Road slope from DGPS sites [vectors]

James_Cameron4 · February 1, 1999, 3:15am

G'day,

I have data from my DGPS survey of a set of tracks on my farm, which
includes a series of points with altitude (ASL) as I walked along the
tracks. I want to prepare a colour map showing the tracks in colours
according to the slope of the track in the direction of travel. The
colours would change along a track.

This is so that I can work out which bits of the track need work.

I've moved the data into a sites file, with the ASL properly formatted
such that I can use s.whatever.to.rast to convert to a raster. I've
then been able to spread the raster slightly before using r.slope.aspect
to generate a raster of the slope in degrees. But of course, as you
move away from the track the numbers fall off ...

So I think I've done it wrongly. How should it be done?

What I get looks like this;

  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 4 5 5 4 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0

What I want should look like this;

  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 5 5 5 5 5 5 5 5 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0

Background information;

- the DGPS data set was from a Trimble unit, with 95% within
30cm accuracy,

- the raster cell size is 5m x 5m,

  - there are often multiple DGPS samples within a cell, because it
    was a reasonably hot day and I walked slowly with all that
    equipment on my back,

  - the resultant tracks drawn on a raster need to be enlarged to
    aid readability, I have not yet chosen the size, though I used
    eight cells in my example above.

--
James Cameron (cameron@stl.dec.com)

OpenVMS, Linux, Firewalls, Software Engineering, CGI, HTTP, X, C, FORTH,
COBOL, BASIC, DCL, csh, bash, ksh, sh, Electronics, Microcontrollers,
Disability Engineering, Netrek, Bicycles, Pedant, Farming, Home Control,
Remote Area Power, Greek Scholar, Tenor Vocalist, Church Sound, Husband.

"Specialisation is for insects." -- Robert Heinlein.

Angus_Carr1 · February 1, 1999, 1:33pm

Out of curiosity, what did you use to "spread the raster slightly"?

What I would do, personally, would be to run a series of r.mapcalc
commands, using mapcalc's ability to refer to cells by relative location.
In essence, this becomes a series of laplacian filters of this form:

0 1 0
0 0 0
0 0 0

0 0 0
0 0 1
0 0 0

0 0 0
0 0 0
0 1 0

0 0 0
1 0 0
0 0 0

repeated until the growth is sufficient.
so:
if mapA is the original with a single line which potentially has holes
(where you jumped 10 metres very quickly because of a bunch of horseflies,
presumably , then this series of mapcalcs will do it...

tempA = mapA

#then loop through the following until you feel you are done...

tempB = if(tempA == 0,tempA[0][-1],tempA)
tempA = tempB
tempB = if(tempA == 0,tempA[1][0],tempA)
tempA = tempB
tempB = if(tempA == 0,tempA[0][1],tempA)
tempA = tempB
tempB = if(tempA == 0,tempA[-1][0],tempA)
tempA = tempB

This should be set up in a shell script.
I have used this for filling in holes in a classified satellite image that
has been clumped and seived to drop polygons that are too small to manage
for. It preserves any existing pixels while changing 0's to the nearest
known value. It tends to preserve shape. To remove some bias, you might
want to start the cycle of four at a random place, but make sure you do
all four directions every time. The overall effect is like a wound
scabbing in from the outside.

Anyway, good luck.
Angus.

On Mon, 1 Feb 1999, James Cameron wrote:

G'day,

I have data from my DGPS survey of a set of tracks on my farm, which
includes a series of points with altitude (ASL) as I walked along the
tracks. I want to prepare a colour map showing the tracks in colours
according to the slope of the track in the direction of travel. The
colours would change along a track.

This is so that I can work out which bits of the track need work.

I've moved the data into a sites file, with the ASL properly formatted
such that I can use s.whatever.to.rast to convert to a raster. I've
then been able to spread the raster slightly before using r.slope.aspect
to generate a raster of the slope in degrees. But of course, as you
move away from the track the numbers fall off ...

So I think I've done it wrongly. How should it be done?

What I get looks like this;

  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 4 5 5 4 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0
  0 0 0 0 0 1 2 2 3 4 4 3 2 2 1 0 0 0 0 0

What I want should look like this;

  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 5 5 5 5 5 5 5 5 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0
  0 0 0 0 0 0 4 4 4 4 4 4 4 4 0 0 0 0 0 0

Background information;

  - the DGPS data set was from a Trimble unit, with 95% within
    30cm accuracy,

  - the raster cell size is 5m x 5m,

  - there are often multiple DGPS samples within a cell, because it
    was a reasonably hot day and I walked slowly with all that
    equipment on my back,

  - the resultant tracks drawn on a raster need to be enlarged to
    aid readability, I have not yet chosen the size, though I used
    eight cells in my example above.

--
James Cameron (cameron@stl.dec.com)

OpenVMS, Linux, Firewalls, Software Engineering, CGI, HTTP, X, C, FORTH,
COBOL, BASIC, DCL, csh, bash, ksh, sh, Electronics, Microcontrollers,
Disability Engineering, Netrek, Bicycles, Pedant, Farming, Home Control,
Remote Area Power, Greek Scholar, Tenor Vocalist, Church Sound, Husband.

"Specialisation is for insects." -- Robert Heinlein.

Randall_H_Craig · February 1, 1999, 8:06pm

check out:

http://www.baylor.edu/~grass/5.0betarelease.html

James_Cameron4 · February 1, 1999, 10:29pm

Angus Carr wrote:

Out of curiosity, what did you use to "spread the raster slightly"?

I can't remember, and it's on the home system, and I'm at work right
now, but I think it was a s.surf.something masked by a single-value
(0/1) raster made with r.grow. Very hazy memory there.

What I would do, personally, would be to run a series of r.mapcalc
commands, using mapcalc's ability to refer to cells by relative
location.

I'm afraid I don't quite understand yet what this means, but I know how
to take the r.mapcalcs and script them. I'll look at what it does,
thanks.

But my question remains, I think ... is the method I used reasonable or
should I be thinking of another method?

--
James Cameron (cameron@stl.dec.com)

OpenVMS, Linux, Firewalls, Software Engineering, CGI, HTTP, X, C, FORTH,
COBOL, BASIC, DCL, csh, bash, ksh, sh, Electronics, Microcontrollers,
Disability Engineering, Netrek, Bicycles, Pedant, Farming, Home Control,
Remote Area Power, Greek Scholar, Tenor Vocalist, Church Sound, Husband.

"Specialisation is for insects." -- Robert Heinlein.