GPS Testing Project

GPS Testing Project

 Well, I have a GPS, and I have a new computer, so lets test this all out.  This has been compiled from emails that went out to the swasc mailing list.  I am trying this out in the Book page format to see how it looks.

Wednsday the 28th of September 2005:-

Well, it's really the heart of my computer controlled boat, but for the moment, it's my GPS PC!

The wires off the left, are the power on and reset switches, HD and power on LED's.

12V Battery power comes in is at the grey terminator block in the middle. 12V Power is jumpered back out for the GPS.

Turn on the power, and 20 seconds later, it's automatically recording the GPS reading once every second into a file on a RAM drive.  This is then saved onto the CF card when it shuts down.

and later... 

Well, here is some good(?) data I collected this afternoon.  There I was walking around with a yellow PC board strapped to my GPS pole.  :-)

It was an interesting exercise, comparing the data, with where I walked! Just a few side notes.  North is to the right.  At interesting places i.e.. Corners, interesting objects, I stopped and took 20 to 60 readings, with the idea that I should be able to average out the readings.  Reading around the Internet, some people say if you want to do that, you should do it for 20 minutes!  I am not that patient!

I have added some purple note points that correlate with the notes and picture  below:-

1. well, what went wrong here!  My guess is that as I turned it on inside, it collected some spurious data points.  I will try pruning them out later.

2.  I was walking down the back fence line, keeping the GPS within 10cm of the fence line.  Interesting to see that it started to waver around this point.

3. A big data error here.  I was initially going to walk around the perimeter of the park, following the curb.  A bit further on, I realised that the PC felt cold, and that I didn't know how long the battery was going to last (it was a very old UPS battery), so I cut across the park, to check out the PC status, back at GPS central.

4. The screen saver was running, so all being well, so I headed back out. This time, I walked down the track, close to the back fence.  I then walked to a Shire pump, and then walked 1.5 times the perimeter of a large peppi tree.  There is a table there, so I stood still for 30 seconds hoping to average out some readings, but instead the GSP went for a virtual walk! Hence the spur heading to the left.

5. yes, this is me just wandering around the corner area looking for a marker peg (that turns out doesn't exist).  I did the same thing after marking the children's play area (bright white/yellow pixels close by). Then I eventually got back to walking the curb down to # 6.

6.  I stopped here for ~ 60seconds, notice how it wandered a bit.  Using the MyGPS program I can zoom in and see each error, very clearly.  From here I walked back to the fence line, and walked it back to our back gate, and into home again.

It's interesting to note here that in general, walking East - West (down-up), has a lot less errors than walking North - South (right-left). Hmm.  I will have to test this out some more at a later stage.  Time to get my web site up and running, instead of posting all these emails.

and later... 

after spending an hour trying to locate all the crappy data, I manged to clean it up somewhat. 

Monday the 26th of September 2005:-

 

and later... 

Sunday the 25th of September:-

Cool!

I have used FlashEarth to grab three large images of my area.  I used GIMP to carefully stitch them together.  I then noted down the coordinates that FlashEarth gave me on some of the surrounding roads junctions.  I then used those to calibrate the GIMP bmp image I had loaded into MyGPS.

I then manually translated the GPS readings into decimal lat/long and manually plotted them into the Parkview MyGPS project.  The calibration is ~ 21m out, but hey, short of asking the GIS guys at work (which I will do in two weeks time), I don't know of another way to get an accurate fix on two points in my neighbourhood.

Looking at the largest spread of data in my 1.25Hr dump of GPS data, it was pretty accurate, in that MyGPS says the BY1 and BY4 are only 2.8m apart, as the crow flies ;-)

To view the data, you have to open MyGPS, then open the map:  Parkview.MyGPS  then Import the points from file:  Parkview_backyard.txt You will have to zoom into about 1000%  (zoom 400%, then press F1 key, to end up at 1000%).

This is absolutely awesome  <very big grin>

These are the programs I initialy used:  FlashEarth     MyGPS

My next step will be trying to automate the lat/long data from the GPS data file (generated by the C program I wrote), data conversion and dumping it in a format the MyGPS can accept.  I can then import thousands of data points in one go.  :-)

Just imagine how this will look with the ASC Mark I journeying around the Bunbury estuary's.

It's been a great GPS-GIS day.

Wednsday the 21st of September 2005:- 

First, I spent a few nights coding up a program to accept data from the VP-200 GPS.