Re: [NTLK] Help with GPS MAP process (VERY,VERY LONG POST)

From: Edward Johnson (radar9999_at_aol.com)
Date: Sat Apr 09 2005 - 22:58:02 PDT


OKAY, After playing with NTK for a couple of hours, I figured out that
the Make Pix Family error goes away when I set the Layout Preferences
for the 2000 instead of the messagepad. I still have not figured out
the file open error noted earlier. I worked on several iterations of
the calibration correction factor problems and cannot get rid of the
NAN's. I will illuminate;
RAW DATA to be input into 'RECT 2 application
Calibration Point 3 (Upper Lft ref pt) --La Luna Dr & Unk dead end road
Lat 34 26 17 (North)
Long -119 17 17 (West)
Pixel coordinates (r/d) 90/29 (vertex is upper left corner of
map-default for Graphic Converter)

Calibration Point 1 (Lft ref pt) -rte 33 & villanova rd
Lat 34 25 33 (North)
Long -119 17 17 (West)
Pixel coordinates (r/d) 91/387

Calibration Point 2 (Lower Rt ref point) villanova rd & unk rd
Lat 34 25 46 (North)
Long -119 16 21 (West)
Pixel coordinates (r/d) 454/276

The raw geographic data was confirmed correct. The Pixel coordinates
are referenced from the vertex at the upper left corner of the graphic
resource. IE movement to the right increases the X ordinate while
downward movement increases the Y ordinate. Shoot me now if Photo$hop
is referencing these coordinates from another location. Graphic
converter does not appear to allow you to change this vertex location
either. The above data is plugged into 'RECT 2 and this is what I get.
(the correction factors are at the bottom).

*******************************************************************
* *
* Coefficient Calculator for Rectangular Projection Type Maps *
* (working with freely selected Calibration Points) *
* V 1.5.1, for use with GPS Map *
* Copyright 1995-98 G. Staudenmaier, Germany *
* (Results are also stored to Clipboard) *
* *
*******************************************************************
Latitude : 34 25 33
Longitude : -119 17 17
  ->X: 91
  ->Y: 387
Latitude : 34 25 46
Longitude : -119 16 21
  ->X: 454
  ->Y: 276
Latitude : 34 26 17
Longitude : -119 17 17
  ->X: 90
  ->Y: 29

(Exit Loops by entering E<CR>)

*** Test Coefficients: Geo -> Pixel ***
Latitude : 34 25 45
Longitude : -119 16 30
X = -2147483648
Y = 285

                       C_X : [ NaN,
NaN, NaN],
                       C_Y : [ 9.6866552029E+05,
-2.9290909091E+04, -3.3603896104E+02],
                       C_Long : [ NaN,
NaN, NaN],
                       C_Lat : [ NaN,
NaN, NaN],
                       A : NaN,

As you can see the X-pixel coordinate for the test point is an unusual
negative number rather than being the expected 401. Why? (Scratch,
Scratch, yawnnnn) Okay try again tomorrow. I really do not want to find
all the ways to do this wrong before I discover the right way.

Ed Johnson

A Smart man learns from his mistakes. A Wise man learns from other
peoples mistakes. Be Wise!
On Apr 6, 2005, at 2:28 PM, newtontalk_at_newtontalk.net wrote:

> From: "Johannes Wolf" <mp2100_at_mail-gw.estec.esa.int>
> Subject: Re: [NTLK] Help with GPS MAP process (VERY,VERY LONG POST)
> Date: Wed, 6 Apr 2005 15:00:06 +0200
>
> YES, but the NaNs are just the next problem ...
>
> Cheers,
> Johannes
>
>> -----Original Message-----
>> From: newtontalk-bounce_at_newtontalk.net
>> [mailto:newtontalk-bounce_at_newtontalk.net]On Behalf Of Michael Briley
>> Sent: Wednesday, April 06, 2005 7:39 AM
>> To: newtontalk_at_newtontalk.net
>> Subject: Re: [NTLK] Help with GPS MAP process (VERY,VERY LONG POST)
>
>>
>> Dear Ed,
>
>> John is essentially correct. To use the MakePixFamily function, you
>> must have NTK version 1.6.4 (Mac) or 1.6.1 (Windows), and you must
>> use the Newton 2.1 platform Project Setting. Look in the Project menu
>> under Project Settings (at the bottom). Pick the Project Setting icon
>> and select "Newton 2.1" as the Platform.
>
>> But also note you have other troubles. When you look at the output of
>> the calibration generator (as shown in your listing of afterScript),
>> you have several NaNs. This is bad as NaN means "Not a Number."
>> Something has gone wrong with the calibration (usually a floating
>> point overflow/underflow). What, I don't know because I don't know
>> what your inputs were. Just remember it asks for latitude (N/S)
>> first, then longitude (E/W), and east longitudes are positive, west
>> longitudes (e.g., California) are negative.
>
>> Cheers,
>> -Mike

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/


This archive was generated by hypermail 2.1.5 : Mon Apr 11 2005 - 06:30:01 PDT