-
公开(公告)号：KR101303660B1
公开(公告)日：2013-09-04
申请号：KR1020070098958
申请日：2007-10-01
Applicant: 삼성전자주식회사
IPC: H04N5/208
CPC classification number: G06T5/002 , G06T5/003 , G06T5/20 , G06T2207/20012 , G06T2207/20192
Abstract: 잡음제거를 고려한 선명도 향상 방법 및 장치를 개시한다. 영상에 대한 선명도 향상 방법은, 중심화소와 주변화소간의 차를 이용하여 잡음 가중치(noise weight)를 계산하는 단계 및 상기 잡음 가중치에 기초하여 상기 중심화소에 대한 대체값을 계산하는 단계를 포함한다.
영상신호, 잡음(noise), 해상도, 선명도, 잡음제거-
公开(公告)号：KR101303647B1
公开(公告)日：2013-09-04
申请号：KR1020070098315
申请日：2007-09-28
Applicant: 삼성전자주식회사
IPC: H04N7/01
Abstract: 영상의 해상도 향상 시스템 및 방법이 개시된다. 영상의 해상도 향상 시스템은 입력 영상을 미리 설정한 크기로 업샘플링(upsampling)하고, 업샘플링된 상기 입력 영상을 보간하여 해상도를 향상시키는 영상 스케일러부, 해상도가 향상된 상기 입력 영상의 프레임을 복수 개의 블록으로 분할하고, 노이즈를 통해 상기 블록마다 블러(blur)값을 예측하여 복원 정보를 추출하는 복원 정보 추출부 및 상기 복원 정보에 따라 상기 입력 영상의 프레임 전체에 대해 적어도 하나의 변환 영역 필터링(transform area filtering)을 병렬적으로 수행하여 상기 입력 영상을 복원하는 영상 복원부를 포함한다.
해상도 향상, 스케일러, 블러, PSF(point spread function), 변환 영역 필터링-
公开(公告)号：KR1020130086437A
公开(公告)日：2013-08-02
申请号：KR1020120007210
申请日：2012-01-25
Applicant: 삼성전자주식회사
IPC: G03H1/04
CPC classification number: G03H1/0808 , G03H1/20 , G03H1/2294 , G03H2001/0825 , G03H2001/205 , G03H2210/30 , G03H2210/452 , G03H2223/23 , G03H2226/02 , G03H2240/61
Abstract: PURPOSE: A method and a device for generating a three-dimensional (3D) hologram at high speed are provided to be suitable for parallel processing using a multi-processer and to minimize calculation delays due to external memory access. CONSTITUTION: A pattern setting unit (110) determines the size of a pattern. A scheduling unit (120) performs scheduling in order to minimize the number of processors in an idle state as well as the number of iterations. A calculation unit (130) performs pattern calculation by regressive interpolation using 3D data input with respect to a part of an area in which a hologram pattern is to be generated. A pattern copying unit (140) copies a pattern in the part of the area to complete the pattern for the entire area. [Reference numerals] (110) Pattern setting unit; (120) Scheduling unit; (130) Calculation unit; (140) Pattern copying unit; (AA) Processor#1; (BB) Processor#2; (CC) Processor#3; (DD) Processor#4
Abstract translation: 目的：提供一种用于以高速生成三维（3D）全息图的方法和装置，以适合于使用多处理器的并行处理，并且由于外部存储器访问而使计算延迟最小化。 构成：图案设定单元（110）确定图案的尺寸。 调度单元（120）进行调度，以使处于空闲状态的处理器的数量以及迭代次数最小化。 计算单元（130）通过使用相对于要生成全息图的区域的一部分的3D数据输入的回归插值来执行图案计算。 图案复制单元（140）在区域的一部分中复制图案以完成整个区域的图案。 （附图标记）（110）图案设定单元; （120）调度单位; （130）计算单位; （140）图案复印单元; （AA）处理器＃1; （BB）处理器＃2; （CC）处理器＃3; （DD）处理器＃4
-
公开(公告)号：KR1020130056803A
公开(公告)日：2013-05-30
申请号：KR1020120006711
申请日：2012-01-20
Applicant: 삼성전자주식회사
IPC: H04N19/597 , H04N19/186 , H04N13/00
CPC classification number: H04N19/597 , H04N13/161
