Jump to content

Averaging a track log


savant9

Recommended Posts

I am building a Wherigo cartridge. It takes place in a local nature park/heavily wooded area. This area is a maze of trails, which I am attempting to map all the trails as a baseline for my layout. Being fall doesn't help that much with the tree cover as most are coniferous trees, so my reported EPE varies quite abit and is relatively poor @ 10-16 meters on my Colorado, and even worse on my Legend.

I can walk all the trails at a brisk pace in about 45 minutes, recording the tracklog. Overlaying both the Colorado track and the Legend track, makes for a real mess. Short of me manually averaging the tracks to come up with a general trail map, is there an application that will allow me to enter multiple tracks and will auto smooth them out for me? I don't have a problem collecting the track logs 10 times if I need to to get a more accurate map.

Link to comment

My colorado is at the current firmware/software 2.8/2.8 And I have tracklog set to auto and most often. I could take a faster sample rate, but due to the poor epe it really wont be any more accurate, I will just have more points of innacuracy to deal with.

Edited by savant9
Link to comment

I asked a similar question a few months ago and someone mentioned JOSM:

 

http://josm.openstreetmap.de/

 

I've never tried it but it might be worth a look. I'd be interested to know if the tracks have to be time correlated or not.

 

There's also something similar going on here but I'm not sure it would be easy to use for what you are looking to do:

 

http://www.trailguru.com/wiki/index.php/Tr..._Trail_Database

Link to comment

That trailguru one looks like it would work, but it seems to be an online application only? I may check out that other app you suggested as well.

 

Do you know of another mapping program that will allow me to drag track points, if i wanted to manually adjust the track.? Mapsource doesn't seem to be ideal for this, unless I just don't know what I am doing.

Link to comment

It seems like it would be a whole lot less work to just use the "Track Draw" tool in Mapsource.

 

Display the two tracks, click "TOOLS" on the menu bar, click "Track Draw" in the dropdown menu and draw a better representation of the tracks. Then save that track as the good one.

 

It's worth a look because it's free, has minimal learning curve (5 or 10 seconds) and there's nothing to uninstall if you hate it. :D

 

If you like it, you can right-click on the Toolbar and add the "Track Edit" stuff to the toolbar so you don't need to go to the menu for it.

 

...ken...

Edited by Ken in Regina
Link to comment

Easier solution....

 

Find a friend with a 60CSx or a 76CSx and an external antenna, and map the trails using that one. Use the active log track from the memory card not a "saved" track.

I would bet that that track would be more accurate than either one you have now. It won't even NEED to be averaged.

Before anyone asks.......Yes, there is that much difference.

Don't stop and stand in one place for any length of time because, with the high sensitivity chip, it will log a scatter of points surrounding that area.

 

Use Expert GPS for any editing you feel is necessary.(like the above mentioned scatter)

Edited by Grasscatcher
Link to comment

It seems like it would be a whole lot less work to just use the "Track Draw" tool in Mapsource.

 

Display the two tracks, click "TOOLS" on the menu bar, click "Track Draw" in the dropdown menu and draw a better representation of the tracks. Then save that track as the good one.

 

It's worth a look because it's free, has minimal learning curve (5 or 10 seconds) and there's nothing to uninstall if you hate it. :D

 

If you like it, you can right-click on the Toolbar and add the "Track Edit" stuff to the toolbar so you don't need to go to the menu for it.

 

...ken...

 

Thanks, Ken. This turned out to be the simplest solution for me right now. I manually drew a track through the discrepancies with the existing 2 separate tracks I had. I think I came up with a fairly accurate representation of the trails, well at least good enough for my purpose. I plan to go walk the trails again in the next couple days and record 2 more track logs which I will manually average with my now single, averaged track.

Thanks to everyone who offered their input on this matter. :)

Edited by savant9
Link to comment

Based on my most recent testing the Colorado using 2.8 GPS firmware is about as good as the 60csx in terms of track accuracy. Two readings on different days from the same unit are going to vary 30-40', even the 60csx.

 

GO$Rs

 

Did you miss the part about the external antenna?

 

Even with the SIRF III chip , a high mounted E antenna (velcro on cap) is a great variable eliminator.

Link to comment

savant9, the data you are collecting is exactly the kind of data I am trying to collect for the Northwest Trails Project. BC is very underrepresented in the data, too, which makes any that you are collecting all the more valuable. I'd be happy to average your tracks for you if you would be willing to have the data included in the mapset!

