-
公开(公告)号：KR1020070079603A
公开(公告)日：2007-08-08
申请号：KR1020060010326
申请日：2006-02-03
Applicant: 경희대학교 산학협력단
IPC: H04N19/89 , H04N19/174 , H04N19/177 , H04N19/70
CPC classification number: H04N19/89 , H04N19/174 , H04N19/177 , H04N19/70
Abstract: A method for generating an XRP(XOR based Redundant Picture) slice and a method for recovering a lost slice by using the same are provided to recover the lost slice although one of the XRP and k+1 slices is lost by compressing a moving picture by moving picture compression algorithm, generating one XR(XOR Redundant) slice for k important slices, and transmitting the generated XR slice. The number of pictures for configuring an XRP and an object are set up. A current picture is encoded and picture data are stored. Data of the current picture are updated in the XRP according as the data of the current picture are used in the XRP. It is configured whether the configuration of the XRP is ended. When the configuration of the XRP is ended, the XRP is stored as a redundant coded picture after a primary coded picture currently encoded. When the configuration of the XRP is not ended, an XRP memory is initialized and the next XRP is configured.
Abstract translation: 提供了一种用于生成XRP（XOR）的冗余图片（slice）的方法和用于通过使用该片段来恢复丢失的片段的方法来恢复丢失的片段，尽管XRP和k + 1片段之一通过压缩运动图像而丢失 运动图像压缩算法，为k个重要片段生成一个XR（XOR冗余）片段，并发送生成的XR片段。 设置用于配置XRP和对象的图片数量。 编码当前图像并存储图像数据。 根据当前图像的数据在XRP中使用，当前图像的数据在XRP中更新。 配置XRP的配置是否结束。 当XRP的配置结束时，XRP在当前编码的主编码图像之后被存储为冗余编码图像。 当XRP的配置未结束时，将初始化XRP内存，并配置下一个XRP。
-
12.发明授权다시점 비디오 코딩에서의 움직임 벡터 예측 방법 및 이를이용한 다시점 영상의 부호화/복호화 방법 및 장치 有权使用预测方法预测多视点视频编码中的运动矢量和多视点视频的编码/解码方法和装置的方法
公开(公告)号：KR101039204B1
公开(公告)日：2011-06-03
申请号：KR1020060132575
申请日：2006-12-22
Applicant: 경희대학교 산학협력단
IPC: H04N19/597 , H04N13/00
Abstract: 본 발명은 다시점 비디오 코딩에서 움직임 벡터 예측값(MV
P )을 효율적으로 구함으로써 다시점 영상에 대한 부호화 효율을 향상시킬 수 있는 다시점 비디오 코딩에서의 움직임 벡터 예측 방법 및 이를 이용한 다시점 영상의 부호화/복호화 방법 및 장치에 관한 것이다. 본 발명의 일 실시예에 따른 움직임 벡터 예측 방법에서는 현재 블록의 주변 블록의 움직임 벡터(MV
N )와 함께, 현재 영상이 속한 뷰와 동일한 뷰에서 시간 방향 참조 영상의 시간 방향 대응 블록의 움직임 벡터(MV
T ) 및/또는 현재 영상의 시간대와 동일한 시간대에서 뷰 방향 참조 영상의 뷰 방향 대응 블록의 움직임 벡터(MV
V ) 중에서 선택된 하나 이상의 움직임 벡터를 이용하여 움직임 벡터 예측값(MV
P )를 구한다. 본 발명에 의하면, 현재 블록이 참조하는 영상의 특성, 즉 시간 방향 참조 영상을 참조하는지 또는 뷰 방향 참조 영상을 참조하는지를 고려하여 움직임 벡터 예측값(MV
P )을 구하기 때문에, 보다 정확한 예측을 함으로써 예측 부호화 효율을 향상시킬 수가 있다.
MPEG, 스케일러블 비디오 코딩, 다시점 비디오 코딩, 시간방향 예측코딩 (Temporal Predictive Coding)-
13.发明公开
公开(公告)号：KR1020070110636A
公开(公告)日：2007-11-20
申请号：KR1020060043338
申请日：2006-05-15
Applicant: 경희대학교 산학협력단
IPC: H04N19/597 , H04N13/00
CPC classification number: H04N19/597 , H04N19/103 , H04N19/137 , H04N19/174
Abstract: A method for predicting an image information disparity value and encoding/decoding method and apparatus of multi-view video using the image information disparity value prediction method are provided to prevent encoding efficiency from being deteriorated when a difference of color and brightness between images inputted from cameras is generated. A method for predicting an image information disparity value for compensating image information of views of multi-view video comprises the following steps of: using an image information disparity value of blocks for constructing a reference image at a different time and current video in the same view; and computing a prediction image information disparity value of current blocks for constructing current videos. The reference video includes a previous image of the current image and a next image of the current image.
