View Full Version : Measuring positions on a map?

10-17-2009, 09:33 AM
Is there any free software which will write to a file the x,y values of a selected position on an image?

I know this can be done using professional GIS software, and I know that there are a few freeware GIS systems, but
a. I don't know if any of the free ones include digitization
b. GIS systems and their installation procedures tend to be rather complicated and would seem to be overkill for this specific problem.

I was a little surprised that, although the forum has lists of software for creating maps, there don't seem to be any lists of software for analyzing maps. Have I overlooked them?

My specific problem is that I have a drawn 2D starmap that I want to convert into a 3D database. X and Y positions can be determined by measuring the stars' positions on the map. (I'll have to come up with reasonable Z distances, constrained by trade routes and the maximum ftl jump distances the traders can manage, but that's a separate issue. ;) )

There are only about 50 or so stars involved, so one option would be to use a drawing program like Paint.net to determine XY coordinates, typing in the values by hand from the on-screen display. I've probably already spent more time searching for a solution than all the typing would have taken, but it'd be nice to have a general solution for the future.

Thanks for whatever help anyone can provide!

10-17-2009, 11:50 AM
Photoshop has a Ruler tool that gives x, y, distance and angle. It doesn't write to a file, though; but you could easily type those coordinates onto any landmark.

10-17-2009, 06:14 PM

Thanks for the suggestion. I don't have photoshop, but paint.net does display the xy coordinates, which is what I needed. I wound up using that and typing in the values by hand. Now I'm fighting with generating reasonable z positions....

10-17-2009, 06:42 PM
You can also just lay a grid over the image, and read any position at will.

A hex grid is often more practical than a square grid ... but it sounds like you are doing something in 3D and maybe I should say no more until you POST YOUR MAPS (hint hint).

10-18-2009, 12:26 PM

If all one has is paper, then a ruler and transparent grid would seem to be appropriate. Once a map has been converted to a digital picture, however, measuring in pixels (as shown by various image display programs) is much more accurate.

My attempts yesterday to write a progrm to place the stars using their numeric coordinates failed. Close examination of the constraints suggest that at least one set of them is not geometrically possible -- it works in 2D but not 3D.

I'm reluctant to publish the original map, since it's copyright by someone else. However, here's a 2D screengrab showing my most recent 3D model. It's still 2D: I haven't modified the Z positions yet.

The polygons representing the stars are colored arbitrarily according to their political affiliation. They were placed as accurately as possible by using the map as a reference image behind them. I've colored the routes between the stars according to the distances between the stars they connect. Yellow = short, orange = medium, red = long.

10-18-2009, 01:18 PM
Ahhh, it's a 3D star map!

Do you not have their positions as a separate database to begin with? I don't understand why you have to measure, if you already know their coordinates.

I still hope to play around with using the color spectrum as the Z indicator in a basic birdseye view.

10-18-2009, 01:52 PM
Another idea would be to download something like Bryce 5.5 (free). You can place objects by entering their x,y and z coordinates. You could then rotate freely to get a good render angle.

very quick and dirty example attached.

10-18-2009, 02:20 PM
But that still doesn't indicate their relative positions in space. That top left star might be very close to the camera, or very far away.

Size would work for the depth dimension, if you are not needing to show the actual relative sizes of the stars. The small ones would be far, and the large ones near, like regular distance perception.

10-18-2009, 03:00 PM
No...that it doesn't do. You could do the ball on a stick thing (bad fast render below) - but that becomes cumbersome if you have lots of bodies. I like your colour idea.

10-18-2009, 03:35 PM
What I started out with was a paper map illustrating a book(*). This map does not include the stars' "depths" (Z values). Their XY positions are available only by measuring the drawn positions of the stars on the map. I took a digital picture of the map to help do that. I wanted a way to digitize the stars' positions so I wouldn't have to type in their XY values by hand. I finally gave up and did the typing. There are somewhat more than 70 stars involved, so it just took a little patience.

The map includes some indications of the lengths of the trade routes (long, medium, short). From those route lengths, one can infer some of the stars' depths (Z values). My simplistic attempts to generate all of the Z values algorithmically didn't work, so I've resorted to placing the stars in Z by hand, using a 3D modeling program that I know reasonably well (Anim8or). That actually has gone reasonably quickly. At any rate, it's taken less time than I spent writing the failed program.

After I've moved all the stars to reasonable Z positions, I'll measure those, too. The values are available in Anim8or's 3D model file, but figuring out which ones are appropriate is more work than typing the Z values shown on the screen.

My goal is to eventually produce a 3D map drawn as an anaglyph so that the depth of the stars can be seen directly, similar to the other map I've been working on. I'll probably try to generate pairs of stereoscopic images, but getting the interocular distance small enough is an issue because of the widths of the maps and the amount of text involved in their labels.

* - Chanur's Venture by C. J. Cherryh, star map drawn by David Cherry, her brother.