ENHANCED REFERENCE REGION UTILIZATION FOR SCALABLE VIDEO CODING
    4.
    发明申请
    ENHANCED REFERENCE REGION UTILIZATION FOR SCALABLE VIDEO CODING 有权
    用于可扩展视频编码的增强参考区域利用

    公开(公告)号:US20140286409A1

    公开(公告)日:2014-09-25

    申请号:US13997935

    申请日:2012-12-28

    IPC分类号: H04N19/61 H04N19/91 H04N19/82

    摘要: Techniques to identify one or more candidate reference blocks used to generate a prediction block to encode a current coding block. The candidate reference blocks can be in the same layer as the current coding block or a different layer. In addition, the candidate reference blocks do not have to be co-located with the current coding block. Motion vectors and shift vectors can be used to identify the one or more candidate reference blocks. In addition, uniform and non-uniform weighting can be applied to the one or more candidate reference blocks to generate the prediction block. Accordingly, an encoder can determine and identify reference blocks to a decoder that can provide desirable rate-distortion cost.

    摘要翻译: 用于识别用于生成预测块以编码当前编码块的一个或多个候选参考块的技术。 候选参考块可以与当前编码块或不同层位于相同的层中。 此外,候选参考块不必与当前编码块共同定位。 运动矢量和移位矢量可用于识别一个或多个候选参考块。 此外,可以对一个或多个候选参考块应用均匀和不均匀的加权以生成预测块。 因此,编码器可以确定并识别能够提供所需速率失真成本的解码器的参考块。