发明名称 MOVING PICTURE CODING METHOD AND A MOVING PICTURE DECODING METHOD
摘要 A moving picture coding apparatus includes a motion estimation unit (101) for performing motion estimation by fixing the one of two reference pictures as a reference picture indicated by an inputted default reference picture number DefRefNo and a variable length coding unit (107) for performing variable length coding on coded residual data ERes, a prediction type PredType, a reference picture number RefNo2 and motion vectors MV1, MV2 on a block-by-block basis, and outputting them as coded moving picture data Str.
申请公布号 US2015312566(A1) 申请公布日期 2015.10.29
申请号 US201514745983 申请日期 2015.06.22
申请人 Godo Kaisha IP Bridge 1 发明人 HAGAI Makoto;KADONO Shinya;KONDO Satoshi;ABE Kiyofumi
分类号 H04N19/105;H04N19/172;H04N19/159;H04N19/573 主分类号 H04N19/105
代理机构 代理人
主权项 1. A picture decoding apparatus which selects two reference pictures from among reference pictures on a block basis, and performs a predictive decoding onto a block in a current picture to be decoded, the picture decoding apparatus comprising: a judging device, when decoding a plural-block image unit made up of a plurality of blocks, configured to judge whether or not information identifying one or two common reference pictures to be commonly referred to is described in a common information area for the plural-block image unit, the one or two common reference pictures being selected from among plural reference pictures and are assigned commonly to each of the plurality of blocks of the plural-block image unit such that reference picture identification information for the one or two common reference pictures can be omitted for at least one of the plurality of blocks of the plural-block image unit; a generating device: configured to generate a predictive image of a current block included in the plural-block image unit, using the one common reference picture and the reference picture specified on a block basis, in the case where it is judged that information identifying only the one common reference picture is described in the common information area, configured to generate a predictive image of the current block included in the plural-block image unit, using the two common reference pictures, in the case where it is judged that information identifying the two common reference pictures is described in the common information area, and configured to generate a predictive image of the current block included in the plural-block image unit, using two reference pictures specified on a block basis, in the case where it is judged that the information identifying the one or two common reference pictures is not described in the common information area; and a decoding device configured to decode the current block using the predictive image, wherein, any of following reference pictures stored in a buffer memory is selected as the one or two common reference picture, (1) a reference picture closest to the current picture in a display order, (2) a reference picture located before the current picture in a display order and closest to the current picture in a display order, (3) a reference picture located after the current picture in a display order and closest to the current picture in a display order, (4) a reference picture closest to the current picture in a decoding order, (5) a reference picture located before the current picture in a display order and closest to the current picture in a decoding order, and (6) a reference picture located after the current picture in a display order and closest to the current picture in a decoding order.
地址 Tokyo JP