公开(公告)号：KR1020140146560A

公开(公告)日：2014-12-26

申请号：KR1020140148730

申请日：2014-10-29

Applicant: 삼성전자주식회사

Inventor： 알쉰알렉산더 ,  알쉬나엘레나 ,  천지엔러 ,  쉬라코프니콜라이 ,  한우진 ,  홍윤미

IPC: H04N19/117 ,  H04N19/186

CPC classification number: H04N19/86 ,  G06F17/17 ,  G06F17/175 ,  G06T3/4007 ,  G06T5/002 ,  H04N19/117 ,  H04N19/182 ,  H04N19/186 ,  H04N19/42 ,  H04N19/43 ,  H04N19/521 ,  H04N19/523 ,  H04N19/59 ,  H04N19/635 ,  H04N19/80 ,  H04N19/82

Abstract: Provided is a method of the present invention to interpolate an image by determining an interpolation filter coefficient. The disclosed image interpolation method considering a smoothing: selects an interpolation filter differently based on the interpolation position and smoothness of a subpixel unit from interpolation filters for generating the pixel value of at least one subpixel unit positioned between pixel units; interpolates pixel values of pixels of a pixel unit by using the selected interpolation filter; and generates the pixel value of at least one subpixel unit.

Abstract translation: 本发明提供了一种通过确定内插滤波器系数来内插图像的方法。 所公开的考虑平滑的图像插值方法：根据内插滤波器，根据位于像素单元之间的至少一个子像素单元的像素值的内插位置和子像素单元的平滑度，不同地选择插值滤波器; 通过使用所选择的内插滤波器来内插像素单元的像素的像素值; 并生成至少一个子像素单元的像素值。

公开(公告)号：KR1020140146559A

公开(公告)日：2014-12-26

申请号：KR1020140148729

申请日：2014-10-29

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더

IPC: H04N19/117 ,  H04N19/14

CPC classification number: H04N19/463 ,  G06T5/001 ,  H04N19/117 ,  H04N19/137 ,  H04N19/14 ,  H04N19/176 ,  H04N19/182 ,  H04N19/186 ,  H04N19/189 ,  H04N19/44 ,  H04N19/50 ,  H04N19/625 ,  H04N19/85 ,  H04N19/89 ,  H04N19/90

Abstract: The present invention proposes a video encoding method, an apparatus therefor, a video decoding method and an apparatus therefor, which generate a restored image having a minimum error with respect to an original image. The video decoding method includes: parsing the offset merge information of a current block from a received bitstream; restoring an offset type and offset values among the offset parameters of the current block based on the offset merge information; determining an edge class or a pixel value band of a restored pixel based on an edge type or a pixel value band type of the current block, which the offset type indicates; and determining an offset value corresponding to the edge class or the pixel value band of the restored pixel from the offset values and adjusting a pixel value of the restored pixel according to the determined offset value.

Abstract translation: 本发明提出了一种视频编码方法，其装置，视频解码方法及其装置，其生成相对于原始图像具有最小误差的恢复图像。 视频解码方法包括：从接收的比特流解析当前块的偏移合并信息; 基于偏移合并信息，恢复当前块的偏移参数中的偏移类型和偏移值; 基于偏移类型表示的当前块的边缘类型或像素值带类型来确定恢复像素的边缘类别或像素值带; 以及根据所述偏移值确定与所述恢复像素的边缘类别或像素值带对应的偏移值，并且根据所确定的偏移值来调整所述恢复像素的像素值。

公开(公告)号：KR1020140122214A

公开(公告)日：2014-10-17

申请号：KR1020140115700

申请日：2014-09-01

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더

IPC: H04N19/117 ,  H04N19/523 ,  H04N19/82

CPC classification number: H04N19/523 ,  G06T3/4007 ,  H04N19/117 ,  H04N19/167 ,  H04N19/182 ,  H04N19/52 ,  H04N19/537 ,  H04N19/59 ,  H04N19/61 ,  H04N19/82 ,  H04N19/80

Abstract: 본 발명은 변환 기반의 보간 필터를 이용하여 부화소단위로 영상을 보간하는 방법을 개시한다. 화소단위 픽셀들 사이에 위치하는 적어도 하나의 부화소단위의 픽셀값을 생성하기 위한 보간 필터들 중에서, 부화소단위의 보간 위치에 따라 개별적으로, 상기 보간 위치를 중심으로 대칭적인 형태의 대칭형 보간 필터 또는 비대칭적인 형태의 비대칭형 보간 필터를 선택하고, 선택된 보간 필터를 이용하여 화소단위 픽셀들의 픽셀값들을 보간하여, 적어도 하나의 부화소단위의 픽셀값을 생성하는 영상 보간 방법이 개시된다.

