UCL CENTRE FOR ADVANCED SPATIAL ANALYSIS
Centre for Advanced Spatial Analysis University College London 1 - 19 Torrington Place Gower St London WC1E 7HB
Tel: +44 (0)20 7679 1782 casa@ucl.ac.uk www.casa.ucl.ac.uk
WORKING
PAPERS
SERIES
Digital Urban - The Visual City
ISSN 1467-1298
Paper 124 - Sept 07
Digital Urban - The Visual City
Dr Andy Hudson-Smith
Abstract
Nothing in the city is experienced by itself for a city’s perspicacity is the sum of its
surroundings. To paraphrase Lynch (1960), at every instant, there is more than we can
see and hear. This is the reality of the physical city, and thus in order to replicate the
visual experience of the city within digital space, the space itself must convey to the user
a sense of place. This is what we term the “Visual City”, a visually recognisable city built
out of the digital equivalent of bricks and mortar, polygons, textures, and most
importantly data.
Recently there has been a revolution in the production and distribution of digital
artefacts which represent the visual city. Digital city software that was once in the
domain of high powered personal computers, research labs and professional software
are now in the domain of the public-at-large through both the web and low-end home
computing. These developments have gone hand in hand with the re-emergence of
geography and geographic location as a way of tagging information to non-proprietary
web-based software such as Google Maps, Google Earth, Microsoft’s Virtual Earth,
ESRI’s ArcExplorer, and NASA’s World Wind, amongst others. The move towards
‘digital earths’ for the distribution of geographic information has, without doubt, opened
up a widespread demand for the visualization of our environment where the emphasis is
now on the third dimension. While the third dimension is central to the development of
the digital or visual city, this is not the only way the city can be visualized for a number
of emerging tools and ‘mashups’ are enabling visual data to be tagged geographically
using a cornucopia of multimedia systems. We explore these social, textual,
geographical, and visual technologies throughout this chapter.
1. The Development of Digital Space
Digital space takes many forms. However in terms of the visual city, we are concerned
with the creation of space that allows us to generate a visual understanding of our built
environment. Knowledge of space is hard wired into us insubstantial and invisible; space
is yet somehow there and here, penetrating all around us. Space for most of us hovers
between ordinary, physical existence, and something given. Thus it alternates in our
minds between the analytical and the absolutely given (Benedikt, 1996). Our
interpretation of space and the resulting sense of location and place that is engendered
influence our perception of space both in real and digital terms.
Bell (1996) identifies three different kinds of space: visual, informational and perceptual.
Visual space is unsurprisingly all that we can see. It is the array of objects that surround
us creating, when viewed collectively, our environment. Each of the objects in any such
space has a multitude of different attributes, from variations in light and colour to
reflectivity. These objects create a reality which is a fully immersive environment in
Cartesian space, space that can be interrupted and explored in three dimensions. If these
objects are broken down to singular levels, then each can be viewed as being made up of
a combination of primitives. Primitives in turn are a collection of graphic tokens such as
points, lines and polygons, forming a two-dimensional or three dimensional
arrangements, and it pays us to think of visual space populated by these tokens (Mitchell,
1994). If these points, lines and polygons can be recreated in digital space, along with
their attributes, then digital space can mimic sufficient aspects of reality in terms of the
urban dimensions necessary to create what we have called the visual city.
Informational space can be seen as an overlay to visual space and it is in this space which
we communicate and receive information. From urban signage to oral communication,
information is communicated in visual space. In terms of the visual city, information
should not be viewed as a separate space but an additional attribute or in more prosaic
terms a new layer. Digital information takes the form of an embedding of data within
digital space. This combination of informational and visual space can be seen as forming
the basis for Google Earth and other digital globes. With the addition of user-friendly
communication to convey such informational space, an overlaps occurs with the third
form of space, that of social or perceptual space. Social space defines the user’s identity
and role in relation to other users in the social environment. In digital space, the social
dimension is increasingly important and this is seen in the rise of social networks such as
MySpace, Facebook and Twitter, to name but a few. Of interest is the fact that these
social spaces allow either the creation of visual space in terms of multi-user, three
dimensional, environments such as the virtual world Second Life or more direct mashups
which combine geo-located photographs of general users as displayed and accessed
through Flickr within Google Maps. These applications are the key to the Visual City and
we will come back to them in more detail later.
