Scikit Learn/Numpy- Tell between a linear path and an anomaly (non-linear) path?

Question

I have an array of lat/lon/time data for an airplane track inside a numpy array:

array([[29.2205, -97.0371, 21373410],
   [29.2554, -97.0201, 21373439],
   [29.2993, -97.0178, 21373469],
    [...]])

Is there any way to tell between a path that's roughly straight- such as a commercial airliner- and an anomaly (a non-linear path)?

(I've already centered the data to make the average lat/lon = 0 so we can just look at the shape instead of worrying about where it is on the graph):

array([[3.2205, -2.0371, 21373410],
       [3.2554, -2.0201, 21373439],
       [3.2993, -2.0178, 21373469],
        [...]])

Example of a commercial aircraft track (roughly linear, but will turn and change direction slightly)

Example of an anomaly (military) aircraft (does crazy turns!)