How to double python numpy array size by adding interpolated points?

Question

Lets suppose that in python I have a N-dimensional square array. for example: x=[[1,2],[3,4]]. How can I double the resolution of this array by inserting (linearly, average)