The idea would be to make unwrap a more general function applicable to problems like quantization overflow (uint8 254+1 -> 0) and allow for min_val and max_val parameters (which would default to give the same behavior as currently implemented).

Reproducing code example:

import numpy as np
np.unwrap([0, 1, 2, 0], min_val=0, max_val=3)

should give array([0., 1., 2., 3.])