FIELD: physics, image processing.
SUBSTANCE: invention concerns video coding technology, particularly method and device for reduction of artifacts related to division into units during residual prediction in layer set-based video coding. Method of video coding based on layer sets involves stages of obtaining difference between predicted unit for second unit in lower layer corresponding to first unit of current layer and second unit; addition of obtained difference to predicted unit for first unit; smoothing of third unit generated by addition, applying smoothing function; and coding of difference between first unit and smoothed third unit.
EFFECT: enhanced coding efficiency.
29 cl, 12 dwg, 1 tbl
Authors
Dates
2009-05-10—Published
2007-01-11—Filed