2. Creating Place and Space
In our research group Centre for Advanced Spatial Analysis (CASA), we have built a 3D
model of Greater London which we consider represents a Visual City. The production of
the model has only been made possible due to the development of 3D GIS and related
relevant tools in our case ESRI’s ArcScene, 3D Studio Max and various online
visualization packages, most notably Google Earth. A reoccurring theme in the
development of such 3D city models is the way in which emerging technologies are
enabling us to query, manipulate and construct our environment remotely. The Visual
City can now effectively be streamed and developed over the internet opening up a range
of possibilities, not only for visualization, but also for displaying attributes of the
population in the form of either socio-economic geographic data, agent-based models of
how cities function or even as actual users engaging with the software. Indeed it is fair to
say that we at a tipping point in city based information systems both in the way they are
used and created.
The goal of our Virtual London project is to develop a truly virtual city which can be
occupied, queried and manipulated by citizens within a collaborative environment. This
development route has entailed a combination of data capture, model development, and
optimisation. The acquisition of suitable digital data is central to the development of
Visual Cities and their use in the emerging online 3D GIS systems. In terms of pure
visualization the production of photorealistic models of the built environment is key to
the creation of visual space, yet it is a time consuming, manual process and one that up
until recently was in the domain of professional photogrammetry. The standard approach
to producing a photogrammetric reconstruction of the city has been through the use of
calibrated images and matching control points. Figure 1 illustrates the development of
one of the key buildings along the north bank of the Thames which is modelled using a
combination of oblique photography from helicopter capture and ground based imagery.
The model took approximately two days to produce.
Figure 1 Photogrammetric Modelling
In today’s Google-led world which is based on releasing free software with high levels of
functionality combined with low levels of required expertise, it is now possible to
considerably reduce the time taken to produce such models. Google SketchUp is a unique
program that is available in both professional and freeware versions. The differences
between the free and professional versions are negligible and are only significant in terms
of importing and exporting data. This now means that the public at large are able to
photomodel through SketchUp and produce their own sections of the city. Google
SketchUp is linked directly with Google Earth which we examine further a little later, and
it is linked for a good reason – for users to develop free content.
Creating a Visual City, one that reflects the actual built form, is a huge task and therefore
time consuming. So far, the only groups that have been able to get close to representing
the city visually are Games companies such as Sony and more recently Microsoft. Games
such as ‘The Getaway 3’ on the Playstation 3 represent the cutting edge in city
visualization. The Getaway originally appeared on the PlayStation 2 as a 3D rendition of
London covering approximately 10 square miles (16 square kilometres). The team behind
the model produced a wire frame model based on a photographic survey of London and
then projected the resulting textures onto the geometry.
The results of such developments are impressive but the costs are typically in order of
tens of millions of dollars to produce, while the models are also only of use for gaming.
They cannot be easily ported into contexts either where geographical analysis is required
or where the public at large can interact with them, largely due to their nature of
construction. Therefore to reduce cost, Google released their SketchUp software so the
users-at-large could produce the city themselves, block by block, building by building. Of
note is the latest version of SketchUp which, at the time of writing, allows users to import
and calibrate their own photography, directly modelling over the imagery. Although not
as accurate as the traditional photogrammetry, it does allow rapid modelling and the
widespread adoption of photorealistic content. Figure 2 illustrates a streetscape modelled
in less than a day using SketchUp. The image is shown without textures to illustrate how
architectural detail can be added to the model quickly and easily.
Figure 2 Rapid Modelling Using Google SketchUp
The release of Google SketchUp has in turn led to the development of the Google 3D
Warehouse, an online repository which is directly linked with the SketchUp program.
Currently there are 252 user submitted buildings in London, ranging from landmarks to
people’s houses. Buildings are uploaded to the warehouse automatically from the
programme, creating a quick and easy way to populate the city. This of course leads to
duplication and worries about quality, ruling it out for real world applications such as
architectural impact analysis or planning applications work, but it does provide a quick
and effective visual insight into the city. The best of the models are selected by Google
for viewing in the Community Layer of Google Earth, thus completing the cycle of the
public-at-large creating the Visual City.
One of the additional techniques we have used extensively in our group at CASA to
communicate a visual sense of the city is panoramic imagery. The use of panoramas is
not a new phenomenon; indeed the first panorama was patented in 1787 (Wyeld, 2006).
Panoramic visualization is not three dimensional per se in that it consists of a series of
photographs or computer rendered views stitched together to create a seamless image.
