[ros-users] Simplest way to segment spatially separated objects in point clouds?

[Björn Giesler]

Hi all,

I'd like to use my robot's Kinect to find objects (obstacles / humans) in 3D. The way the sensor is integrated, it cannot see the ground plane, so segmentation should be very easy because no objects are connected by any structure. What I'm looking for is the simplest way to segment the point cloud into non-connected sub clouds or convex hulls thereof. I've found the border extraction tutorial, which kind of seems to do what I want... Is that it or is there an even simpler way?

Thanks a lot & regards,
Björn

Von meinem iPad gesendet