I do not think such a thing exists, but I am starting.
The advantage is that you can execute some analytic methods ... for example, instead of hacking in openCV or even Matlab indefinitely, you can calculate analytically the number and see that the method leading to this matrix is ββnumerically unstable as a function of input variables. This way you do not need to hack, as that would be pointless.
As for the opencv wrap, this does not seem to make sense. The correct procedure would be to fix bad implementations in opencv based on your analysis in Mathematica and on paper.
peter karasev
source share