Rigg (2000) defines a panorama as an unusually wide picture that shows at least as much
width-ways as the eye is capable of seeing. As such, it provides greater left-to-right views
than we can actually see (i.e., it shows content behind the viewer as well as in front). It
was not until 1994 and the introduction of Quick Time Virtual Reality (QTVR) for the
Apple Macintosh that panoramic production became available on home computers.
Software was released that allowed a series of photographs to be seamlessly stitched to
form a single complete 360 x 180 degree view, and we illustrate an example panorama in
Figure 3.
Figure 3 Swiss Re 360x180 Degree Panorama
Although panoramas are essentially two dimensional, they can be inserted into a three
dimensional scene to provide an instant sense of location and place. The field of view of a
panorama equates to the coverage of a sphere. As such by draping a panoramic view onto
a sphere and then moving the viewing field to the spheres centre, or nodal point, the view
straightens out the lines in the image providing an exact replica of the human eye’s line
of sight from the location. The ability to drape onto a sphere allows the panorama in to be
depicted in x-y-z three dimensional space, and it can be embedded in other models of the
Visual City. Figure 4 illustrates a panoramic sphere embedded within Google Earth.
Figure 4 Panoramic Images in Google Earth
The images are placed on the reverse face of the sphere allowing the user to look inside;
while wrapping around a user when they enter the nodal point of the view. The
panoramas or ‘Urban Spheres’ as we call them, are open source files linking to imagery
outside of Google Earth on sites such as Flickr which quickly enables us to create a sense
of location and place. Google Earth has been fundamental to the development of the
Visual City and it is this to that we now turn.
4. Visual Cities and the Visual Earth
The World Wide Web has provided a revolution in the way we obtain, distribute and
react to information and we now take for granted the ability to search, edit and publish
information regardless of location. The first commercially available browser, Netscape
based on the earlier Mosaic, was released in 1994 and much has happened in terms of the
way we now distribute, manipulate and visualize data since that time. It is arguable that
we also take for granted to ability to zoom into any location on the globe and view
various levels of informational and visual data in a three dimensional environment. Yet it
is barely 24 months, at the time of writing, since the original Keyhole Earth Browser was
re-branded and launched as Google Earth.
Google Earth is the current buzz word in terms of geographic information and is covered
in detail by Michael Goodchild in Chapter 2. The importance of Google Earth to the
Visual City is three fold. Firstly, is the ability to view the city in two dimensions via high
resolution aerial imagery. Levels of detail vary according to location with ‘Googleplex’
(the Google Campus) providing the highest current resolution at 2.54cm per pixel. The
use of high resolution digital imagery allows the user to gain a visual overview of a city
from the air. This is a new emergence and as such has led to the rise of sites that track
locations of sightseeing within Google Earth. Although the highest resolutions are limited
to urban areas, Google Earth sightseeing is a global phenomenon. To hone this discussion
back to the city, the move to the third dimension has probably the largest impact in terms
of visualization of geographic information than any other. Although predominantly US
and Japan based due to copyright issues on data which we return to later in terms of our
own model, Google’s three dimensional cities are fundamental to the idea of the Visual
City. They represent a significant development in the visualization of city environments,
not only in terms of our ability to view building outlines and polygons but also due to
their location in true geographical space. Thus geographical location provides the third
issue of importance in Google Earth which involves the ability to add data and visualize
information using the three dimensional Visual City as a backdrop or canvas to other data
sources. The ability to visualize and overlay information opens up a number of
applications for the Visual City, applications which were once in the domain of the
professional user, it is these to which we now turn.
4.1 Applications in the Visual City
With the rise of computing power has come an increase in publicly accessible GIS
information and with it the ability to visualize in three dimensions leading to a massive
demand for city models. In terms of Virtual London, the fully functional complete model
has been developed in different ways for different audiences. This is of some importance
as each audience requires a different level of interaction and interface. While the visual
use of the city is almost universally similar between different users, what changes is the
level of data mining possible, the delivery method, and the interface. Broadly it is
possible to identify two main categories of use, firstly we have fully professional usage
which includes the use of the model by architects, developers, planners and other
professionals who are anxious to use its full data query and visualization capabilities. For
example, an architect might place a building within the model and use this to assess a
variety of issues from its basic visualization to the impact it might have on traffic and
surrounding land use.
In terms of our London model, the fully professional application has been our main focus.
The 3D model has been rolled out to all 33 London Boroughs, providing London with its
first city wide 3D GIS system. This raises a number of issues in terms of software,
hardware and expertise required to manage and view the model. As such we have rolled
out along side the professional model a customised version written to dynamically load
according to a viewpoint in Google Earth. Figure 5 illustrates a section of the model in
Google Earth.
