-
1.发明专利
公开(公告)号：ES2392437T3
公开(公告)日：2012-12-10
申请号：ES05712072
申请日：2005-01-26
Applicant: PANASONIC CORP
Inventor： LU JIUHUAI , CHEN TAO , KASHIWAGI YOSHIICHIRO , KADONO SHINYA , LIM CHONG SOON
Abstract: Procedimiento de decodificación de imágenes para decodificar cada una de una pluralidad de imágenescodificadas por bloque utilizando una matriz de cuantificación, comprendiendo dicho procedimiento las etapassiguientes:obtener, a partir de un flujo codificado, una matriz de cuantificación distinta de una matriz de cuantificaciónpredefinida y un ID de matriz para identificar la matriz de cuantificación, y reservar la matriz de cuantificación y elID de matriz;extraer, a partir del flujo continuo codificado, un ID de matriz que se añade a los datos generados codificandouna imagen actual y que se utiliza para identificar una matriz de cuantificación que se ha utilizado para codificarla imagen actual;identificar, a partir de las matrices de cuantificación reservadas en dicha etapa de reserva, una matriz decuantificación correspondiente al ID de matriz ydecodificar los datos de la imagen codificada actual utilizando la matriz de cuantificación identificada,componiéndose cada imagen de un componente luma, un primer componente croma y un segundo componentecroma, y estando el procedimiento de decodificación de imágenes caracterizado porque comprende las etapassiguientes:una primera etapa, en la que en caso de que exista una matriz de cuantificación para el componente luma,una matriz de cuantificación para el primer componente croma y una matriz de cuantificación para el segundocomponente croma, por separado, en la matriz de cuantificación identificada por el ID de matriz extraído, lamatriz de cuantificación para el componente luma se identifica como matriz de cuantificación para uncomponente luma de la imagen actual, la matriz de cuantificación para el primer componente croma seidentifica como matriz de cuantificación para un primer componente croma de la imagen actual, y la matriz decuantificación para el segundo componente croma se identifica como matriz de cuantificación para unsegundo componente croma de la imagen actual,una segunda etapa, en la que, en caso de que la matriz de cuantificación para el primer componente cromano esté presente y la matriz de cuantificación para el segundo componente croma esté presente en la matrizde cuantificación identificada por el ID de matriz extraído, se identifica la matriz de cuantificación para elsegundo componente croma, en lugar de la matriz de cuantificación predefinida, como matriz decuantificación para el primer componente croma de la imagen actual, yuna tercera etapa, en la que en caso de que tanto la matriz de cuantificación para el primer componentecroma y la matriz de cuantificación para el segundo componente croma no estén presentes en la matriz decuantificación identificada por el ID de matriz extraído, se identifica la matriz de cuantificación para elcomponente luma, en lugar de la matriz de cuantificación predefinida, como la matriz de cuantificación para elprimer componente croma de la imagen actual y el segundo componente croma de la imagen actual.
-
公开(公告)号：AU2004214795A9
公开(公告)日：2009-05-28
申请号：AU2004214795
申请日：2004-02-18
Applicant: PANASONIC CORP
Inventor： KOZUKA MASAYUKI , LU JIUHUAI , KADONO SHINYA , KASHIWAGI YOSHIICHIRO
Abstract: The picture coding method of the present invention is a picture coding method for coding a picture on a block-by-block basis, comprising: a selection step of selecting one of at least two sizes as a size of a block on which orthogonal transformation should be performed; a transformation step of performing orthogonal transformation on a block having the selected size; a coding step of coding data of said block obtained in the transformation step; and a generation step of generating a coded stream that includes the coded data of the block and size information concerning the size selected in the selection step, wherein the size information indicates whether or not the size is a fixed block size within a predetermined section in the coded stream, and the predetermined section is one of a sequence, a group of pictures, a picture, a slice, and a macroblock.
-
3.发明公开
公开(公告)号：EP1714484A4
公开(公告)日：2009-03-18
申请号：EP05712071
申请日：2005-01-26
Applicant: PANASONIC CORP
Inventor： CHEN TAO , LU JIUHUAI , KASHIWAGI YOSHIICHIRO , KADONO SHINYA , LIM CHONG SOON
CPC classification number: H04N19/42 , H04N19/126 , H04N19/46 , H04N19/61
Abstract: The picture decoding method according to the present invention is a decoding method for decoding coded pictures by inverse quantization (IQ1) and inverse orthogonal transformation, in which a quantization matrix which defines a scaling ratio of a quantization step for each component is multiplied by a multiplier (IQ11), which is a coefficient for frequency transformation or a quantization step, and also, a result of the multiplication is multiplied by a quantized value, as a process of inverse quantization.
