9 replies to this topic

[Matthew Hampton]

I have been working on a set of maps to accompany a manuscript based on a pioneer that lived in Montana at the mouth of the Musselshell River and kept a daily diary from 1865-1872. While scouring the internets for historic maps I came across detailed surveys the Missouri River Commission published in 1893.

 Screen_shot_2011_06_30_at_4.14.31_PM.png   804.42K   103 downloads

I also found shapefiles of the maps that had been vectorized, so I created a 3D version in ArcScene.

 2011_06_30_1623.png   192.52K   123 downloads

I've flipped historic maps into 3-demensions before, but this is the first opportunity I've had to try vectors. I'm not too sure where this will go, but I thought it was an interesting blend of history and modernity.

[frax]

It looks gorgeous, I really like the look - although the type doesn't really work. I guess that got converted to shape file as well?

[DaveB]

It does look very cool, modern and old school at the same time.

[Charles Syrett]

Fascinating. Looks like someone first vectorized the original using some sort of autotrace, set to digitize the centre of each "line" detected. Then this "jumble of lines" was georeferenced, and a shapefile made. (This accounts for text appearing as lines -- which it was in the original, since everything was hand drawn.) And then -- what did you do, Matthew, add the z values? Details, please. Truly a labour of love.

Charles Syrett
Map Graphics
http://www.mapgraphics.com

[Matthew Hampton]

I'm glad you like the image. It's pretty cool to fly through too!

If you come to the annual NACIS conference in Madison, Wisconsin Oct. 12-14th perhaps I will be able to share in greater detail.

Fortunately I was able to stand on the shoulders of others - including the draughtsman in the 1880's, Dr. David Miller from SUNY Cortland who scanned, vectorized and projected the maps, and the folks at USGS CERC who made them all accessible. Every bit of linework was vectorized (even anno) on a single layer.

I applied the elevation values to the vectors in ArcScene using the base heights from some NED data. Here are a couple more screenshots.

 2011_07_01_0956.png   347.82K   41 downloads

 2011_07_01_1004.png   302.64K   51 downloads

[Charles Syrett]

I suspected the process was something like that, and the screenshots very nicely add the missing pieces. Interesting to see how accurate the hachure-depicted relief actually was. Of course what's lost in the vectorization is the variation of thickness in the hachure strokes. Still -- very entertaining to carto-geeks!

As usual, I'll likely pass on the conference; it's just too close to Canadian Thanksgiving.

Charles Syrett
Map Graphics
http://www.mapgraphics.com

[frax]

Imagine a documentary about historic event x.

First the camera shows just the original 1880 map in 2d, then moves into an oblique view of the map, still all flat.

Then the heights are slowly extruded to create the 3d oblique view - like the image you have in your first post.

Finally the camera flies through the scene - alternatively it morphs into a satellite image, or oblique aerial shot of what the area looks like today.
(or if one can locate a satellite image from the 1880's, one could use that!)

[frax]

I know why I like this too - it also reminds of 3d wire-frame views - as in 80's computer displays, CAD and video games. Like Elite or the original Star Wars Arcade game.

Oh, it is 1980's computers, not 1880's computers.

[Jacques Gélinas]

Mathew,

I must say I like the old version better (hachuring work is wonderful)... but I like your idea of blending of history and modernity. Revisiting old maps with new technology sound like a great idea.

Keep us posted,

Regards,

[Charles Syrett]

I must say I like the old version better (hachuring work is wonderful)...

Wonderful indeed. From what I can tell, this hachuring was done using the Lehmann method (determined by slope), rather than the Dufour method (determined by northwest illumination).

Charles Syrett
Map Graphics
http://www.mapgraphics.com