WO2012142777A1 - 一种视频编码8x8整数变换方法 - Google Patents

一种视频编码8x8整数变换方法 Download PDF

Info

Publication number
WO2012142777A1
WO2012142777A1 PCT/CN2011/073586 CN2011073586W WO2012142777A1 WO 2012142777 A1 WO2012142777 A1 WO 2012142777A1 CN 2011073586 W CN2011073586 W CN 2011073586W WO 2012142777 A1 WO2012142777 A1 WO 2012142777A1
Authority
WO
WIPO (PCT)
Prior art keywords
transform
integer
matrix
integer transform
transformation
Prior art date
Application number
PCT/CN2011/073586
Other languages
English (en)
French (fr)
Inventor
徐苏珊
Original Assignee
深圳市融创天下科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市融创天下科技股份有限公司 filed Critical 深圳市融创天下科技股份有限公司
Publication of WO2012142777A1 publication Critical patent/WO2012142777A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/12Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
    • H04N19/122Selection of transform size, e.g. 8x8 or 2x4x8 DCT; Selection of sub-band transforms of varying structure or type
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock

Definitions

  • the present invention relates to the field of video coding technologies, and in particular, to a video coding 8x8 integer transform method.
  • a complete video coding system consists of a series of algorithm modules such as prediction, transform, quantization, and entropy coding.
  • intra-frame interframe prediction is compression data using spatial and temporal correlation of video data.
  • the transform module transforms the image residual data outputted by the prediction module from the time domain to another space, so that the data in the time domain is uniformly concentrated in the transform space, and the energy of the data is mostly concentrated in the low frequency region of the space. .
  • the transformed data distribution features are effectively utilized to further compress the data.
  • the transform module is an important part of the video coding system, and the performance of the transform directly affects the performance of the video coding system.
  • the DCT Discrete Cosine Transform
  • the energy concentration of the transform is very good, which can greatly eliminate the correlation of video data in the transform space.
  • this algorithm is insufficient.
  • the DCT transformation matrix is represented by floating point. A large number of floating-point multiplications and additions are used in the calculation process, which takes up a lot of hardware resources and is computationally complex.
  • the floating-point operation has a truncation error, and the floating-point calculation accuracy is not high.
  • the DCT forward-reverse transform cannot be completely matched, and the transform is irreversible. Summary of the invention
  • the purpose of the embodiment of the present invention is to provide a video coding 8x8 integer transform method, which aims to solve the problem that the prior art DCT transform matrix adopts floating point multiplication and addition, which occupies many hardware resources, is computationally complex, and has low calculation precision, and the DCT is positive.
  • the inverse transformation cannot be completely matched, and the transformation is irreversible.
  • a video coding 8x8 integer transform method includes:
  • the 8x8 general integer transformation matrix is: kO kO kO kO kO kO kO kO kO kO kO kO kO kO
  • kO, k5, k6 take values 1, 2, 1, kl k3 4 respectively [ is an integer transform base; positive transform, integer transform of 8x8 image residual data , the transformation formula is yo 3 ⁇ 4P ; wherein P is the transformation matrix; is the image residual data;
  • the integer transform base is one of the following values: [15, 15, 10, 3], [6, 7, 5, 1].
  • step of "selecting an integer transform base and constructing the transform matrix P" may further include the following steps:
  • step of "positive transformation” performs integer transformation on 4x4 image residual data
  • transformation formula may further include steps:
  • is the point multiplication operation of the matrix
  • Y is the data after 8x8 positive transformation
  • PF is the scaling matrix, which is the output value after the scaling process, representing the information of the video data transformed into the frequency domain
  • the scaling matrix is:
  • the transform method of the embodiment of the present invention has excellent energy concentration performance, and the energy concentration is close to 8x8DCT, which can greatly eliminate the transformation of video data. Spatial correlation. Low computational complexity, can be added And shift to achieve, easy to implement hardware, and high precision of operation, there is no problem of positive and negative transform mismatch.
  • FIG. 1 is a flowchart of a method of a preferred embodiment of a video coding 8x8 integer transform method of the present invention
  • FIG. 2 is a diagram of an integer transform base selection of a preferred embodiment of the video coding 8x8 integer transform method of the present invention [15, 15, 10, 3] a schematic diagram of the corresponding integer transformation calculation method;
  • FIG. 3 is a calculation method diagram of the calculation module M1 in the schematic diagram of the integer transformation calculation method of FIG. 2;
  • FIG. 4 is a selection of integer transformation bases for the preferred embodiment of the video coding 8x8 integer transformation method of the present invention [6, 7, 5, 1] a schematic diagram of the corresponding integer transformation calculation method;
  • FIG. 5 is a diagram showing a calculation method of the calculation module M2 in the schematic diagram of the integer transformation calculation method of FIG. 4.
  • FIG. 1 is a video coding 8x8 integer transform method according to Embodiment 1 of the present invention, where the method includes the following steps:
  • S101 Receive 8 ⁇ 8 image residual data output by the prediction module in the encoder.
  • S102 selecting an integer transform base, constructing a transform matrix P;
