| Re: Azimuth and Offset XYZ position correction [message #48013] |
Fri, 24 March 2006 15:08 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
Dave wrote:
> Plenty fast enough....next step is to try and get the elevation angle
> incorporated.
That's no big deal, if the distances you're measuring are small
compared to the radius of the earth (which seems pretty likely for
line-of-sight laser ranging to positions on the surface of the earth).
Just multiply the dx and dy formulas I gave you by
cos(elevation*!DTOR). Of course, you have to add
distance*sin(elevation*!DTOR) to the altitude of your known site to get
the altitude of the other site.
|
|
|
|
| Re: Azimuth and Offset XYZ position correction [message #48017 is a reply to message #48013] |
Fri, 24 March 2006 13:07  |
Dave[2]
Messages: 6
Registered: June 2005
|
Junior Member |
|
|
Plenty fast enough....next step is to try and get the elevation angle
incorporated.
Thanks again,
Dave
kuyper@wizard.net wrote:
> Dave wrote:
>> Oops, your right I forgot the angle, we do have that information as an
>> output. The scanner we use collects data at 250 measurements/second and
>> scanes 270 degrees vertical and 360 horizontal.
>>
>> Hypothetically speaking, if I was to feed ~5 million data points into
>> the methodology you described, what are the chances of it choking? Or
>> speed issues?
>
> The key step in that process was the map projections. I just tried it
> with 5 million randomly generated lat/lon pairs, and MAP_PROJ_FORWARD()
> took about 5 seconds to process them. Is that fast enough for your
> application?
|
|
|
|
| Re: Azimuth and Offset XYZ position correction [message #48020 is a reply to message #48017] |
Fri, 24 March 2006 06:12 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
Dave wrote:
> Oops, your right I forgot the angle, we do have that information as an
> output. The scanner we use collects data at 250 measurements/second and
> scanes 270 degrees vertical and 360 horizontal.
>
> Hypothetically speaking, if I was to feed ~5 million data points into
> the methodology you described, what are the chances of it choking? Or
> speed issues?
The key step in that process was the map projections. I just tried it
with 5 million randomly generated lat/lon pairs, and MAP_PROJ_FORWARD()
took about 5 seconds to process them. Is that fast enough for your
application?
|
|
|
|
| Re: Azimuth and Offset XYZ position correction [message #48022 is a reply to message #48020] |
Thu, 23 March 2006 16:42 |
Dave[2]
Messages: 6
Registered: June 2005
|
Junior Member |
|
|
Oops, your right I forgot the angle, we do have that information as an
output. The scanner we use collects data at 250 measurements/second and
scanes 270 degrees vertical and 360 horizontal.
Hypothetically speaking, if I was to feed ~5 million data points into
the methodology you described, what are the chances of it choking? Or
speed issues?
All in all though....great approach!!!
Dave
kuyper@wizard.net wrote:
> Dave wrote:
>> Hi All,
>>
>> Quick question for anyone out there. I am looking to write a script and
>> I'm wondering if anyone out there has done it already. I have a known
>> precise position, say X and Y (UTM coordinate) in meters. From that
>> precise position I am using a laser range finder to calculate a
>> distance measurement from my original XY location. Along with this
>> distance measure I obtain the exact azimuth.
>>
>> What I would like to do is calculate the precise XY location at the end
>> of the laser measurement.
