Mapping software and GPS.

[Neil]

I have NG Topo! version 3.42 and a Garmin Rino110.
I downloaded a tracklog from internet for the Sewards and couldn't get it in my Rino. The pop-up window said that 500 points was max. That dosn't seem right, does it?

So I figured I'd save or download the tracklog as a route or something. No can do.
Then I copy-pasted 500 consecutive wp's (only a portion of the total tracklog) into a new .tpo file and uploaded the whole shebang into my Rino. Great! I had all these WP's that I didn't feel like adding one at a time to a new route so I deleted them all.

I tried this workaround which worked but it is inelegant and clumsy:
Draw a freehand route over top of the tracklog. Delete tracklog. Create a GPS route from the freehand one and upload that into the GPS.

Ideally there must be a way to select every nth wp (downsample)from a list of wp's.
Does anybody out there have a solution?

 

[Jay H]

Wow, nobody responded... I thought I'd hold out since I don't have NG Topo. However, I'd suggest to look into your preferences since I use Maptech's Terrain Navigator which has a bunch of tools to make routes, there is a freehand tool which is basically just that, a freehand pencil and there's also a connected line tool where you can make segments of a straight line.

Either way, in TN, to make a route, you basically use a pointer to trace the route you want and then you convert it to an actual route that can be uploaded to a GPS. In such a fashion, in the preferences, there are options on how to convert the trace to a GPS route. In those preferences, I can select the max. # of waypoints (TN calls them "markers", same thing) to use, thereby downsampling the curves.

I would look into the preferences to see if there is something in there that you can set the limit of waypoints...

Jay

 

[imarchant]

I use the Garmin Mapsource Topo software with a e-trex legend. It has a filter function. You can filter the track points to a minimum spacing, minimum time interval of maximum number of points. Your software may have a similar function.

I have found I do not like too many waypoints on my route. It just makes finding the important ones, like trail junctions, more difficult to identify. You do sacrifice a little in the distance displays.

 

[wally2q]

tracks vs. plots

Just to clarify a few points.

A "TRACK" is a so-called bread-crumb trail, a set of position points stored in memory of the unit as you hike. The feature must be turned on in order for the track to be stored. The Rino110 can store 2048 points. These points can then be downloaded to your computer. These points can also be used in the unit for establishing a route (say if you want to "retrace your steps"), but only in "track following mode" which has certain limitations.

A "ROUTE" is a set of waypoints, which establish a path that you will follow in the future. You set the points using the unit, or using a mapping tool, upload the points to the unit, and then turn on "route following" mode. The Rino100 can store 500 waypoints per route.

Certain software can take a track log, and "Parse it" into fewer points, depending on maximum confugurable allowance for track point vs. position deviation. One such program, which I consider to be the best, is OZI EXPLORER. The only problem is that it may not be able to read your map files, depending on the format.

you may be able to find something else on the web - maybe a simple GPS tool, shareware style.

good luck......

the wall.....

 

[mavs00]

Must be a Rino thing. I just (within the day or so) downloaded over 3000 tracklog points (not waypoints, as stated, those are different) into my NG program from my eTrex Vista for my last bushwhack (up Lewey Peak). My "track" intervals were ridiculously close, seeing as I had the unit operating for about 5 out of the 8 we hiked.

I calculated 5280 (feet) x 5 (miles) divided by 3000 (track points) and it gave me a track point interval of 9 feet for the whole time my unit was operating. Needless to say, connecting the dots gave me a pretty darn acurate route, probably much more acurate than the map it displayed it on. I bet if I zoomed close enough, I'd be able to see where I veered off course to "melt some snow"

Don't ask my how I did it, I'm a total idiot on the computer. I just plugged the unit in and told it to download "routes, tracks and waypoints", and the program did the rest. I use a Mac OS X platform too.

 

[Neil]

Since first posting I have fiddled around and read a couple of things. It seems my problem was that I was attempting to take someone’s tracklog and put it into my gps as a route.
Furthermore, in NG Topo! it is possible to configuring the gps settings to accept an upload up to 3000 trackpoints per log.
In spite of understanding the above comments differentiating a trackpoint from a waypoint they are indistinguishable when displayed in the “gps waypoint list “. The only difference I can see is in the way they are entered into the gps.