  • the 8x8 general integer transformation matrix is:
  • kO, k5, and k6 take values 1, 2, and 1, respectively.
  • the definition k2 k3 is constructed as a transformation matrix P for an integer transform base.
  • the integer transform base may select one of the following values:
  • r [l/
  • m . is the row vector of the transformation matrix P, i takes a value of 1-8;
  • is the vector norm of the row vector;
  • ® is the point multiplication of the matrix
  • Y is the 8x8 positive transformed data
  • PF is the scaling matrix
  • output value after the scaling process is the information of the video data transformed into the frequency domain.
  • the transform method of the embodiment of the invention has excellent energy concentration performance, and the energy concentration is close to 8x8DCT, which can greatly eliminate the correlation of video data in the transform space; the computational complexity is low, and can be realized by addition and shifting, and is easy to implement in hardware. Moreover, the operation precision is high, and there is no problem that the forward and reverse transforms do not match.
  • Embodiment 2
  • Embodiment 2 of the present invention is an integer transform base of an 8x8 general integer transform matrix, and the method includes the following steps:
  • the general integer transformation matrix of 8x8 is: k0 kO kO kO kO kO kO kO kO kO kO kO
  • k0, k5, and k6 take values 1, 2, and 1, respectively.
  • the definition k2 k3 is constructed as a transformation matrix P for an integer transform base.
  • [15, 15, 10, 3] is selected as an integer transform base, and the transform matrix Pc is constructed.
  • the transform matrix P is:
  • Change matrix P construct: scaling matrix PF ;
  • PF Z z ⁇ l ⁇ m2 ⁇ l/
  • j [yl y2 y3 y4 y5 y6 yl yS] T ⁇ xl xl x3 x4 x5 x6 xl x8 represents the input eight-point value of the one-dimensional forward transform of the integer transform; ⁇ 2 3 4 5 ⁇ is the eight-point output value of the forward transform ;
  • xl x2 x3 x4 x5 x6 xl x8 represents the input eight-point value of the one-dimensional forward transform of the integer transform
  • yl y3 y4 y5 y6 yl y8 is the eight-point output value of the positive transform
  • the data processing direction is from left to right, intersecting
  • the value of the intermediate variable of a dot is equal to the data represented by the line on the left side of the dot; the square means the data is multiplied by a coefficient;
  • the "-” means taking a negative value;
  • the "2” means multiplying by 2, that is, one bit to the left;
  • "4" represents multiplying by 4, that is, shifting two digits to the left;
  • “8” means multiplying by 8, that is, shifting 3 digits to the left, and
  • "16" means multiplying by 16, that is, shifting 4 digits to the left.
  • ® is the point multiplication of the matrix
  • Y is the 8x8 positive transformed data
  • PF is the scaling matrix
  • output value after the scaling process is the information of the video data transformed into the frequency domain.
  • the transform method of the embodiment of the invention has excellent energy concentration performance, and the energy concentration is close to 8x8DCT, which can greatly eliminate the correlation of video data in the transform space; the computational complexity is low, and can be realized by addition and shifting, and is easy to implement in hardware. Moreover, the operation precision is high, and there is no problem that the forward and reverse transforms do not match.
  • Embodiment 3
  • Embodiment 3 of the present invention is an integer transform base of an 8x8 general integer transform matrix, and the method includes the following steps: 5301. Receive 8 ⁇ 8 image residual data X output by the prediction module in the encoder.
  • the general integer transformation matrix of 8x8 is: kO kO kO kO kO kO kO kO kO kO kO kO kO
  • kO, k5, and k6 take values 1, 2, and 1, respectively.
  • k2 k3 as an integer 3 to construct a transformation matrix? .
  • [6, 7, 5, 1] is selected as an integer transform base, and is constructed into a transform matrix P.
  • the forward transform fast calculation method of this transform base is shown in FIG. 4, wherein the calculation module M2 is as shown in FIG. 5.
  • xl x2 x3 x4 x5 x6 xl x8 represents the input four-point value of the one-dimensional forward transformation of the integer transformation
  • yl y3 y4 y5 y6 yl y8 is the four-point output value of the positive transformation
  • the data processing direction is from left to right, intersecting
  • the value of the intermediate variable of a dot is equal to the data represented by the line on the left side of the dot; the square means the data is multiplied by a coefficient;
  • the "-” means taking a negative value;
  • the "2” means multiplying by 2, that is, one bit to the left;
  • "4" Indicates multiplied by 4 that is, shifted to the left by two.
  • ® is the point multiplication of the matrix
  • Y is the 8x8 positive transformed data
  • PF is the scaling matrix
  • output value after the scaling process is the information of the video data transformed into the frequency domain.
  • the transform method of the embodiment of the invention has excellent energy concentration performance, and the energy concentration is close to 8x8DCT, which can greatly eliminate the correlation of video data in the transform space; the computational complexity is low, and the addition can be performed by adding It is realized by shifting, easy to implement in hardware, and has high precision of operation, and there is no problem that the forward and reverse transforms do not match.
  • all or part of the steps of the foregoing embodiments may be implemented by a program instruction related hardware, and the program may be stored in a computer readable storage medium.
  • the storage medium may be a ROM, a RAM, a magnetic disk, an optical disk, or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Discrete Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Description