Abstract translation: 公开了一种使用基于变换的内插滤波器来内插子像素单元中的图像的方法。 用于内插图像的方法包括：在用于生成至少一个子像素的像素值的内插滤波器中，基于根据子像素单元的插值位置的插值位置，单独选择对称内插滤波器或非对称内插滤波器 位于整数像素单元的像素之间的像素单元; 以及通过使用所选择的内插滤波器内插所述整数像素单元的像素的像素值来生成所述至少一个子像素单元的像素值。

公开(公告)号：KR1020140117326A

公开(公告)日：2014-10-07

申请号：KR1020140115699

申请日：2014-09-01

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더 ,  세레긴바딤 ,  쉬라코프니콜라이 ,  코로티프막심

IPC: H04N19/119 ,  H04N19/105 ,  H04N19/11

CPC classification number: H04N19/119 ,  H04N19/105 ,  H04N19/11

Abstract: A method of decoding an image is disclosed. The method of decoding an image includes: determining encoding units of a hierarchical structure by using information of whether an encoding unit parsed from a received bitstream is divided; obtaining a first predictive value of the current encoding unit by using neighboring pixels of the current encoding unit among the determined encoding units of hierarchical structure; obtaining a second predictive value of each pixel adjacent to the upper and left boundaries of the current encoding unit by using at least one neighboring pixel of each pixel; and obtaining a final predictive value of the current encoding unit by using the first predictive value and the second predictive value.

Abstract translation: 公开了解码图像的方法。 解码图像的方法包括：通过使用分割从接收比特流解码的编码单位的信息来确定分层结构的编码单位; 通过使用所确定的层次结构的编码单元中的当前编码单元的相邻像素来获得当前编码单元的第一预测值; 通过使用每个像素的至少一个相邻像素来获得与当前编码单元的上边界和左边界相邻的每个像素的第二预测值; 以及通过使用所述第一预测值和所述第二预测值来获得所述当前编码单元的最终预测值。

公开(公告)号：KR1020140043037A

公开(公告)日：2014-04-08

申请号：KR1020130116915

申请日：2013-09-30

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더 ,  김찬열 ,  이태미

IPC: H04N19/50 ,  H04N19/00

CPC classification number: H04N19/46 ,  H04N19/117 ,  H04N19/14 ,  H04N19/182 ,  H04N19/33 ,  H04N19/59 ,  H04N19/82 ,  H04N19/86

Abstract: Provided are a method of coding a scalable video which compensates an interlayer prediction error by using a sample adaptive offset (SAO), an apparatus thereof, an apparatus for decoding a scalable video and various embodiments thereof. Provided are a method of decoding a scalable video and an apparatus thereof, in which SAO using information of an interlayer between a base layer image and an enhancement layer image and an SAO parameter representing an SAO type and an offset for an enhancement layer prediction image is obtained from a received enhancement layer stream, an SAO type of the enhancement layer prediction image and an offset obtained by classifying an prediction error according each of category are determined from the obtained SAO parameter, and a current sample is compensated by the offset determined according to each category so that an enhancement layer restored image is generated by using the enhancement layer prediction image. [Reference numerals] (22) Interlayer SAO parameter obtaining unit; (24) Interlayer offset determining unit; (26) Enhancement layer decoding unit; (28) Base layer decoding unit; (AA) Enhancement layer stream; (BB) Base layer stream; (CC) Enhancement layer image sequence; (DD) Base layer image sequence

Abstract translation: 提供了一种通过使用采样自适应偏移（SAO），其装置，用于解码可伸缩视频的装置及其各种实施例来补偿层间预测误差的可分级视频的编码方法。 提供一种解码可伸缩视频的方法及其装置，其中SAO使用基本层图像和增强层图像之间的中间层的信息和表示SAO类型的SAO参数和用于增强层预测图像的偏移的SAO参数是 从获得的SAO参数确定从接收到的增强层流获得的SAO类型的增强层预测图像和通过根据类别分类预测误差而获得的偏移，并且通过根据所获得的SAO参数确定的偏移来补偿当前样本 每个类别，使得通过使用增强层预测图像来生成增强层恢复图像。 （参考号）（22）层间SAO参数获取单元; （24）层间偏移确定单元; （26）增强层解码单元; （28）基层解码单元; （AA）增强层流; （BB）基层流; （CC）增强层图像序列; （DD）基层图像序列

公开(公告)号：KR1020110065092A

公开(公告)日：2011-06-15

申请号：KR1020090121937

申请日：2009-12-09

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더 ,  세레긴바딤

IPC: H04N19/61 ,  H04N19/176 ,  H04N19/122 ,  H04N19/157 ,  H04N19/625 ,  H04N19/124 ,  H04N19/13

CPC classification number: H04N19/157 ,  H04N19/122 ,  H04N19/176 ,  H04N19/61 ,  H04N19/124 ,  H04N19/13 ,  H04N19/625

