| 摘要 |
Processing is carried out assuming that a picture structure is fixed to I picture and P picture and that motion compensation value is fixed. Header analyzing section 11 identifies the type of picture of input video code. In case of P picture, IDCT section 14 carries out IDCT to calculate a differential value of each block. An adder 15 adds the differential value and a motion compensated block obtained by motion compensating, by forward prediction section 18, a video code of preceding frame stored in a preceding frame section 16. Thus, an image is expanded and reproduced. By fixing the picture structure to I picture and P picture, it is possible to dispense with judgement whether or not the picture is B picture and backward prediction or bidirectional prediction in processing of the B picture. <IMAGE> |
