FP-SarahVanWart-RazvanCarbunescu

From CS294-10 Visualization Fa08

Jump to: navigation, search

Contents

Group Members

  • Razvan Carbunescu
  • Sarah Van Wart

Problem and Motivation

Currently there are not easy ways to view how Shapefiles -- geospatial vector data files for geographic information systems software -- change over time. Since a Shapefile is simply a snapshot of how a set of geographic features looks at a particular point in time, many Shapefiles must be generated to represent changes over time. Yet, few applications exist to allow users to visualize and explore temporal changes in geography. Our idea is to create a spatial-temporal mapping application that will attempt to unify Shapefile snapshots by:

Solving a Visualization Problem

By creating an easy way to visualize geometric changes across Shapefile snapshots through tweening, geographic changes over time can be better visualized.

  • Potential applications include:
  • Visualizing how glacial structures recede.
  • Visualizing how political boundaries change.
  • Visualizing the environmental impact or cumulative effects of development.

Solving an Exploration Problem

By creating an easy way to query geometric changes across Shapefile snapshots, temporal patterns in spatial data can be more easily exposed. Potential questions that could be answered include:

  • What areas have suffered the biggest lost or gain over time?
  • How has the increase in road construction affected a particular wetland or habitat area?

Background

Work has been done in 1) creating non-proprietary shapefile readers that work with Flash, 2) tweening simple shapes and developing algorithms to tween complex shapes, and 3) visualizing spatio-temporal data.

Shapefile Readers

Tweening

Spatio-Temporal Applications

  • Time-Integrative Geographic Information Systems : Management and Analysis of Spatio-Temporal Data
    • Ott, Thomas, and Frank Swiaczny.

New Applications Are Still Needed

Although work has been done within this spatial-temporal application domain, there is space for improvement:

  • There is still room to develop an easy, generic, and non-proprietary way to visualize how Shapefiles change over time over the web.
  • There is still room to develop tweening and animations to better conceptualize Shapefile snapshots.

Our Spatial-Temporal Map Viewer will attempt to address this problem.

Key Technical Challenges and Approach

We plan to approach this problem from two angles:

  1. Designing a user interface and configuration scheme that allows users to analyze multiple Shapefile snapshots for a given geographic entity.
  2. Design a set of shape tweening implementations that allows a user to animate Shapefile snapshots for a selected date range.

User Interface Challenges

  • Develop a way to map individual shapes in a Shapefile snapshot to corresponding shapes in a different snapshot.
  • Creating an obvious user interface that allows user to:
    • explore shape information across Shapefile snapshots.
    • Specify data search parameters easily
  • Provide basic map tools, such as pan and zoom, to facilitate relevant queries.

Tweening Challenges

  • Polygon/polylines point matching will be addressed starting from the nonlinear assignment problem for an initial matching between polygons/polylines.
  • For matching nonexistent points from either initial or final polygon the algorithm will calculate the distance to the nearest 2 points and create a ‘new’ overlapping point to tween or be tweened to either at the closest point or at the middle of the segment between the 2 closest points
  • For matching nonexistent points of polylines the algorithm will find the polylines which the new line originates and create the line from there or in case there are no intersections start building the line from the middle point of the segment
  • Basic viewing modifications (pan, zoom, rotate, stretch) will be implemented using affine transformations on the shape points structure.
  • Depending on the speed with which the polygon drawing functions perform an algorithm to split the polygon into simple triangle/square shapes might be used for performance

Technologies to be Used

The final product will be a web-based interactive map. Technologies to be used:�

  • Flash
  • Flex
  • Flare
  • Vanrijkom’s Shapefile Libraries (for Flash)
  • ERSI Shapefile Technical Description Whitepaper
  • Modest Maps

Milestones and Division of Work

The programming, algorithmic, and data acquisition tasks were divided as follows:

Razvan

  • Polygon tweening algorithms and implementation.
  • Polygon pair point matching (mapping a source polygon onto a destination polygon).
  • Optimization.
  • Generating slides and writing the report.

Sarah

  • Data acquisition and transformation (finding sample Shapefile data for proof of concept).
  • Integrating Flex / Flare polygons with Modest Maps framework.
  • Implementing the polygon matching algorithm and creating the data structures that map polygon and vertex adjustments across shapefiles.
  • Interacting With Shapefiles (integrating tweening functions into application).
  • Generating slides and writing the report.

Initial Presentation Slides

Slides

Midterm Presentation Slides (Status Update)

Slides

Final Paper and Application

Final Projects Page

Back To Final Project Page


[add comment]
Retrieved from "http://vis.berkeley.edu/courses/cs294-10-fa08/wiki/index.php/FP-SarahVanWart-RazvanCarbunescu"
Personal tools