-
公开(公告)号：KR1020120089404A
公开(公告)日：2012-08-10
申请号：KR1020100124396
申请日：2010-12-07
Applicant: 전자부품연구원
IPC: H04N19/80 , H04N19/176 , H04N19/86
CPC classification number: H04N19/86 , H04N19/174 , H04N19/176
Abstract: PURPOSE: A two stage deblocking filtering device based on a macroblock and a method thereof are provided to perform two stage deblocking filtering of SVC(Scalable Video Coding) immediately after each macroblock is decoded. CONSTITUTION: A secondary deblocking filtering unit(120) determines a boundary side of a block and an adjacent block as a secondary deblocking filtering area in a current macroblock. The boundary side of the block is located in the slice boundary of some adjacent macroblocks. The adjacent block does not need to pass a secondary deblocking filtering process. Some adjacent macroblocks include a left macroblock, an upper left macroblock, and an upper macroblock in a current macroblock. The second deblocking filtering unit performs secondary deblocking filtering in the determined secondary deblocking filtering area.
Abstract translation: 目的：提供一种基于宏块的两阶段去块滤波装置及其方法，用于在每个宏块解码之后立即执行SVC（可缩放视频编码）的两阶段去块滤波。 构成：二次去块滤波单元（120）将块和相邻块的边界侧确定为当前宏块中的二次去块滤波区域。 块的边界位于一些相邻宏块的片边界中。 相邻的块不需要通过二次去块滤波处理。 一些相邻宏块包括当前宏块中的左宏块，左上宏块和上宏块。 第二去块滤波单元在确定的二次去块滤波区域中进行二次去块滤波。
-
公开(公告)号：KR1020120072718A
公开(公告)日：2012-07-04
申请号：KR1020100134597
申请日：2010-12-24
Applicant: 전자부품연구원
CPC classification number: H04N19/30 , H04N19/159 , H04N19/174 , H04N19/176 , H04N19/426
Abstract: PURPOSE: A device for decoding scalable video coding using an access unit and a method thereof are provided to shorten decoding time to avoid an unnecessary memory access, thereby reducing a memory consumption amount. CONSTITUTION: A residual finding unit(221) analyzes a header of a first slice NALU(Network Abstraction Layer Unit) of all layers. All layers are included in an extracted access unit. The residual finding unit finds necessary residual data among TCL(Transform Coefficient Level) data, TC(Transform Coefficient) data, and PR(Pixel Residual) data to enable an upper layer to perform interlayer prediction. A decoding unit(222) stores the found residual data in a common residual memory when a reference layer is decoded.
Abstract translation: 目的：提供一种用于使用访问单元解码可缩放视频编码的装置及其方法，以缩短解码时间以避免不必要的存储器访问，从而减少存储器消耗量。 构成：残差查找单元（221）分析所有层的第一片NALU（网络抽象层单元）的头。 所有层都包含在提取的存取单元中。 剩余查找单元在TCL（变换系数水平）数据，TC（变换系数）数据和PR（像素残差））数据中找到必要的残留数据，以使上层执行层间预测。 当解码参考层时，解码单元（222）将找到的残差数据存储在公共的剩余存储器中。
-
公开(公告)号：KR101531036B1
公开(公告)日：2015-06-23
申请号：KR1020130110005
申请日：2013-09-12
Applicant: 전자부품연구원
Abstract: 본발명은디폴트값으로복호화파라미터중 컬러영상의 HEVC 데이터스트림에대해복호화처리하는제1 디코더의쓰레드개수를 N개설정하고, 상기복호화파라미터중 뎁스영상의 HEVC 데이터스트림에대해복호화처리를하는제2 디코더의쓰레드개수를 N개보다적은 K개로설정하고, 상기복호화파라미터별 HEVC 데이터스트림의헤더정보에서화상정보를파악하며, 상기복호화파라미터별 HEVC 데이터스트림의화상정보차이를파악한후 상기복호화처리시간이긴 화상정보를가진 HEVC 데이터스트림에대응하는디코더의쓰레드개수를상기디폴트된쓰레드개수에더하여증가시킴으로써, 3D 복호화기의성능을향상시킨다.