一种视频编码 8x8整数变换方法 技术领域
本发明涉及视频编码技术领域, 尤其涉及一种视频编码 8x8 整数变换方 法。
背景技术
一套完整的视频编码系统是由预测、 变换、 量化、熵编码等一系列算法模 块组成的。其中帧内帧间预测是利用视频数据在空间和时间上的相关性压縮数 据。 变换模块是将预测模块输出的图像残差数据由时域变换到另一个空间上, 使时域上平坦分布的数据在变换空间上能量集中,将数据的能量大部分集中在 空间的低频区域中。在随后的量化和熵编码模块中,会有效的利用变换后的数 据分布特征, 进一步压縮数据。显而易见, 变换模块是视频编码系统中重要的 组成部分, 变换的性能直接影响视频编码系统的性能。
在现有视频编码系统中, DCT (离散余弦变换: Discrete Cosine Transform) 变换得到了广泛的应用, 其变换的能量集中性很好, 可以极大地消除视频数据 在变换空间上的相关性。但是这种算法存在不足, DCT变换矩阵是浮点表示的, 计算过程中使用大量的浮点乘法和加法, 占用硬件资源多, 计算复杂的大。 同 时由于处理器运算位数影响,浮点运算存在截断误差,浮点计算精度不高, DCT 正反变换不能完全匹配, 变换不可逆。 发明内容
本发明实施例的目的在于提出一种视频编码 8x8整数变换方法,旨在解决 现有技术 DCT变换矩阵采用浮点乘法和加法,占用硬件资源多,计算复杂的大, 计算精度不高, DCT正反变换不能完全匹配, 变换不可逆的问题。
本发明实施例的方法是这样实现的, 一种视频编码 8x8整数变换方法, 包 括:
接收编码器中预测模块输出的 8x8的图像残差数据 ;
选取整数变换基, 构建变换矩阵,
8x8的通用整数变换矩阵为: kO kO kO kO kO kO kO kO
kl kl k3 k4 Dk4 Dk3 Uk2 Uk\
k5 k6 Uk6 Uk5 Uk5 Uk6 k6 k5
kl Dk4 Uk\ Dk3 k3 k\ k4 Uk2
C=
kO DkO UkQ kO kO UkQ DkO kO
k3 Uk\ k4 kl Uk2 Dk4 kl Dk3
k6 Uk5 k5 Uk6 Uk6 k5 Uk5 k6
k4 Dk3 kl Uk\ k\ Uk2 k3 Dk4 矩阵 C中, kO、 k5、 k6分别取值 1、 2、 1, kl k3 4[为整数变换基; 正变换, 对 8x8的图像残差数据 进行整数变换, 变换公式为 yo ¾P ; 其中 P为所述的变换矩阵; 为所述的图像残差数据;
所述整数变换基为以下数值中的一个: [15, 15, 10, 3]、 [6, 7, 5, 1]。
进一步地, 所述步骤 "选取整数变换基, 构建变换矩阵 P"之后还可以包 括步骤:
根据变换矩阵 P, 构建縮放矩阵 PF,
PFUZ T^Z 其中,
zTull\\m\\\ \l\\m2\\ l/||m3|| l/\\m4\\ l/\\m5\\ l/\\m6\\ l/\\m7\\ l/||m8|| [; m. 为变换矩阵 P的行向量, i取值为 1-8; 为行向量^的向量范数;
进一步地, 所述步骤 "正变换, 对 4x4的图像残差数据 进行整数变换, 变换公式为 之后还可以包括步骤:
对经过 8x8正变换后的数据 Y进行縮放处理,
Y UYU PF
□是矩阵的点乘运算, Y是经过 8x8正变换后的数据, PF 是縮放矩阵, 是本縮放处理后的输出值, 代表视频数据变换到频域上的信息; 其中, 所述正变换中的变换的基本单元是形如 y = Px的 8点一维变换, 其 中 x = [xl x2 x3 x4 x5 x6 xl xSf , 输出的 j = [yl y2 y3 y4 y5 y6 yl j8] , l x2 x3 x4 x5 x6 χΊ J8为正变换的 输入八点数值; yl yl y3 y4 y5 ye yl 8为正变换的八点输出数值。
当整数变换基取值为 [15, 15, 10, 3]时:
1 1 1 1 1 1 1 1
15 15 10 3 -3 -10 -15 -15
2 1 -1 -2 -2 -1 1 2
15 -3 -15 -10 10 15 3 -15
:述变换矩阵 P=
1 -1 -1 1 1 -1 -1 1
10 -15 3 15 -15 -3 15 -10
1 -2 2 -1 -1 2 -2 1
3 -10 15 -15 15 -15 10 -3
[0.35355 0.029907 0.22361 0.029907 0.35355 0.029907 0.22361 0.029907]
" 0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
PF
0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
"对 8x8的图像残差数据 进行整数变换, 变换公式为 y = P W计; 过程如下:
bl=xl-x8; b2=x2 x3~x6; b4=x4~x5; b5=xl+x8; b6=x2+x7; b7=x3+x6; b8=x4+x5; al=b5+b8; a2=b6+b7; a3=b5-b8; a4=b6_b7; yl=al+a2; y5=al~a2; y3=a3«l+a4; y7=a3— a4《l
Figure imgf000006_0001
cl=bl«4-bl; c2=b2«4-b2; c3=b3〈〈4_b3; c4=b4〈〈4_b4;
dl=bl«3+bl«l; d2=b2«3+b2«l; d3=b3«3+b3«l; d4=b4«3+b4«l el=bl«l+bl; e2= b2«l+b2; e3= b3«l+b3; e4= b4«l+b4
y2=cl+c2+d3 +e4; y4=cl-e2-c3_d4; y6=dl_ c2 + e3 + c4;
