Numpy using multidimensional array to index a 1D array

Question

I do not understand the following code, i.e. the last part of it.

max = np.max(rel_coords, axis=0)
min = np.min(rel_coords, axis=0)
bins = [np.arange(low, hig