33:26 -0500

larsb gravatar image

Hi

Im have a lidar from Sick, LMS, this has a ROS driver https://github.com/SICKAG/sick_scan , my goal is to use the 2D lidar to localise a snowpole (circle, half circle). The lidar is to be mounted on a truck, se image https://imgur.com/a/TYIM7hM Image of snowpole https://imgur.com/a/Mko6E2E , its about 20mm in diameter.

Does anybody have some idea to where i could find some packege that could help me achive my goal?

I have found: (but not sure these would work):

  1. https://github.com/praveen-palanisamy...

  2. https://github.com/tysik/obstacle_det...

  3. https://github.com/yangfuyuan/obstacl...

My main consern is the update rate and the detection of such a small object, it is no problem with the size og the object when using the built in functionality of the scanner.

Thanks

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

answered 2021-03-22 02:26:42 -0500

Loy gravatar image

20mm is really really narrow. Depending on distance to the scanner and the resolution, such a pole will result in only a few laser points, if at all. My suggestion would be to do some basic calculations first to see the number of points the lidar will 'see' a snowpole with at what range.

My bet is that that is relatively low and thus that a lidar is not really a great way to achieve this goal. If I'm wrong and you do get eg. tens of points on a snowpole, you could do a Hough transform to fit circles on those points. Or just clustering could be sufficient.

If lidar does turn out to not be useful, a stereo-camera that filters out red blobs (which should be do-able against a snowy background) could be an option worth considering.

edit flag offensive delete link more

Question Tools

2 followers

Stats

Asked: 2021-03-16 03:33:26 -0500

Seen: 186 times

Last updated: Mar 22 '21