ossim-correl question

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

ossim-correl question

William G. Miller
Greetings!

I am trying to co-register an aerial photograph with a DOQQ mosaic
without success. I have:

        1. converted the color DOQQ mosaic to grey scale
        2. resampled the DOQQ from 1 meter to 8 meters (the scale of the
photo).
        3. created a GeoTiff of the photo from its metadata. There is no
rotation but about 200 meters of translation easily seen when the photo
is swiped against thee DOQQ,

The photo was taken in 1981 and the DOQQ was created in 1994.

The following is an example of how I am approaching the problem.
However, I am unable to co-register the DOQQ with itself! The images
work fine with imagelinker. The harris corner filter seems to be working
with the data. Any suggestions would be appreciated.


William G. Miller, computer scientist
U.S. Geological Survey
12201 Sunrise Valley Drive

=============================================================


>>ossim-correl -r 1.0 -d 0.003 -t 7 -e 15 -c 0.8 -o gross_8m.xml
doq_8_tiled.tiff doq_8_tiled.tiff
Getting tie points...
100%
100%
Statistics for tie point displacements
number of tie points=0
Mean displacements: mx=nan my=nan
Correlation coeffs: Cx2=nan Cxy=nan Cy2=nan
RMS along eigen axes : rmax=nan rmin=nan
Major axis angle with x axis (degrees): nan
ossimCommon.cpp:118
Unhandled scalar type:  0
ossimCommon.cpp:178
Unhandled scalar type:  0
ossimCommon.cpp:118
Unhandled scalar type:  0
ossimCommon.cpp:178
Unhandled scalar type:  0
>>

=======================================================

ossim-info doq_8.tiff

image0.band0.max_value:  254
image0.band0.min_value:  1
image0.band0.null_value:  0
image0.decimal_degrees_per_pixel_lat:  7.20149868270585e-05
image0.decimal_degrees_per_pixel_lon:  9.20892684463113e-05
image0.decimations:  (1,1) (0.5,0.5) (0.25,0.25) (0.125,0.125)
(0.0626971608832808,0.0626971608832808)
(0.0315457413249211,0.0315457413249211)
(0.0157728706624606,0.0157728706624606)
image0.entry:  0
image0.gsd:  (7.99785997455651,7.9978597938936)
image0.ll_lat:  38.656134971513
image0.ll_lon:  -77.675727606177
image0.lr_lat:  38.6612329371964
image0.lr_lon:  -77.4428739621368
image0.lr_x:  2535
image0.lr_y:  2535
image0.meters_per_pixel_x:  8
image0.meters_per_pixel_y:  8
image0.number_decimation_levels:  7
image0.number_input_bands:  1
image0.number_lines:  2536
image0.number_output_bands:  1
image0.number_samples:  2536
image0.projection.central_meridian:  -75
image0.projection.datum:  WGE
image0.projection.elevation_lookup_flag:  0
image0.projection.ellipse_code:  WE
image0.projection.ellipse_name:  WGS 84
image0.projection.false_easting_northing:  (500000,0)
image0.projection.false_easting_northing_units:  meters
image0.projection.hemisphere:  N
image0.projection.major_axis:  6378137
image0.projection.minor_axis:  6356752.3142
image0.projection.origin_latitude:  0
image0.projection.pcs_code:  32618
image0.projection.pixel_scale_units:  meters
image0.projection.pixel_scale_xy:  (8,8)
image0.projection.srs_name:  EPSG:32618
image0.projection.tie_point_units:  meters
image0.projection.tie_point_xy:  (267168,4302296)
image0.projection.type:  ossimUtmProjection
image0.projection.zone:  18
image0.radiometry:  8-bit
image0.tie_point_lat:  38.8386929631196
image0.tie_point_lon:  -77.6825551954605
image0.type:  ossimImageGeometry
image0.ul_lat:  38.8386929631196
image0.ul_lon:  -77.6825551954605
image0.ul_x:  0
image0.ul_y:  0
image0.ur_lat:  38.8438241018836
image0.ur_lon:  -77.4491088999491
number_entries:  1



------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer
Reply | Threaded
Open this post in threaded view
|

Re: ossim-correl question

Hernán De Angelis
Hi William,

I would guess that your airphoto and mosaic have very different
quality and spatial and spectral resolutions. I do not mean pixel size
only, but the texture and how well the different details are resolved,
Also if the older airphoto is black-and-white and the more modern
mosaic is colour (i.e. 3 band or more) you will need to specify a
single channel in the color mosaic that has comparable spectral
resolution than the older airphoto. Beware as well of the radial
distortions typical of airphotos. If the surveys where made with
different cameras at different altitudes you are very likely to have
different images.
I would suggest that you first try filtering the images so to achieve
visually similar resolutions and textures. Then play with the
parameters -d and -e and see what happens.