 

I have looked at your site before, and actually have your mapset loaded on my mapsource. Although this park is a very small area (~50 acres) I will be sure to upload my tracks. It is supposed to be a nice day tomorrow, with minimal cloud cover, so I plan on collecting more tracks tomorrow. Will let you know how it pans out.

Link to comment

Did you miss the part about the external antenna?

 

Even with the SIRF III chip , a high mounted E antenna (velcro on cap) is a great variable eliminator.

 

You can use an external antenna with the Colorado as well.....

 

That's right, the Oregon doesn't accept an external antenna and I do miss that sometimes.

 

GO$Rs

Link to comment

That's right, the Oregon doesn't accept an external antenna and I do miss that sometimes.

 

GO$Rs

 

That is what kill the Oregon for me. I do a lot of map making using my truck and the difference with and without an external antenna in low reception areas is night and day.

Link to comment

savant9, the data you are collecting is exactly the kind of data I am trying to collect for the Northwest Trails Project. BC is very underrepresented in the data, too, which makes any that you are collecting all the more valuable. I'd be happy to average your tracks for you if you would be willing to have the data included in the mapset!

If you want a whole bunch of track files for hiking/biking trails in southern British Columbia, just search for the "trans canada trail GPS tracks bc" in your favorite search engine.

 

Another useful search is "Kettle Valley Railway GPS tracks". There are tracks for all three major sections of it available.

 

...ken...

Link to comment

I agree with Ken: manually creating an "average" track is the most accurate solution. I've walked some trails more than 10 times, and there is no way to systematically find the best fit of the tracks. Since I walked the trail, I can apply that knowledge to my manually drawn line.

 

Antennas and better chips help for sure, but even then there will be error. My methodology is to walk the trail at least four times: two in each direction. If the four tracks have too much variation, I will walk again. (I've found some instances in heavily canopied areas where the distance between two track points at the same location can be more than 100 feet apart.)

 

I wish there were a better solution!

Link to comment

I don't think anyone has mentioned this yet, but if its visible on an aerial photo, that's perhaps the most accurate way to draw a track.

 

RMB,

True, or at least reasonably correct, as long as it IS NOT aerial photos from Google Earth.

 

GE's geo-referencing in many locations is absolutely atrocious. Looks like it was done with a "H.A. guess " method.

 

Try Expert GPS by Topografix.

Link to comment

The problem with satellite photos is that they typically do not show the trails where they are most needed: in tree canopy. In open terrain: sure they're great, but then who needs to average? I find the most difficulty in obtaining accurate trail data is in situations where the satellites can't see.

Link to comment

The problem with satellite photos is that they typically do not show the trails where they are most needed: in tree canopy. In open terrain: sure they're great, but then who needs to average? I find the most difficulty in obtaining accurate trail data is in situations where the satellites can't see.

 

My "real world" track logging has found that with my 76CSx if I use my high (cap) mounted external antenna (I always use one when mapping a trail) then even in difficult conditions the tracks repeat within less than the mapping tolerance of the USGS maps. In that case, why bother averaging?

 

In very isolated instances (slot canyons,tunnels, etc), nothing works. I have never had a problem with even the densest canopy with the 76CSx and ext antenna. That's not to say it's impossible, just that I haven't experienced it after mapping several thousand miles of trails and ditches.

 

Once however, I did turn my GPS off while taking a break on a hot Summer hike, and forgot to turn it back on when I started again. It failed to log the last half of that hike and I had to go back and re-do that portion the next day. Ungrateful piece of equipment! It should have known that I MEANT to have it on.....and of course that was the steepest portion of the hike.

Link to comment

This is something I'd be interesting in finding out as well. I have some tracks that I would like to average together, if there is any software out there that will do it.

 

I see Rich mentioned TopoFusion for tracing tracks using aerial photos, but it also has a "Network" feature that can be used to combine GPS tracks:

 

http://topofusion.com/network.php

 

The main function is to create a network of trails, but a side bonus is that it averages tracks when there are multiple representations of the same trail in the network:

 

net-detail.jpg

 

(Average in cyan). You might give it a try.

Edited by bike_climber
Link to comment

The OP is in Canada. We don't have access to nice free aerial photos like in the US.....

 

That is also the issue with the Wherigo builder. Setting up the zones has to be done somewhat blindly as the builder does not have image coverage for Canada. When you choose to edit the zones though, it pulls from Google aerial photos, but thankfully the image is fairly close for where I am. Just have to make the zones a little more generous, as mentioned in a proir post. The price we pay for living in americas hat....

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...