Hi, I used to write  UKF and CKF algorithm which involves the N-dimensional Simplex, I am not sure that whehter it is what you want. 

I am not very familiar with those filters but  I don't think that they are the thing what I am looking for. My actual goal is the following:

I have an image which I can analytically describe by an equation consisting of N parameters (N can vary) and I have a second image which is my target. What I want to do is, finding the correct parameters for image one that bring me as close as possible to my target image.

Two answers:

1) Use the code available at the Numerical Recipes book:

http://www.nr.com/

2) Also you can download and use the function ANMS available here:

https://www.researchgate.net/profile/Lixing_Han/contributions

And explained here:

https://www.researchgate.net/publication/225691623_Implementing_the_Nelder-Mead_simplex_algorithm_with_adaptive_parameters

Article Implementing the Nelder-Mead simplex algorithm with adaptive...