Good luck,


Hernán


2009/12/15 William G. Miller <[hidden email]>:

> Greetings!
>
> I am trying to co-register an aerial photograph with a DOQQ mosaic
> without success. I have:
>
>        1. converted the color DOQQ mosaic to grey scale
>        2. resampled the DOQQ from 1 meter to 8 meters (the scale of the
> photo).
>        3. created a GeoTiff of the photo from its metadata. There is no
> rotation but about 200 meters of translation easily seen when the photo
> is swiped against thee DOQQ,
>
> The photo was taken in 1981 and the DOQQ was created in 1994.
>
> The following is an example of how I am approaching the problem.
> However, I am unable to co-register the DOQQ with itself! The images
> work fine with imagelinker. The harris corner filter seems to be working
> with the data. Any suggestions would be appreciated.
>
>
> William G. Miller, computer scientist
> U.S. Geological Survey
> 12201 Sunrise Valley Drive
>
> =============================================================
>
>
>>>ossim-correl -r 1.0 -d 0.003 -t 7 -e 15 -c 0.8 -o gross_8m.xml
> doq_8_tiled.tiff doq_8_tiled.tiff
> Getting tie points...
> 100%
> 100%
> Statistics for tie point displacements
> number of tie points=0
> Mean displacements: mx=nan my=nan
> Correlation coeffs: Cx2=nan Cxy=nan Cy2=nan
> RMS along eigen axes : rmax=nan rmin=nan
> Major axis angle with x axis (degrees): nan
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
>>>
>
> =======================================================
>
> ossim-info doq_8.tiff
>
> image0.band0.max_value:  254
> image0.band0.min_value:  1
> image0.band0.null_value:  0
> image0.decimal_degrees_per_pixel_lat:  7.20149868270585e-05
> image0.decimal_degrees_per_pixel_lon:  9.20892684463113e-05
> image0.decimations:  (1,1) (0.5,0.5) (0.25,0.25) (0.125,0.125)
> (0.0626971608832808,0.0626971608832808)
> (0.0315457413249211,0.0315457413249211)
> (0.0157728706624606,0.0157728706624606)
> image0.entry:  0
> image0.gsd:  (7.99785997455651,7.9978597938936)
> image0.ll_lat:  38.656134971513
> image0.ll_lon:  -77.675727606177
> image0.lr_lat:  38.6612329371964
> image0.lr_lon:  -77.4428739621368
> image0.lr_x:  2535
> image0.lr_y:  2535
> image0.meters_per_pixel_x:  8
> image0.meters_per_pixel_y:  8
> image0.number_decimation_levels:  7
> image0.number_input_bands:  1
> image0.number_lines:  2536
> image0.number_output_bands:  1
> image0.number_samples:  2536
> image0.projection.central_meridian:  -75
> image0.projection.datum:  WGE
> image0.projection.elevation_lookup_flag:  0
> image0.projection.ellipse_code:  WE
> image0.projection.ellipse_name:  WGS 84
> image0.projection.false_easting_northing:  (500000,0)
> image0.projection.false_easting_northing_units:  meters
> image0.projection.hemisphere:  N
> image0.projection.major_axis:  6378137
> image0.projection.minor_axis:  6356752.3142
> image0.projection.origin_latitude:  0
> image0.projection.pcs_code:  32618
> image0.projection.pixel_scale_units:  meters
> image0.projection.pixel_scale_xy:  (8,8)
> image0.projection.srs_name:  EPSG:32618
> image0.projection.tie_point_units:  meters
> image0.projection.tie_point_xy:  (267168,4302296)
> image0.projection.type:  ossimUtmProjection
> image0.projection.zone:  18
> image0.radiometry:  8-bit
> image0.tie_point_lat:  38.8386929631196
> image0.tie_point_lon:  -77.6825551954605
> image0.type:  ossimImageGeometry
> image0.ul_lat:  38.8386929631196
> image0.ul_lon:  -77.6825551954605
> image0.ul_x:  0
> image0.ul_y:  0
> image0.ur_lat:  38.8438241018836
> image0.ur_lon:  -77.4491088999491
> number_entries:  1
>
>
>
> ------------------------------------------------------------------------------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>



