Jump to content

How Do They Map Coordinates To Named Locations?


007
Followers 0

Recommended Posts

Does anyone out there know how GPS coordinates get associated with roads, businesses, and addresses? Tools like Google Earth that zoom to an address have to know what address is at what location. There are an uncountable number of places on this earth. I can see how parks, railroads, schools, restaurants, etc (all of those things that you can turn on and off on your mapping software) can be seen by satellite and mapped. However, that in formation has to somehow transition from "a building at xxxx N yyyy W" to "Joe's bar and Grill at xxxx N, yyyy W". If I need to find the coordinates of my parents house in Florida for example, where would I go to find that GPS coordinate listing? I'm sure that Surveyor information is around somewhere that is done on buildings/homes for building permits, insurance purposes (like the distance from a fire hydrant) etc etc. Magellan, Garmin, Google Earth must have some place to retrieve the information, but where?

 

:smile:

Link to comment

I'm fairly certain that the process involves math at some level, therefore to me it's a lot like magic.

 

If I were to hazard a guess, I'd say that it all starts with some sophisticated maps that know the location of every address. From there, those addresses are cross referenced to fixed points on earth and the coordinates are extrapolated from there. The very fact that I had to type words like "cross referenced" and "extrapolated" means that they might as well do it by rubbing a duck on a globe until he quacks in just the right tone as far as I'm concerned. Either way, I'm just as amazed as you are. :smile:

 

Bret

Link to comment

For purposes of mapping addresses, that information generally comes from the US Census. It's not coordinate-based, however.

Think of an intersection as a dot, and each road connecting dots as a link. What you might see on Google Earth as "1st Street" is actually a whole bunch of those links connected together. Now each of these links has information on the lowest and highest house numbers on each side of the road. For example:

 

300 O===============O 386

 

So the numbers on one side of this imaginary link go from 300 to 386. What most mapping software does is if you're looking for, say, 340, it'll just mathematically guess where that is (that is, it'll go 40/86 of the way from 300 to 386).

 

The simply answer is that for most of the country, specific coordinate data doesn't really exist. For the most part, local land boundaries and such are tied locally and not globally.

 

SLer

 

Edit: Clarification

Edited by Shorelander
Link to comment

For purposes of mapping addresses, that information generally comes from the US Census. It's not coordinate-based, however.

Think of an intersection as a dot, and each road connecting dots as a link. What you might see on Google Earth as "1st Street" is actually a whole bunch of those links connected together. Now each of these links has information on the lowest and highest house numbers on each side of the road. For example:

 

300 O===============O 386

 

So the numbers on one side of this imaginary link go from 300 to 386. What most mapping software does is if you're looking for, say, 340, it'll just mathematically guess where that is (that is, it'll go 40/86 of the way from 300 to 386).

 

The simply answer is that for most of the country, specific coordinate data doesn't really exist. For the most part, local land boundaries and such are tied locally and not globally.

 

SLer

 

Edit: Clarification

Thanks Shorelander. I understand what you're saying about segments between nodes. There still has to be a process that knows what businesses /homes/etc lie between the points on the line segment. I can't figure out how my software knew that JB's grill was on the corner of an intersection. Google earth can take addresses and map them. There has to be some database for correlation.

T

Link to comment

Thanks Shorelander. I understand what you're saying about segments between nodes. There still has to be a process that knows what businesses /homes/etc lie between the points on the line segment. I can't figure out how my software knew that JB's grill was on the corner of an intersection. Google earth can take addresses and map them. There has to be some database for correlation.

T

 

That's just what shorelander was explaining. The program knows the coordinates of the roads, and it knows the street adresses at each end of the road, where the intersections are. It then guesses on the location of adresses in the middle.

Link to comment

For purposes of mapping addresses, that information generally comes from the US Census. It's not coordinate-based, however.

Think of an intersection as a dot, and each road connecting dots as a link. What you might see on Google Earth as "1st Street" is actually a whole bunch of those links connected together. Now each of these links has information on the lowest and highest house numbers on each side of the road. For example:

 

300 O===============O 386

 

So the numbers on one side of this imaginary link go from 300 to 386. What most mapping software does is if you're looking for, say, 340, it'll just mathematically guess where that is (that is, it'll go 40/86 of the way from 300 to 386).