In order to convert a tracklog consisting of thousands of points into a route with 20 or 30 key waypoints a program must surely exist that will downsample the original tracklog. I tried GPS Utility from GPSU but it doesn’t seem to have this feature.

 

[arghman]

tangential issue:
anyone know of any GPS's which have a "real" interface to organizing/downloading waypoints? I have about 300 waypoints now on my Garmin eTrex Legend and it is getting to be a pain to manage them. They all share the same namespace (if I call a waypoint "oak01" and there is already an "oak01" from a completely different area I've marked, I can't organize these into separate "folders"/"projects" whatever), and you have no ability to selectively download waypoints to a PC. If I mark 5 new waypoints, I have to download all 305 instead of the most recent 5, or the 15 waypoints within one mile of point X. (And I need to keep all those waypoints on the GPS for future use.)

I hate it when consumer electronics is limited not by hardware functionality but by software that could be powerful & extensible if the manufacturer felt like putting in a decent effort.

 

[spaddock]

In order to convert a tracklog consisting of thousands of points into a route with 20 or 30 key waypoints a program must surely exist that will downsample the original tracklog. I tried GPS Utility from GPSU but it doesn’t seem to have this feature.

GPSU's full version lets you upload an unlimited number of waypoints. The trial version limits you to 500. PM me for more details.


-Shayne

 

[Mark]

I use GPS TrackMaker (http://www.gpstm.com) free version. You can upload and download waypoints, tracks and routes. I use GPS TrackMaker to manage everything. Then I can just wipe out everything on my GPS and selectively upload whatever I need.

GPS Trackmaker does not come with any maps, but you can scan in maps that you need and use them in the program.

 

[wally2q]

Downsampling & Waypoint organization

Neil,

OZI Explorer does exactly that - down sampling. The function is called: Track Filter. It allows you to set the Min and Max "leg distance", as well as the Max Deviation in distance and Max Deviation in angle / direction. Super simple!


Arghman,

Once again, the answer is Ozi Explorer. It allows you to save "groups of waypoints" on a PC. You download them all, and then parse them into groups and save them on your computer. You can selectively upload certain points or groups of them into your GPS. Unfortunately, the GPS does not have a directory listing for points, so the only thing you can do is create name prefixes, so that in the alphanumeric order, they would be grouped. For example:
1-Oak01
1-Bluesky
1-Pt2
....
2-Oak01
2-waypoint3
2-waypoint4
....

The names can be easily edited in Ozi Explorer - much easier than on the GPS unit itself.

good luck.......

the wall.......

 

[Neil]

For what I specifically set out to do, ie. put someone else's hike into my GPS, I'm thinking that maybe my workaround is the simplest method after all.
Here it is again:
Import tracklog into mapping software.
Draw a freehand route over top of tracklog.
Delete the tracklog. Do a save as if you want to keep tracklog.
Create a GPS route with the software (using select criteria).
Export (upload) into the GPS.
Go hiking!

As for organizing all my waypoints and such, I'll cross that bridge when I find its coordinates.

 

[wally2q]

whatever works

at the end of the day - whatever works..... works........

the wall........

 

[arghman]

Arghman,

Once again, the answer is Ozi Explorer. It allows you to save "groups of waypoints" on a PC. You download them all, and then parse them into groups and save them on your computer. You can selectively upload certain points or groups of them into your GPS. Unfortunately, the GPS does not have a directory listing for points, so the only thing you can do is create name prefixes, so that in the alphanumeric order, they would be grouped. For example:
1-Oak01
1-Bluesky
1-Pt2
....
2-Oak01
2-waypoint3
2-waypoint4
....

The names can be easily edited in Ozi Explorer - much easier than on the GPS unit itself.

Thanks for the suggestion, it sounds like it has some interesting features that I don't have w/ my existing software (EasyGPS, works pretty well for shareware. I like it because it stores data in plaintext (actually in XML) and is very easy to manipulate the data in scripts. (I'm a hacker-type person.) Once I get the data on my PC, it's a snap.)

But my complaint is more with the GPS itself, again, the two pain factors (aside from entering text within the GPS, that's going to be hard no matter what as long as all you have is a "joystick") are keeping different sets of data organized on the GPS itself, and being able to download just a portion of the GPS's waypoints. Can't solve that with PC software if the weak link is the GPS receiver unit. It would make my life a whole lot easier if I could tell my GPS to put its waypoints into a particular folder/project (e.g. "Mountain Road Sitewalk" or "Japanese knotweed survey") to give the waypoints a particular context.

