Location-aware Mobile Devices & Landscape Reading Ruben JOYE, Joris VERBEKEN, Steven HEYDE,Harlind LIBBRECHT Accepted as fully reviewed paper in: “Buhmann/Ervin/Palmer/Tomlin/Pietsch (Eds.): Peer Reviewed Proceedings Digital Landscape Architecture 2011: Teaching & Learning with Digital Methods & Tools”, Anhalt University of Applied Sciences. Germany Abstract This paper examines how „smartphones‟ – a type of advanced and location-aware mobile phone (e.g. Apple iPhone® or phones running Google Android®) - can be useful for landscape architecture and planning, and more specifically how they can share knowledge about our surroundings using „Mobile Augmented Reality‟ (MAR). We examined how a smartphone-based information system can influence reading and understanding the landscape, and to what degree it influences landscape valuation and imaging by its users. A quantitative survey was conducted with students to determine the educational possibilities of such tool. The main part of this paper however, is more technical in nature. Using the „Layar® Reality Browser‟ as a framework for our MAR, we wanted to facilitate the management of a „Layar® 3D‟ system. We did so by building a graphical front -end to the administrative part of the system by using a combination of ESRI ArcGIS® and Microsoft Access®. This to facilitate adding, editing or removing „Points Of Interest‟ (POI), even to people with a limited technical background. 1 Introduction With smartphones (e.g. Apple iPhone ® or phones running Google Android ® ) becoming more popular every month, in the third quarter of 2010 accounting worldwide for 19,3% of overall mobile phone sales compared to 9,85% in 2009 (COZZA ET AL. 2010; GARTNER 2010), it‟s safe to conclude that location-aware mobile devices are getting more and more widespread. These types of devices - equipped with both GPS technology and mobile internet connectivity – are taking GIS information mobile, making it possible to associate digital media to a geographical location (VARNELIS & FRIEDBERG 2008). Handheld location-aware mobile devices are becoming the interface to the „geospatial Web‟, that delivers on the spot georeferenced information to its users. This allows people to be present in both the physical and networked (digital) place (ITO 2008). 2008 – the year in which for the first time in history, mobile access to the internet exceeded desktop computer-based access (ITU 2009) – turned out to be the start of an internet revolution, quickly named „the mobile web‟. An interesting application for these location-aware devices is „Mobile Augmented Reality‟ (MAR): the superpositioning of rich media elements (audio, video, images and even 3D- models) on top of a real-time view from the built in camera lens of the portable device. The technical aspect of a MAR-system has already been considered by JOYE ET AL. (2010) and
12
Embed
Location-aware Mobile Devices & Landscape Reading › landschaftsinformatik › file...Location-aware Mobile Devices & Landscape Reading 267 the opportunities for education in landscape
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Location-aware Mobile Devices & Landscape Reading
Ruben JOYE, Joris VERBEKEN, Steven HEYDE,Harlind LIBBRECHT
Accepted as fully reviewed paper in: “Buhmann/Ervin/Palmer/Tomlin/Pietsch (Eds.): Peer
Reviewed Proceedings Digital Landscape Architecture 2011: Teaching & Learning with Digital
Methods & Tools”, Anhalt University of Applied Sciences. Germany
Abstract
This paper examines how „smartphones‟ – a type of advanced and location-aware mobile
phone (e.g. Apple iPhone® or phones running Google Android®) - can be useful for
landscape architecture and planning, and more specifically how they can share knowledge
about our surroundings using „Mobile Augmented Reality‟ (MAR). We examined how a
smartphone-based information system can influence reading and understanding the
landscape, and to what degree it influences landscape valuation and imaging by its users. A
quantitative survey was conducted with students to determine the educational possibilities
of such tool. The main part of this paper however, is more technical in nature. Using the
„Layar® Reality Browser‟ as a framework for our MAR, we wanted to facilitate the
management of a „Layar® 3D‟ system. We did so by building a graphical front-end to the
administrative part of the system by using a combination of ESRI ArcGIS® and Microsoft
Access®. This to facilitate adding, editing or removing „Points Of Interest‟ (POI), even to
people with a limited technical background.
1 Introduction
With smartphones (e.g. Apple iPhone® or phones running Google Android
®) becoming
more popular every month, in the third quarter of 2010 accounting worldwide for 19,3% of
overall mobile phone sales compared to 9,85% in 2009 (COZZA ET AL. 2010; GARTNER
2010), it‟s safe to conclude that location-aware mobile devices are getting more and more
widespread. These types of devices - equipped with both GPS technology and mobile
internet connectivity – are taking GIS information mobile, making it possible to associate
digital media to a geographical location (VARNELIS & FRIEDBERG 2008). Handheld
location-aware mobile devices are becoming the interface to the „geospatial Web‟, that
delivers on the spot georeferenced information to its users. This allows people to be present
in both the physical and networked (digital) place (ITO 2008). 2008 – the year in which for
the first time in history, mobile access to the internet exceeded desktop computer-based
access (ITU 2009) – turned out to be the start of an internet revolution, quickly named „the
mobile web‟.
