211 The 3D map for obstacles recognition by guide dog robot

The purpose of this study is making 3D maps for obstacles recognition of guide dog robots. It is necessary that the robots recognize surrounding obstacles for working on the real environment. In order to recognize surrounding obstacles, the robots should have 3D maps of concerning the anticipated route where the robots will pass. We used a laser range sensor (LSR) for making 3D maps. The LSR is made rotated by a servo motor, so 3D data is measured by this rotating LSR. It was identified whether it is space that can be passed with the remodeled LSR for each individual space, and kept the aggregation as 3D grid data. By this way, we could make a 3D grid map of real environment such as the garden in university. And we have confirmed to be able to recognize the height information, for example the height of a stone wall and branches over the head.