y8=el - d2 + c3 - c4; 其中 "〈〈"表示向左移位运算, 优先级高于加减法, 基本单元 y = Px共需
40个加法和 14个移位。
当整数变换基取值为 [6, 7, 5, 1]时:
1 1 1 1 1 1 1 1
6 7 5 1 -1 -5 -7 -6
2 1 -1 -2 -2 -1 1 2
7 -1 -6 -5 5 6 1 -7
:述变换矩阵 P=
1 -1 -1 1 1 -1 -1 1
5 -6 1 7 -7 -1 6 -5
1 -2 2 -1 -1 2 -2 1
1 -5 7 -6 6 -7 5 -1
= [0.35355 0.067116 0.22361 0.067116 0.35355 0.067116 0.22361 0.067116]
縮放矩阵为:
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
PF =
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045 的图像残差数据 进行整数变换, 变换公式为 r = p W计 过程如下:
bl=xl-x8; b2=x2-x7; b3=x3-x6; b4=x4-x5 ; b5=xl+x8; b6=x2+x7 b7=x3+x6; b8=x4+x5 ; al=b5+b8; a2=b6+b7: a3=b5-b8; a4=b6-b7; yl=al+a2; y5=al-a2; y3=a3«l+a4; y7=a3-a4«l;
Figure imgf000007_0001
cl=bl«2+bl«l+b4; c2=b2«2+b2«l-a3 ; c3=b2+b3«2+b3«l; c4 =-al+b4«2+b4«l; y2=cl+c2+c3 ; y4=cl-c3-c4; y6=cl-c2+c4; y8=-c2+c3-c4; 其中 "<<"表示向左移位运算, 优先级高于加减法, 基本单元 y = Px共需 要 32个加法和 10个移位。
本发明的有益效果:
本发明实施例提出的二组的 8x8的整数变换矩阵, 并且给出了变换的快速 算法, 本发明实施例的变换方法能量集中性能优秀, 能量集中度接近 8x8DCT, 可以极大地消除视频数据在变换空间上的相关性。计算复杂度低,可通过加法 和移位来实现, 易于硬件实现, 并且运算精度高, 不存在正反变换不匹配的问 题。
附图说明
下面将结合附图及实施例对本发明作进一步说明, 附图中:
图 1是本发明的视频编码 8x8整数变换方法的优选实施例的方法流程图; 图 2是本发明的视频编码 8x8整数变换方法的优选实施例的当整数变换基 选取 [15, 15, 10, 3]时相应的整数变换计算方法示意图;
图 3是图 2中整数变换计算方法示意图中的计算模块 Ml的计算方法图; 图 4是本发明的视频编码 8x8整数变换方法的优选实施例的当整数变换基 选取 [6, 7, 5, 1]时相应的整数变换计算方法示意图;
图 5是图 4中整数变换计算方法示意图中的计算模块 M2的计算方法图。
具体实施方式 为了使本发明的目的、技术方案及优点更加清楚明白, 以下结合附图和实 施例, 对本发明进行进一步详细说明, 为了便于说明, 仅示出了与本发明实施 例相关的部分。 应当理解, 此处所描写的具体实施例, 仅仅用于解释本发明, 并不用以限制本发明。 本发明通过接收编码器中预测模块输出的 8x8的图像残差数据 选取不 同整数变换基 ([15, 15, 10, 3]或 [6, 7, 5, 1] ); 对图像残差数据 ^进行正变换 { Y = PXPT ) , 并根据不同的整数变化基, 采用相应的整数^换计算方法; 构 建变换矩阵 Ρ, 根据变换矩阵 Ρ, 构建縮放矩阵 PF ( PF = z ' z ); 对经过 8x8 正变换后的数据 Y进行縮放处理。 实施例一
如图 1所示为本发明实施例 1一种视频编码 8x8整数变换方法,所述方法 包括以下步骤:
S101 , 接收编码器中预测模块输出的 8x8的图像残差数据 ; S102, 选取整数变换基, 构建变换矩阵 P;
8x8的通用整数变换矩阵为:
Figure imgf000009_0001
矩阵 C中, kO、 k5、 k6分别取值 1、 2、 1。 定义 k2 k3 为整数变换 基构建成变换矩阵 P。
本发明实施例中, 整数变换基可选取以下数值中的一个:
[15, 15, 10, 3]、 [6, 7, 5, 1] 。
5103, 根据变换矩阵 P, 构建縮放矩阵 PF; 其中,
r = [l/||ml|| \l\\m2\\ l/||m3|| \l\\mA\\ \l\\m5\\ \l\\m6\\ \l\\ml\\ l/||m8||]; m. 为变换矩阵 P的行向量, i取值为 1-8; ||^ 为行向量 的向量范数;
5104, 正变换, 对 8x8 的图像残差数据 进行整数变换, 变换公式为 Y = PXPT; 其中 Ρ为所述的变换矩阵; 为所述的图像残差数据;
变换的基本单元是形如 y = Px的 8点一维变换, 其中
x = [xl x2 x3 x4 x5 x6 xl x8 , 输出的
j = [yl y2 y3 y4 y5 y6 yl , xl x2 x3 x4 x5 x6 xl x8表不整 数变换的一维正变换的输入八点数值; yl y3 y4 y5 y6 yl y8为正 变换的八点输出数值; 根据不同的整数变化基, 采用相应的整数变换计算方法。
S105, 对经过 8x8正变换后的数据 Y进行縮放处理; Y' = Y®PF
®是矩阵的点乘运算, Y是经过 8x8正变换后的数据, PF 是縮放矩阵, 是本縮放处理后的输出值, 代表视频数据变换到频域上的信息。
