This work considers the determination of the relative movements betwee
n consecutive images, taken from the same scene (registration), a prob
lem arising in many application fields. With the exception of speciali
zed techniques for specific applications, the most reliable registrati
on algorithms are based on the use of 2-D Fourier transforms. They typ
ically require a relevant computational effort. The computational bulk
iness typical of such methods can be overcome by 1-D techniques, needl
ess to say, at the price of a degraded performance in presence of dist
urbances. This paper proposes a new 1-D method, which exploits the ava
ilability of several frames, typical of practical contexts, in order t
o determine the translation of an object moving between two endpoints
on a static background. The noise performance of the new technique is
shown to be quite superior to that of the reported 1-D methods. For su
fficiently long sequences the noise performance favourably compares wi
th that of 2-D techniques, with considerable computational savings. Th
e possibilities of the method are exemplified by a number of practical
examples.