Motion vector derivation method, moving picture coding method and moving picture decoding method

摘要

A motion vector derivation unit includes a comparison unit for comparing a parameter TR1 for a reference vector with a predetermined value to determine whether it exceeds the predetermined value or not; a switching unit for switching selection between the maximum value of a pre-stored parameter TR and the parameter TR1 according to the comparison result by the comparison unit; a multiplier parameter table (for multipliers); and a multiplier parameter table (for divisors) for associating the parameter TR1 with a value approximate to the inverse value (1/TR1) of this parameter TR1.

主权项

1. An image coding method for coding a current block in a current picture, the method comprising:
obtaining a reference motion vector of a reference block, the reference motion vector being used for deriving a motion vector of the current block to be coded; calculating a first parameter corresponding to a difference between a display order of a picture including a reference block and a display order of a reference picture of the reference block, wherein said reference block is motion-compensated using the reference motion vector, and said reference picture is referred to by the reference motion vector; calculating a second parameter corresponding to a difference between a display order of a current picture and a display order of the reference picture, wherein said current picture includes the current block; judging if (i) the reference motion vector of the reference block refers to a picture having a display order located after a display order of a picture including the reference block and (ii) the first parameter is a negative value within a predetermined range; deriving the motion vector of the current block by scaling the reference motion vector based on a predetermined negative value and the second parameter, when the first parameter is a negative value out of the predetermined range as a result of said judging, and by scaling the reference motion vector based on the first parameter and the second parameter, when the first parameter is a negative value within the predetermined range as a result of said judging; generating a motion compensated image of the current block using the derived motion vector and a reference picture corresponding to the derived motion vector; and coding a difference image between the current block and the motion compensated image of the current block.