-
公开(公告)号：EP1709801A4
公开(公告)日：2010-11-24
申请号：EP05712072
申请日：2005-01-26
Applicant: PANASONIC CORP
Inventor： LU JIUHUAI , CHEN TAO , KASHIWAGI YOSHIICHIRO , KADONO SHINYA , LIM CHONG SOON
CPC classification number: H04N19/188 , H04N19/126 , H04N19/159 , H04N19/169 , H04N19/186 , H04N19/61 , H04N19/70
-
5.发明公开PICTURE CODING METHOD, PICTURE DECODING METHOD, PICTURE CODING APPARATUS, PICTURE DECODING APPARATUS, AND PROGRAM THEREOF 审中-公开图像编码方法，图像解码方法; 图像编码装置; 图像解码装置和程序THEREFOR
公开(公告)号：EP1665133A4
公开(公告)日：2009-05-13
申请号：EP05711937
申请日：2005-01-19
Applicant: PANASONIC CORP
Inventor： KADONO SHINYA , KASHIWAGI YOSHIICHIRO , LU JIUHUAI , LIM CHONG SOON , TAO CHEN
CPC classification number: H04N19/136 , H04N19/126 , H04N19/176 , H04N19/196 , H04N19/463 , H04N19/61 , H04N19/91
Abstract: The picture coding method according to the present invention is a picture coding method for coding a picture on a block-by-block basis through orthogonal transformation (Fig. 3, T) and quantization (Fig. 3, Q), and coding a quantization matrix (Fig. 2) that is used to derive quantization steps for frequencies of orthogonal transformation coefficients, the method comprising: calculating a difference value between each of frequency components included in the quantization matrix (Fig. 2) and a predetermined value corresponding to said each of the frequency components; and coding the difference value into a variable length code (Fig. 3, VLC), wherein a code length of the variable length code is shorter as the difference value is smaller, or equal to a code length of a neighboring difference value of said difference value (Fig. 4).
-
Patent Agency Ranking
公开(公告)号：JP2011109711A
公开(公告)日：2011-06-02
申请号：JP2011023311
申请日：2011-02-04
Applicant: Panasonic Corp , パナソニック株式会社
Inventor： CHEN TAO , LU JIUHUAI , KASHIWAGI YOSHIICHIRO , SUMINO SHINYA , CHON SUN RIMU
CPC classification number: H04N19/42 , H04N19/126 , H04N19/46 , H04N19/61
Abstract: PROBLEM TO BE SOLVED: To provide an image decoding method which reduces computing loads of quantization processing and inverse quantization processing. SOLUTION: In the image coding method, a component value which is the i-th in a vertical direction and the j-th in a horizontal direction of a weighting matrix is multiplied by a normalization value and a level scale value is obtained. The normalization value is a predetermined value according to (1) whether the i and the j are both even numbers, both odd numbers, or either one is an odd number and (2) a remainder obtained by dividing the quantization parameter by 6. The quantized orthogonal transformation coefficient is multiplied by the level scale value, and an inversely quantized orthogonal transformation coefficient is calculated by shifting the result of the multiplication for the number of bits corresponding to the quotient obtained by dividing the quantization parameter by 6. A block image is created by the inverse orthogonal transformation of the inversely quantized orthogonal transformation coefficient and the image is decoded. The decoded image is stored as a reference image. COPYRIGHT: (C)2011,JPO&INPIT
Abstract translation: 解决的问题：提供减少量化处理和逆量化处理的计算负荷的图像解码方法。 解决方案：在图像编码方法中，将加权矩阵的垂直方向上的第i个第i个和水平方向上的第j个分量值乘以归一化值，并获得电平比例值 。 归一化值是根据（1）i和j是偶数，奇数还是奇数，或者是奇数，和（2）通过将量化参数除以6得到的余数的预定值。 量化正交变换系数乘以电平标度值，并通过将与通过将量化参数除以6获得的商相对应的位数乘以乘法结果来计算逆量化的正交变换系数。块图像是 通过逆量化正交变换系数的逆正交变换创建的图像被解码。 解码图像被存储为参考图像。 版权所有（C）2011，JPO＆INPIT
