94
Marian de Vries and Peter van Oosterom
example ‘g,’ ‘i,’ and ‘j,’ are merged, this is done in two steps: first ‘i’ and ‘j’ are
merged (see Fig. 5.9), then edge ‘g’ is merged (see Fig. 5.10).
When two BLG-trees are joined a new top-level tolerance value has to be assigned. This can be done using a ‘worst-case’ estimation or by computing the
Fig. 5.9. First step in the merging of edges ‘g,’ ‘i,’ and ‘j’: the BLG-trees of ‘i’ and ‘j’ are
joined (worst-case estimation for new top-level tolerance value is used – ‘1.4’)
Fig. 5.10. Second step in the merging of edges ‘g,’ ‘i,’ and ‘j’: the BLG-trees of ‘ij’ and ‘g’
are joined (note that for BLG-tree ‘ij’ the worst-case estimation for the tolerance value ‘1.4’
was used and therefore the tolerance band does not touch the polyline in the middle as might
be expected)
Marian de Vries and Peter van Oosterom
example ‘g,’ ‘i,’ and ‘j,’ are merged, this is done in two steps: first ‘i’ and ‘j’ are
merged (see Fig. 5.9), then edge ‘g’ is merged (see Fig. 5.10).
When two BLG-trees are joined a new top-level tolerance value has to be assigned. This can be done using a ‘worst-case’ estimation or by computing the
Fig. 5.9. First step in the merging of edges ‘g,’ ‘i,’ and ‘j’: the BLG-trees of ‘i’ and ‘j’ are
joined (worst-case estimation for new top-level tolerance value is used – ‘1.4’)
Fig. 5.10. Second step in the merging of edges ‘g,’ ‘i,’ and ‘j’: the BLG-trees of ‘ij’ and ‘g’
are joined (note that for BLG-tree ‘ij’ the worst-case estimation for the tolerance value ‘1.4’
was used and therefore the tolerance band does not touch the polyline in the middle as might
be expected)