my GIS paper summary

[aya awad]

One of these papers presented an approach to generate a geo-database
customized to the needs of pedestrian navigation. The main idea was to
use different existing geo-datasets like (1) topographic, (2)
cadastral and (3) indoor vector map data to develop methods for
automatically deriving an area-wide geo-dataset that serves pedestrian
needs. Then the separate data layers from the different maps were
analyzed to see which parts are relevant for pedestrians. The relevant
parts are then merged into a geometrically consistent dataset using
Conflation methods and finally, a connected graph is built to
calculate shortest-path algorithms for pedestrians using the tailored
geo-database. The application used in the paper was to help
pedestrians find the shortest route from the Hannover train station to
a shopping mall.
 Step #1: Data Sources include:
 ATKIS- Authoritative Topographic-Cartographic Information Systems
from the national mapping agencies of Germany (transportation network,
such as roads and rail tracks, waters, and vegetation)
 ALK- the real estate cadastral map data (information about real
estate boundaries, buildings, their owners and use)
 Indoor vector Map data of the main train station of Hannover
provided by the Deutsche Bahn AG.
 Step #2: Data Selection
 The 3 given datasets are explored and relevant pedestrian data are
extracted
 ATKIS à accessible areas for an average pedestrian (no wheelchair or
baby carriages)
 ALK à buildings and accessible areas in a city like streets and
plazas
 Indoor map à the public shopping area and entrances onto the train
platform are chosen. Different kinds of entrance/ exits points, such
as staircase, escalators, or lifts are also taken into consideration.
 Step #3: GIS Analysis is applied to provide the needed navigational
data structure in the form of a connected linear graph network. There
are 2 aspects to deal with:
 Firstly, the given accessible areas have to be transformed into a
linear structure for the routing application using medial axis or
topological skeleton approach
 Secondly, point-like information pieces, such as entrance/ exit
points have to be connected to the linear network graph ex: stops for
public transportation, which is linked using the shortest
perpendicular distance with the nearest edge of the network
 Step #4: Geometric Integration
 The different data parts have to be merged into a single geometric
dataset using Conflation algorithms to establish the necessary
connectivity between all information parts. This step is produced
automatically using ArcGIS from ESRI, which integrates the different
data layers into Network Dataset [1].
References
[1] Elias, Birgit, Pedestrian Navigation—creating a tailored geo-
database for routing, in Institute of Cartography and Geo-informatics,
Leibniz University of Hannover, 2007
