Intel® Integrated Performance Primitives (Intel® IPP) Developer Guide and Reference
Tricubic Interpolation
Short Description
The tricubic interpolation algorithm uses source image intensities at 64 pixels in the neighborhood of the point (x s, y s, z s) in the source image:

First, for each z sk the algorithm determines 16 cubic polynomials F nm (z), 0 ≤ m, n ≤ 15 :
F nm (z) = a mn z 3 + b nm z 2 + c nm z + d nm
For fixed values z s , the function F nm (z snm) is equal to:

There are sixteen points with fixed z s F 00 (z), F 01 (z), F 02 (z), F 03 (z), F 10 (z), F 11 (z), F 12 (z), F 13 (z), F 20 (z), F 21 (z), F 22 (z), F 23 (z), F 30 (z), F 31 (z), F 32 (z), F 33 (z). The need is to fix y s and x s. This situation is described in the Cubic Interpolation, refer to this page to visualize the operation of the algorithm as well.

To use the cubic interpolation, set the interpolation parameter to IPPI_INTER_CUBIC.