PREDICTING ERROR ADAPTIVE VIDEO SIGNAL COMPRESSIVE ENCODING/ DECODING

摘要

On a transmitting side, a digital video signal comprising respective pixel samples (each of which may include eight bits) is received, and a set of elementary pixel samples (o/ &cir& ) determined from the received pixel samples at a predetermined rate are transmitted. The values of respective subject pixel samples (o, &squ& , DELTA , x) other than the elementary ones (o/ &cir& ) are predicted, and respective predictive errors ( epsilon ) of the predicted values are detected from present values of the received pixel samples. First flags are then generated individually for each subject pixel sample whose predictive error ( epsilon ) is greater than a threshold value (TH), and second flags are generated individually for each subject pixel sample whose predictive error is less than or equal to the threshold value. The present values of the subject pixel samples having the first flags are transmitted, and compressed data (including the second flags) is transmitted for the subject pixel samples having the second flags. On a receiving side, values of the subject pixel samples represented by the compressed data with the second flags are predicted, and the subject pixels are interpolated with pixel samples having present values equal to the predicted values. A decoded video signal, including the transmitted elementary pixel samples, the respective subject pixel samples having the first flags, and the respective interpolated subject pixel samples, is then outputted.