search

hku-mars / Point-LIO

Other
783 stars 131 forks source link

Questions about the parameter point_filter_num and space_down_sample #46

Open wlxing1901 opened 1 year ago

wlxing1901 commented 1 year ago

The work on Point-LIO is truly impressive, elevating lidar SLAM to an entirely new level. I have some questions about this work and hope the authors can address them.

  1. I've noticed that the default value for the space_down_sample parameter is true, which means the point cloud in the map generated when the Lidar is static will concentrate in the central grid area. However, when the space_down_sample parameter is set to false, the speed of mapping becomes very slow, making real-time operation impossible. I'd like to know which specific part of the algorithm is causing this slowdown? Is it possible to only use the downsampled points for calculating the Lidar position, but still add all points to the map?

  2. Is the point_filter_num parameter the one that controls the number of points included in the calculation? When point_filter_num is set to 3, only 33.3% of the points are involved in the computation. But I've observed that when space_down_sample is set to false, any value of point_filter_num other than 0 makes mapping not working, with no point cloud or odometry output.

Thank you for your time and assistance.

Joanna-HE commented 1 year ago

Thanks for your attention on our work!

  1. When the space_down_sample parameter is true, the current LiDAR frame would be downsampled to a smaller size based on the space distribution of the current LiDAR points in the body frame. On the other hand, when the space_down_sample parameter is false, the current LIDAR frame would not be downsampled based on their space distribution. And the processing time in Point-LIO for each LiDAR point of such an increased number of points would increase without doubts. It is not possible to use downsampled LiDAR points for calculating but still add all points to map, due to the principle of the Point-LIO which calculates the position of LiDAR at each sampled point, resulting in that for those points that are not used to calculate the LiDAR position, the position of them in the global map is also unknown.
  2. First of all, when space_down_sample is set to false, no matter which value the point_filter_num is, mapping will still work but with longed processing time. Maybe there are some problems of your observations. Then the point_filter_num is a parameter to downsample the current LiDAR frame by sequence of sampling time. When the point_filter_num is set to n, only the LiDAR points that are numbered in sampling sequence as integral multiples of n will be kept for processing in Point-LIO.

I hope that your questions are settled well by the above content.

Best regards,

Dongjiao He

寄件者: Forrest @.> 寄件日期: 2023年8月26日 下午 05:26 收件者: hku-mars/Point-LIO @.> 副本: Subscribed @.***> 主旨: [hku-mars/Point-LIO] Questions about the parameter point_filter_num and space_down_sample (Issue #46)

The work on Point-LIO is truly impressive, elevating lidar SLAM to an entirely new level. I have some questions about this work and hope the authors can address them.

  1. I've noticed that the default value for the space_down_sample parameter is true, which means the point cloud in the map generated when the Lidar is static will concentrate in the central grid area. However, when the space_down_sample parameter is set to false, the speed of mapping becomes very slow, making real-time operation impossible. I'd like to know which specific part of the algorithm is causing this slowdown? Is it possible to only use the downsampled points for calculating the Lidar position, but still add all points to the map?

  2. Is the point_filter_num parameter the one that controls the number of points included in the calculation? When point_filter_num is set to 3, only 33.3% of the points are involved in the computation. But I've observed that when space_down_sample is set to false, any value of point_filter_num other than 0 makes mapping not working, with no point cloud or odometry output.

Thank you for your time and assistance.

— Reply to this email directly, view it on GitHubhttps://github.com/hku-mars/Point-LIO/issues/46, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ANJUIXQJCHIXBKW4JRBMHC3XXG6MFANCNFSM6AAAAAA37NSP54. You are receiving this because you are subscribed to this thread.Message ID: @.***>