Abstract: PURPOSE: In consideration of image properties, an image encoding and decoding method and apparatus using rotational transform are provided to adjust the size of the maximum encoding unit. CONSTITUTION: An image encoding apparatus creates a first frequency coefficient matrix(1810). The image encoding apparatus decides an angle parameter(1820). The image encoding apparatus creates a secondary frequency coefficient matrix with one or more partial exchange between the rows or columns of the first frequency coefficient matrix(1830). The image encoding apparatus quantizes the secondary frequency coefficient matrix(1840). The image encoding apparatus performs the entropy encoding of the quantized secondary frequency coefficient matrix(1850).

Abstract translation: 目的：考虑到图像特性，提供使用旋转变换的图像编码和解码方法和装置来调整最大编码单元的大小。 构成：图像编码装置创建第一频率系数矩阵（1810）。 图像编码装置决定角度参数（1820）。 图像编码装置创建具有在第一频率系数矩阵（1830）的行或列之间的一个或多个部分交换的次频系数矩阵。 图像编码装置对次频系数矩阵（1840）进行量化。 图像编码装置执行量化次频系数矩阵（1850）的熵编码。

公开(公告)号：KR1020110017303A

公开(公告)日：2011-02-21

申请号：KR1020090074898

申请日：2009-08-13

Applicant: 삼성전자주식회사

Inventor： 알쉬나엘레나 ,  알쉰알렉산더 ,  세레긴바딤

IPC: H04N19/176 ,  H04N19/13 ,  H04N19/91 ,  H04N19/119 ,  H04N19/147 ,  H04N19/122 ,  H04N19/61 ,  H04N19/18 ,  H04N19/19 ,  H04N19/124 ,  H04N7/24

CPC classification number: H04N19/119 ,  H04N19/122 ,  H04N19/147 ,  H04N19/176 ,  H04N19/18 ,  H04N19/19 ,  H04N19/61 ,  H04N7/24 ,  H04N19/124 ,  H04N19/13 ,  H04N19/91

Abstract: PURPOSE: A method and an apparatus for encoding and decoding an image by using rotational transform are provided to performing quantization and entropy encoding for a second frequency coefficient matrix by performing partial exchange between the columns and rows of a first frequency coefficient matrix. CONSTITUTION: A frequency transform unit generates a first frequency matrix by performing discrete cosine transform for a current block. In addition, the frequency transform unit generates a second frequency coefficient matrix by performing partial exchange between the columns and rows of the first frequency coefficient matrix based on an angle parameter. A quantization unit quantizes the second frequency coefficient matrix, and an entropy encoder performs an entropy encoding operation for a second coefficient matrix and the information of the angle parameter. The angle parameter indicates the partial exchange degree between the columns and rows.

Abstract translation: 目的：提供一种通过使用旋转变换对图像进行编码和解码的方法和装置，用于通过执行第一频率系数矩阵的列和行之间的部分交换来对第二频率系数矩阵进行量化和熵编码。 构成：频率变换单元通过对当前块执行离散余弦变换来生成第一频率矩阵。 此外，频率变换单元通过基于角度参数执行第一频率系数矩阵的列和行之间的部分交换来生成第二频率系数矩阵。 量化单元量化第二频率系数矩阵，并且熵编码器对第二系数矩阵和角度参数的信息执行熵编码操作。 角度参数表示列和行之间的部分交换度。

公开(公告)号：KR1020110014839A

公开(公告)日：2011-02-14

申请号：KR1020090072403

申请日：2009-08-06

Applicant: 삼성전자주식회사

Inventor： 알쉰알렉산더 ,  알쉬나엘레나 ,  쉬라코프니콜라이 ,  한우진

IPC: H04N19/61 ,  H04N19/105 ,  H04N19/176 ,  H04N19/157 ,  H04N19/18 ,  H04N19/44 ,  H04N19/122

CPC classification number: H04N19/18 ,  H04N19/105 ,  H04N19/132 ,  H04N19/136 ,  H04N19/157 ,  H04N19/17 ,  H04N19/176 ,  H04N19/61 ,  H04N19/895 ,  H04N19/122 ,  H04N19/44

Abstract: PURPOSE: A method and an apparatus for encoding a video, and a method and an apparatus for decoding a video are provided to reduce a data amount of a conversion coefficient of a low frequency component to be encoded, thereby increasing compression efficiency of the video. CONSTITUTION: A maximum encoding unit dividing part(110) divides a current picture or a current slice. An encoding depth determining part(120) determines a different maximum depth at every maximum encoding unit. A video data encoding unit(130) encodes video data of the maximum encoding unit. An encoded information encoding part(140) encodes information about an encoding mode by depth at every maximum encoding unit.

