I want to use two cameras to track a marker on a large area. I have one Kinect and one webcam to do it. I have succeed to run two arpose node to generate two armarker tf from both cameras. Now I want to use arposeekf to get more accuracy when both cameras detect the marker together. But I don't understand how can I use this node. I tried to run as the documentation says but I'm getting some errors:

roscd ar_pose_ekf
./runekf.sh

(I have compiled arposeekf with the package bfl as dependency instead of orocos-bfl which was not working.)

ERRORS:

1.173 [ Info   ][PropertyLoader:load] Loading properties into Service 'PoseEkf' with '/media/Data/fuerte_workspace/rosstacks/camera_pose_estimation/ar_pose_ekf/cpf/ekf.cpf'.
1.176 [ ERROR  ][PropertyLoader:load] Could not compose unknown type 'ColumnVector'.
1.176 [ ERROR  ][PropertyLoader:load] Could not compose unknown type 'ColumnVector'.
1.176 [ ERROR  ][PropertyLoader:load] Could not compose unknown type 'ColumnVector'.
1.176 [ ERROR  ][PropertyLoader:load] Could not compose unknown type 'SymmetricMatrix'.
Segmentation fault (core dumped)

Does anyone know how to use this node ?