Some of the newer GPS's have SD cards, that would help somewhat in that the transfer rate is so quick that who cares whether I download all waypoints. But you still have the problem of keeping your GPS's data & the PC data in sync, when you can modify them on either end.

Garmin Legend also does not seem to store date/time of when waypoints are gathered. (or at least, I haven't figured out how to download that part of the data yet.) That's a big minus for me, some of the things I do w/ waypoint data would be a lot easier if I had the date/timestamp there also.

 

[arghman]

Ideally there must be a way to select every nth wp (downsample)from a list of wp's.
Does anybody out there have a solution?

What format is the data? if it's in a plaintext format, I could whip up a script pretty quickly to do that.

 

[Neil]

What format is the data? if it's in a plaintext format, I could whip up a script pretty quickly to do that.

Its a .txt file and its right here:
http://alavigne.net/newHomePage/FileDownloads/index.jsp?dirselection=GPSData/Adirondacks/Sewards2004

 

[imarchant]

.....Garmin Legend also does not seem to store date/time of when waypoints are gathered. (or at least, I haven't figured out how to download that part of the data yet.) That's a big minus for me, some of the things I do w/ waypoint data would be a lot easier if I had the date/timestamp there also.

The Garmin legend records the time on TRACKPOINTS not waypoints. (See the earlier post be wally2q for the diffrence) They can be viewed with the Garmin Mapsource software.

 

[arghman]

Its a .txt file and its right here:
http://alavigne.net/newHomePage/FileDownloads/index.jsp?dirselection=GPSData/Adirondacks/Sewards2004

Great, got it, done. I'll PM you with my email address & you can let me know where to send the processed file to.

If you need to do it again for files of this format, here's a short "awk" script which should do what you want. Save this as "downsample.awk" and run "gawk -f downsample.awk textfile.txt >textfile_d.txt" in a Command Prompt on Windows or a shell in unix. "gawk" is Gnu AWK and for Windows you can get the .exe file in UnxUpdates.zip from http://unxutils.sourceforge.net/

BEGIN {
T_number = 0;
T_downsample = 10;
}

# Show every N'th row starting with "T"
# where N=T_downsample
$1 == "T" {
if (T_number % T_downsample == 0)
print $0
T_number++;
}
$1 != "T" { print $0; }

 

[DougPaul]

While this awk script will reduce the number of points in a track, it is hardly optimal. (In signal processing this is called subsampling, the errors are known as aliasing.)

For example, let us assume one has a many-point track going north to a point, then going east (ie two straight lines). Now reduce the track to 3 points by selecting 3 from the original: one at the beginning, one in the middle, and one at the end. Depending on which point is chosen for the middle point, the resultant track can range from exact to rather inaccurate.

Sub-sampling will tend to round off corners (depending on which points happen to be chosen), but if it happens to select an outlying (erroneous?) point (frequently obvious in the original) it can yield a misleading track.

Optimal choice of which points to use in subsampling is more complex and difficult.

Not trying to suggest that you not use the awk scrpit, just pointing out its limitations.

Doug

 

[arghman]

Optimal choice of which points to use in subsampling is more complex and difficult.

Not trying to suggest that you not use the awk scrpit, just pointing out its limitations.

Right, call what I posted algorithm#0, algorithm#1 would do something like print out a point if it's a distance X from the last printed-out point or if the past N-1 points have not been printed out, algorithm #2 could calculate curvature of the track & print out more points when the route gets more squiggly vs. when it's straight. (I've been working on something like this for parametric function graphing, a very similar problem)

all a question of how much effort you want to put into solving the problem & what the customer wants/needs.

 

[DougPaul]

Right, call what I posted algorithm#0, algorithm#1 would do something like print out a point if it's a distance X from the last printed-out point or if the past N-1 points have not been printed out, algorithm #2 could calculate curvature of the track & print out more points when the route gets more squiggly vs. when it's straight. (I've been working on something like this for parametric function graphing, a very similar problem)

all a question of how much effort you want to put into solving the problem & what the customer wants/needs.

Ayup!

Iterative, genetic, or simulated annealing algorithms come to mind. Then how does one define optimum?

Doug