Verfahren zur Suche von Bewegungsvektoren für ein Videosignal

摘要

The invention relates to a method for the search of motion vectors for a video signal to make use of motion vectors in video signal processing applications, such as video signal compression. In the method, an error value is calculated for at least part of the image blocks (24) on the basis of each motion vector candidate to be used, and the motion vector candidate giving the smallest error value in each particular case is selected as the motion vector of the image block (24). To obtain a motion vector field more homogeneous than that obtained by conventional motion vector search methods, the error value is changed (25) before the final selection of the motion vector by a weighting value determined by the vector to be tested and/or by the ratio between the vector to be tested and motion vectors possibly previously selected from its neighbourhood for obtaining a more homogenous motion vector field, said selection being carried out on the basis of the changed error values.