-
7.发明专利Coding method of quantization matrix, coding apparatus of moving image and moving image signal stream 有权量化矩阵的编码方法，移动图像的编码设备和移动图像信号流
公开(公告)号：JP2011101420A
公开(公告)日：2011-05-19
申请号：JP2011010283
申请日：2011-01-20
Applicant: Panasonic Corp , パナソニック株式会社
Inventor： SUMINO SHINYA , KASHIWAGI YOSHIICHIRO , LU JIUHUAI , CHON SUN RIMU , CHEN TAO
CPC classification number: H04N19/136 , H04N19/126 , H04N19/176 , H04N19/196 , H04N19/463 , H04N19/61 , H04N19/91
Abstract: PROBLEM TO BE SOLVED: To provide a coding method for improving the coding efficiency of a quantization matrix. SOLUTION: The method is for coding the quantization matrix in which respective components take an integer value from 1 to 255. A difference between the Ist element of a numerical sequence in which the components of the quantization matrix are arranged in a prescribed order and the I-1st element of the numerical sequence is calculated. A value to which the difference and +256 are added is turned to an offset difference value when the difference is smaller than -128, a value for which the difference and -256 are added is turned to the offset difference value when the difference is the same as or larger than +128, and the difference is generated as the offset difference value when the difference is the same as or larger than -128 and is smaller than +128, and the generated offset difference value is variable-length coded. COPYRIGHT: (C)2011,JPO&INPIT
Abstract translation: 解决的问题：提供一种用于提高量化矩阵的编码效率的编码方法。 解决方案：该方法用于编码量化矩阵，其中各个分量取从1到255的整数值。数字序列的Ist元素之间的差异，其中量化矩阵的分量以规定的顺序排列 并计算数值序列的I-1st元素。 当差值小于-128时，将差和+256相加的值转换为偏差差值，当差为...时，将差和-256相加的值转换为偏移差值 相同或者大于+128，并且当该差值等于或大于-128并且小于+128时产生差异作为偏差差值，并且生成的偏移差值是可变长度编码的。 版权所有（C）2011，JPO＆INPIT
-
8.发明专利Moving picture encoding method, moving picture encoder, program, and recording medium 有权移动图像编码方法，移动图像编码器，程序和记录介质
公开(公告)号：JP2011091848A
公开(公告)日：2011-05-06
申请号：JP2010276437
申请日：2010-12-10
Applicant: Panasonic Corp , パナソニック株式会社
Inventor： LU JIUHUAI , CHEN TAO , KASHIWAGI YOSHIICHIRO , SUMINO SHINYA , CHON SUN RIMU
CPC classification number: H04N19/188 , H04N19/126 , H04N19/159 , H04N19/169 , H04N19/186 , H04N19/61 , H04N19/70
Abstract: PROBLEM TO BE SOLVED: To provide a moving picture encoding method capable of reducing the amount of encoding to efficiently perform encoding and decoding. SOLUTION: The method for encoding moving pictures, includes the steps of: quantization matrix encoding step for encoding a second quantization matrix and associating a matrix ID which specifies the second quantization matrix with the second encoded quantization matrix to output as a code train, if the second quantization matrix is used when a plurality of pictures are encoded; picture encoding step for encoding the current picture using at least one of a default quantization matrix and the second quantization matrix to outputs picture encoded data as a code train; and matrix ID grant step for associating a matrix ID which specifies a quantization matrix used when the current picture is encoded with the picture encoded data. COPYRIGHT: (C)2011,JPO&INPIT