本发明实施例的变换方法能量集中性能优秀, 能量集中度接近 8x8DCT, 可以极大地消除视频数据在变换空间上的相关性; 计算复杂度低,可通过加法 和移位来实现, 易于硬件实现, 并且运算精度高, 不存在正反变换不匹配的问 题。 实施例二
以下为本发明实施例 2的一种视频编码 8x8整数变换方法,本发明实施例 2选取 [15, 15, 10, 3]为 8x8通用整数变换矩阵的整数变换基, 所述方法包 括以下步骤:
5201, 接收编码器中预测模块输出的 8x8的图像残差数据 ;
5202, 选取整数变换基, 构建变换矩阵 P;
8x8的通用整数变换矩阵为: k0 kO kO kO kO kO kO kO
k\ k2 k3 k4 -k4 -k3 -kl -kl
k5 k6 -k6 -k5 -k5 -k6 k6 k5
k2 -k4 -kl -k3 k3 kl k4 -kl
C=
kO -kO -kO kO kO -kO -kO kO
k3 -kl k4 kl -kl -k4 kl -k3
k6 -k5 k5 -k6 -k6 k5 -k5 k6
k4 -k3 kl -kl kl -kl k3 -k4 矩阵 C中, k0、 k5、 k6分别取值 1、 2、 1。 定义 k2 k3 为整数变换 基构建成变换矩阵 P。 本发明实施例选取 [15, 15, 10, 3]为整数变换基, 构建成变换矩阵 Pc 当变换基为 [15, 15, 10, 3]时, 变换矩阵 P为:
Figure imgf000011_0001
换矩阵 P, 构建:缩放矩阵 PF;
PF = Z z
Figure imgf000011_0002
\l\\m2\\ l/||m3|| l/\\m4\\ l/\\m5\\ l/\\m6\\ 1/ 1| m71| l/||m8||J; Ifli 为变换矩阵 P的行向量, i取值为 1-8; ||^ 为行向量 的向量范数;
当变换矩阵 P的变换基为 [15, 15, 10, 3]时, = [0.35355 0.029907 0.22361 0.029907 0.35355 0.029907 0.22361 0.029907] 縮放矩阵为:
0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
PF
0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
S204, 正变换, 对 8x8 的图像残差数据 进行整数变换, 变换公式为 Y = PXPT; 其中 Ρ为所述的变换矩阵; 为所述的图像残差数据; 变换的基本单元是形如 y = Px的 8点一维变换, 其中
X = [xl x2 x3 x4 x5 x6 xl xS]T ^ 输出的
j = [yl y2 y3 y4 y5 y6 yl yS]T ^ xl xl x3 x4 x5 x6 xl x8表示整数变 换的一维正变换的输入八点数值; ^ 2 3 4 5 ^为正变换的八 点输出数值;
所述整数变换的计算过程如下:
bl=xl-x8; b2=x2-x7; b3=x3_x6; b4=x4_x5;
b5=xl+x8; b6=x2+x7; b7=x3+x6; b8=x4+x5;
al=b5+b8; a2=b6+b7; a3=b5_b8; a4=b6_b7;
yl=al+a2; y5=al_a2; y3=a3«l+a4; y7=a3-a4«l ;
其中, bl、 b2、 b3、 b4、 b5、 b6、 b7、 b8, al、 a2、 a3、 a4 为整数变换 的中间变量;
再完成以下计;
Figure imgf000012_0001
cl=bl«4-bl ; c2=b2«4-b2 ; c3=b3〈〈4_b3; c4=b4〈〈4_b4;
dl=bl«3+bl« l ; d2=b2«3+b2«l ; d3=b3«3+b3«l ; d4=b4«3+b4«l ; el=bl«l+bl ; e2= b2«l+b2 ; e3= b3«l+b3; e4= b4«l+b4;
y2=cl+c2+d3+e4; y4=cl-e2-c3_d4; y6=dl_c2+e3+c4;
y8=el-d2+c3-c4;
其中, cl、 c2、 c3、 c4, dl、 d2、 d3、 d4, el、 e2、 e3、 e4 为整数变换 的中间变量; 其中 "〈〈"表示向左移位运算, 优先级高于加减法, 基本单元 y = ¾共需 要 40个加法和 14个移位。 此变换基的正变换快速计算方法如图 2所示,其中计算模块 Ml如图 3所 示。 其中 xl x2 x3 x4 x5 x6 xl x8表示整数变换的一维正变换的输入八点 数值; yl y3 y4 y5 y6 yl y8为正变换的八点输出数值;数据处理方向 是从左向右, 相交于一圆点的中间变量值等于圆点左侧线表示的数据相加; 方 形表示数据乘一个系数; "-"表示取负值; "2"表示乘 2, 即左移一位; "4" 表示乘 4, 即左移两位; "8"表示乘 8, 即左移 3位, "16"表示乘 16, 即左 移 4位。
"inl" "in2" "in3" "in4"表示 Ml模块的输入数据, 分别与 b4、 b3、 b2、 bl相连; "outl" "out2" "out3" "out4"表示 Ml模块的输出数据,分别与 y2、 y4、 y6、 y8相连。
S205, 对经过 8x8正变换后的数据 Y进行縮放处理;
Y' = Y®PF
®是矩阵的点乘运算, Y是经过 8x8正变换后的数据, PF 是縮放矩阵, 是本縮放处理后的输出值, 代表视频数据变换到频域上的信息。
本发明实施例的变换方法能量集中性能优秀, 能量集中度接近 8x8DCT, 可以极大地消除视频数据在变换空间上的相关性; 计算复杂度低,可通过加法 和移位来实现, 易于硬件实现, 并且运算精度高, 不存在正反变换不匹配的问 题。 实施例三
