In the PR2 Interactive Manipulation tutorial, it has been mentioned that after taking the snapshot using the camera in PR2, we can segment the image in order to detect objects on the table. But the segment command returns a segmentation fault. What might be the probable reason?