Abstract translation: 要解决的问题：提供能够减少编码量以有效执行编码和解码的运动图像编码方法。 解决方案：用于编码运动图像的方法包括以下步骤：量化矩阵编码步骤，用于对第二量化矩阵进行编码，并将指定第二量化矩阵的矩阵ID与第二编码量化矩阵相关联，以作为代码序列输出 如果在编码多个图像时使用第二量化矩阵; 图像编码步骤，用于使用默认量化矩阵和第二量化矩阵中的至少一个对当前图像进行编码，以输出图像编码数据作为代码串; 以及矩阵ID授权步骤，用于将指定当前图像被编码的量化矩阵的矩阵ID与图像编码数据相关联。 版权所有（C）2011，JPO＆INPIT
-
9.发明专利Moving picture decoding method, moving picture decoder, program, and recording medium 有权移动图像解码方法，移动图像解码器，程序和记录介质
公开(公告)号：JP2011091847A
公开(公告)日：2011-05-06
申请号：JP2010276436
申请日：2010-12-10
Applicant: Panasonic Corp , パナソニック株式会社
Inventor： LU JIUHUAI , CHEN TAO , KASHIWAGI YOSHIICHIRO , SUMINO SHINYA , CHON SUN RIMU
CPC classification number: H04N19/188 , H04N19/126 , H04N19/159 , H04N19/169 , H04N19/186 , H04N19/61 , H04N19/70
Abstract: PROBLEM TO BE SOLVED: To provide a moving picture decoding method capable of reducing the amount of encoding to efficiently perform encoding and decoding. SOLUTION: The method for decoding moving pictures, includes the steps of: holding step for acquiring a second quantization matrix and a matrix ID from a code train to hold them; step for extracting a matrix ID, which accompanies picture encoded data and specifies a quantization matrix used when the current picture is encoded, from the code train; and quantization matrix specifying step for specifying a quantization matrix corresponding to the extracted matrix ID. A picture is composed of a luminance component and two color difference components. In the quantization matrix specifying step, when a quantization matrix for a color difference component does not exist in a quantization matrix specified by the matrix ID, a quantization matrix for another color difference component is specified as a quantization matrix used for color difference component. COPYRIGHT: (C)2011,JPO&INPIT
Abstract translation: 要解决的问题：提供能够减少编码量以有效执行编码和解码的运动图像解码方法。 解决方案：用于解码运动图像的方法包括以下步骤：从代码串获取第二量化矩阵和矩阵ID以保存它们的步骤; 步骤，用于提取伴随图像编码数据的矩阵ID，并且从代码串中指定当前图像被编码时使用的量化矩阵; 以及量化矩阵确定步骤，用于指定与所提取的矩阵ID相对应的量化矩阵。 图像由亮度分量和两个色差分量组成。 在量化矩阵确定步骤中，当在由矩阵ID指定的量化矩阵中不存在用于色差分量的量化矩阵时，将用于另一色差分量的量化矩阵指定为用于色差分量的量化矩阵。 版权所有（C）2011，JPO＆INPIT
-
10.发明专利
公开(公告)号：JP2009044772A
公开(公告)日：2009-02-26
申请号：JP2008298829
申请日：2008-11-21
Applicant: Panasonic Corp , パナソニック株式会社
Inventor： SUMINO SHINYA , LU JIUHUAI , KASHIWAGI YOSHIICHIRO , KOZUKA MASAYUKI
CPC classification number: G06T9/004 , H04N19/117 , H04N19/159 , H04N19/172 , H04N19/423 , H04N19/46 , H04N19/61 , H04N19/80 , H04N19/82
Abstract: PROBLEM TO BE SOLVED: To provide an image encoding method capable of effectively encoding images without damaging an image quality in a movie picture and the like, and to provide an image decoding method. SOLUTION: The image encoding method performs prediction encoding process to the picture in reference to the picture obtained from the decoded picture after encoding. Further, the method performs filtering to the decoded picture, stores the picture after filtering to a memory as a reference picture in two pictures after and before filtering one decoded picture, and stores the picture before filtering in the two pictures as an output picture to the memory. COPYRIGHT: (C)2009,JPO&INPIT
Abstract translation: 要解决的问题：提供一种能够有效地编码图像而不损害电影图像等中的图像质量的图像编码方法，并提供图像解码方法。 解决方案：图像编码方法参照从编码后的解码图像获得的图像对图像执行预测编码处理。 此外，该方法对解码图像进行滤波，将滤波后的图像作为参考图像存储在作为参考图像的两个图像中，并且在对一个解码图像进行滤波之前和之后，并将滤波前的图像作为输出图像存储到两个图像中 记忆。 版权所有（C）2009，JPO＆INPIT
-
-
-
-
-
-
-
-
-
Search Results
10
records
(took 0.028 seconds)
