How example "Perform SLAM Using 3D lidar point clouds"collects 3D lidar data?ud

2 views (last 30 days)
When I was learning this routine, I didn't know how to collect the 3D Lidar data. Is it a 1*240 cell array formed by scanning 240 instances and then extracting their location attributes?

Accepted Answer

Ryan Salvo
Ryan Salvo on 16 Feb 2022
Hi Wenjun,
For the Perform SLAM Using 3-D Lidar Point Clouds example, the lidar data stored in the pClouds MAT-file was collected on a Clearpath Husky robot moving around a parking garage. Each cell in the 1-by-240 cell array corresponds to a different timestep and each cell contains an n-by-3 matrix, where n is the number 3-D points in the captured lidar data, and the columns represent xyz-coordinates associated with each captured point.
Thanks,
Ryan
  2 Comments
Wenjun Li
Wenjun Li on 17 Feb 2022
Hi Ryan Salvo,
Thank you for your patience!But I still have a question, does this robot scan 240 times, does each scan take a different time?I‘m looking forward to your reply,your answer will help me next step.Thank you so much!
Wenjun
Ryan Salvo
Ryan Salvo on 17 Feb 2022
Hi Wenjun,
In the example, the lidar takes 240 successive scans, however, there is a varying number of points detected in each scan due to the number of obstacles in the immediate environment. That is why the scan for each timestep is logged in a different cell of the cell array. The scans take the same amount of time, however, successive scans are only accepted when the robot has moved a certain distance, specfied by the distanceMovedThreshold parameter.
Thanks,
Ryan

Sign in to comment.

More Answers (1)

Wenjun Li
Wenjun Li on 18 Feb 2022
Hi Ryan,
Thank you for your patience!I still have a question,I can't understand why we need 240 successive scans.And I want to ask, can I record the .bag package multiple times to achieve the effect of the routine?We don't have a Husky robot in our lab yet, so I don't understand how the 240 successive scans were recorded.I'm looking forward to your reply.Please forgive my repeated interruptions.
Thank you,
Wenjun
  2 Comments
Ryan Salvo
Ryan Salvo on 18 Feb 2022
Hi Wenjun,
You do not need 240 successive scans, that is just the number of scans in the MAT-file provided in the example.
The example assumes that the provided lidar data is from a robot that is moving through a static space, so if you can re-create that when recording the BAG-file, then you should be able to use that data with the example.
Thanks,
Ryan
Wenjun Li
Wenjun Li on 18 Feb 2022
Hi Ryan,
I tried loading the recorded .bag package into the example today, and it works! But the created occupancy map is not as dense as the example, so I plan to improve it tomorrow for the next path planning work. Thank you very much for your patience in answering my questions. Your answers have taught me a lot. I take the liberty to ask, if I have any questions in the future, can I ask you again?
Thank you,
Wenjun

Sign in to comment.

Categories

Find more on Labeling, Segmentation, and Detection in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!