Abstract translation: 提供了使用图像信息视差值预测方法来预测图像信息差异值和多视点视频的编码/解码方法和装置的方法，以防止当从相机输入的图像之间的颜色和亮度之间的差异时编码效率降低 被生成。 一种用于预测用于补偿多视点视频的图像信息的图像信息差异值的方法，包括以下步骤：在相同视图中使用用于在不同时间构建参考图像的块的图像信息差异值和当前视频 ; 以及计算用于构建当前视频的当前块的预测图像信息差异值。 参考视频包括当前图像的先前图像和当前图像的下一图像。
-
公开(公告)号：KR101560182B1
公开(公告)日：2015-10-15
申请号：KR1020080087435
申请日：2008-09-04
Applicant: 삼성전자주식회사 , 경희대학교 산학협력단
IPC: H04N19/597 , H04N19/103 , H04N13/00
CPC classification number: H04N19/44 , H04N19/597 , H04N19/70
Abstract: 본발명은다시점비디오의부호화및 복호화에관한것이다. 본발명의일 실시예는, 현재픽처내의현재블록의움직임벡터의결정방식을나타내는블록타입을판독하고, 현재픽처가인터예측을위한기준픽처타입인지여부를나타내는제 1 픽처타입을판독하여, 현재블록의블록타입및 현재픽처의제 1 픽처타입중 적어도하나에기초하여, 현재픽처에대한추가정보를판독하고, 현재픽처에대한추가정보를이용하여현재블록및 현재픽처를복원함으로써, 보다정확한앵커픽처를이용하여다시점비디오전체의코딩성능및 효율을향상시킬수 있다.
-
公开(公告)号：KR1020110014507A
公开(公告)日：2011-02-11
申请号：KR1020100019893
申请日：2010-03-05
Applicant: 경희대학교 산학협력단
CPC classification number: H04N19/31 , H04N19/187 , H04N19/36 , H04N19/597 , H04N19/61
Abstract: PURPOSE: A video encoding apparatus, capable of facilitating the conversion and transmission of processed video information is provided to integrately process video information to existing two-dimensional terminals and a tangible terminal in ubiquitous environment. CONSTITUTION: An acquisition unit(310) acquires a plurality of video images of the different viewpoints. An encoding unit(320) applies multi-viewpoint video coding to one of multi-viewpoint video images. The encoding unit applies scalable video coding to a plurality of viewpoints. An output unit(330) generates output bit-stream by the combination of itstreams by viewpoint and layer of the encoder.
Abstract translation: 目的：提供一种能够促进经过处理的视频信息的转换和传输的视频编码装置，以将视频信息整合到现有的二维终端和无处不在的环境中的有形终端。 构成：获取单元（310）获取不同视点的多个视频图像。 编码单元（320）将多视点视频编码应用于多视点视频图像之一。 编码单元对多个视点应用可分级视频编码。 输出单元（330）通过编码器的视点和层的组合来产生输出比特流。
-
Patent Agency Ranking
公开(公告)号：KR100922275B1
公开(公告)日：2009-10-15
申请号：KR1020070109323
申请日：2007-10-30
Applicant: 경희대학교 산학협력단
IPC: H04N19/80 , H04N19/117 , H04N19/51
CPC classification number: H04N19/00909 , H04N19/117 , H04N19/137 , H04N19/14 , H04N19/159 , H04N19/176 , H04N19/61 , H04N19/80 , H04N19/86
Abstract: Provided are a bS derivation process and a deblocking filtering method and apparatus using illumination compensation and/or chrominance compensation, such as a multi-view video coding, in a picture coding/decoding process in which a prediction coding is performed. If neither of the two blocks are intra-coded, neither of the two blocks do contain a none-zero transformed coefficient, and motion compensation for the two blocks is performed based on the identical reference frame and the absolute difference between the horizontal or vertical components of the two motion vectors of the two blocks is less than 1, the bS of the two blocks is derived by considering whether the two blocks are coded by illumination compensation (IC) according to an embodiment of the present invention,. Specifically, if both of the two blocks are coded by IC mode or neither of the two blocks are coded by IC mode, bS is set as such a value that filtering is unnecessary. On the other hand, if both of the two blocks are coded by IC mode and the blocks have different illumination change values, or only one of the blocks is coded by IC mode, the bS is derived using difference of illumination change values or the illumination change value.
-
公开(公告)号：KR1020090076762A
公开(公告)日：2009-07-13
申请号：KR1020080087435
申请日：2008-09-04
Applicant: 삼성전자주식회사 , 경희대학교 산학협력단
IPC: H04N19/597 , H04N19/103 , H04N13/00
CPC classification number: H04N19/44 , H04N19/597 , H04N19/70
Abstract: A multi-view video encoding method and an apparatus therefor and a multi-view video decoding method and an apparatus therefor are provided to restore a current block and a current picture by using additional information about the current picture, thereby improving coding performance and efficiency of the entire multi-view video. A block type determiner(110) determines a block type. The block type shows a determination method of a motion vector of a current block within a current picture. The first picture type determiner(120) determines the first picture type. The first picture type shows whether the current picture is a reference picture type for inter prediction. An additional information determiner(130) determines additional information about the current picture. An additional information encoder(140) encodes additional information about the current picture.
