maptalk.co.nz News   |   Events   |   Forum   |   Photos   |   Links    

  Forum

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  

Mapping

Show Profile  Michael Posted: 9 December 2004, 1:18 AM  
The Czech Republic has published a critique of the draft specification for sprint-o maps, due to come into effect next year. Its on the Cz website maps.orientacnibeh.cz, and many of the points seem to make good sense. Ales Hejna is involved in the critique, he was here a couple of years ago and did some mapping in Auckland including the sprint champs map.

Show Profile  Michael Posted: 21 July 2005, 7:07 PM  
The OA Tech Newsletter reports a field trial of mapmaking using a tablet PC with integrated GPS unit running OCAD 8. Main limitations are batteries and cost. Full report promised in Dec edition Australian Orienteer. No mention of GPS reception problems, gum trees might be more open than NZ forests, and of course topography is generally less severe.

Show Profile  Michael Posted: 29 December 2006, 3:25 PM  
The topic "Mapping" under "Events and Training" is getting unweildy whereas this one under "Technical" is underused. I carry on the discussion of Google Earth as a source of aerial photos. GE is a free internet application which provides amazingly good aerial photos of the entire earth. The resolution and age of the photos vary, but eg I can see that there's a white car outside my house. There are however "holes" in the coverage with only reduced resolution, or cloud in the way. You need broadband.

Thanks to Martin I have improved my understanding of how to use these as templates in OCAD. First of all download the latest version 4 which is significantly better than v3. Secondly switch off the "terrain" layer which seems to distort things somehow. Thirdly switch off the sidebar and you'll see more at a time.

There are two ways to grab the image which I think is always 1024 X 693 pixels with the sidebar off: File --> Save Image which makes a jpg of about 130k, and Edit --> Copy which puts the image onto the clipboard where you can paste it into anything else.

A reason for taking it somewhere other than OCAD is that the higher resolution the smaller the area within the fixed window. So if you want to see every gorse bush you'll have heaps of images and you might want to join them together before opening them as a template in OCAD. For 1:15,000 maps where its a generalised edge of the gorse that's of interest then not zooming in so far will result in more in the picture and fitting the separate images in OCAD may be fine.

In my original posting on this I couldn't understand why I couldn't see as much in OCAD. The answer is that the resolution was "frozen" at the time of capture from GE, and no amount of zooming in OCAD can get any more detail. So you make your decision in GE at the time you save the image.

Finally you'll need a basemap of some sort to fit the images to. Most existing maps have an arbitrary coordinate origin and are unrelated to "the real world". Starting at an arbitrary point and tacking bits on the side is bound to create distortion which will eventually cause problems. You could use a low-resolution GE image from which you trace say the road network (after rotating to magnetic north of course). I favour the LINZ orthophotos, free on the LINZ website. They have less resolution than GE but are fine for setting up a framework in terms of the NZ Map Grid which then allows other sources of info (eg coords from a GPS) to be used in OCAD. But that's another story. Would love to hear from others who have used GE.

Show Profile  Greg Posted: 29 December 2006, 6:13 PM  
The topics are probably because you are using an old link or bookmark as Fraser has done away with the 4 different topics and is all under one large forum... refresh to www.maptalk.co.nz

Show Profile  Svend Posted: 31 December 2006, 2:34 PM  
I have only recently started using Google Earth images for mapping
and I was surprised how good the photos are compared to the ones I
have been getting from the city council. The images I have looked at so far are less than 2 years old and it is amazing the informatio you can get from from a photo afer enhancing it at the editing stage.
However, GE has some limitationg, the map I am working on at the moment is obscured by a large cloud and I need the aerial photo to
map the many patches of gorse.

I can certainly see GE as being of great value to mappers in the future. Some day they may even provide us with free photogrammetry.





Show Profile  Michael Posted: 2 January 2007, 11:57 PM  
Thanks Greg you are right. Therefore I would encourage mappers to use the "Mapping" topic with the (at present) smaller number of replies. What sort of enhancing do you do Svend?