-
公开(公告)号：KR101364086B1
公开(公告)日：2014-02-21
申请号：KR1020120101033
申请日：2012-09-12
Applicant: 전자부품연구원
IPC: H04N19/00
CPC classification number: H04N19/436 , H04N19/117 , H04N19/169
Abstract: The present invention relates to a method for decoding coding units, by an image decoding apparatus, through a decoding process consisting of multiple steps. The method of the present invention comprises the steps of: receiving a frame composed of multiple coding units; determining the decoding order between a first coding unit and a second coding unit based on the dependent relationship upon the decoding process performed by the first coding unit and the second coding unit; and decoding each of the coding units according to the determined order. The decoding process is applied to a horizontal deblocking filter, a vertical deblocking filter, a sample adaptive offset filter, and an adaptive loop filter, in the listed order. [Reference numerals] (S210) Perform LCU parsing and LCU decompression on any LCU; (S220) Apply horizontal deblocking filter (Hor-DF) to any LCU; (S230) Apply vertical deblocking filter (Ver-DF) to any LCU; (S240) Apply sample adaptive offset filter (SAO) to any LCU; (S250) Apply adaptive loop filter (ALF) to any LCU
Abstract translation: 本发明涉及通过图像解码装置通过由多个步骤组成的解码处理对编码单元进行解码的方法。 本发明的方法包括以下步骤：接收由多个编码单元组成的帧; 基于由所述第一编码单元和所述第二编码单元执行的解码处理的依赖关系，确定第一编码单元和第二编码单元之间的解码顺序; 并根据确定的顺序对每个编码单元进行解码。 解码过程按照所列顺序应用于水平去块滤波器，垂直去块滤波器，采样自适应偏移滤波器和自适应环路滤波器。 （参考号）（S210）对任何LCU执行LCU解析和LCU解压缩; （S220）将水平去块滤波器（Hor-DF）应用于任何LCU; （S230）将垂直去块滤波器（Ver-DF）应用于任何LCU; （S240）将样本自适应偏移滤波器（SAO）应用于任何LCU; （S250）将自适应环路滤波器（ALF）应用于任何LCU
-
公开(公告)号：KR1020110052203A
公开(公告)日：2011-05-18
申请号：KR1020090109148
申请日：2009-11-12
Applicant: 전자부품연구원
IPC: H04N7/24
CPC classification number: H04N19/33 , H04N19/105 , H04N19/117 , H04N19/159 , H04N19/176 , H04N19/186 , H04N19/46 , H04N19/59 , H04N19/61 , H04N19/70 , H04N19/80
Abstract: PURPOSE: A scalable video coding method and apparatus thereof are provided to enhance the execution speed of a scalable video codec are provided to reduce the capacity of necessary memory necessary for space layer prediction of a scalable image. CONSTITUTION: A base layer and an enhancement layer are generated from an input image signal(S110). A macro block is separated from the enhancement layer. The block of the reference layer is up-sampled by using a coordinate value of a reference block in reference with an up-sampling of a current macro block. The up-sampled macro block is coded.
Abstract translation: 目的：提供一种可伸缩的视频编码方法及其装置，以提高可伸缩视频编解码器的执行速度，以减少可伸缩图像的空间层预测所必需的存储空间。 构成：从输入图像信号生成基本层和增强层（S110）。 宏块与增强层分离。 通过使用参考块的坐标值参考当前宏块的上采样来对参考层的块进行上采样。 上采样的宏块被编码。
-
Patent Agency Ranking
公开(公告)号：KR1020100079597A
公开(公告)日：2010-07-08
申请号：KR1020080138130
申请日：2008-12-31
Applicant: 전자부품연구원
CPC classification number: H04N19/433 , H04N19/176 , H04N19/436 , H04N19/96
Abstract: PURPOSE: A motion vector extractor based array and method thereof are provided to minimize delay time by reading motion and writing motion of a memory. CONSTITUTION: A comparison operation unit(20) includes a plurality of comparators. The plurality of comparators receive a current SAD(Sum of Absolute Difference) value outputted from a relevant SAD tree block among a plurality of SAD tree blocks. The plurality of comparators compares the current SAD value and existing calculated SAD value. A motion vector merger(22) combines a motion vector values of pixels based on a comparison result values of the plurality of comparators and arranged the motion vector values.
