Forum   |   Links    


Forum Home   Start New Topic   Edit Profile   Register  

1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60   61   62   63   64   65   66   67   68   69   70  


Show Profile  mcroxford Posted: 7 June 2012, 12:07 PM  
Cheers Selwyn, I'm from the Nelson club with access to the Tasman DC data. We currently have Jason Markham doing a map near St Arnaud based on Lidar data and he's recently discovered that at a cost we can potentially source some of the data beyond the Council purchased swaths that were captured when flown. Exciting as we are starting the process of mapping a significant area around the Lake Rotoiti area as a result.

Show Profile  Michael Posted: 7 June 2012, 2:18 PM  
Big files. I got 18gb of contours from a local authority because they were talking about charging for the time to extract stuff, so I asked for "everything". Took a bit of work with FGIS and QGIS to extract what I wanted.

I think our work is changing from using undermapped photogrammetry (we add extra wiggles in the field) to overmapped LIDAR (we take superfluous wiggles off, by manual tracing, or mathematical smoothing, or whatever).

In the old days, if you weren't perfect (who is?) there might have been missed reentrants. The course planner worked round them, they weren't fatal if you kept controls away from them. In the new way, if you aren't perfect, there are extra reentrants and you may have to look at pretty much everything. There are some papers on the web, I think an IOF Mapping Conference, that talk about the tendency for excessive detaill on maps produced from LIDAR contours.

The significance of Routegadget's experiments, as I understand it, is in applying automated logic to these contours-with-spurious-wiggles to produce something that is like what a good mapper would produce. Is that what he's up to?

Show Profile  mcroxford Posted: 9 June 2012, 11:20 AM  
I think so. But also he is applying some other programmes which when there is a certain closeness to the contours then it is interpreting them as cliffs but more interesting when they capture Lidar there is top of canopy info and also bottom of canopy info. He seems to be using the difference between the two to map vegetation. Where the valley are essentially the same then no vegetation but when there is a differential then he's mapping the vegetation. Also there seems to be some sought of capture of density of the vegetation which he is using to determine runability.

Show Profile  Michael Posted: 9 June 2012, 4:38 PM  
Ah, the holy grail. Push a button and the map is made:-))

Actually, I think that some of Stewart Hyslop's basemaps could be run on without any fieldwork, where the vegetation is consistent eg white forest.

Show Profile  Michael Posted: 11 June 2012, 3:47 AM  
Revisiting the Hydrabad map (QB long) reminds me that I made the wrong decision when I mapped it in 1993. The excellent photogrammetry from Stewart Hyslop was at 2.5m. There were some (only a couple of) steep hills and I worried about contours touching, so I produced the map at 5m using the 2.5's as a guide for formlines.

I now think that 2.5m contours would be better, for at least two reasons. I am sure that dashed lines are harder on the eye than continuous lines. And because they are used on flatter and not steeper areas they remove one of the cues that we use - the between-line spacing. I'm sure that the variable white space between the lines is almost as important as the lines themselves in giving us the picture of the land.

We did the same at Watchtower in the same forest. When we came to put it in the computer we added in the 2.5s and I'm sure its much better for it.

Show Profile  Paul I Posted: 11 June 2012, 8:12 AM  
IMO 2.5m contours for sure on Hydrabad and most dune maps. Incidentally I havn't seen info anywhere about the scale and contour interval for oceania/wc races. From my observations there does seem to be a move away from 2.5 to 5 m contours in scandinavia,(not sandune terrain though), which probably explains why the need for more formlines on todays detailed terrain maps.
Re Vegetation; I think there is a need to create a new symbol in ISOM20XX (or is that now 2XXXX) for scattered ground vegetation such as we encountered on 'Prickly Sands' that is easier to read contours through and has better edge definition, it has to be the same on open land as for forest, as does the stripes but will get rid of that once and for all! What I imagine could work would be a random green dot symbol, where the dots are of very small diameter and spaced apart, although not sure how much without experimenting. I think the random placement,like stoney ground is self explanitary (an area symbol would have to be made to save time). Other green symbos such as the fight blobs would compliment the representation. In marginal areas there could be an ability to place the dot manually where if the vegetation got more dense or more scattered.
There again it could be rubbish but so is the line screen quite often.

Michael, your idea of the Solid Yellow on the clear non vegetated parts, with Rough Open where there is veg. sounds good until you picture terrain where the map is all rough open with the same vegetation type, then it goes back to the default green stripe. So there's no consistancy.