An interesting application for these location-aware devices is „Mobile Augmented Reality‟
(MAR): the superpositioning of rich media elements (audio, video, images and even 3D-
models) on top of a real-time view from the built in camera lens of the portable device. The
technical aspect of a MAR-system has already been considered by JOYE ET AL. (2010) and
Location-aware Mobile Devices & Landscape Reading 267
the opportunities for education in landscape architecture have been discussed by VERBEKEN
ET AL. (2010).
A prebuilt and free client-application called
„Layar® Reality Browser‟ is available for both
the Apple iPhone® and phones running Google
Android®. One of the harder parts is setting up
your own Layar® POI server that communicates
with the Layar® client on the smartphone.
Luckily there are a few different development
tools available online (CAMERON 2010), with
PorPOISe (DE SMIT 2010) – a PHP based POI-
server - being one of the most comprehensive.
Some adjustment to the programming code is
needed, in order to have it connect to the record
with the POI‟s you want to show. This can
either be an XML formatted file, or a MySQL®
relational database.
PorPOISe (version 1.0a was used) comes
prebuilt with a textual web interface called
„PorPOISe server dashboard‟ (Fig. 1) that
allows authorized persons with no programming
skills to add new „points of interest‟ (POI‟s).
However, when you‟re planning to have
numerous POI‟s, this can be cumbersome since
you need to look up (and enter) the lat- lon-
coordinates for each POI manually.
One of the main reasons for fractional use of
landscape visualization tools (in general) -
according to BISHOP & LANGE (2005) – is the
lack of user-friendliness for easy manipulation
(BISHOP & LANGE 2005). It was our intention to
meet this
Fig. 1: Web interface of the
„PorPOISe server dashboard‟
constraint by improving the usability, by linking the underlying MySQL® database with
ESRI Arcmap. A new POI can be added more easily by use of and at the same time it
becomes possible to import both the spatial as well as the attribute data of already existing
GIS-datasets (e.g. ESRI shapefiles). It appears we are not alone in seeing great value in a
tighter integration of GIS-data in the Layar® Reality Browser. A similar idea was carried out
by EMGE & PRASAD (2010). Details on their approach are missing though.
The need for a better way to manage POI‟s arose from the fact that an arrow-less GPS-
based touristic route will be created at the end of our landscape research project. In order to
evaluate our new workflow involving ESRI ArcMap, we‟ve put together a small test case.
As part of the research project – which studies change in a former World War One region -
a pilot study on the subject of landscape evaluation has just been completed in close
cooperation with historians from the „Memorial Museum Passchendaele 1917‟ (MMP) and
R. Joye, J. Verbeken, S. Heyde, and H. Libbrecht 268
the Flemish heritage institute (BOSTYN ET AL. 2010). This preparatory research work served
as a base for the landscape walk.
The second part of the paper will focus on how location based information can influence
landscape perception and -valuation. During the process of mental imaging, the observer
interprets the perceived landscape. Therefore, individual landscape valuation is not only
contextual and time bound, but also strongly personal or subjective (JACOBS 2006) since the
plethora of visual stimuli gets filtered, before final imaging takes place. Unconsciously, a
selection of stimuli is performed in terms of usefulness for the given situation, partially
based on prior knowledge (cognitive aspects) (DIJKSTRA & KLIJN 1992). For instance - in
the case of cultural heritage landscape valuation - an observer with little to no knowledge
about the landscape and cultural heritage will be able – at best - to distinguish the main
landscape structures, but will overlook more detailed information (COETERIER 1995; VAN
DEN BERG & CASIMIR 2002). As landscape experience and –valuation is much more than
just an aesthetic consideration (DIJKSTRA & KLIJN 1992), this detailed information may have
a strong influence on one‟s final assessment of the whole. By pointing the observer‟s
attention to these details (e.g. cultural remnants), as well as providing background
information (e.g. about historic events that took place), the appreciation of landscape
experience will be more balanced, more strongly founded, and surpass a purely aesthetic
appraisal.
A quantitative survey, was carried out with a group of 20 students enrolled in a one-year
advanced study in landscape development. They gave their opinion on both the ease of use
of the smartphone system, the educational possibilities, as well as their impressions of the
physical environment and how it had changed their appraisal and overall assessment.
2 Material & Methods
2.1 Technical
After installing the ODBC MySQL® Connector (version 5.1.8. was used), linking a
shapefile to a MySQL® database is easily feasible from within ArcGIS, using ESRI
ArcCatalog to setup a 'database connection' by means of an 'OLE DB Connection'. The
additional columns and corresponding values that are stored in the external MySQL®
database get added to the attribute table in ArcMap without any problems. But the drawback
is that this external database cannot be edited from within ArcMap (as this is a read-only
process), making this approach useless for our stated goal. To overcome this problem, an
intermediate step was taken that at the same time adds functionality to the workflow. A
Microsoft Access® form linked to the main POI shapefile was created, in which all the
necessary data about each of the POI's can easily be entered.
