TOPONYMS 26th International Cartographic Conference August 25-30, - - PowerPoint PPT Presentation

TOPONYMS PLACEMENT ON WEB MAPS

University of Zagreb Faculty of Geodesy Croatia

Robert Župan and Stanislav Frangeš rzupan@geof.hr sfranges@geof.hr

26th International Cartographic Conference August 25-30, 2013 | Dresden Germany

analysis with proposed solutions

Presentation content

2



Introduction



Common errors of label placement



Assumption and rules



Research process



Research findings



Conclusion and additional research

Introduction



Geographic names are the heritage of material and spiritual culture of a people, and the set of all toponyms makes heritage of all mankind



Map labelling – attaching text (toponyms) to geographic features



Features must be labelled to convey information.

3

Problem Definition

4

The Map Label Placement Problem



Placing map feature labels



Legibly



Without overlap (features / other labels)



Maintaining visual association between features and their labels

Map without toponyms



Maps without toponyms or their labels give little information



The same holds for charts and graphs

5

Common errors of label placement for some most popular and global web map servers

Google Earth and common errors of label placement.(left fig. – shows overlaps and the right fig. shows duplicate toponym)

6

Left – Google Earth with many overlaps with all layers turned on. Right – Google Maps with different label generalization degree between Slovenia and Croatia.

Common errors of label placement for some most popular and global web map servers

7

Ask Maps web map service with map from Nokia and Microsoft and their misrepresentation and toponym visualization.

Common errors of label placement for some most popular and global web map servers

8

Large scale map in Openstreetmap and some toponyms are shown on the map but some are not visible until zooming in although there is enough space?

Common errors of label placement for some most popular and global web map servers

9

Two possibilities



Label Size Maximization

Maximize the size of the placed labels The labeling has to be complete



Label Number Maximization

Maximize the number of placed labels Labels have fixed size



Most literature focuses on Label Number Maximization, so will this presentation

10

Three kinds of labels

Town 1 Town 2 Town 3 Town 4 Straight, horizontal name Curved name Curved name with spacing between characters

11

Assumptions



toponyms related item or coordinate (oikonyms, oronyms) and



toponyms following the line (oronyms, regionyms, territoryonyms, hydronyms)



toponyms inside the closed polygon or an area that provide information about the that area (regionyms and territoryonyms or the names of national parks, military training grounds, etc.)



there is a hierarchy showing the various types of toponyms according to the criterion such as the type of toponyms, the importance of objects for which provide attribute information, categorization of subspecies of toponyms, color, font size of toponym, etc. Hierarchy is needed due to the significance of one kind of toponym over the other and automatic decision on the choice

• f toponym placement, as one of the processes of map

generalization.

12

Elastic labels

According to first assumption, solution is:

13

Proposal for geometric solution

How to represent a rectangle?

X1 <X2 , Y1 <Y2 , and so forth.

14

Horizontal collision detection

4 cases for collision detection

15

Vertical collision detection

4 cases for collision detection

16

Collision detection

No edges overlap, but still a collision for map label of

• toponyms. This case

happens whenever X1′ < X1 and X2′ > X1.

17

Collision detection

Except in that special case, a collision only exists if there is both a vertical and horizontal collision.

18

Possible rectangle movements?

Sliding label presenting possible positions for toponym, software choice for priority placement, and priorities for label placement according to different authors.

19

Detection of the intersection of lines and rectangle

Test of collision for moving rectangles (left) and case with no collision in final position although there is intersection.

20

Detection of the intersection of lines and rectangle

Case where paths do not intersect

21

Detection of the intersection of lines and rectangle

Paths of rectangle r1 and r2 intersects without collision at the final point.

22

Detection of the predicted trajectory

r2 is below r1 when the left edge

• f r1 is collinear

with the right edge of r2 so r1 and r2 do not collide

23

Detection of the predicted trajectory

r2 is initially to the left of r1 and finishes below and to the right of A, colliding with r1 along the way.

24

Using rectangles and rhombs depending on the type of letters

25

Collision prediction along curve or trajectory approximating letters with rectangulars or rhombs.

Using rectangles and rhombs for

• verlap detection

Conclusions



The overlap of toponym with other cartographic

• bjects and map content in this case is been ignored,

though it also must be taken into account



New algorithmic framework for fast and consistent labelling



To be tested how the proposed model behaves in the real computing server environment.

27

TOPONYMS PLACEMENT ON WEB MAPS

University of Zagreb Faculty of Geodesy Croatia

Robert Župan and Stanislav Frangeš rzupan@geof.hr sfranges@geof.hr

26th International Cartographic Conference August 25-30, 2013 | Dresden Germany

analysis with proposed solutions

THANK YOU FOR YOUR ATTENTION!