I was searching for a 2D interpolation function in IPP image processing library or MKL, which can do the following...

Given an input matrix of let say size 2000x15, lets say xaxis corresponds to integers from 0 to 4000. yaxis corresponds to floating point numbers between 0 to 1.0. The zaxis is also floating point with values varying from 0.0 to 2.0

I want to get the values of znew, at (xnew,ynew) where xnew is one of the values found in xaxis ie. between 0 to 4000. But ynew is a floating point value between 0 to 1.0, which may or may not be found in the yaxis value vector. I need to do this thousands of times where ynew are always varying in from previous ynew values. (ie ynew is not always equally spaced, but can be any value between 0 to 1.0)

I looked at the image processing functions, and the ones which I can find, would resize the image only at a given new grid which is some factor of previous input grid. Given in my case the ynew is varying and can be any value between 0 to 1.0, i cant define a fixed factor between ynew and input yaxis.