以下为本发明实施例 3的一种视频编码 8x8整数变换方法,本发明实施例 3选取 [6, 7, 5, 1]为 8x8通用整数变换矩阵的整数变换基,所述方法包括以下步 骤: 5301, 接收编码器中预测模块输出的 8x8的图像残差数据 X。
5302, 选取整数变换基, 构建变换矩阵 P;
8x8的通用整数变换矩阵为: kO kO kO kO kO kO kO kO
k\ k2 k3 k4 -k4 -k3 -kl -kl
k5 k6 -k6 -k5 -k5 -k6 k6 k5
k2 -k4 -kl -k3 k3 kl k4 -kl
C=
kO -kO -kO kO kO -kO -kO kO
k3 -kl k4 kl -kl -k4 kl -k3
k6 -k5 k5 -k6 -k6 k5 -k5 k6
k4 -k3 kl -kl kl -kl k3 -k4 矩阵 C中, kO、 k5、 k6分别取值 1、 2、 1。 定义 k2 k3 为整数 3 构建成变换矩阵?。
本发明实施例选取 [6, 7, 5, 1]为整数变换基, 构建成变换矩阵 P。
变换基为 [6, 7, 5, 1]时, 变换矩阵 P为:
1 1 1 1 1 1 1 1
6 7 5 1 -1 -5 -7 -6
2 1 -1 -2 -2 -1 1 2
7 -1 -6 -5 5 6 1 -7
1 -1 -1 1 1 -1 -1 1
5 -6 1 7 -7 -1 6 -5
1 -2 2 -1 -1 2 -2 1
1 -5 7 -6 6 -7 5 -1
S303, 根据变换矩阵 P, 构建縮放矩阵 PF;
其中,
[1/ II ml II \l\\m2\\ l/||m3|| \l\\mA\\ \l\\m5\\ \l\\m6\\ \l\\ml\\ l/||m8 换矩阵 P的行向量, i取值为 1-8; || ||为行向量 的向量范数; 当变换矩阵 P的变换基为 [6, 7, 5, 1]时, 0.35355 0.067116 0.22361 0.067116 0.35355 0.067116 0.22361 0.067116] 矩阵为:
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
PF
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
S304 , 正变换, 对 8x8 的图像残差数据 ^进行整数变换, 变换公式为 Y = PXPT ; 其中 Ρ为所述的变换矩阵; 为所述的图像残差数据;
变换的基本单元是形如 y = Px的 8点一维变换, 其中
Figure imgf000015_0001
换的一维正变换的输入八点数值; ^ y2 y3 y4 y5 y6 y1 ^为正变换的八 点输出数值;
所述整数变换的计算过程如下:
bl=xl-x8; b2=x2-x7; b3=x3-x6; b4=x4-x5 ;
b5=xl+x8; b6=x2+x7; b7=x3+x6; b8=x4+x5 ;
al=b5+b8; a2=b6+b7; a3=b5-b8; a4=b6-b7;
yl=al+a2; y5=al-a2; y3=a3«l+a4; y7=a3-a4«l;
其中, bl、 b2、 b3、 b4、 b5、 b6、 b7、 b8, al、 a2、 a3、 a4 为整数变换 的中间变量;
再完成以下计算步骤:
Figure imgf000016_0001
cl=bl«2+bl«l+b4; c2=b2«2+b2«l-a3 ; c3=b2+b3«2+b3«l; c4 =-al+b4«2+b4«l;
y2=cl+c2+c3 ; y4=cl-c3-c4; y6=cl-c2+c4; y8=-c2+c3-c4; 其中, cl、 c2、 c3、 c4为整数变换的中间变量; 其中 "<<"表示向左移位运算, 优先级高于加减法, 基本单元 y = Px共需 要 32个加法和 10个移位。 此变换基的正变换快速计算方法如图 4所示,其中计算模块 M2如图 5所 示。 其中 xl x2 x3 x4 x5 x6 xl x8表示整数变换的一维正变换的输入四点 数值; yl y3 y4 y5 y6 yl y8为正变换的四点输出数值;数据处理方向 是从左向右, 相交于一圆点的中间变量值等于圆点左侧线表示的数据相加; 方 形表示数据乘一个系数; "-"表示取负值; "2"表示乘 2, 即左移一位; "4" 表示乘 4, 即左移两位。
"inl " "in2" "in3 " "in4"表示 M2模块的输入数据, 分别与 b4、 b3、 b2、 bl相连; "outl " "out2" "out3 " "out4"表示 M2模块的输出数据,分别与 y2、 y4、 y6、 y8相连。
S305 , 对经过 8x8正变换后数据 Y进行縮放处理;
Y' = Y ® PF
®是矩阵的点乘运算, Y是经过 8x8正变换后的数据, PF 是縮放矩阵, 是 本縮放处理后的输出值, 代表视频数据变换到频域上的信息。
本发明实施例的变换方法能量集中性能优秀, 能量集中度接近 8x8DCT, 可以极大地消除视频数据在变换空间上的相关性; 计算复杂度低,可通过加法 和移位来实现,易于硬件实现,并且运算精度高,不存正反变换不匹配的问题。 本领域的普通技术人员可以理解,实现上述实施例方法中的全部或部分步 骤是可以通过程序指令相关硬件来完成的,所述的程序可以存储于一计算机可 读取存储介质中, 所述的存储介质可以为 R0M、 RAM, 磁盘、 光盘等。
以上所述仅为本发明的较佳实施例而已, 并不用以限制本发明, 凡在本发 明的精神和原则之内所作的任何修改、等同替换和改进等, 均应包含在本发明 的保护范围之内。

