Your question is very broad to be answered fully here. There are many tutorials on Zernike moments that you can find online with a simple search.
At a high-level - Zernike moments are essentially descriptors for quantifying shape of an object (typically in a binary image). They are rotational invariant, are can be adapted to be translation (shift) and scale invariant.
I don't see a 'tow' value anywhere in this submission. Where do you see it?