This was done by making use of the ArcScript „ArcMap Hyperlink to Filtered Microsoft
Access® Form‟ provided by CALLAHAN & CARSON (2003) on the ESRI support page. This
filters all of the entries in order to show only the corresponding data in a clear Access form
(Fig. 1). Although programmed to work in conjunction with ArcGIS 8.2 and Microsoft
Location-aware Mobile Devices & Landscape Reading 269
Access® 2000, this script works just fine using ESRI ArcGIS 9.3.1 and the newer versions
of Microsoft Access® (both 2007 and 2010 have been tested).
The use of this script requires working with a personal geodatabase, which can be created in
ESRI ArcCatalog. Once we have a personal geodatabase set up, we add a new (point)
feature class which we name „POI‟. As coordinate system „WGS 1984‟ – the international
standard for use in cartography, geodesy, and navigation – is chosen. The resulting feature
class has two mandatory fields: „OBJECTID‟ and „SHAPE‟. We add an extra field
„accessid‟ (text) which will mirror „OBJECTID‟, but is necessary to link with our Microsoft
Access® form since the ArcScript requires the data type of the field to be a text value.
A minimal MySQL® database for use with PorPOISe - that enables basic Layar
®
functionality - consists of one table named „POI‟ with the under mentioned structure. Note
that in the overview schema (Fig. 2) we added the three fields necessary for use with
ArcGIS (indicated by an asterix), but that these do not negatively influence the functioning
of either PorPOISe or Layar®.
POI id
attribution
imageURL
lat
lon
line2
line3
line4
title
type
doNotIndex
showSmallBiw
showBiwOnClick
layerID
dimension
alt
relativeAlt
SHAPE(*)
OBJECTID(*)
accesid(*)
Action id
uri
label
poiId
contentType
method
activityType
params
closeBiw
showActivity
activityMessage
autoTriggerRange
autoTriggerOnly
Object poiId
baseURL
full
reduced
icon
size
Transform angle
rel
scale
poiID
Layer layer
refreshInterval
refreshDistance
fullRefresh
showMessage
id
Fig. 2: Overview of PorPOISe database schema with the five tables making the structure
for use with Layar® (*) = added for use in conjunction with ArcGIS
If you want to make use of more advanced features in Layar® like 2D/3D objects and
actions, you need to add additional information to the POI‟s: what is the URL to the 3D-
model, should it be scaled, rotated,… Are there actions connected to the POI like a link to
an external webpage, audio, video and should these trigger automatically or not? All this
information gets stored in three tables separate from the POI table. A fourth additional table
was introduced to support the Layar® v4 API features (DE SMIT, 2010), totaling five
tables: „POI‟, „Object‟, „Transform‟, „Action‟ and „Layer‟. This last table isn‟t of much
interest to us, and can be left blank to have the Layar®-client use the default values.
R. Joye, J. Verbeken, S. Heyde, and H. Libbrecht 270
The functionality of each of these fields has already been discussed in detail by (DE SMIT,
2010; WANG, 2010; WANG, 2011) and therefore is not repeated in this paper. Albeit, the
meaning of each field was included in the final version of our Microsoft Access® form as a
tooltip displayed when the mouse hovers over a field.
Fig. 3: The Microsoft Access® form linked to the ESRI Arcmap document. This form
combines data from five different tables and constitutes the MySQL® database
which PorPOISe uses to serve to the Layar® client-application.
To get the proper database schema for all the required tables, a SQL script file 'database.sql'
comes with PorPOISe. This file was executed to our external MySQL®-server using the
MySQL® Workbench (version 5.2.31a) which easily created all the proper tables and data
types settings for the different fields. Having already installed the ODBC MySQL®
Connector, we were also able to connect to our MySQL®-server using Microsoft Access®,
and have the database schema imported to our personal geodatabase (opened as a Microsoft
Access® document). The first thing we need to do for it to work, is to add a new connection
by using the Microsoft ODBC Data Source Administrator which can be found in the
„control panel‟ under „administrative tools‟. Once this connection has been added under the
„file-DSN‟ tab, it will become available in Microsoft Access®. With our personal
geodatabase opened in Microsoft Access, we chose „external data‟ and picked „ODBC-
database‟. This allows to import the database schema, and have a functional database in
accordance with the PorPOISe structure. The process of importing a MySQL® database in
Location-aware Mobile Devices & Landscape Reading 271
Microsoft Access® is described more in depth in a „White Paper‟ from SUN
MICROSYSTEMS (2009). After defining the relations between the different tables, and
executing the query, a Microsoft Access® form was created (Fig. 3).
2.2 Content & use
To put this workflow into practice, a small quantitative survey was conducted. Not only to
test the ease of use, but at the same time to evaluate the influence of location based
information on landscape perception and valuation. The chosen area for our test case, is an
old castle park which has undergone complete destruction during the First World War. Now
the castle park serves a public function as a museum. In future developments this will be the
starting point of a complete landscape route using a smartphone application. We chose a
total of 39 historical photos and drawings (Fig. 4-7) that accurately depict the gradual decay
of the beautiful center of „Zonnebeke‟ (near Ypres) as it underwent successive