Claims

权 利 要 求
1、 一种视频编码 8x8整数变换方法, 包括
接收编码器中预测模块输出的 8x8的图像残差数据 ;
选取整数变换基, 构建变换矩阵 P,
8x8的通用整数变换矩阵为: k0 kO kO kO kO kO kO kO
k\ k2 k3 k4 -k4 -k3 -kl -kl
k5 k6 -k6 -k5 -k5 -k6 k6 k5
k2 -k4 -kl -k3 k3 kl k4 -kl
kO -kO -kO kO kO -kO -kO kO
k3 -kl k4 kl -kl -k4 kl -k3
k6 -k5 k5 -k6 -k6 k5 -k5 k6
k4 -k3 kl -kl kl -kl k3 -k4 矩阵 C中, k0、 k5、 k6分别取值 1、 2、 1, -k k2 k3 为整数变换 正变换, 对 8x8的图像残差数据 进行整数变换, 变换公式为 Γ = Ρ Ρ 中 P为所述的变换矩阵; 为所述的图像残差数据;
其特征在于:
所述整数变换基为以下数值中的一个:
[15, 15, 10, 3]、 [6, 7, 5, 1]。
2、 如权利要求 1所述的视频编码 8x8整数变换方法, 其特征在于, 所 i "选取整数变换基, 构建变换矩阵 P"之后还可以包括步骤:
根据变换矩阵 P, 构建縮放矩阵 PF, 其中, ml|| \l\\m2\\ l/||m3|| \l\\mA\\ \l\\m5\\ \l\\m6\\ \l\\ml\\ l/||m8||]; 为变换矩阵 P的行向量, i取值为 1-8; ||^ 为行向量 的向量范数。
3、 如权利要求 2所述的视频编码 8x8整数变换方法, 其特征在于, 所述步 骤 "选取整数变换基, 构建变换矩阵 P"之后还可以包括步骤: 对经过 8x8正变换后的数据 Y进行縮放处理, Y' = Y®PF
®是矩阵的点乘运算, Y是经过 8x8正变换后的数据, PF 是縮放矩阵, 是本縮放处理后的输出值, 代表视频数据变换到频域上的信息。
4、 如权利要求 1所述的视频编码 8x8整数变换方法, 其特征在于, 所述正 变换中的变换的基本单元是形如 y = Px的 8点一维变换, 其中 x = [xl x2 x3 x4 x5 x6 xl 8f , 输出的 y = [yl yl y3 y4 y5 y6 yl ySf , xl x2 x3 x4 x5 x6 xl J8为所述正变 换的输入八点数值; yl yl y3 y4 y5 y6 yl y8为所述正变换的八点输出数 值。
5、 如权利要求 2所述的视频编码 8x8整数变换方法, 其特征在于, 所述 整数变换基取值为 [15, 15, 10, 3]时:
1 1 1 1 1 1 1 1
15 15 10 3 -3 -10 -15 -15
2 1 -1 -2 -2 -1 1 2
15 -3 -15 -10 10 15 3 -15
:述变换矩阵 P=
1 -1 -1 1 1 -1 -1 1
10 -15 3 15 -15 -3 15 -10
1 -2 2 -1 -1 2 -2 1
3 -10 15 -15 15 -15 10 -3
= [0.35355 0.029907 0.22361 0.029907 0.35355 0.029907 0.22361 0.029907]
0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
PF =
0.125 0.010574 0.079057 0.010574 0.125 0.010574 0.079057 0.010574
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
0.079057 0.0066875 0.05 0.0066875 0.079057 0.0066875 0.05 0.0066875
0.010574 0.00089445 0.0066875 0.00089445 0.010574 0.00089445 0.0066875 0.00089445
6、 如权利要求 1所述的视频编码 8x8整数变换方法, 其特征在于, 所述 当整数变换基取值为 [15, 15, 10, 3]时, "对 8x8的图像残差数据 进行整 数变换, 变换公式为 r = P W计算过程如下:
bl= =xl-x8; b2= =x2~x7; b3= =x3~x6; b4= =x4~x5; b5= =xl+x8; b6= =x2+x7; b7= =x3+x6; b8= =x4+x5; al= =b5+b8; a2= =b6+b7; a3= =b5-b8; a4= =b6-b7; yl= =al+a2; y5= =al~a2; y3= =a3«l+a4; y7=a3-a4«l; 再; 成以下计駕 〔步骤:
Figure imgf000020_0001
cl=bl«4-bl; c2=b2«4-b2; c3=b3«4-b3; c4=b4〈〈4_b4; dl=bl«3+bl«l; d2=b2«3+b2«l; d3=b3«3+b3«l; d4=b4«3+b4«l el=bl«l+bl; e2= b2«l+b2; e3= b3«l+b3; e4= b4«l+b4 y2=cl+c2+d3 +e4; y4=cl-e2-c3_d4; y6=dl_ c2 + e3 + c4; y8=el - d2 + c3 - c4;
其中 "〈〈"表示向左移位运算, 优先级高于加减法。
7、 如权利要求 2所述的视频编码 8x8整数变换方法, 其特征在于, 所述 整数变换基取值为 [6, 7, 5, 1]时:
1 1 1 1 1 1 1 1
6 7 5 1 -1 -5 -7 -6
2 1 -1 -2 -2 -1 1 2
7 -1 -6 -5 5 6 1 -7
:述变换矩阵 P=
1 -1 -1 1 1 -1 -1 1
5 -6 1 7 -7 -1 6 -5
1 -2 2 -1 -1 2 -2 1
1 -5 7 -6 6 -7 5 -1 = [0.35355 0.067116 0.22361 0.067116 0.35355 0.067116 0.22361 0.067116] 縮放矩阵为:
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
PF
0.125 0.023729 0.079057 0.023729 0.125 0.023729 0.079057 0.023729
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
0.079057 0.015008 0.05 0.015008 0.079057 0.015008 0.05 0.015008
0.023729 0.0045045 0.015008 0.0045045 0.023729 0.0045045 0.015008 0.0045045
8、 如权利要求 1所述的视频编码 8x8整数变换方法, 其特征在于, 所 当整数变换基取值为 [6, 7, 5, 1]时, "对 8x8的图像残差数据 进行整数变 变换公式为 y = PXPT "计算过程如下: bl=xl-x8; b2= =x2-x7; b3=x3-x6; b4=x4-x5;
b5=xl+x8; 6 =x2+x7; b7=x3+x6; b8=x4+x5;
al=b5+b8; a2= =b6+b7; a3=b5-b8; a4=b6-b7;
yl=al+a2; y5= =al-a2; y3=a3«l+a4; y7=a3-a4«l; 再完成以下计算步骤:
Figure imgf000021_0001
Figure imgf000022_0001
y y y y;;;ccc3c2l24lc3c6ccc4l24∞cc3c2++++-----====
cabb4l424l++««-
Cbl=
PCT/CN2011/073586 2011-04-20 2011-04-29 一种视频编码8x8整数变换方法 WO2012142777A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN 201110101417 CN102227135A (zh) 2011-04-20 2011-04-20 一种视频编码8x8整数变换方法
CN201110101417.3 2011-04-20

