Prediction
Use a visibility tree and building database for fast prediction.
Figure 1 illustrates the
visibility relations computed in the preprocessing in the shape of a visibility
tree
. Only the relations in the first layer of the tree must be computed in the
prediction which can be done very fast - all other visibility relations are determined
during the preprocessing stage and can be read from file. The stored visibility relations
(except the first layer) can be used for all predictions with the same database.
For the prediction shown in Figure 2, a resolution of 1 m for the discretization of the building database was selected. However, it is possible to increase the resolution to some extent without reducing accuracy.