Prev: understanding imadjust function
Next: Array of structs. Column vector
From: bhavin on 9 Jul 2010 09:42
Did you manage to see the images?

I await your response
From: Image Analyst on 9 Jul 2010 10:11
So the small image is fully contained within the larger image? And it's just rotated, and at an unknown location?

That's a pattern recognition problem. I'm not too familiar with that field.
You might look at some of the techniques listed in sections 12-14 here:
http://iris.usc.edu/Vision-Notes/bibliography/contents.html

Another thought . . . this probably isn't the best way, but it might be fast since you're only dealing with a few points of data rather than the whole image. Find some landmarks in the images, for example bright peaks. Then use a star matching program like the Groth algorithm or this one (http://nedwww.ipac.caltech.edu/level5/Stetson/Stetson5_2.html) to find where the landmarks of your small image are located in your larger image.

Or maybe you can use/adapt something like the "image content engine"
https://www-eng.llnl.gov/sens_img_comm/sens_img_comm_imagery.html
From: bhavin on 9 Jul 2010 11:14
yes the small image is fully contained within the larger image and it's just rotated, and at an unknown location.

I will have a look at the links. My algorithm does the pattern recognition part which involves finding matching points and so on followed by finding the transformation between the two images.

So I was wondering if there is any freeware or matlab code available that does that so i can compare compare my outcome to its outcome transformation coefficients.

Thanks for the help
First  |  Prev  | 
Pages: 1 2
Prev: understanding imadjust function
Next: Array of structs. Column vector