Coding and Decoding a Video Signal Including Generating and Using a Modified Residual and/or Modified Prediction Signal

摘要

Coding a block of a video signal (10) comprises generating a prediction signal (i.e. a motion vector) between it and a reference frame (20), and calculating a residual (50) based on a difference between the video signal and the prediction signal. A transformation module (200) calculates an enhanced modified residual (80) (shown in the figure) and/or modified prediction signal along with transform parameters (60) based on the residual and/or prediction signal. Additional compression of residuals and/or prediction signals is thereby performed prior to the conventional compression process (300 and 400). The transformation used in the transformation module may be defined by one or more parameters, and the transformation function used may be a shifting transformation function. The transformation function and/or the parameters used may reduce or minimise an error metric (e.g. SAD) of the residual.