|
TrekBuddy www.trekbuddy.net Outdoor companion.
|
- internal / bluetooth / simulator GPS
- offline raster maps
- smart GPX / raw NMEA logs
- waypoints and simple navigation
- custom views
|
- MIDP and Symbian phones
- Blackberry
- Android
|
Visit wiki to see all features, guides and howtos.
Project tracker.
|
|
|
|
|
|
|
View previous topic :: View next topic |
Author |
Message |
kruch Site Admin
Joined: 02 Jul 2006 Posts: 5676
|
Posted: Thu Feb 18, 2016 3:38 pm Post subject: |
|
|
MMPLL are always WGS84 lat/lon, or specified datum? Not sure...
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Thu Feb 18, 2016 3:50 pm Post subject: |
|
|
kruch wrote: |
MMPLL are always WGS84 lat/lon, or specified datum? Not sure... |
Good question...
And another question,
Why does Ozi give:
Code: |
Point01,xy, 0, 0,in,
Point02,xy, 5119, 0,in,
Point03,xy, 5119, 5119,in,
Point04,xy, 0, 5119,in,
|
for Points but give:
Code: |
MMPXY,1,0,0
MMPXY,2,5120,0
MMPXY,3,5120,5120
MMPXY,4,0,5120
|
for MMPXY points?
In Mobac it is:
Code: |
MMPXY, 1, 0, 0
MMPXY, 2, 5119, 0
MMPXY, 3, 5119, 5119
MMPXY, 4, 0, 5119
sbMap.append(String.format(mmpxLine, 1, 0, 0));
sbMap.append(String.format(mmpxLine, 2, width - 1, 0));
sbMap.append(String.format(mmpxLine, 3, width - 1, height - 1));
sbMap.append(String.format(mmpxLine, 4, 0, height - 1));
|
Which gives better results...
|
|
Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 6030
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Fri Feb 19, 2016 7:00 am Post subject: |
|
|
Ranger wrote: |
What it was for sample coordinates ?
|
What sample coordinates?
|
|
Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 6030
|
Posted: Sat Feb 20, 2016 10:37 am Post subject: |
|
|
On previous page we can see sets of coords in Northing/Easting as well as Lat/Long.
Question was about ones from OziE, where you have coords value, but we know nothing about map picture precision MM1B (in meters per pixel).
Calibration precision in OziE couldn't be mostly better than map itself is
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Sat Feb 20, 2016 6:34 pm Post subject: |
|
|
It is still the same test map (5x5 tiles) that you renamed to A10. The northing and easting values are calculated from the tile numbers. I used these values in OziExp to calibrate the map. The lat/lon values were calculated by OziExp when the calibration file was saved. So OziExp gets different lat/lon values than the Karney method using the same input.
|
|
Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 6030
|
Posted: Sun Feb 21, 2016 10:03 pm Post subject: |
|
|
Yep, in topic
http://www.trekbuddy.net/forum/viewtopic.php?t=8121&start=4
you attached two S 10 maps with calibration files, one (Test3m) as original from Mobac where MM1B is 146.488825 (means more than 146m per pixel precision), second modified (Test3) have MM1B 16.000072 m/pixel. Difference of MM1B comes from ?
JNT wrote: |
Perhaps it is accurate enough, but my preliminary tests show that I am about 30m off compared to a known point... |
It explains how original map picture was changed, but you should consider that this way map own errors can be enlarged too
If 30m error is compared with original calibration it means less than 1 pixel in first case, or can be 2 pixels for second
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Mon Feb 22, 2016 7:23 am Post subject: |
|
|
Ranger wrote: |
Difference of MM1B comes from ?
|
The first map comes directly out of mobac and it still treats it as spherical Mercator so it is in completely the wrong place and at the wrong scale. The second one was calibrated afterwards using OziExplorer (if I remember correctly). For this tile level it is exactly 16m/pixel.
The solution to get the correct calibration is to always use the upper left corner to calculate the northing and easting values. Then to use the Karney method to calculate the lat /lon values (using WGS84 as the datum)
I have now modified mobac to do this during atlas creation. The result seems to be perfect, but I will do more testing before I can say for sure...
|
|
Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 6030
|
Posted: Mon Feb 22, 2016 10:13 am Post subject: |
|
|
Exactly
|
|
Back to top |
|
 |
drivetrdr
Joined: 29 Sep 2018 Posts: 8
|
Posted: Wed Oct 03, 2018 3:34 am Post subject: |
|
|
If I omitted MMPLL, MMPXY, how many points TB can use from Point01,.. section?
Can I ommite MMPLL, MMPXY but not delete this data from file? Can I comment this lines?
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Wed Oct 03, 2018 7:15 am Post subject: |
|
|
kruch wrote: |
JNT wrote: |
So the MMPLL lines must also be there? Do they have to give lat/lon values? |
Unfortunately yes.
In fact, I just checked, shortened map format used for built-in world map only works because it uses lat-lon in PointXY lines, otherwise it would not work. And also because it uses default WGS 84 datum and does not need projection setup parameters.  |
It has been a long time since I struggled with this...
This was kruch's reply at that time. So, it depends on what projection and datum you are using, whether you need those MMPLL points or not.
|
|
Back to top |
|
 |
drivetrdr
Joined: 29 Sep 2018 Posts: 8
|
Posted: Wed Oct 03, 2018 3:16 pm Post subject: |
|
|
Hmm. I create map with MOBAC, so I think there is Mercator, WGS 84.
I deleted MMPLL, MMPXY, left only points section (4 points) and TB works fine. I have hard map and I need to calibrate it. So 4 corner points is not enaugh.
I need to know how many points TB can handle. I read Ozi wiki, there is about 100 points, so TB can handle all of them or it have points limitation?
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Wed Oct 03, 2018 3:31 pm Post subject: |
|
|
By "hard map" I presume that you have a paper map that you then scanned in and want to calibrate...
What software are you using to do this calibration? OziExplorer works well.
I'm not 100% sure about the limits, but I believe even 2 points would theoretically work. More is better. I normally used 4 points (as far apart as possible). I'm not sure there is any upper limit, but too many becomes unpractical.
|
|
Back to top |
|
 |
drivetrdr
Joined: 29 Sep 2018 Posts: 8
|
Posted: Wed Oct 03, 2018 10:47 pm Post subject: |
|
|
I learned ozi, but now I spent some time to my own calibration tool based on php script (Iam too lazy to learn other hard languages;), where I need to choose 2 coords (replace), then script will handle all maps files recursive. I dont needed in full raster map, and I can use sliced TB tiles. But for some calculations (XY global to raster) needs starting data (4 corner points) so I need save this data in .map files but not use it. This is why I asked about ability to comment MMPLL, MMPXY lines. The main idea is to make the calibration automatically, as possible.
Script is not done yet, but I work on it.
|
|
Back to top |
|
 |
JNT
Joined: 10 Jan 2009 Posts: 416 Location: Sweden
|
Posted: Thu Oct 04, 2018 9:44 am Post subject: |
|
|
Unfortunately, I don't really understand your use-case, and my knowledge of TrekBuddy is only based on using it for many years, so my usefulness in answering your questions is limited...
|
|
Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You can attach files in this forum You can download files in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|