Abstract translation: 提供一种多视点视频编码方法及其装置以及多视点视频解码方法及其装置，用于通过使用关于当前图像的附加信息来恢复当前块和当前图像，从而提高编码性能和效率 整个多视图视频。 块类型确定器（110）确定块类型。 块类型示出当前图像内的当前块的运动矢量的确定方法。 第一图像类型确定器（120）确定第一图像类型。 第一图像类型示出当前图像是否是用于帧间预测的参考图像类型。 附加信息确定器（130）确定关于当前图片的附加信息。 附加信息编码器（140）编码关于当前图片的附加信息。
-
公开(公告)号：KR1020090045035A
公开(公告)日：2009-05-07
申请号：KR1020080106035
申请日：2008-10-28
Applicant: 경희대학교 산학협력단
CPC classification number: H04N19/40 , H04N19/117 , H04N19/14 , H04N19/176 , H04N19/61 , H04N19/86
Abstract: Disclosed are a transcoding apparatus and method between two codecs each including a deblocking filter. The transcoding method between first and second codecs each including a deblocking filter, may include decoding input data encoded according to the first codec, according to the first codec so as to generate decoded data; and encoding the decoded data according to the second codec. The decoded data may include data on which deblocking filtering is not performed by the first codec, or data on which deblocking filtering is adaptively performed by the first codec. The decoded data may further include data on which deblocking filtering is performed by the first codec. The decoded data may be used as input data when the second codec performs encoding and/or when the second codec performs motion estimation.
-
公开(公告)号：KR1020080097914A
公开(公告)日：2008-11-06
申请号：KR1020080026315
申请日：2008-03-21
Applicant: 삼성전자주식회사 , 경희대학교 산학협력단
IPC: H04N19/597 , H04N13/00 , H04N19/105
CPC classification number: H04N19/597 , H04N19/172 , H04N19/176
Abstract: A method and an apparatus of encoding and decoding a multi view moving picture are provided to apply a skip block mode to an anchor picture allowing referring only between views when coding the multi view moving picture, thereby improving image quality and coding efficiency. It is determined whether a picture is an anchor picture. When the picture is an anchor picture, a skip block mode is not applied when coding about blocks comprising the picture. Encoding about the block to which the skip block mode is applied is skipped(1040). In case that the picture is an anchor picture, it is determined whether a skip block mode is allowed when encoding the picture. According to the result, the skip block mode is selectively applied during the coding about blocks comprising the picture.
Abstract translation: 提供了对多视图运动图像进行编码和解码的方法和装置，以便在锚定图像上应用跳过块模式，从而仅允许在对多视图运动图像进行编码时在视图之间进行参考，从而提高图像质量和编码效率。 确定图片是否是锚图片。 当图像是锚图像时，当对包括图像的块进行编码时，不应用跳过块模式。 忽略应用跳过块模式的块的编码（1040）。 在图像是锚图像的情况下，确定当对图像进行编码时是否允许跳过块模式。 根据该结果，在关于包括图像的块的编码期间，有选择地应用跳过块模式。
-
公开(公告)号：KR101460362B1
公开(公告)日：2014-11-14
申请号：KR1020080014999
申请日：2008-02-19
Applicant: 삼성전자주식회사 , 경희대학교 산학협력단
IPC: H04N19/85 , H04N19/567 , H04N13/00
CPC classification number: H04N19/196 , H04N19/105 , H04N19/159 , H04N19/176 , H04N19/197 , H04N19/198 , H04N19/46 , H04N19/597 , H04N19/61
Abstract: 본 발명은 다시점 비디오 코딩에서 휘도 보상 방법에 있어서, 휘도 보상을 수행하는 현재 블록이 스킵 모드인 경우, 상기 현재 블록의 주변 블록의 정보를 이용하여 휘도 보상 예측 값을 유도하는 과정과, 유도한 휘도 보상 예측 값이 0인 경우, 현재 블록의 휘도 보상을 수행하지 않도록 설정하는 과정과, 유도한 휘도 보상 예측 값이 0이 아닌 경우, 현재 블록의 휘도 보상을 수행하도록 설정하는 과정을 포함한다.
MPEG, 스케일러블 비디오 코딩(Scalable Video Coding), 다시점 비디오 코딩(Multi-view Video Coding), 휘도 보상(Illumination Compensation)
-
-
-
-
-
-
-
-
-
Search Results
43
records
(took 0.034 seconds)