Abstract translation: 目的：提供一种基于运动矢量提取器的阵列及其方法，通过读取运动和写入运动来最小化延迟时间。 构成：比较运算单元（20）包括多个比较器。 多个比较器接收从多个SAD树块中的相关SAD树块输出的当前SAD（绝对差值之和）值。 多个比较器比较当前SAD值和现有计算的SAD值。 运动矢量合并（22）基于多个比较器的比较结果值组合像素的运动矢量值并且布置运动矢量值。
-
公开(公告)号：KR101212227B1
公开(公告)日：2012-12-13
申请号：KR1020100134597
申请日：2010-12-24
Applicant: 전자부품연구원
Abstract: 액세스유닛단위처리를이용한 SVC의복호화방법이개시된다. 액세스유닛단위처리를이용한 SVC의복호화방법은 SVC 영상스트림에서액세스유닛(access unit)을추출하는단계, 추출된액세스유닛에포함되는모든레이어를분석하여상위레이어(enhancement layer)가레이어간예측(inter layer perdiction)을위해필요로하는참조레이어(reference layer)의레지듀얼(residual) 데이터를파악하는단계, 및참조레이어복호화시참조레이어의레지듀얼데이터들중 파악된레지듀얼만을공용레지듀얼메모리에저장하는복호화단계를포함한다. 이에의해메모리접근이줄어들어복호화시간이단축되며, 메모리소비또한줄어든다.
-
公开(公告)号：KR100973083B1
公开(公告)日：2010-07-29
申请号：KR1020080138130
申请日：2008-12-31
Applicant: 전자부품연구원
Abstract: 본 발명의 움직임 벡터 추출기는 프레임 데이터를 입력받기 위하여 외부 메모리와 인터페이스싱하는 인터페이스/라우터 블록과, 읽기와 쓰기 동작이 동시에 수행되는 내부 메모리 및 레지스터 어댑터를 포함한다. 이러한 본 발명에 의하면, 움직임 추정을 수행하기 위해 필요한 프레임 데이터를 보다 속도가 빠른 내부 메모리로 복사함과 동시에 SAD(Sum of Absolute Difference) 연산을 하기 위한 프레임 데이터를 내부 메모리에서 Register 어댑터로 복사한다. 따라서 메모리의 읽기 동작 및 쓰기 동작에 의한 지연시간을 최소화함으로써, 연산효율을 향상시킬 수 있다. 또한, 각 연산 블록에 가변적으로 제어되는 파이프라인 레지스터를 구비함으로써, 기존의 일관된 파이프라인 구조에 따른 전력 소모량을 줄일 수 있다.
-
公开(公告)号：KR1020100018810A
公开(公告)日：2010-02-18
申请号：KR1020080077481
申请日：2008-08-07
Applicant: 전자부품연구원
IPC: H04N19/186 , H04N19/51
CPC classification number: H04N19/186 , H04N19/105 , H04N19/109 , H04N19/11 , H04N19/176 , H04N19/61
Abstract: PURPOSE: A recording media which is readable with a computer is provided to improve compression efficiency using correlation among G, B, and R planes. CONSTITUTION: A first plane encoder(100) codes a first plane. A second plane encoder(200) encodes the second plane. A second plane encoder comprises an inter-plane/intra-plane predicting unit, inter-frame predicting unit, subtracter, and coding unit. The subtracter generates residue. The coding unit codes the residue.
Abstract translation: 目的：提供使用计算机可读取的记录介质，以利用G，B和R平面之间的相关性提高压缩效率。 构成：第一平面编码器（100）对第一平面进行编码。 第二平面编码器（200）编码第二平面。 第二平面编码器包括平面内/平面内预测单元，帧间预测单元，减法器和编码单元。 减法器产生残留物。 编码单元对残留物进行编码。
-
公开(公告)号：KR101517806B1
公开(公告)日：2015-05-06
申请号：KR1020140018453
申请日：2014-02-18
Applicant: 전자부품연구원
IPC: H04N19/82 , H04N19/117 , H04N19/86
CPC classification number: H04N19/86 , H04N19/117
Abstract: 영상복호화방법및 이를적용한영상기기가제공된다. 본발명의실시예에따른영상복호화방법은, 특정블럭을복호화하고, Hor-DF를수행한후에, '특정블럭의일부'와 '특정블럭의주변에위치하는블럭들중 적어도하나의일부'에대해 Ver-DF하고, 이후 SAO를수행한다. 이에의해, 동기화딜레이를줄여, 영상복호화속도를획기적으로향상시킬수 있게된다.
Abstract translation: 提供了一种用于视频解码的方法及其使用方法。 根据本发明的用于视频解码的方法对特定块进行解码，执行Hor-DF，然后对特定块的一部分和围绕特定块的块的至少一部分执行Ver-DF，然后执行 SAO。 从而，可以减少同步延迟，并且可以显着地提高视频解码速度。
-
-
-
-
-
-
-
-
-
Search Results
13
records
(took 0.021 seconds)
