Understanding deskewing of lidar pointclouds - ashBabu/Utilities GitHub Wiki

I am taking the example of a Livox Mid360 lidar. Its frame rate is 10 Hz. This means that every 1/10 = 0.1 s = 100 ms, it is rotating one full circle or 360 degrees. When it completes one full rotation, the points that are reflected back are captured by the receiver and packed into a array. During this short time interval of 100 ms if the robot carrying the lidar moves, then the points wont be captured from the position where they were emitted causing skewing, meaning straight walls appear curved

As per livox Mid360 spec, there are 200,000 points emitted per second. This means in a 0.1 s (100 ms), it emits 20,000 points. So when you run the command ros2 topic echo --once /lidar/points, you can see height: 1 and width: 19968. This means that out of 20,000 points emitted, 19968 were reflected back. So this 20,000 is the theoretical limit of the max points in the width field.