Figure 5 Virtual London in Google Earth
The Google Earth version is specifically developed for the non GIS user. In terms of
professional use, the level of functionality is compromised but the ability to navigate and
overlay other datasets is increased. This is a common trade-off for functionality versus
cost and ease of use. As such the choice to roll out a Google Earth version is important as
it allows any local government employee to view the model. This links in to our second
level of user – the concerned citizen for public participation. Initially this was seen as the
main focus for using the model but has had to be restricted due to issues of copyright with
the Ordnance Survey base data used to present this version. The restrictions on data use
have been central to city visualization and GIS in general, especially outside of the
academic community. In short, data costs money to collect and therefore license to use
the data is often restrictive in terms of further distribution. In terms of Virtual London,
this led to the withdrawal of the public access version illustrating the difficulty faced by
Ordnance Survey in adapting its licensing policies for the new age (Cross, 2007).
The public face of Virtual London is therefore currently limited to movie files and as
such indirect visualization of data within the city model. While this is restrictive in terms
of public participation and allowing access to the data, it does result in improved visual
output as with movie files; one is not concerned with real-time visualization. A good
example of this is how a three dimensional city model can effectively communicate data
in the visualization of air pollution where such levels of visualization are currently not
possible in real-time but possible offline as we show in Figure 6.
Figure 6 Air Pollution Rendering
Figure 6 illustrates air pollution data from the Environmental Research Group at Kings
College London where a pollutant surface based on nitrogen dioxide in three dimensions
is draped over the cityscape. The move to visualize data in three dimensions is
controversial and often seen as mere ‘eye candy’ by some specialists in the field. Yet in
terms of a communication tool, it illustrates the areas of intense air pollution arguably
more effectively than any two dimensional map. This may partly be due to the visual
nature of the medium allowing a stronger sense of location and place to be obtained than
a top down two dimensional view. As such any amount of data can be visualized with the
model. Figure 7 illustrates how the city can flooded as the result of sea level rise. With
the animation file, it is possible to watch the water level rise and therefore identify which
areas are more at risk according to the degree of rise. Again, this is a use of the Visual
City in offline mode where we can sensibly embed data to visualize important outcomes.
Figure 7 London Flooding
The Visual City does not necessarily need to be three dimensional. Indeed as we argue
later, there are a number of emerging two dimensional technologies that create a Visual
City. Neither should a Visual City be seen as purely data or informational space; for
social space is becoming increasingly important in its development as we will now show.
5. The Development of Virtual Social Space
Terms and phases come into and out of fashion. Cyberspace, a once common term for
describing the Internet is now passé, as is the term Metaverse for the description of multiuser
worlds. Yet it is Stephenson’s (1992) textual definition of the Metaverse which is
closest to today’s visual virtual cities. Stephenson’s novel Snow Crash depicts life in the
Metaverse as the following:
“As Hiro approaches the Street, he sees two young couples, probably using their
parents’ computer for a double date in the Metaverse, climbing down out of Port
Zero, which is the local port of entry and monorail stop. He is not seeing real
people of course. This is all part of the moving illustration drawn by his computer
according to the specifications coming down the fiber-optic cable. The people are
pieces of software called avatars.” Neal Stephenson, Snow Crash (1992, p.35).
Avatars are an individual’s embodiment in the Visual City, providing the all-important
visual and social presence in the digital environment. They are the citizens, the
occupants, and the commuters of the digital realm; indeed they are the inhabitants of the
Visual City in all but a real physical presence. The term avatar – for use in terms of
digital environments – was first used by Chip Morningstar, the creator of Habitat, the first
networked graphical virtual environment, developed on the Internet in 1985. The term
‘Avatar’ originates from the Hindu religion as an incarnation of a deity; hence an
embodiment or manifestation of an idea or greater reality. Figure 8 illustrates typical
designs for avatars in a virtual world, in this case in Second Life.
Figure 8 Avatars in Second Life
Second Life, launched in 2003, currently represents the most successful social/visual
space on the Internet. It differs from other more game-based systems such as the popular
World of Warcraft as it does not have any quests or goals. The system is purely a social
geographic space within which its users are able to construct the environment entirely
themselves. From the elevation of the landscape to the scale of a city, every part of
Second Life’s visual space is editable. It is as close to the Metaverse that current
technology allows and provides a unique insight into the future of the Visual City.
