A comparison of different methods for determining optical flow based o
n local information is presented. It is shown that to overcome the ape
rture problem, second order differential terms of the image function a
re required. It is also argued that the fundamental problem of measuri
ng optical flow is to determine the velocity in as small an aperture a
s possible. It is concluded that the technique of Lucas and Kanade (se
e Barren et al., Performance of optical flow techniques, Int. J. Compu
ter Vision, 12 (1) (1994) 43-77) is the best generalised second order
method for achieving this objective.