--

Hernán De Angelis
Linux user # 397217

------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer
Reply | Threaded
Open this post in threaded view
|

Re: ossim-correl question

j artieda

I'm getting exactly the same results,
I have also tried with the the same image in both sides, so the image is not very similar, it is the same image.
any clue?
regards,
Jorge


From: Hernán De Angelis <[hidden email]>
To: [hidden email]
Cc: [hidden email]
Sent: Fri, December 18, 2009 12:10:36 PM
Subject: Re: [OSSIM] ossim-correl question

Hi William,

I would guess that your airphoto and mosaic have very different
quality and spatial and spectral resolutions. I do not mean pixel size
only, but the texture and how well the different details are resolved,
Also if the older airphoto is black-and-white and the more modern
mosaic is colour (i.e. 3 band or more) you will need to specify a
single channel in the color mosaic that has comparable spectral
resolution than the older airphoto. Beware as well of the radial
distortions typical of airphotos. If the surveys where made with
different cameras at different altitudes you are very likely to have
different images.
I would suggest that you first try filtering the images so to achieve
visually similar resolutions and textures. Then play with the
parameters -d and -e and see what happens.


Good luck,


Hernán


2009/12/15 William G. Miller <[hidden email]>:

> Greetings!
>
> I am trying to co-register an aerial photograph with a DOQQ mosaic
> without success. I have:
>
>        1. converted the color DOQQ mosaic to grey scale
>        2. resampled the DOQQ from 1 meter to 8 meters (the scale of the
> photo).
>        3. created a GeoTiff of the photo from its metadata. There is no
> rotation but about 200 meters of translation easily seen when the photo
> is swiped against thee DOQQ,
>
> The photo was taken in 1981 and the DOQQ was created in 1994.
>
> The following is an example of how I am approaching the problem.
> However, I am unable to co-register the DOQQ with itself! The images
> work fine with imagelinker. The harris corner filter seems to be working
> with the data. Any suggestions would be appreciated.
>
>
> William G. Miller, computer scientist
> U.S. Geological Survey
> 12201 Sunrise Valley Drive
>
> =============================================================
>
>
>>>ossim-correl -r 1.0 -d 0.003 -t 7 -e 15 -c 0.8 -o gross_8m.xml
> doq_8_tiled.tiff doq_8_tiled.tiff
> Getting tie points...
> 100%
> 100%
> Statistics for tie point displacements
> number of tie points=0
> Mean displacements: mx=nan my=nan
> Correlation coeffs: Cx2=nan Cxy=nan Cy2=nan
> RMS along eigen axes : rmax=nan rmin=nan
> Major axis angle with x axis (degrees): nan
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
>>>
>
> =======================================================
>
> ossim-info doq_8.tiff
>
> image0.band0.max_value:  254
> image0.band0.min_value:  1
> image0.band0.null_value:  0
> image0.decimal_degrees_per_pixel_lat:  7.20149868270585e-05
> image0.decimal_degrees_per_pixel_lon:  9.20892684463113e-05
> image0.decimations:  (1,1) (0.5,0.5) (0.25,0.25) (0.125,0.125)
> (0.0626971608832808,0.0626971608832808)
> (0.0315457413249211,0.0315457413249211)
> (0.0157728706624606,0.0157728706624606)
> image0.entry:  0
> image0.gsd:  (7.99785997455651,7.9978597938936)
> image0.ll_lat:  38.656134971513
> image0.ll_lon:  -77.675727606177
> image0.lr_lat:  38.6612329371964
> image0.lr_lon:  -77.4428739621368
> image0.lr_x:  2535
> image0.lr_y:  2535
> image0.meters_per_pixel_x:  8
> image0.meters_per_pixel_y:  8
> image0.number_decimation_levels:  7
> image0.number_input_bands:  1
> image0.number_lines:  2536
> image0.number_output_bands:  1
> image0.number_samples:  2536
> image0.projection.central_meridian:  -75
> image0.projection.datum:  WGE
> image0.projection.elevation_lookup_flag:  0
> image0.projection.ellipse_code:  WE
> image0.projection.ellipse_name:  WGS 84
> image0.projection.false_easting_northing:  (500000,0)
> image0.projection.false_easting_northing_units:  meters
> image0.projection.hemisphere:  N
> image0.projection.major_axis:  6378137
> image0.projection.minor_axis:  6356752.3142
> image0.projection.origin_latitude:  0
> image0.projection.pcs_code:  32618
> image0.projection.pixel_scale_units:  meters
> image0.projection.pixel_scale_xy:  (8,8)
> image0.projection.srs_name:  EPSG:32618
> image0.projection.tie_point_units:  meters
> image0.projection.tie_point_xy:  (267168,4302296)
> image0.projection.type:  ossimUtmProjection
> image0.projection.zone:  18
> image0.radiometry:  8-bit
> image0.tie_point_lat:  38.8386929631196
> image0.tie_point_lon:  -77.6825551954605
> image0.type:  ossimImageGeometry
> image0.ul_lat:  38.8386929631196
> image0.ul_lon:  -77.6825551954605
> image0.ul_x:  0
> image0.ul_y:  0
> image0.ur_lat:  38.8438241018836
> image0.ur_lon:  -77.4491088999491
> number_entries:  1
>
>
>
> ------------------------------------------------------------------------------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>