Benedikt (1996) states that virtual worlds are not real in the material sense, many of the
axioms of topology and geometry so compellingly observed to be an integral part of
nature can therefore be violated or reinvented as can many of the laws of physics. It is
this reinvention that allows attributes to be enhanced and emphasised and the laws of
gravity, density and mass to be excluded, allowing buildings to be moved or deleted with
the click of a mouse and allowing the user to fly above or anywhere within the
environment.
As such Second Life is a Visual City which does not collate to the cities in Google Earth.
It is a landscape of fictional space existing only on one of the 3000 servers that power
Second Life. The lack of gravity and the ability of Avatars to fly or teleport to locations
creates a cityscape which differs considerably from the real world. With a combination of
limited design control - there are no planners or architects - simply the ability of any user
to create a virtual sprawl of spiralling urbanity mixed with eccentric retail areas and
recreational land use parcels.
In terms of the Visual City, you would not necessarily expect textual information to allow
the creation of a cityscape. Yet combined with a social network, text-based
communication can provide a uniquely visual view of the city as a whole. Text-based
messages via mobile phones are now part of everyday life. The first text message was
sent in December 1992, while SMS (short messaging service) was launched
commercially for the first time in 1995 (Wilson, 2005). Text-based messaging is, in
general, a one-to-one communication system. To create a social space, the SMS needs to
be shared via a wider network and thus it becomes one-to-many in its communicative
potential through newly emerging services such as Twitter.
Twitter is representative of the recent trend in social networking sites allowing people to
connect and communicate. Where it differs from sites such as MySpace is that it is purely
based on the SMS format of 140 maximum characters with the text entry box via Twitter
asking the simple question of ‘What are you doing’? As such, the system is applicable to
short and often pithy updates on a person’s activity sent via mobile phone, instant
messaging device, or via the Twitter website. The question is how can this text based
information source create a Visual City? The answer is partly due to the shear number of
users on Twitter (in excess of 200,000) and the ability to include a user’s location in the
messages. Combining the location of Twitter posts, known as Tweets, with a Google
Maps Mashup generates the ability to visualize what people are doing at different
locations in a city in real-time. We illustrate the location of Tweets in Central London in
Figure 9.
Figure 9 Tweets in Central London
New visualizations of Tweets are currently emerging on an almost daily basis allowing
the concept to scale to the global level with the ability to visualize in real-time feeds of
people’s thoughts and ‘what they are doing’. Using Microsoft Live (Microsoft’s webbased
mapping service), it is possible to visualize these one way conversation flows
updated every five seconds with either a global or street level view. Developed using a
system known as Atlas and GeoRss, a startup company ‘Freshlogic’ have developed a
mapping system that updates these feeds geographically. Of interest in terms of the city is
the overview which is gained when viewing from above. By simply letting the system
run, it will zoom into each new location, complete with address, users’ photographs and
Tweet every 5 seconds. The system also works with geotagged photographs via Flickr.
Using the same Atlas system, the map will update with new photography of places at
each predetermined time interval. These again are live feeds into a web-based
geographical visualization system, something that was unheard of and hardly imaginable
a mere 18 months ago.
The key to this rise in geographical information is data but not data as we would
traditionally view it in large information sets from a central repository, often governmentbased,
but personally gathered data. The move towards low cost, yet powerful, software
such as SketchUp creates a geographically tagged database of three dimensional objects,
mainly in relation to our built environment. Yet this is only one aspect, as we have seen.
The move towards the increasing miniaturisation of hardware and the demand for remote
access to information is pushing forward hand-held personal digital assistants (PDA’s)
and more importantly the mobile phone market. These hardware innovations come in
waves with each new wave adding increasingly complex functionality within increasingly
easy-to-use interfaces. In the late 90’s, PDA’s were the ‘must have’ gadget for remote
access to information. Functionality was limited to email and internet access, firstly via
slow modem connections linked to mobile phones and then later via wi-fi hotspots. Such
devices allow access to information but not in the geographic sense per se. As with all
waves of innovation, PDA’s fell out of favour and are only just re-emerging, this time
integrated within mobile phones, making available a portable digital tool kit for the data
capture of Visual Cities available to the public at large.
The latest of these devices is the Nokia N95, a phone which features a 5 mega pixel
camera, wi-fi and more importantly a built in GPS. As such it makes the perfect tool for
both capturing and communicating within the built environment – a portable tool to
create the Visual City. The camera has a high enough resolution for use in
photomodelling and SketchUp as well as holding up the possibilities of panoramic
capture. The GPS unit allows tracking of routes and the uploading of data to Google
Earth.
