| 摘要 |
A sequence of video frames (<B>A, B</B>) of data representing a moving image is encoded by obtaining a value representing the difference between a first (<B>X</B>) and a second block of pixels (<B>X'</B>) from sequential frames. This value is obtained by comparing corresponding pixels from the first and second block. An output indicates that either the pixel values are the same, or produces a difference value if they are not. The binary difference value is produced by placing a binary 1 at the bit position where there is a difference in bit value, followed by 0's for the remaining bits (see <B>Table A</B> on pages 12 and 13 for examples). The thus generated difference value is itself used to generate an error measurement indicating the amount of difference between the two blocks. This method of generating the difference value reduces the computational load as compared with standard block comparison techniques. |