 

The simply answer is that for most of the country, specific coordinate data doesn't really exist. For the most part, local land boundaries and such are tied locally and not globally.

 

SLer

 

Edit: Clarification

Thanks Shorelander. I understand what you're saying about segments between nodes. There still has to be a process that knows what businesses /homes/etc lie between the points on the line segment. I can't figure out how my software knew that JB's grill was on the corner of an intersection. Google earth can take addresses and map them. There has to be some database for correlation.

T

 

Your GPSr knows the address of "JB's grill" is 340 Main Street, Anytown, Any State. :blink: It finds the address just the same way as if you were looking for the address and then chooses the best route to get you there. the Software also knows that even numbers are on one side of the street and odd numbers on the other.

 

My maping softeware (Garmin City Select NA), has Point of Interest names, addresses and phone numbers all stored on the DVD. When I download the data onto the GPSr, the same is saved for future reference. :blink:;)

 

teald024

Link to comment

Thanks Guys. Again, I understand that estimating is required and not every building is in the databse, but BigWhiteTruck says that theGPSr know the coordinates of the roads and knows the streets addresses at the end of each road and figures things out - HOW? Someone or some program has to correlate the road info to GPS coords and then has to know the end points of the roads to make guesses of points in between. I guess some govt mapping agency does this or something. :P

Link to comment

Thanks Guys. Again, I understand that estimating is required and not every building is in the databse, but BigWhiteTruck says that theGPSr know the coordinates of the roads and knows the streets addresses at the end of each road and figures things out - HOW? Someone or some program has to correlate the road info to GPS coords and then has to know the end points of the roads to make guesses of points in between. I guess some govt mapping agency does this or something. :P

The TIGER (Topologically Integrated Geographic Encoding and Referencing) database is (I believe) what Shoreline was talking about in his example above. (There can't be that many address databases maintained by the Census Dept. :P) The missing link is that in addition to the beginning/ending addresses for a section of road, the TIGER database also contains the (approximate) lat/lon for the beginning and end of each section of road. So once you've estimated how far down a road segment some location is using its street address, you can then go and do a similar estimate on how far it is between the two lat/lon points, and come up with (approximate) coords corresponding to that address.

Link to comment

I saw this on TechTV back when it was cool. :anibad:

The companies have people driving in a GPS-outfitted car marking the POIs, road conditions (speed limits, etc.) and the like. One such company is Navteq.

 

There's the missing link!!

People in cars drive around and give precise (and probably continuous) feeds back to a database of where they are and what coordinates map to that. How else can a GPSr see that a road has a bend in it or how cloverleafs are built, etc. I can see how that can work for a lot of things but the software maps are far more comprehensive than people driving around in cars can handle. It's still a bit of a mystery to me.

 

This calls for some research!

Link to comment

In the episode I saw, it was two people in the car... one driving, one recording into a laptop. They had a pretty elaborate GPS setup in there and the passenger would punch info into the laptop for POI's. The driver probably just maintained the speed limit to have it feed that data into the program. Pretty cool.

 

Let us know what your research finds.

 

:anibad:

Link to comment

I saw this on TechTV back when it was cool. :anibad:

The companies have people driving in a GPS-outfitted car marking the POIs, road conditions (speed limits, etc.) and the like. One such company is Navteq.

 

There's the missing link!!

People in cars drive around and give precise (and probably continuous) feeds back to a database of where they are and what coordinates map to that. How else can a GPSr see that a road has a bend in it or how cloverleafs are built, etc. I can see how that can work for a lot of things but the software maps are far more comprehensive than people driving around in cars can handle. It's still a bit of a mystery to me.

 

This calls for some research!

 

The maps in the GPS are created from survey maps. A statring ppoint is given for an "object" like a road and a directuion is given. When the road turns/curves the point of the change is given and the angle/curve radius is given to the point that the road goes straight again. Thus a road the goes absolutely straight for a very long distance need only take up a very small amount of data to describe. Tracks for roads are not gathered by people driving around all day.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Followers 0
×
×
  • Create New...