VERFAHREN UND VORRICHTUNG ZUR VERARBEITUNG VON BILDSEQUENZEN MIT MASKIERUNG

摘要

The invention concerns a method and a device for processing image sequences which consists in: differentiating between a current image (11) and a reference image (12); distinguishing in said image difference (13), movement zones and immobility zones, said zones being distinguished by means of at least one motion threshold, masking the image difference by imposing null values in the immobility zones and encoding the masked image difference (15); determining the movement zones and the immobility zones with a Markov field modelling. Preferably, the image difference is thresholded prior to said modelling.