Show Profile  Svend Posted: 3 January 2007, 10:24 PM  
Re: enhancement of GE images
I edit the image in "Arc Soft Photo Studio" before saving it to OCAD.
I always enhance the image with "Brightness" and "Contrast" so I can actually see the details. Other enhancement depend on what I want to map. If I want to map the clearings in a forest I use the "Special
Filter" called "Find Contour" (nothing to do with contours), I select the green colour and have the blue and red turned off. With a click of the mouse and a little adjustment all vegetation boundaries will be shown with a green line.Nothing else is shown.
Since the template is totally dark apart from the green lines it is
necessary to draw the vegetation boundaries off the map or use registration marks to adjust the template.
There are about eighteen different functions in the "Enhance Menu"
and some of them I haven't tried out yet. The "Sharpen Filter",
"Sharpen Vertically" has proven very usefull in identifying
large trees in native bush.


Show Profile  robbie Posted: 5 January 2007, 11:58 PM  
Very interesting Svend.
Tania has told my that you can scan the original photogrammetry in to illistrator (mac) and the programme will convert a bip map into contours which you can theb save as jpeg and import into ocad.I really want to try this. Has anyone done it?

Show Profile  Michael Posted: 20 January 2007, 12:34 PM  
Last year I suggested a meeting as below. There wasn't any interest then. I always learn something when I visit Jim Lewis and he has agreed to host a mappy-chat at his place on the Sunday evening of Waitangi weekend. No doubt the role of Google Earth in the process will be included. Let me know if you want to join us.

Last year's posting: I've been grappling with using NZ Map Grid coordinates and LINZ orthophotos as a way of providing a framework for map extensions or joining maps together. And removing distortions caused by less than perfect joins in the past, which have led to fudges at overlaps or underlaps. If there's anyone else who has done this I wouldn't mind a chat at the Wgtn Champs at Labour Weekend - one of the evenings - at Wanganui somewhere. RSVP to michael.wood (at) mapsport.co.nz.

Show Profile  Paul I Posted: 1 March 2007, 8:07 PM  
I would love a mass debate on symbol sizes... I am concerned that mappers are being over-ruled and bullied into conforming to a set of specifications that are less than perfect.
Every area we map has its own uniqueness, and when a mapper carries out his/her survey they get a feeling for that terrain and end up with a representation that should be consistant over the entire map. I understand that there are varying mapper styles and thats just human nature. We all try to create our interpretation using the same rules. An experiment was made overseas with several different mappers. Every map ended up quite different from the other, there was probably no right or wrong. ISOM 2000 was made to set some rules as required, with the emphisis on legibility which is fine. Now days Orienteering has changed with the normal map being made at 1:10000 and even larger for sprints etc. Smaller maps are being made using more complicated and interesting terrain, with the emphisis on accuracy, creating a challenging map for the runners.
I'm sure everyone mostly agrees so far...
The problems start to arise when we get in certain types of terrain that have complex arrangments of one or more particular feature. Such as lots of bolders, intricate vegetation boundary lines, limestone cliffs such as Pio Pio and of course sand-dune terrain with lots of knolls etc.
If the ISOM 2000 rules that a 1:10000 map be drawn using 1:7500 symbol sizes was applied to some of our best detailed maps the result would be maps that would have to be ieter undermapped by excessive generalisation or the result would be an illegible, hidious to look at map. Using the theory that if an area is too complicated to be legible using these specs, is unsuitable for orienteering (as the ISOM 2000 states)is bollocks (in my opinion) therefore we have an issue to solve on these type of maps.
I wish to use White Lightening as my example because it is recent and has some relivant issues. Both myself and Mike B, the other mapper, both requested that the knoll dimensions be slightly reduced, consequently this was not done by the cartographer because she had too much criticism from doing so on other occasions. This dissapointed me because it made some areas of knolls look terrible because they bled into each other and are hard to read. They also did not represent what I had mapped as the dimensions of the knoll on paper was ridiculous when applied to actual size on the ground. The white space between symbols is also very important as this gives the runner information about distance and overall feeling. When you are at these sites the map can give you the impression of expecting something larger than was intended. One way to try and remedy this is to slightly expand the mass of symbols and give cartographers the right to move contour lines etc to fit. This is currently what often happens to make a complex area easy to read, the consequence of this, however is that the end result is often not how the mapper intended the terrain to be represented. I strongly feel this should not be done unless the cartographer is also the mapper and the final result still works on the ground. In my opinion, an easy solution would be to allow slight deviation from ISOM 2000 in extreme cases.
I know that rules are rules, but it does not mean they are always right.
I stumbled on a website by Orest Kotylo, he is one of the worlds most prolific and best mappers and also an advisor to the IOF Technical department, it is very clear that in his respected opinion there needs to be rule changes in the future specification and, in the mean time mapper should be given some small lee-way in regard to symbol size and line thickness when required. I request that the cartographer be allowed to reduce to size of knolls on White Lightening just enough to fit without making them too small as were the previous specifications. A compromise would be that a new, slightly smaller knoll size symbol was created and used only on the problem areas where needed. Check it out http://www.o-maps.com

