Conclusions

We have presented a next best view planner for gaging and digitalization of 3D indoor environments without any intervention. The next best view planner computes next nominal poses under the conditions: Maximizing the information gain, reducing the overall robot path length, minimizing the rotation angles, and avoiding obstacles, including tricky ones like jutting out edges. Furthermore we presented a collision free trajectory planner with a closed loop stable motor controller.

3D laser range finders on mobile robots enable the automatic acquisition, surveying and mapping of entire interiors. Future work will concentrate on building semantic maps from the digitalized environment, i.e., a high level description of the scanned scene.