Hi Alain,
The counting was made by a transportation engineering company and uses
a completely different method. It is based in main roads (not exactly
motoways) and 12 hours countings (which are much more precise).
The problem with this paper is not with the countings, but with the
same problem of Liu, it was to adapt the axial and continuity map to
vehicular movements, which are much more associated to physical
constraints. Different adaptations (or decisions, such as how to
reduce the number of lanes) would lead to different results. This need
more investigation.
The figure 2 illustrates one of the problems that one can find by
drawing an axial maps. This is not only about creating a "elastic"
simplification. These difference would also arise from the resolution
of the base map used to create the axial map. Low resolution maps lead
to simplified axial maps and small angles (in their continuations)
while high resolution lead to detailed axial maps and wider angles.
Regards,
Lucas
On 05/01/07, Alain Chiaradia <[log in to unmask]> wrote:
> Lucas
>
> How many cordon counts were used in your paper? Were the cordon counts done for a range of road (motorway, main roads, secondary roads) or just motorways?
>
> Seeing figure 2, it quite interesting to see in black and white that the understanding of what "the minimum set of longest line that minimise depth" means, seems simple but is not quite easily understood. If minimizing depth, which is the number of change of direction, is what the formalism is supposed to measure, obviously the encoding should take care of not introducing extraneous depth. Then there is only one figure that does it within the little bit of space that we can see.
>
> Otherwise, it seems alike wanting to measure shortest distance with an elastic band.
>
>
> ________________________________________
>
> Alain
> ________________________________________
>
>
> -----Original Message-----
> From: Lucas Figueiredo [mailto: Sent: 05 January 2007 13:36
> To: Subject: Re: How to draw Axial map for traffic analysis?
>
> Dear Chengke Liu,
>
> What axial maps capture the "potential" of the grid, which is related
> to the hierarchical structure of the street network. The actual
> physical properties (width, number of lanes) and constraints
> (directions, traffic flows, traffic lights) are not taken in account.
> However, we expect that most cases present a clear correspondence, for
> instance, long lines tend to have more lanes and larger width than
> short lines.
>
> Therefore I would recommend representing the urban grid in the most
> simplified way, reducing all lanes to a single axial line unless when
> there is a clear physical barrier spliting the lanes, such as a canal
> or a wall.
>
> I had similar problems when applying this kind of representation for
> vehicular movements. You may find some help at:
>
> Figueiredo, L., Amorim, L., 2004, "Continuity lines: aggregating axial
> lines to predict vehicular movement patterns"
> http://mindwalk.com.br/papers/
>
>
> Regards,
> Lucas Figueiredo
>
> On 05/01/07, Chengke LIU <[log in to unmask]> wrote:
> > Dear all,
> >
> > Thanks Professor Bill Hiller, Alan Penn, and Rui Carvalho for your valuable
> > replies. I really gain a lot from your words. As Professor Alan Penn
> > mentioned in his reply, the road system in Hong Kong is quite complicated,
> > for example, on some road, there are several traffic lines (drive ways),
> > which have the same road name, should I use one axial line to represent them
> > or is that oversimplified? Please see the attached figure, is that suitable
> > to use the highlighted axial line to represent that road?
> >
> > Many thanks in advance!
> >
> >
> >
> >
>
>
> --
> Lucas Figueiredo
>
> Mindwalk
> http://www.mindwalk.com.br
>
--
Lucas Figueiredo
Mindwalk
http://www.mindwalk.com.br
|