For UTM, the conversion is complicated by the fact that UTM is projected differently in different areas (zones) so it's not a simple mathematical conversion algorithm.
If you are operating in a small enough area of a single zone it should be possible to use one of the various online converters to come up with enough points to interpolate. I know one of the prior AVC winners did bi-linear interpolation in a parking lot sized area as an approximation.
I suppose one could work out the level of error in the conversion for larger areas. I would probably try to find some library code in C or something that would do the conversion, and then work things out with Octave.
In my own robot I didn't use UTM, I generated my own cartesian system using 4 GPS points and translated the GPS coordinates to that. (Same parking lot). In this parking lot, the distances are so small that lat/lon are cartesian to a fairly decent precision.
Replies
I did a search just now and ultimately found this: http://geographiclib.sourceforge.net/ There are others out there I am sure.
I think tinygps will not do that but maybe it can help you a little?
http://arduiniana.org/libraries/tinygps/
Cheers
Kraut Rob
For UTM, the conversion is complicated by the fact that UTM is projected differently in different areas (zones) so it's not a simple mathematical conversion algorithm.
If you are operating in a small enough area of a single zone it should be possible to use one of the various online converters to come up with enough points to interpolate. I know one of the prior AVC winners did bi-linear interpolation in a parking lot sized area as an approximation.
I suppose one could work out the level of error in the conversion for larger areas. I would probably try to find some library code in C or something that would do the conversion, and then work things out with Octave.
In my own robot I didn't use UTM, I generated my own cartesian system using 4 GPS points and translated the GPS coordinates to that. (Same parking lot). In this parking lot, the distances are so small that lat/lon are cartesian to a fairly decent precision.
I have no idea what MGRS is or how it works. :)