METHOD OF ESTIMATING MOTION IN A PICTURE SIGNAL

In a method of estimating motion per picture portion in an image of a picture signal, wherein, starting from a first (x0=x1, y0=y2) and a second (x0=x3, y0=y4) starting vector a motion vector is determined, a first and a second candidate motion vector, respectively, are determined on the basis of the first and second starting vector in accordance with a predetermined criterion, a motion vector is selected from both candidate motion vectors, and as components of the starting vectors, respective corresponding components of candidate motion vectors (x1, y1), (x2, y2), (x3, y3) and (x4, y4) already previously determined are taken, which candidate motion vectors correspond respectively to the starting vectors.