Vario-scale data: developments in the server side

1
Vario-scale data developments in the server
side
Technical University of Delft GIS-technology
section
MobiMap Consortium meeting
2
Contents

1. Background tGAP structure
2. Results from first implementation
3. Changes and additions

3
tGAP face tree
2n-1
n
4
Constructing tGAP face tree
Step 0
Importance(u) Area(u)Class-Weight(u)
5
Constructing tGAP face tree
Step 1
Comp(u,v) Length(Bnd(u,v)) Class-Similarity(u,
v)
6
Constructing tGAP face tree
Step 1
7
Constructing tGAP face tree
Step 1
8
Constructing tGAP face tree
Step 2
9
Constructing tGAP face tree
Step 2
10
Constructing tGAP face tree
Step 2
11
Constructing tGAP face tree
Step 3
12
Constructing tGAP face tree
Step 3
13
Constructing tGAP face tree
Step 3
14
Constructing tGAP face tree
Step 4
15
Constructing tGAP face tree
Step 4
16
Constructing tGAP face tree
Step 4
17
Constructing tGAP face tree
Step 5
18
Constructing tGAP face tree
Step 5
19
Constructing tGAP face tree
Step 5
0.9
10
0.6
9
0.4
8
0.6
7
0.3
1
2
3
4
5
6
0.3
0.2
0.4
0.35
0.35
0.2
20
tGAP face tree
21
Using tGAP face tree
Importance 0.38
22
tGAP edge forest
Step 1
23
tGAP edge forest
Step 2
d
24
tGAP edge forest
Step 3
25
tGAP edge forest
Step 4
26
tGAP edge forest
Step 5
27
tGAP edge forest
28
BLG edge trees
29
Joined BLG trees
err_ijdist(point(ij), line(b_i,e_j))
max(err_i, err_j) 0.50.91.4
30
Implementation of tGAPcombination of tables
structures
GAP face tree allow face selection
GAP edge forest allow line selection
BLG tree allow line simplification
3D R-tree allow fast selection
31
UML diagram for tGAP tables
32
Source topology and tGAP tables
33
tGAP storage requirements

• Several test datasets (small/medium/large)
  cadastral and topographic data (11.000-110.000)

34
tGAP storage improvements

• tgap_edge explodes 17 times more than base
  edges, many versions of same edge (at different
  impl levels). However only few attributes change
  left, right, imp ? All versions of edge in same
  recordvarrays for variable attributes
• ? Decrease number of rows or columns at the
  expense of more calculations

35
tGAP storage improvements
edge table
face table
36
tGAP improvements

• Selection merging of areas based only on area
• Importance calculated from iteration number
• Use class weight and class similarity functions
• Calculate importance from weights of classes

37
Updating tGAP

• Local update, control the propagation effect

• Delete old edges
• Delete their BLG (joined)
• Create new from splitting
• Add the new edge
• Create BLG trees for all new

38
Updating tGAP

• Local update, control the propagation effect

12
13