--

Hernán De Angelis
Linux user # 397217

------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer


------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer
Reply | Threaded
Open this post in threaded view
|

ossim-correl question

j artieda

I have a problem with ossimNCC_FFTW. It doesn't get into the foor loops at line 141.
startx, starty, endx , endy are all equal to 44.
it maybe is something like for (i=0;i<endy; ++i)...

ossimNCC_FFTW lines 141:

   int startx = _mx-1;
   int starty = _my-1;
   int endx   = _sx-1;
   int endy   = _sy-1;
   double* pcc   = _NCC.refBuffer() + startx + starty * _pcx;
   for(i=starty;i<endy;++i)
   {
      for(j=startx;j<endx;++j)
      {
         //get std & average for master window over PADDED slave
         psi=i-_my;


From: j artieda <[hidden email]>
To: [hidden email]
Cc: [hidden email]
Sent: Wed, February 24, 2010 4:50:56 PM
Subject: Re: [OSSIM] ossim-correl question


I'm getting exactly the same results,
I have also tried with the the same image in both sides, so the image is not very similar, it is the same image.
any clue?
regards,
Jorge


From: Hernán De Angelis <[hidden email]>
To: [hidden email]
Cc: [hidden email]
Sent: Fri, December 18, 2009 12:10:36 PM
Subject: Re: [OSSIM] ossim-correl question

Hi William,

I would guess that your airphoto and mosaic have very different
quality and spatial and spectral resolutions. I do not mean pixel size
only, but the texture and how well the different details are resolved,
Also if the older airphoto is black-and-white and the more modern
mosaic is colour (i.e. 3 band or more) you will need to specify a
single channel in the color mosaic that has comparable spectral
resolution than the older airphoto. Beware as well of the radial
distortions typical of airphotos. If the surveys where made with
different cameras at different altitudes you are very likely to have
different images.
I would suggest that you first try filtering the images so to achieve
visually similar resolutions and textures. Then play with the
parameters -d and -e and see what happens.


