| 摘要 |
An interframe coding method performs predictive coding by dividing an input video signal into blocks each of which has a predetermined size, wherein a second memory for storing background data is arranged in addition to a first memory for storing coded picture data, and a block having a minimum prediction error among blocks read out from the first and second memories is detected as an optimal predictive block for every block of the input video signal, whereby predictive coding is performed by using the optimal predictive block. According to this method, after a moving element in the picture is displaced, the background area is monitored by using as predictive value data stored in the second memory. A prediction error becomes small, and coding efficiency is improved. In particular, motion compensated interframe coding is performed in addition to detection of the background area, thereby further improving the coding efficiency. The background image stored in the second memory is properly updated in accordance with the input video signal. |
