| 摘要 |
A method of video signal processing, such as motion estimation, in which an output motion vector [vector(p)] is required to be associated with a picture element p, there being N trial vectors [vectorn(p)] for that picture element, wherein the output motion vector is derived according to (I), where [biasn(p)] is a bias value associated with the n'th trial vector. The bias value can be related to a pixel match error, a probability function from the vector identification and, for block based processes, a function of block position. |
