Automatic Typographic Maps
-
Upload
niklas-elmqvist -
Category
Technology
-
view
998 -
download
1
description
Transcript of Automatic Typographic Maps
![Page 1: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/1.jpg)
Shehzad Afzal*, Ross Maciejewski†, Yun Jang‡, Niklas Elmqvist*, David S. Ebert* Purdue University*, Arizona State University†, Sejong University in Seoul ‡
Spatial Text Visualization Using Automatic Typographic Maps
![Page 2: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/2.jpg)
2
Motivation
AxisMaps.com
![Page 3: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/3.jpg)
3
Motivation
• Typographic Map: Map made entirely of the geographical labels (i.e., “Type”)
• Aesthetically pleasing
• Exists only for handful of cities
• Printed map sold from $30-$150’
AxisMaps.comSan Francisco
![Page 4: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/4.jpg)
4
MotivationHow are these maps designed?• Manual text placement using Adobe Illustrator over an
‘OpenStreetMap’ image• Text options are modified based on underlying spatial
features• Process takes several weeks to complete one map
• Focus of their current approach is ‘purely aesthetic’
• Our work “Automates the Typographic Map Generation Process”
• Potential of visualizing data using spatialized text
Image Courtesy: AxisMaps.com
![Page 5: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/5.jpg)
5
Related Work• Maharik et al. (2011) introduced ‘calligrams’
(digital micrography images)
• ‘TagMaps’ by Yahoo: Word clouds on top of graphical features
• ‘Wordle’ by Viegas et al (2009), ‘ManiWordle’ by Koh et al (2010) & ‘SparkClouds’ by Lee et al (2010)
![Page 6: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/6.jpg)
6
System Overview
OSM Parser
Data Cleaning/Filtering
Build Graphical Objects & Layers
Optimizations
Region Generation
Path/Road Generation
Open Street Map(OSM) File
Open Street Map(OSM) File
Visual Properties/Style Sheet
SVG FileSVG File
SVG Renderer
Typographic Map
SVG Code Generation
![Page 7: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/7.jpg)
7
System Overview
OSM Parser
Data Cleaning/Filtering
Build Graphical Objects & Layers
Optimizations
Region Generation
Path/Road Generation
Open Street Map(OSM) File
Open Street Map(OSM) File
Visual Properties/Style Sheet
SVG FileSVG File
SVG Renderer
Thema-Typographic
MapSpatial Statistical Dataset
SVG Code Generation
![Page 8: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/8.jpg)
8
Data Model• Layers: Particular class of geographical objects
• e.g., Highways, Primary roads, park etc.• Visual Attributes such as font size, color, weight etc.
• Graphical Objects belong to exactly one layer • 1D paths(roads) or 2D paths(polygons)
• Ordering Layers: • Layers are drawn in ascending order of priority
![Page 9: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/9.jpg)
9
Data Model• Ordering Layers:
• 1D Paths have higher priority than 2D Paths• Ordering Graphical Objects within Layers:
• Horizontal Paths have higher priority than vertical paths
![Page 10: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/10.jpg)
10
Optimizations• Divided lanes having the same label are merged together to
form a single lane
• Font size for polygonal areas adjusted according to the area
![Page 11: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/11.jpg)
11
Limitations• Definitions of polygonal areas are not always consistent in
OSM. e.g., Rivers & lakes boundaries
• Segments of same road have different names or category and they often overlap
• OSM data is not completely defined for some geographic regions
![Page 12: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/12.jpg)
12
Paths as Text• Rendering Path using Text:
• Fit Text to path and repeat it for the duration of path’s length• Rotate characters to align with path normal• Path thickness is controlled by font size
![Page 13: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/13.jpg)
13
Paths as Text• Visual Clutter - Path Overlap:
Clutter from Label Overlap Character Mask Character Halo
![Page 14: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/14.jpg)
14
Regions as Text
Bounding Box filled with Oriented Text Clipped Text using Region Path
![Page 15: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/15.jpg)
15
Regions as Text
Adjacent Regions: Vary the orientation of the lines in adjacent regions resulting in visual continuity between regions
![Page 16: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/16.jpg)
16
Comparison with AxisMap Output
AxisMaps Map of San Francisco, CA Automatic Typographic Map of San Francisco, CA
![Page 17: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/17.jpg)
17
Comparison with AxisMap Output
AxisMaps Map of San Francisco, CA Automatic Typographic Map of San Francisco, CA
![Page 18: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/18.jpg)
18
Comparison with AxisMap Output
AxisMaps Map of San Francisco, CA Automatic Typographic Map of San Francisco, CA
![Page 19: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/19.jpg)
19
Comparison with AxisMap Output
AxisMaps Map of San Francisco, CA Automatic Typographic Map of San Francisco, CA
![Page 20: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/20.jpg)
20
Typographic Map - Seattle
Automatic Typographic Map – Seattle near VisWeek 2012 Venue
Visweek Venue
![Page 21: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/21.jpg)
21
Typographic Map - Chicago
Automatic Typographic Map – Chicago, IL
![Page 22: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/22.jpg)
22
Thema-Typographic Maps
Thematic Maps:
Geographic Maps where Geospatial variable is visually encoded on the map
Thema-Typographic Maps: Modify Font attributes on per character level to convey the value of a statistical variable at each character’s spatial location
Font Attributes: Typically Size, but color, intensity etc.
![Page 23: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/23.jpg)
23
Thema-Typographic Maps Showing Crime Rate
KDE Map for West Lafayette, INshowing Crime Activity
Thema-Typographic Map Statistical variable visualized is
Crime Rate
![Page 24: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/24.jpg)
24
Thema-Typographic Maps in SVG
• Scaling individual characters: Internally we need to calculate where characters end on a path in 2D Space. This helps in following ways:
• Correct Lookup of Mapping Variable in Spatial Dataset• # of characters required to fill the path can be calculated• Stroke width of the background mask is now defined as an average of
the minimum and maximum font size
Other Applications: Traffic Intensity, Demographics, political data can be overlaid on a typographic map
![Page 25: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/25.jpg)
25
Conclusions and Future Work
• Automatic Generation of Typographic Maps for any Geographic Region within seconds
• ‘Thema-Typographic Maps’: Combines Typographic Maps technique with spatial datasets
Future Work:• Spatial data features as a means of visualizing data• Support Navigation, drilling down and changing map layout
![Page 26: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/26.jpg)
26
Acknowledgements• AxisMaps for their helpful feedback/discussions and permission
to use their Typographic Map image in our paper
• This work was supported in part by the U.S. Department of Homeland Security’s VACCINE Center under Award no.
103659 / 2009-ST-061-CI0001 and the Defense Threat Reduction Agency under Award no. HDTRA 1-10-1-0083
![Page 27: Automatic Typographic Maps](https://reader030.fdocuments.net/reader030/viewer/2022020206/546cc380af7959ec228b4b30/html5/thumbnails/27.jpg)
Thank you
[Web Service Client]
http://web.ics.purdue.edu/~safzal/typomaps.html
Shehzad [email protected]
Spatial Text Visualization Using
Automatic Typographic Maps