How much of an error are you seeing caused by poor convergence? Are these highly distorted lenses?

]]>The default number of columns in the app is 7, however, the template picture you have on the boofcv website for printout has 8 columns! I think many people won’t notice this. ]]>

Can I email you a Screenshot and describe the problem? ]]>

The iteration will then stick to nearest minimum, which may not be the original (undistorted) value.

I know this does not pose a problem with simple applications, but estimating lens distortion from points (e.g. camera calibration, image registration, bundle adjustment), with some noise added may cause such issue.

I tried to solve this by splitting the distortion function in in part that is monotonic to a certain point and then is replaced by linear function. This ensures the function is differentiable and invertible everywhere.

However, the solution is quite messy and the “breaking point” itself depends on kappa parameters. This causes my unit tests failing because partial derivatives w.r.t. kappa are badly defined.

Any ideas?

]]>we you have a samples then it will be great for me.

Thank you so much for your help.

Tom

]]>isn’t sympy also an option? Sympy is pure python, although it doesn’t run in jython – but this may be more the fault of jython than sympy. The last time i tried sage was a bit heavy on the dependecies and since there are so many of them, it seems the sage devs aren’t even aiming for cross-compatibilty…

]]>BoofCV does come with a function for computing the homography directly, see FactoryMultiView.computeHomography(). Technically you could use a fundamental matrix instead, but it would require more work and be less reliable. There are a few more steps left..

For BoofCV releated questions could you post to the discussion board instead of posting here?

]]>I appreciate your engagement and efforts in boofCV very much. I would like to use boofCV in a little project and came up with this question: I have 4 points in a 2D (source-)image that build up an arbitrary convex(!) shape. I would like to warp the image, such that the 4 points constitute a perfect rectangle with known width/height centered in the middle of the (destination-)image. My idea is to take the 8 known points and to compute a homography matrix H, such that p1 = H*p2. Can this be done using the computeFundamental methods of boofCV? If yes, how does the homography matrix relate to the fundamental matrix? (Also found this one on my googling journey, which I find very useful: http://www.youtube.com/watch?feature=endscreen&v=fVJeJMWZcq8&NR=1).

It would be very nice if you could take some of your precious time to answer this silly question of mine.

Thanks in advance

Hendrik

(Sorry for the bad english)

]]>