Extracting 3D Occupancy Grid

For path planning, we need an occupancy grid to represent the 3D environment the robot is navigating within. This can be extracted from our SLAM implementation ( BLAM ). This is because BLAM is built upon an efficient 3D mapping framework known as OctoMap.

Fig 1 :: BLAM ROS graph

Using the dataset collected in the green house we can conduct an initial test on extracting the 3D occupancy grid. The extracted 3D occupancy grid is depicted in Fig. 2.

Fig 2 :: Initial construction of 3D occupancy grid

Tomorrow, I will work on compressing the 3D graph into a 2D cost map through prunning for initial navigation testing.