Apparatuses for image coding and decoding

摘要

An input video signal (101) is divided into units of an MC block. An encoding is performed for each division unit. A motion compensation section (107) determines the amount of motion for predetermined MC blocks to generate a motion compensation prediction image (106a). A smoothing filter section (124) smoothes a pixel of the prediction image at a boundary of adjacent MC blocks according to a predetermined judgment criterion, generates a prediction image (106b) by the smoothing, and encodes a prediction residual signal (108) produced by the difference between the prediction image and the input image (input video signal). Though a discontinuous waveform is produced between MC blocks of a prediction frame image generated by the motion compensation frame-to-frame prediction (MC) for each block, the discontinuous waveform is smoothed by a relatively simple calculation, and the encoding efficiency of a low bit rate using a frame-to-frame MC is improved.