One thing I mentioned that I know you don't really agree with Michael is the need for Index contours on flater maps. To me they portray an overall picture of the terrain that is lost without them no matter if it's hilly or not. I think they should be mandatory.

With the wc and supporting races closing in, but not yet last minute, and having just run on similar terrain I think it's a good opportunity to comment and highlight things, good or bad, to the organisers that may make a big or small difference to all the events.

I really enjoyed the Long and Loops races despite things not being perfect(didn't go to the sprint). I did feel that the beginning of the long in the green low vis was a bit too bingo for a championship race, mapping would need to be spot on to make it fairer. In the Prickly Sands area in particular, I felt there was a real need for more slope line tags.

I was also very surprised at the short distances for the long but don't know if it was intentional or min/km quite underestimated.

My loops race was such fun even though the complex terrain beat me at times. The map highlights the difficulty in depicting such detailed sanddune terrain at 1:10000 with some map symbols, such as depressions and dot knolls having been reduced in size. It's very difficult be we need to strive to stop this if we can. Lot's of us are guilty of this including myself. It's fine to look at on paper but much harder to see while running.

Show Profile  Bryan Posted: 11 June 2012, 9:23 AM  
Some detail info about maps and event notes on the Oceania/WC 2013 website.

Specifically for elite courses only:
4 Jan: 1:10000 2.5m (sand dunes)

5 Jan: 1:10000 2.5m (sand dunes)
(model: 1:10000 2.5m)

6 Jan: Oceania/World Cup Middle: 1:10000 2.5m (sanddunes)

7 Jan: World Cup sprint qualification: 1:4000 2.5m (urban)

8 Jan: Oceania sprint / World Cup sprint final: 1:4000 2.5m (urban)

10 Jan: Oceania Long Distance: 1:15000 5m (gully-spur)

11 Jan: Oceania Relay: 1:10000 5m (gully-spur)

12 Jan: 1:10000 5m (gully-spur)

13 Jan: World cup prologue: 1:10000 5m (gully-spur)

Decisions not made yet on scales for other courses. Maps for World Cup events for elites will be offset printed as required. (Unlike for WRE events, offset printing for World Cup events is mandatory).

Show Profile  Michael Posted: 11 June 2012, 12:22 PM  
Some more thoughts on the Hydrabad vegetation. Only thru discussion do we work towards a solution. Be good to get more views, too.

First, what were the flat paddocks? Read the spec for "rough open", none of it applies to paddocks grazed by cows. So shouldn't there be 401 bright yellow right up the "inlets" into the boxthorn until something changed?

Then when we get slowed down in the boxthorn we want to portray that something is different. If we regard it as "generally open land" the first level of slowing down is 403 rough open. Putting 407 on top of 403 is the next level, and 409 on top of 403 is the next level. But I don't think we were slowed down all that much, hence 403. Remember that you have to weave around trees in white forest too. And we're not talking about the impassable blobs, they were separately mapped.

Which made me think of... forest. Most of the bushes were tall enough to be called trees. It could well be depicted as 405 white forest. Or in some places shades of green, and the dark green that was in fact used for the impassable blobs.

We will still need the green stripes though Paul, why don't you do some experiments and give us the symbol definitions?

Show Profile  nh Posted: 11 June 2012, 2:34 PM  
I agree with Paul I in that index contours should be shown on flatter maps too. They give a really good overall view of the terrain which makes it easier to understand the contours. When you are orienteering you can build up a picture of the landforms faster which therefore makes it easier to navigate. It means less tag lines are necessary to show the shape of the land which gives a less cluttered map.

Show Profile  Michael Posted: 11 June 2012, 4:25 PM  
There's reason behind my madness. As the land gets flatter the features highlighted by the index contours become random. It is possible for one 4-contour hill to make the cut, and another one to miss out.

Show Profile  nh Posted: 12 June 2012, 1:43 PM  
That is exactly the same in steeper terrain. A 4-contour hill can and also cant make the cut there too

Show Profile  NW Posted: 12 June 2012, 2:28 PM  
How about using this technique?

Show Profile  mcroxford Posted: 14 June 2012, 1:38 PM and to read more on the holy grail approach to map generation... Perhaps more alchemy?

Show Profile  richardh Posted: 20 June 2012, 1:02 PM  
Latest from routegadget. Will be great if he releases his processes for generating a base map.

Show Profile  mcroxford Posted: 22 June 2012, 11:49 AM Testing of a computer generated map.

1   2   3   4   5   6   7   8   9   10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60   61   62   63   64   65   66   67   68   69   70  

Ruffneck Productions © Ruffneck Productions