I hate to put my oar in to this discussion, but I did
want to clarify one matter. Although Alasdair does
rightly deserve a *huge* amount of credit for his
excellent work on isovist integration aka VGA, Alan is
wrong when he states that Alasdair was the first
person to automate this process.
When working on the Intellegent Architecture Project
in 1995/96 I created an "isovist generating camera",
which automatically generated isovists from a list of
given points (as well as a camera which produced
series of isovists along axial lines, which I called
the 'axovist" camera). One variant of the isovist
camera exported the isovist polygons into "Pesh"
allowing me to perform isovist integration analysis of
them - in other words automated isovist integration.
The draw back of this system was that Pesh was limited
by the number of items it could process, which is why
I never developed it much further (until the creation
of the "OmniVista" Machintosh application that I
co-authored with Sheep for my doctoral work). If
anyone is interested in the Intellegent Architecture
work, it was published in the First Space Syntax
Symposium, which include details of this early isovist
camera.
--- Alan Penn <[log in to unmask]> wrote: > Mike Batty
wrote:
>
> > Some guys down here in CASA have shown me some
> correspondence about
> > isovists and isovist fields and the idea of
> visibility graphs. I have some
> > points.
> >
>
> Of course Mike is right, the thread of the
> discussion I guess he is
> referring to is one in which Jake - quite rightly in
> my view - suggests that
> the space syntax community should standardise on
> terminology and call the
> new kind of analysis, first presented by Alasdair at
> the Brasilia
> conference, 'Visibility Graph Analysis'. The
> alternative which has gained
> some currency and certainly predates the term VGA in
> the syntax community
> is 'Isovist Integration'. As Mike points out
> Visibility Graphs have been
> around for a long time and it is in recognition of
> this that we suggested
> that the name for the new variant in space syntax
> analysis be called after
> that. To add some of the geneology in the syntax
> field that Mike and others
> may not be aware of, although visibility graphs have
> been around for a long
> time, the idea that gave rise to VGA came from
> Benedikts work on isovists
> and isovist fields. Benedikt's E&PB paper describes
> a number of possible
> measures of the shape and boundary properties of the
> polygon describing the
> visual field from a particular point within an
> environment. Many of these
> measures are interesting in their own right, but all
> are local in the sense
> we use the term in syntax, and none are global. I
> think it is probably for
> this reason that although there has been a lot of
> interest in isovists
> within the syntax community, that nothing much has
> emerged from their use
> considered in Benedikt's original form. The main
> measures that syntax
> analysis has found useful (ie. that have predictive
> or explanatory capacity)
> are based on the idea of global measures of the
> graph, such as 'integration'
> and 'choice', or on the relation between local and
> global such as the
> measures of intelligibility or local area effect.
> When we first started
> thinking about how one might construct global
> measures of isovist fields in
> the mid 80's the obvious thing to do was to use the
> isovist polygon as the
> basic element of the representation map and then to
> measure integration
> values for the set of isovist polygons that covered
> the environment. Early
> exercises in this were by students at the time such
> as Tad Grajewski and
> Mehmet Fikilier, who computed these measures
> manually - a daunting task -
> and the method went under the name of 'Isovist
> Integration'. However, when
> Alasdair automated the procedure it became clear
> that although we had been
> thinking about isovists, the basic mathematical
> entity we were dealing with
> was the visibility graph. Although these graphs have
> been around for some
> time the idea of measuring their global properties
> and using them directly
> as spatial representations is as Mike suggests not
> the common currency in
> the traditional fields that have made use of
> visibility graphs, and their
> use in this sense by the syntax community is a new
> step. Of course the only
> thing that matters in the long run is whether the
> VGA explains or predicts
> anything of interest, and for that the jury is still
> out.
>
> Alan
>
>
>
>
>
=====
+-----------------------+
| Ruth A Conroy |
| [log in to unmask] |
| |
| tel. +44 20 7419 4255 |
| fax. +44 20 7419 4233 |
+-----------------------+
__________________________________________________
Do You Yahoo!?
Get Yahoo! Mail - Free email you can access from anywhere!
http://mail.yahoo.com/
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|