Good luck,


Hernán


2009/12/15 William G. Miller <[hidden email]>:

> Greetings!
>
> I am trying to co-register an aerial photograph with a DOQQ mosaic
> without success. I have:
>
>        1. converted the color DOQQ mosaic to grey scale
>        2. resampled the DOQQ from 1 meter to 8 meters (the scale of the
> photo).
>        3. created a GeoTiff of the photo from its metadata. There is no
> rotation but about 200 meters of translation easily seen when the photo
> is swiped against thee DOQQ,
>
> The photo was taken in 1981 and the DOQQ was created in 1994.
>
> The following is an example of how I am approaching the problem.
> However, I am unable to co-register the DOQQ with itself! The images
> work fine with imagelinker. The harris corner filter seems to be working
> with the data. Any suggestions would be appreciated.
>
>
> William G. Miller, computer scientist
> U.S. Geological Survey
> 12201 Sunrise Valley Drive
>
> =============================================================
>
>
>>>ossim-correl -r 1.0 -d 0.003 -t 7 -e 15 -c 0.8 -o gross_8m.xml
> doq_8_tiled.tiff doq_8_tiled.tiff
> Getting tie points...
> 100%
> 100%
> Statistics for tie point displacements
> number of tie points=0
> Mean displacements: mx=nan my=nan
> Correlation coeffs: Cx2=nan Cxy=nan Cy2=nan
> RMS along eigen axes : rmax=nan rmin=nan
> Major axis angle with x axis (degrees): nan
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
> ossimCommon.cpp:118
> Unhandled scalar type:  0
> ossimCommon.cpp:178
> Unhandled scalar type:  0
>>>
>
> =======================================================
>
> ossim-info doq_8.tiff
>
> image0.band0.max_value:  254
> image0.band0.min_value:  1
> image0.band0.null_value:  0
> image0.decimal_degrees_per_pixel_lat:  7.20149868270585e-05
> image0.decimal_degrees_per_pixel_lon:  9.20892684463113e-05
> image0.decimations:  (1,1) (0.5,0.5) (0.25,0.25) (0.125,0.125)
> (0.0626971608832808,0.0626971608832808)
> (0.0315457413249211,0.0315457413249211)
> (0.0157728706624606,0.0157728706624606)
> image0.entry:  0
> image0.gsd:  (7.99785997455651,7.9978597938936)
> image0.ll_lat:  38.656134971513
> image0.ll_lon:  -77.675727606177
> image0.lr_lat:  38.6612329371964
> image0.lr_lon:  -77.4428739621368
> image0.lr_x:  2535
> image0.lr_y:  2535
> image0.meters_per_pixel_x:  8
> image0.meters_per_pixel_y:  8
> image0.number_decimation_levels:  7
> image0.number_input_bands:  1
> image0.number_lines:  2536
> image0.number_output_bands:  1
> image0.number_samples:  2536
> image0.projection.central_meridian:  -75
> image0.projection.datum:  WGE
> image0.projection.elevation_lookup_flag:  0
> image0.projection.ellipse_code:  WE
> image0.projection.ellipse_name:  WGS 84
> image0.projection.false_easting_northing:  (500000,0)
> image0.projection.false_easting_northing_units:  meters
> image0.projection.hemisphere:  N
> image0.projection.major_axis:  6378137
> image0.projection.minor_axis:  6356752.3142
> image0.projection.origin_latitude:  0
> image0.projection.pcs_code:  32618
> image0.projection.pixel_scale_units:  meters
> image0.projection.pixel_scale_xy:  (8,8)
> image0.projection.srs_name:  EPSG:32618
> image0.projection.tie_point_units:  meters
> image0.projection.tie_point_xy:  (267168,4302296)
> image0.projection.type:  ossimUtmProjection
> image0.projection.zone:  18
> image0.radiometry:  8-bit
> image0.tie_point_lat:  38.8386929631196
> image0.tie_point_lon:  -77.6825551954605
> image0.type:  ossimImageGeometry
> image0.ul_lat:  38.8386929631196
> image0.ul_lon:  -77.6825551954605
> image0.ul_x:  0
> image0.ul_y:  0
> image0.ur_lat:  38.8438241018836
> image0.ur_lon:  -77.4491088999491
> number_entries:  1
>
>
>
> ------------------------------------------------------------------------------
> This SF.Net email is sponsored by the Verizon Developer Community
> Take advantage of Verizon's best-in-class app development support
> A streamlined, 14 day to market process makes app distribution fast and easy
> Join now and get one step closer to millions of Verizon customers
> http://p.sf.net/sfu/verizon-dev2dev
> _______________________________________________
> www.ossim.org
> Ossim-developer mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/ossim-developer
>



--

Hernán De Angelis
Linux user # 397217

------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer



------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
www.ossim.org
Ossim-developer mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/ossim-developer