Hi,

since I wasn't satiesfied with the result of lensfun's distortation  
correction (at short focus lengthes),

I approached it the problem a different way. To sum up the results are  
optimal - in a certain sense, see below.

This is what I did:

- prepare a sheet of paper with some labeled grid with thin lines

- take photographs of this at different focal lengthes

- IMPORT : adjust the FZ1000 to store both, the JPEG and the raw file  
(.RW2)

- ufraw-batch  --lensfun=none --out-depth=8 --out-type=jpeg  
--output=Test.RW2.jpeg Test.RW2

- open both, Test.JPG (from the camera) and Test.RW2.jpeg in Gimp
- note several coordinates of unique points in both images
- compute the radii of all points and scale them to half of the height  
of an image ( = 1.0/1824 for the FZ1000)

- using Octave I fitted a polynomial of the form    
r*(d+r*(c+r*(b+r*a))) to the data using a smoothed maximum norm

- check this polynomial by using convert (ImageMagick) e.g.
                                                                                
                                           
a            b              c          d
   convert Test.RW2.jpeg -define distort:viewport=5472x3648+8+8  
-distort Barrel 0.026153,-0.102825,0.021310,0.997913 New.jpeg

- open both, Test.JPG  and New.jpeg as layers in Gimp, desatured both  
and switch to difference layer mode

Result: this images are identical (except the intensities of gray  
tones). This means, that the polynom computed by me produces excatly  
the same
            lens correction as the camera (FZ1000) does it itself.

Unfortunately the coefficient d differs from 1 which is an assumption  
in lensfun.
Is there a way to use or patch lensfun to use all 4 parameters (e.g. a  
general d)?

Here are my results:

The focal lengthes here are scaled by the crop factor

focal length     d          c            b           a
     25         0.997913   0.0213096   -0.102825    0.0261528
     28         1.00083    0.019843    -0.0899886   0.0237594
     35         0.998356   0.0194224   -0.0581771   0.0137862
     50         1.00373   -0.00242727  -0.0103302   0.00130064
     70         1.0011     0.00645453  -0.0102541   0.00258168
     90         1.00838   -0.0245663    0.0221847  -0.00623619


Thanks for your comments,
Helmut


------------------------------------------------------------------------------
What NetFlow Analyzer can do for you? Monitors network bandwidth and traffic
patterns at an interface-level. Reveals which users, apps, and protocols are 
consuming the most bandwidth. Provides multi-vendor support for NetFlow, 
J-Flow, sFlow and other flows. Make informed decisions using capacity planning
reports. http://pubads.g.doubleclick.net/gampad/clk?id=1444514421&iu=/41014381
_______________________________________________
Lensfun-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lensfun-users

Reply via email to