Finding Geographical Obstructions between two points

[Welshbird]

I don't know very much about Mapping I must confess.

Howver, I currently calcuate the linear distance between two points using simple Pythagorus and the XY coordinates. I don't allow for curvature of the earth or anything flash, as these distances are ALWAYS less than 10 miles.

This project is all based around 'journeys', and my overall model is being affected by obstructions between these two points. These might be motorways, 'A' roads, rivers etc.

I'm trying to find a relatively easy and hopefully relatively cheap method of doing this. The project isn't big enough to justify a huge outlay, and also Ijust smply don't have the budget for it.

I've had some suggestions about using a full mapping layer (using half a terabyte on Oracle) but I'm just not is a position to do that.

So, any suggestions of a way to do this would be great. It's a model, so it doesn't have to be impecably accurate - if it found 'most' of the cases of obstructions then that would probably be enough.

Thanks very much in advance.

Fee

The question should be [red]Is it worth trying to do?[/red] not [blue] Can it be done?[/blue]

 

[gmmastros]

Are you storing a list of possible obstructions? How are these stored?

-George

Strong and bitter words indicate a weak cause. - Fortune cookie wisdom

 

[sedj]

I think thats the point - she does not have this data (but will need to buy it / licence it in order to actually perform the alogorithm).

--------------------------------------------------
Free Java/J2EE Database Connection Pooling Software

http://www.primrose.org.uk

 

[Welshbird]

Yip - that is exactly the problem.

I need to find out which two pairs of points have a possible 'obstruction' between them.

Thanks Sedj!

Fee

The question should be [red]Is it worth trying to do?[/red] not [blue] Can it be done?[/blue]

 

[philhege]

For ground transport, anything beyond as-the-crow-flies calculations will have to involve street routing. A perusal of commercial solutions will show that it's not cheap. You may be able to piggyback on one of the mapping sites if you have any geocoding data available.

Good luck!

Phil Hegedusich
Senior Programmer/Analyst
IIMAK

http://www.iimak.com

-----------
I'll have the roast duck with the mango salsa.

 

[PC888]

I believe it is possible, depending on the accuracy you are looking for, and the availability of data.

If you possess, or can obtain, a list of the obstacles, which can be modelled as line segments, then all you have to do is to verify if the line joining point A to point B crosses any line segment.

Autoroutes and rivers in a city can probably be modelled in a few hundred to a few thousand segments, if you are looking for an accuracy of 10 to 100 feet.

The question boils down to: are you able to get the obstacles as line segments, in the same coordinate system as your points? How critical if you miss an obstacle? (Estimate the cost of error in dollars). Does some gets killed, the truck will arrive 20 minutes late, or it will cost you 100$ in penalties? You'd have to accept that you will miss obstacles unless you model the obstacles with very short line segments.

 

[sedj]

PC888 :

Of course its possible,if not fairly easy, - but, as discussed in the thread, the problem is a lack of data !

--------------------------------------------------
Free Java/J2EE Database Connection Pooling Software

http://www.primrose.org.uk

 

[PC888]

The question should be Is it worth trying to do? not Can it be done?"

I agree. However, someone interested in solving the problem will simplify the problem and digitize some data to make a test work.

 

[Welshbird]

I'm going to try using a proporitery routing software and data set to see if a 'routed' distance will help me. My first plan was to use my linear distance and a flag for an obstruction. IF this other method works then I'll be happy!

Thanks for all advice guys.

Fee

The question should be [red]Is it worth trying to do?[/red] not [blue] Can it be done?[/blue]