How are the empty and occupied nodes specified?

[NoufFarhoud99]

Hello,

I am trying to create an algorithm that creates an occupancy map made of three types of nodes: Empty, Occupied and a third option that basically requires more information to be classified into empty and occupied.
I would like to understand how the empty and occupied nodes are specified, therefore, i did the following:
I am currently using the following to create the occupancy map:
#construct octree
tree = octomap.OcTree(0.01) #specifies the resolution or voxel size of the octree.
tree.insertPointCloud(points, origin=np.array([0, 0, 0], dtype=float))
occupied, empty = tree.extractPointCloud()

I have cheked the Extract pointcloud(): https://github.com/wkentaro/octomap-python/blob/feed59d2070a46b3208c486e125d1c48d5e8f89b/octomap/octomap.pyx#L491 which led me to isNodeOccupied():

octomap/octomap/include/octomap/AbstractOccupancyOcTree.h

Line 114 in 8178b4f

inline bool isNodeOccupied(const OcTreeNode* occupancyNode) const{

. Now i understand that getlogodds returns a value (

octomap/octomap/include/octomap/OcTreeNode.h

Line 68 in 8178b4f

inline float getLogOdds() const{ return value; }

) and compares it to probability threshold. I found how the threshold is calculated however, i am unable to figure out how the value returned is calculated.
I believe the value is found using compute update as well as update node (

octomap/octomap/include/octomap/OccupancyOcTreeBase.hxx

Line 93 in 8178b4f

computeUpdate(scan, sensor_origin, free_cells, occupied_cells, maxrange);

).

I have reached a dead end and am in need of assistance.
Can someone help me figure out how value is computed?

[ahornung]

Best have a look at the paper linked from https://octomap.github.io/ - it should explain everything behind it.