Point cloud from multiple sensors can not match


Hi there,

I am using five structure core sensors in my system, and I am using the ROS driver in SDK 0.7.2. When I tried to calibrate the transformations of them, it seems like I can never get the point clouds from different sensors matched (e.g. if the ground is aligned, the same table has different height in different point clouds). Is that caused by bad intrinsic matrices? If so, how should I fix this?


The problem can be many. Usually inbuilt intrinsics are good enough. You cameras must be looking from different viewpoints maybe with some height difference in between each other. Can you explain your setup more? Maybe post a pic?