Figure 10 Personal GPS Tracking Data
Figure 10 Illustrates my route into Waterloo Station, London, tracked using the N95. The
height of the route represents speed, providing a unique insight into my own travel into
the city. The integration of GPS into devices such as mobile phones allows them to be
used outside of the traditional car-based environment and thus they become part of our
navigational abilities on foot. The ability to navigate through the physical city while
capturing digital data in real-time or sending Tweets or geotagged photographs to Flickr,
represents a key development in the Visual City. People generate data, data which up
until now has generally not been logged, let alone sent to a digital earth for visualization.
In terms of the Visual City, it should not be assumed that there is one Visual City for each
urban area. Indeed we can identify numbers from one or two full three dimensional city
models to hundreds of thousands for individual city visualizations. There is not as such a
single platform or database for the increasing amount of information that can be captured.
Google Earth provides a good basis with its Community Layer which provides
information gathered by the public-at-large. The shear amount of information can
however be overwhelming and in general, this layer is left switched off and therefore
unseen by the majority of users. The shear density of population in a city, and thus the
amount of information that could be input into system such as Google Earth, is resulting
in vast amounts of data of varying quality. While such data is of interest on a number of
levels for display, the move seems to be towards one of the personal, yet shared Visual
City rather than a single collaborative database.
6. The Future: The Personal City
A familiar theme is the decrease in knowledge required to create and present
geographical information which is leading to a direct increase in the amount of
information available. As we have seen, user created data can be visualized within a
global system such as Tweets and Flickr via Atlas and Microsoft Live or as personal
tracks via mobile devices within Google Earth. While all these data streams can by built
into one Visual City, such as our Virtual London, there is also a move to more
personalised geographic data. The editing of Google Maps to create a location previously
involved the manual editing of code and a moderate knowledge of XML, but with the
release of Google’s My Maps it is now possible to create one’s own map in a matter of
minutes. The My Map’s application is a web-based service which allows the user to add
points, lines and polygons as an overlay to Google Maps. This again is a significant
addition to the visualization of the cityscape, both in two and three dimensions, as the
overlays created can be exported to Google Earth or indeed any KML viewer. In addition
to the ability to add points, polygons and lines to the map is the integration of video via
either Google Video or YouTube.
In essence, we are but at the beginning of what will be a revolution in social, visual and
informational data plotted geographically by general users. The ability to create one’s
own map of the cityscape is of prime importance as these maps can be either public or
private. If the user chooses the public option which is the default, the map becomes
searchable within the Google general search engine. Information embedded in the map
thus, if searched for, directly links to the map. As such the map, be it city based or
otherwise, becomes the key interface to informational space.
The rise of social networks provides us with the ability to look down on the city and view
the activities that its citizens are involved in. This ability provides unique social data and
an insight into how the citizens are thinking, working, and socialising. At the moment,
Twitters are two dimensional but it is a short step to move these data streams into a three
dimensional world such as Google Earth. If you then combine this with avatars as in
Second Life, then you not only have a Visual City with visual and informational space,
you also introduce perceptual space into the context. This is more than a Visual City for
we now stand at the threshold of a Visual Earth.
7. References
Bell, J. (2006), Virtual Spaces and Places; Cyberspace; Space and Place in Computing;
Presence Research,http://pegasus.cc.ucf.edu/~janzb/place/virtual.htm
Benedikt, M. (1996), Information in Space is Space in Information, in Michelson, A., and
Stjernfelt, F. (eds) Images form Afar. Scientific Visualisation – An Anthology.
Copenhagen: Akademisk Forlaf, 161-171.
Cross, M. (2007) Copyright Sinks Virtual Planning, The Guardian, 24/01/2007.
Lynch, K. (1960) The Image of the City, Cambridge, MA and London: MIT Press.
Mitchell, W.J. (1994) Picture Theory, Chicago: University of Chicago Press.
Rigg, J. (2007) What is a Panorama, PanoGuide,
http://www.panoguide.com/reference/panorama.html
Stephenson, N. (1992) Snowcrash, Bantram Spectra, New York.
Wilson, F. R, (2005) A History of SMS,
http://prismspectrum.blogspot.com/2005_11_01_archive.html
Wyeld, G. T., and Andrew, A. (2006) The Virtual City: Perspectives on the Dystopic
Cybercity, The Journal of Architecture, 11, No.5, 613-620.
Examples of the Visual City can be found at the author’s blog –
http://www.digitalurban.blogspot.com