>
> You don't mention the elevation angle of the laser finder. Without that
> information, the best we can do is assume that it's 0. I would imagine
> that the distances you measure are much smaller than the radius of the
> earth, which would allow the use of certain approximations, but I'm
> going to use a method that will give accurate results for any distance
> measured along a great circle on the surface of the earth all the way
> up to half the circumference of the Earth. First, set up the utm map
> projection:
>
> utm = MAP_PROJ_INIT('UTM', CENTER_LONGITUDE=clon, $
> CENTER_LATITUDE=clat, ZONE=z);
>
> ;If you don't know what the zone is for your UTM projection, you
> probably aren't using
> ;one, and you can drop that argument. Next, convert your known position
> to longitude
> ;and latitude:
>
> lonlat = MAP_PROJ_INVERSE(known, MAP_STRUCTURE=utm);
>
> ;Next, set up an azimuthal equidistant projection centered at your
> known position:
>
> azeq = MAP_PROJ_INIT('Azimuthal Equidistant',
> CENTER_LONGITUDE=lonlat[0], $
> CENTER_LATITUDE=lonlat[1]);
>
> ;The next step depends upon how you define azimuth - there is a
> standard convention; ;unfortunately, there are many different mutually
> incompatible conventions. 0 degrees
> ;can represent either due North, or due East. A positive azimuth can
> represent either a
> ;clockwise or a counterclockwise rotation, as seen from above. I will
> assume that due
> ;North is 0 degrees, and that due East is +90 degrees. Then calculate
> appropriate
> ;offsets:
>
> dx = distance*sin(azimuth*!DTOR)
> dy = distance*cos(azimuth*!DTOR)
>
> ;Now convert those offsets to a latitude and a longitude.
>
> newlonlat = MAP_PROJ_INVERSE(dx, dy, MAP_STRUCTURE=azeq);
>
> ;And finally, back to UTM XY coordinates:
>
> newxy = MAP_PROJ_FORWARD(newlonlat, MAP_STRUCTURE=utm)
>
> There are other ways to do this, and some are more efficient, but this
> one hides all of the spherical trig inside the map projection code,
> allowing you to concentrate on other issues.
>
> There should be a standard function to do this; it's sort-of the
> inverse of map_2points(). However, I couldn't find one using the
> built-in help.
|
|
|
|
| Re: Azimuth and Offset XYZ position correction [message #48024 is a reply to message #48022] |
Thu, 23 March 2006 15:42 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
Dave wrote:
> Hi All,
>
> Quick question for anyone out there. I am looking to write a script and
> I'm wondering if anyone out there has done it already. I have a known
> precise position, say X and Y (UTM coordinate) in meters. From that
> precise position I am using a laser range finder to calculate a
> distance measurement from my original XY location. Along with this
> distance measure I obtain the exact azimuth.
>
> What I would like to do is calculate the precise XY location at the end
> of the laser measurement.
You don't mention the elevation angle of the laser finder. Without that
information, the best we can do is assume that it's 0. I would imagine
that the distances you measure are much smaller than the radius of the
earth, which would allow the use of certain approximations, but I'm
going to use a method that will give accurate results for any distance
measured along a great circle on the surface of the earth all the way
up to half the circumference of the Earth. First, set up the utm map
projection:
utm = MAP_PROJ_INIT('UTM', CENTER_LONGITUDE=clon, $
CENTER_LATITUDE=clat, ZONE=z);
;If you don't know what the zone is for your UTM projection, you
probably aren't using
;one, and you can drop that argument. Next, convert your known position
to longitude
;and latitude:
lonlat = MAP_PROJ_INVERSE(known, MAP_STRUCTURE=utm);
;Next, set up an azimuthal equidistant projection centered at your
known position:
azeq = MAP_PROJ_INIT('Azimuthal Equidistant',
CENTER_LONGITUDE=lonlat[0], $
CENTER_LATITUDE=lonlat[1]);
;The next step depends upon how you define azimuth - there is a
standard convention; ;unfortunately, there are many different mutually
incompatible conventions. 0 degrees
;can represent either due North, or due East. A positive azimuth can
represent either a
;clockwise or a counterclockwise rotation, as seen from above. I will
assume that due
;North is 0 degrees, and that due East is +90 degrees. Then calculate
appropriate
;offsets:
dx = distance*sin(azimuth*!DTOR)
dy = distance*cos(azimuth*!DTOR)
;Now convert those offsets to a latitude and a longitude.
newlonlat = MAP_PROJ_INVERSE(dx, dy, MAP_STRUCTURE=azeq);
;And finally, back to UTM XY coordinates:
newxy = MAP_PROJ_FORWARD(newlonlat, MAP_STRUCTURE=utm)
There are other ways to do this, and some are more efficient, but this
one hides all of the spherical trig inside the map projection code,
allowing you to concentrate on other issues.
There should be a standard function to do this; it's sort-of the
inverse of map_2points(). However, I couldn't find one using the
built-in help.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
