VIDEO SCALABLE ENCODING METHOD, VIDEO SCALABLE DECODING METHOD, DEVICES THEREFOR, PROGRAMS THEREFOR, AND RECORDING MEDIUM WHERE PROGRAM IS RECORDED

摘要

A video scalable encoding method in which a weight coefficient composed of a proportional coefficient and an offset coefficient and indicating a variation of the brightness between an image region to be encoded in an upper layer and an image region to be referenced is calculated, a motion vector is calculated by estimating a motion by operation of the weight coefficient and the image signal of the image region to be referenced which is a search object, motion compensation is carried out by operation of a decoded signal of the image region to be referenced and indicated by the motion vector and the weight coefficient, and thereby a prediction signal is generated. The data structure of the weight coefficient is determined on the basis of the encoding information on the immediately-below image region present in spatially the same position as that of the image region to be encoded in the immediately-below layer. When the immediately-below image region has performed frame-to-frame prediction in the immediately-below layer, the immediately-below layer image region to be referenced which is used as the reference of motion prediction performed by the immediately-below image region is identified. Assuming that the result of operation of the DC component of the image region in the upper layer present in spatially the same position as that of the immediately-below layer image region to be referenced and the weight coefficient used for weighted motion prediction by the immediately-below image region as the DC component of the immediately-below image region, the weight coefficient is calculated.