Method and apparatus for video compression
    1.
    发明申请
    Method and apparatus for video compression 有权
    视频压缩的方法和装置

    公开(公告)号:US20040233993A1

    公开(公告)日:2004-11-25

    申请号:US10844054

    申请日:2004-05-12

    IPC分类号: H04N007/12

    摘要: A unified solution to coding/decoding of different video formats such as 4:2:0, 4:2:2 and 4:4:4 is provided. A method of video coding includes transforming a first mnulln macro block of residual chrominance pixel values of moving pictures by a first integer-transform function generating a corresponding second mnulln macro block of integer-transform coefficients, further transforming DC values of the integer-transform coefficients by a second integer-transform function to generate a third block of integer-transformed DC coefficients. The method further includes generating the second mnulln macro block of integer-transform coefficients by utilizing a knullk integer-transform function on each knullk sub-block of the first mnulln macro block, wherein n and m are each a multiple of k, and generating the third block of coefficients by utilizing a second inullj integer-transform function on the DC values resulting in a (m/k)null(n/k) third block of integer-transformed DC coefficients.

    摘要翻译: 提供了对4:2:0,4:2:2和4:4:4等不同视频格式进行编码/解码的统一解决方案。 一种视频编码方法包括:通过产生对应的整数变换系数的第二m×m宏块的第一整数变换函数来变换运动图像的残余色度像素值的第一m×m宏块,进一步变换整数变换系数的DC值 通过第二整数变换函数来生成整数变换DC系数的第三块。 该方法还包括通过在第一m×n宏块的每个kxk子块上利用kxk整数变换函数来生成整数变换系数的第二m×n宏块,其中n和m分别为k的倍数,并且生成 通过利用第二个ixj整数变换函数,得到第二个整数变换的直流系数的第(m / k)×(n / k)个第三块的DC值的第三个系数块。

    Early detection of zeros in the transform domain
    2.
    发明申请
    Early detection of zeros in the transform domain 有权
    在变换域中早期检测零

    公开(公告)号:US20040264575A1

    公开(公告)日:2004-12-30

    申请号:US10831158

    申请日:2004-04-26

    发明人: Gisle Bjontegaard

    IPC分类号: H04N007/12

    摘要: A method detects blocks that are to be indicated as skipped at an earlier stage of the encoding process, than would be the case with other implementations of the ITU H.263 and H.264 standards. The method includes transforming 4null4 blocks in the macro blocks having a skip vector of zero with a binary-transform function. The blocks having values of the four uppermost left binary-transform coefficients less than a predefined threshold, are defined as skipped, thus, minimizing the need for computationally demanding block transformation or quantization.

    摘要翻译: 一种方法检测在编码过程的较早阶段被跳过的块,与ITU H.263和H.264标准的其他实施方式相比。 该方法包括利用二进制变换函数来变换具有零跳跃向量的宏块中的4×4块。 将具有小于预定阈值的四个最上面的二进制变换系数的值的块定义为跳过,从而最小化对计算要求高的块变换或量化的需要。

    Method for vector prediction
    3.
    发明申请
    Method for vector prediction 有权
    矢量预测方法

    公开(公告)号:US20040146110A1

    公开(公告)日:2004-07-29

    申请号:US10722479

    申请日:2003-11-28

    发明人: Gisle Bjontegaard

    IPC分类号: H04N007/12

    CPC分类号: H04N19/57 H04N19/52 H04N19/56

    摘要: A method for prediction of the motion vector of a pixel block in a video picture that is to be coded. The actual motion vectors of two adjacent blocks close to the uppermost left corner of the block are selected as candidates for the prediction. One additional block, also adjacent to the block, is selected to decide which of the motion vectors to be used as the prediction. The vector difference to the motion vector of the decision block is decisive for the final selection.