Publications (1)

Publication Number Publication Date
WO2012142777A1 true WO2012142777A1 (zh) 2012-10-26

Family

ID=44808088

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/073586 WO2012142777A1 (zh) 2011-04-20 2011-04-29 一种视频编码8x8整数变换方法

Country Status (2)

Country Link
CN (2) CN102227135A (zh)
WO (1) WO2012142777A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104486629B (zh) * 2014-12-16 2017-10-20 北京金山云网络技术有限公司 H.265视频编码中变换与反变换的快速实施方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1564602A (zh) * 2004-03-18 2005-01-12 华中科技大学 视频编码的整数变换矩阵选择方法及相关的整数变换方法
CN101931817A (zh) * 2010-04-12 2010-12-29 中山大学 一种变换基自适应的变换编码方法
CN101977321A (zh) * 2010-11-04 2011-02-16 深圳市融创天下科技发展有限公司 一种视频编码的整数变换方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3129392B2 (ja) * 1996-02-02 2001-01-29 日本電気株式会社 2次元idct回路
CN1209926C (zh) * 2003-03-26 2005-07-06 浙江大学 运用于图像编码和视频编码的整数变换方法和装置
CN1210960C (zh) * 2003-07-03 2005-07-13 浙江大学 运用于图像编码和视频编码的8×8整数变换方法和装置
US8069201B2 (en) * 2003-11-25 2011-11-29 Texas Instruments Incorporated 8×8 transform and quantization
CN100477796C (zh) * 2005-12-27 2009-04-08 中国科学院计算技术研究所 用于视频转换的变换系数块的转换方法
CN100563337C (zh) * 2007-10-24 2009-11-25 上海广电(集团)有限公司中央研究院 基于avs的整数变换量化和整数反变换反量化方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1564602A (zh) * 2004-03-18 2005-01-12 华中科技大学 视频编码的整数变换矩阵选择方法及相关的整数变换方法
CN101931817A (zh) * 2010-04-12 2010-12-29 中山大学 一种变换基自适应的变换编码方法
CN101977321A (zh) * 2010-11-04 2011-02-16 深圳市融创天下科技发展有限公司 一种视频编码的整数变换方法

Also Published As

Publication number Publication date
CN102227135A (zh) 2011-10-26
CN102281445A (zh) 2011-12-14

Similar Documents

Publication Publication Date Title
KR102149828B1 (ko) 비디오의 부호화 방법 및 장치, 복호화 방법 및 장치
AU2005237142B2 (en) Reversible transform for lossy and lossless 2-D data compression
CN108200439B (zh) 提高数字信号变换性能的方法及数字信号变换方法和装置
AU2002259268C1 (en) Apparatus and method for encoding and computing a discrete cosine transform using a butterfly processor
JP2005513588A5 (zh)
WO2002093750A2 (en) Apparatus and method for decoding and computing an inverse discrete cosine transform using a butterfly processor
Brahimi et al. An efficient fast integer DCT transform for images compression with 16 additions only
WO2012142777A1 (zh) 一种视频编码8x8整数变换方法
JP2010081618A5 (zh)
CN113592966B (zh) 图像处理方法及装置、电子设备和存储介质
KR101722215B1 (ko) 이산 여현 변환 장치 및 방법
TWI468015B (zh) 視訊解碼的平行處理裝置及方法
WO2012142776A1 (zh) 一种视频编码4x4整数变换方法
KR101395143B1 (ko) 영상처리의 정수변환 방법 및 정수변환 장치
WO2012175033A1 (zh) 图像变换处理方法、设备和系统
JP4740992B2 (ja) オーバーラップフィルタリングとコア変換を行うための方法及び装置
CN206962992U (zh) 用于数字视频编解码的3乘3整数dct变换量化器
JP2006065678A5 (zh)
CN113286150B (zh) 用于视频编解码的变换编码硬件实现方法、装置及设备
WO2012142792A1 (zh) 一种视频编码16×16整数变换方法
KR100790846B1 (ko) 영상처리의 정수 변환 방법 및 그의 프로세서
Saleh Efficient mid-band exchange coefficient watermarking system
Li et al. Efficient SIMD acceleration of DCT and IDCT for high efficiency video coding
WO2014060637A1 (en) Image processing method, devices and system
Udomsiri et al. Comparative study on recent integer DCTs

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11864077

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11864077

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 14/03/14)

122 Ep: pct application non-entry in european phase

Ref document number: 11864077

Country of ref document: EP

Kind code of ref document: A1