Abstract translation: 目的：提供一种用于编码视频的方法和装置，以及用于对视频进行解码的方法和装置，以减少要编码的低频分量的转换系数的数据量，从而提高视频的压缩效率。 构成：最大编码单元划分部分（110）划分当前图像或当前切片。 编码深度确定部分（120）确定每个最大编码单元处的不同最大深度。 视频数据编码单元（130）对最大编码单元的视频数据进行编码。 编码信息编码部分（140）在每个最大编码单位深度编码关于编码模式的信息。

公开(公告)号：KR1020090038303A

公开(公告)日：2009-04-20

申请号：KR1020070103724

申请日：2007-10-15

Applicant: 삼성전자주식회사

Inventor： 세레긴바딤 ,  알쉰알렉산더 ,  코로티프막심 ,  알쉬나엘레나 ,  네스테로바예카테리나

IPC: H04N19/90

CPC classification number: H04N19/625 ,  H04N19/12 ,  H04N19/126 ,  H04N19/129 ,  H04N19/13 ,  H04N19/146 ,  H04N19/147 ,  H04N19/176 ,  H04N19/19 ,  H04N19/192 ,  H04N19/46 ,  H04N19/61

Abstract: A method and an apparatus for encoding and decoding an image based on an exponential-Golomb code by using an optimal threshold are provided to determine thresholds for an exponential-Golomb binarization, consequently improving a gain compared with an image encoding result through a conventional encoding method in a high bit rate environment. A circular coding unit(310) receives an input image, and performs an entropy encoding operation for the input image through an exponential-Golomb binarization which uses plural thresholds about the absolute value of an inputted and quantized DCT coefficient. In addition, the circular coding unit outputs a bit stream generated for each threshold to a bit rate measurement unit(320). The bit rate measurement unit measures the bit rate of the generated bit stream by performing an entropy encoding operation based on each threshold, and then outputs the measured bit rate to an optical threshold determining unit(330). The optical threshold determining unit compares the bit rate for each threshold to determine an optimal threshold, and then outputs the determined the optimal threshold to a bit stream output unit(340).

Abstract translation: 提供了一种通过使用最佳阈值对基于指数Golomb码对图像进行编码和解码的方法和装置，以确定指数Golomb二值化的阈值，从而通过常规编码方法提高与图像编码结果相比的增益 在高比特率环境中。 循环编码单元（310）接收输入图像，并且通过使用关于输入和量化的DCT系数的绝对值的多个阈值的指数Golomb二值化对输入图像执行熵编码操作。 此外，循环编码单元将针对每个阈值生成的比特流输出到比特率测量单元（320）。 比特率测量单元通过基于每个阈值执行熵编码操作来测量所生成比特流的比特率，然后将测量的比特率输出到光阈值确定单元（330）。 光阈值确定单元比较每个阈值的比特率以确定最佳阈值，然后将确定的最佳阈值输出到比特流输出单元（340）。

公开(公告)号：KR1020090034574A

公开(公告)日：2009-04-08

申请号：KR1020070099883

申请日：2007-10-04

Applicant: 삼성전자주식회사

Inventor： 코로티프막심 ,  알쉰알렉산더 ,  알쉬나엘레나 ,  세레긴바딤 ,  네스테로바예카테리나

IPC: H04N19/124 ,  H04N19/60

CPC classification number: H04N19/18 ,  H04N19/126 ,  H04N19/14 ,  H04N19/176 ,  H04N19/44 ,  H04N19/46

Abstract: A quantization coefficient control method and an apparatus thereof for improving the compression efficiency by regulating the quantization coefficient are provided to control quantization coefficient by using non uniform statistical distribution after converting video data in frequency domain. An image transfer data stream is input to an extracting unit(110) from a coding terminal. The statistical property value and the quantized coefficient are extracted. A coefficient control value decision unit(120) inputs coefficient and pixel location much statistical property from an extracting unit. The coefficient control value is especially determined by a block by using coefficients depending on the statistical property. A coefficient control part(130) inputs the coefficient control value from the coefficient control value decision unit.

Abstract translation: 提供了一种通过调节量化系数来提高压缩效率的量化系数控制方法及其装置，以便在频域转换视频数据之后通过使用非统一的统计分布来控制量化系数。 图像传送数据流从编码终端输入到提取单元（110）。 提取统计特性值和量化系数。 系数控制值判定单元（120）从提取单元输入系数和像素位置多的统计属性。 系数控制值特别由块通过使用依赖于统计特性的系数来确定。 系数控制部（130）从系数控制值决定部输入系数控制值。