22 replies to this topic

[Polaris]

I'd like to teach my computer how to draw topographic hachures.

I guess we've all seen "Desktop Hachure Maps from DEMs" by Kennelly and Kimerling - which is a great start...

Has anyone else out there made any similar attempts? Any successes? Anyone want to share technique?

Eric

[Martin Gamache]

There was a set of plug-ins for FH that were developed by Tobias Dahinden in Zurich and a series of articles on the subject by he and Lorenz Hurni. Tobias also did some work evaluating a bunch of methods and presented at the Bohinj mtn. cart . workshop http://www.geod-is.si/mcws06/4_3.pdf. If you look back to the Mt. Hood wrokshop you should be able to find his earlier work on this topic.

Jurg Gilgen of Swiss topo wrote a paper on current Swiss Topo practices for hatchuring. It is only available in German, let me know if it's something you are interested in and I'll post it.

Swiss topo now does this work digitally using raster cartographic software by a company called nuage. It is still done manually and not automated although they are switching to a automated scree depiction in the new version of their topo maps.

This is a not very active area of cartographic research but one of the last great problems to be solved and people joke that the first nobel prize in cartography will likely be awarded to she who solves it....

[Hans van der Maarel]

I've been trying to accomplish this in VNS about 18 months ago. Left it simmering for all that time, but I just picked it up again.

Bottom line is: it's still very much work-in-progress and my guess so far is that it really depends on the quality of the data that you put into it.

Having said that, this doesn't look too bad from a distance...

 hatchures.png   651.86K   390 downloads

I really should try to come up with a set of guidelines on how and when to use this component. Do some more testing and see if I can get some consistent results out of it.

[Polaris]

Thanks Martin for the leads on Msrs Dahinden and Hurni - I checked into it and it looks like they are doing some important and exacting work... the Xtras look interesting, but I don't read german - I may try to figure out a few. Are you familiar with these? any favorites?

What I'm after (right now) is a more 'simplistic', but perhaps more difficult to achieve - a 'hand drawn' look, with the mountains appearing somewhat like caterpillars or amoebas (or is it protozoas?) -

I was thinking that by manipulating the underlying terrain and mixing and matching resolutions, I would have enough control using the 'montana' method ... a stumbling block is that though AV (and Surfer) will plot point symbols with length and angle according to specification, I want lines, not point symbols, and I wonder where the algorithms would place the origin of the line (I think this can be overcome, just haven't put enough time into it yet) - I was also thinking it would be neat to use the AI tapered brush that I have seen discussed here - my impression though is that it can only be applied 'object by object' - but I'm not (yet) and Illy user....

Hans - I am a WCS user (though not much lately), and thought today that I should take a look to see what can be done there. Any hints?

[Hans van der Maarel]

Eric,

You'll want to get the latest version and look for the "aspect" dynamic parameter in the texture editor. I've used that to drive the orientation of a simple stripes component. The width is driven by the slope angle.

So far, it takes a lot of trial and error to make it work 'okay-ish' in a new project and it's still not giving the output I'm really looking for

[Maisie]

I recently had a project where the customer needed to have maps from 1776 redrawn in Illustrator. The originals were placed in the document so I only had to reproduce the caterpillars, not create them from scratch.

I tried unsuccessfully to reproduce them using the pen and brushes tools, even with a pen/tablet they just didn't look right. For one thing, I wanted a clean hilltop with all the edges seeming to come off a line. I ended up using type:

I drew a text path in Illlustrator and used an "!" in a font that was thick at the top and thin at the bottom. I used a wider character spacing for the sunny side of the hill and a denser spacing for the shady side. Then converted to outlines, deleted the dot and dragged each resulting polygon to the right length.

The result seemed to look best when the strokes were evenly spaced at both the top and the bottom.

I think you're looking for an automated method and this certainly isn't; very time consuming as a matter of fact, but it's given be the best results so far.

Martha

[Maisie]

...here's the image...

Attached Files

•  Hachure_samples.pdf   640.65K   344 downloads

[bentley]

Has anyone figured this one out yet?
I am currently working on a historic map project and would like to recreate the 'Lehmann" hachures common to many of the 19th century topographic maps. I have looked through many presentations and papers (including the German ones mentioned by Martin), but I can't find any technique that replicates this specific type of hachuring. I have also been playing around with drawing them by hand...argh.
I thank everyone in advance for their input:)
elbie

[MapMedia]

Martha, Can you show us how this looked? It sounds interesting!

[Hans van der Maarel]

Has anyone figured this one out yet?
I am currently working on a historic map project and would like to recreate the 'Lehmann" hachures common to many of the 19th century topographic maps. I have looked through many presentations and papers (including the German ones mentioned by Martin), but I can't find any technique that replicates this specific type of hachuring. I have also been playing around with drawing them by hand...argh.
I thank everyone in advance for their input:)

I haven't really looked into this recently (though I'd love to revisit it). The main problem is that whereas most software products that I know would treat a terrain as a set of discrete faces, it really is a more natural, flowing, shape. This is exactly what the hachures portray and also exactly what's making it so difficult to reproduce them...

[pghardy]

Has anyone figured this one out yet?

One of the guys at ESRI had done some research a few years ago on automating hachuring, but it never got to a useable state.

The best paper I know of describing an automation algorithm was done by Nicolas Regnauld - "Automated relief representation for visualisation of archaeological monuments and other anthropogenic forms", published in Computers, Environment and Urban Systems 26. I don't think it's available as an online resource. I'll ping Nicolas and point him at this thread to see if he has anything to offer.

[François Goulet]

One of the guys at ESRI had done some research a few years ago on automating hachuring, but it never got to a useable state.

The best paper I know of describing an automation algorithm was done by Nicolas Regnauld - "Automated relief representation for visualisation of archaeological monuments and other anthropogenic forms", published in Computers, Environment and Urban Systems 26. I don't think it's available as an online resource. I'll ping Nicolas and point him at this thread to see if he has anything to offer.

There's a copy on a personal page of the University of Edingurgh: http://www.geos.ed.ac.uk/homes/wam/

Now, if it's protected, someone should tell the author...

[François Goulet]

Just found this on ESRI's website: New Tools Produce Classic Cartographic Effects

There's a part on hachures. Look interesting...

[Nicolas Regnauld]

Has anyone figured this one out yet?

One of the guys at ESRI had done some research a few years ago on automating hachuring, but it never got to a useable state.

The best paper I know of describing an automation algorithm was done by Nicolas Regnauld - "Automated relief representation for visualisation of archaeological monuments and other anthropogenic forms", published in Computers, Environment and Urban Systems 26. I don't think it's available as an online resource. I'll ping Nicolas and point him at this thread to see if he has anything to offer.

I did develop this algorithm a few years ago. It's not currently available, but I might resurrect it if I can find a bit of time. I was working well for depicting embankments, and archeological sites at large scale. I'll have to check if I can make it available too, I can't remember who owns the IP over it (probably jointly owned by Ordnance Survey Great Britain and the University of Edinburgh, but I need to check).

[Hans van der Maarel]

Been tinkering with it some more tonight and lo and behold!

 hachures_test2.jpg   123.05K   135 downloads

This is actually starting to look a bit like the real thing I think. It's far from perfect still, but there's definately progress. I do find there is a subtle relationship between the size (line width/length) and scale.

This was done in VNS 3.01, using a strip texture which has its width driven by a greyscale image I prepared earlier (a standard lit-from-top-left hillshade).