Any elegant way of concatenating Pointcloud from multiple solid state Lidars ?
I am attempting to combine data from 4 lidars (Front , back , left and right )
Tried a simple concatenation of the lidar data , taking into consideration the extrinsics of the lidars .
But the resulting lidar pointcloud is not having data similar to a traditional mechanical lidar , we see the scan lines overlapping .
This is causing issue with odometry https://github.com/cuge1995/SC-FLOAM , I am testing with FLOAM , and as FLOAM primarily works by extracting the Edge points and Planar surfaces , this kind of concatenated pointcloud is not doing too well .
Its eventually resulting in a crooked odometry which is not useful .
Kindly shed some light on an alternative approach .
add a comment