Show Profile  Michael Posted: 1 March 2007, 10:47 PM  
I've been prominent in the criticism of undersize symbols.

The problem with a "small" leeway is that there is no smallest knoll, depression or contour wrinkle in nature. Reduce the symbol size, and the mapper can put more on. And history shows they will do just that.

Show Profile  Greg Posted: 2 March 2007, 1:44 AM  
But there is a minimum physical size in the terrain, if this is stuck to there should be a problem with what Paul is suggesting.

This message was edited by Simon Addison on 2 March 2007, 2:24 AM

Show Profile  addison Posted: 2 March 2007, 3:29 AM  
I do agree with Paul to a certain extent. On occasions you do get features that are too close together. Michael I know that you stick very closely to the rules, but I do notice on a few of your maps that you do distort the map so that you can fit features, isn't this bending the rules anyway as you expand the detail compared to what it actually is in relation to the the whole area?

Show Profile  addison Posted: 2 March 2007, 3:30 AM  
Sorry for bad English, too many beers and not enough thinking.

Show Profile  Greg Posted: 2 March 2007, 11:07 AM  
Simon piss off and dont edit other peoples post you tosser

Show Profile  Paul I Posted: 2 March 2007, 12:11 PM  
Yes I understand what you are saying here Michael and am quite aware that some people are still making 1:10000 maps using 1:10000 specs which is not correct, and your'e having a tough time getting heard. It took me a while to be convinced, but now I agree that generally the 1:7500 dimensions are for the good. I do have an open mind.
Currently one of two paths seem to be travelled down with regard to symbol sizing.
A: The rules are adheared to by some cartographers - and [with complex maps only] we get a map that really is quite messy and hard to read.
or B: the cartographer completely over-rides the rules and ignors Michaels suggestions, which is getting us nowhere.
It may be ironic I know but maybe if we want more consistancy we need more flexibity!
Back to my example - there are probably some knolls mapped on White Lightening that could be removed,we all no sand-dune terrain is hard to get right 100%, throwing up constant questions to the mapper however, generally that will only solve a few delemmas. We are still left with many legitamate knolls of large preportion close together that have to be drawn clearly somehow.
Remembering that our brains only take in information easily when the picture is clear. Back in the day before Ocad some maps were works of art, the cartographer had the ability to manipulate almost any problem to get a great, easily understood drawing. Almost all international event maps on complex terrain that I have stowed away and covered in dust still look amazing today, they are so easy to read and comprehend, and I dolbt that these maps would come close today using Ocad with no flexibity and this flat 150% enlargement rule.
One of my pet hates currently experienced on some maps, and I've seen it mentioned on this forum is vegetation boundaries, where the dot size is so big and far apart that it both covers up contour detail as well as not following the edge very well, resulting in one ugly, hard to read mess. Another example of this is where the sandy ground symbol is again so large it camoflages the contours.

In some ways I would be keen to attend the Australian mapping thing at JWOC in order to get their take on the matter, but I also feel that with rocks it is far easier to chose what to leave out when many of the boulders are clearly undersized. That's where the "pick me" concept works well. We have other issues.

I'll reapeat, mostly I think the new rule works well, it is ONLY when we experience spacific problems in trying to create a coherant and pleasing-to-the-eye finished map that we need to figure out what should be allowed to happen.
I definately don't believe in a free for all, but think we could do much better than present in solving some of these issues whist at the same time working as close as possible to the guidelines of ISOM 2000.



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  


Ruffneck Productions © Ruffneck Productions maptalk.co.nz