Abstract: PURPOSE: A device for encoding and decoding a multiview depth image by using a color image and a method thereof are provided to divide a static area by analyzing a color image and to prevent the encoding of the static area during the encoding of a depth image. CONSTITUTION: A color image encoding unit(212) encodes a color image block. A static area determining unit(220) determines the encoded color image block. If the encoded color image block is a static area, a depth image encoding unit(214) does not encode a depth image block corresponding to the color image block. If the encoded color image block is a dynamic area, the depth image encoding unit encodes the corresponding depth image block. [Reference numerals] (210) Encoding device; (212,232) Color image encoding unit; (214,234) Depth image encoding unit; (220) Static area determining unit; (230) Decoding device; (AA) Color image block; (BB) Depth image block
Abstract translation: 目的：提供一种通过使用彩色图像对其进行编码和解码多视点深度图像的装置及其方法，用于通过分析彩色图像来分割静态区域，并且在深度图像的编码期间防止静态区域的编码。 构成：彩色图像编码单元（212）对彩色图像块进行编码。 静态区域确定单元（220）确定编码的彩色图像块。 如果编码的彩色图像块是静态区域，则深度图像编码单元（214）不对与彩色图像块相对应的深度图像块进行编码。 如果编码的彩色图像块是动态区域，则深度图像编码单元对相应的深度图像块进行编码。 （附图标记）（210）编码装置; （212,232）彩色图像编码单元; （214,234）深度图像编码单元; （220）静态区域确定单元; （230）解码装置; （AA）彩色图像块; （BB）深度图像块
-
公开(公告)号：KR1020130056441A
公开(公告)日：2013-05-30
申请号：KR1020110122057
申请日：2011-11-22
Applicant: 삼성전자주식회사
IPC: H04N19/105 , H04N19/51 , H04N19/117
CPC classification number: H04N19/597 , H04N19/105 , H04N19/117 , H04N19/51
Abstract: PURPOSE: A motion prediction encoding/decoding device for depth images and a method thereof are provided to predict a motion by generating an expanded reference screen and using the reference screen. CONSTITUTION: A reference screen selecting unit(110) selects one or more reference screens corresponding to a block of an encoding target image. A reference screen filtering unit(120) applies one or more preset filters to the reference screens. A reference screen expanding unit(130) reforms one or more of the reference screens or filtered screens. A motion prediction unit(140) searches the expanded reference screen for an area matched with the block to predict a motion. An encoding unit(150) encodes generated motion information. [Reference numerals] (110) Reference screen selecting unit; (120) Reference screen filtering unit; (130) Reference screen expanding unit; (140) Motion prediction unit; (150) Encoding unit
Abstract translation: 目的：提供一种用于深度图像的运动预测编码/解码装置及其方法，用于通过生成扩展参考屏幕并使用参考屏幕来预测运动。 构成：参考画面选择单元（110）选择与编码对象图像的块对应的一个或多个参考画面。 参考屏幕过滤单元（120）将一个或多个预设滤波器应用于参考屏幕。 参考屏幕扩展单元（130）对一个或多个参考屏幕或滤波屏幕进行改造。 运动预测单元（140）在扩展参考屏幕中搜索与块匹配的区域来预测运动。 编码单元（150）对生成的运动信息进行编码。 （附图标记）（110）参考画面选择单元; （120）参考屏过滤单元; （130）参考屏幕扩展单元; （140）运动预测单元; （150）编码单元
-
Patent Agency Ranking
公开(公告)号：KR101265956B1
公开(公告)日：2013-05-22
申请号：KR1020070111716
申请日：2007-11-02
Applicant: 삼성전자주식회사
CPC classification number: G06T5/002 , G06T5/003 , G06T5/10 , G06T2207/20056 , G06T2207/20064 , G06T2207/20192
Abstract: 블록기반의영상복원시스템및 방법이개시된다. 블록기반의영상복원시스템은영상의경계영역에서컬러쉬프팅(color shifting)을수행하여상기영상의경계를처리하는경계처리부및 상기영상을적어도하나의영역으로분할하고, 상기분할된영역에포함된블록별로복원파라미터를추출하는복원파라미터추출부및 상기복원파라미터에따른블록기반의변환영역필터링을적용하여상기영상에대해복원을수행하는영상복원부를포함한다.
-
公开(公告)号：KR1020130029572A
公开(公告)日：2013-03-25
申请号：KR1020110092911
申请日：2011-09-15
Applicant: 삼성전자주식회사
IPC: H04N19/85 , H04N19/117 , H04N19/13
CPC classification number: H04N19/85 , H04N19/117 , H04N19/82
Abstract: PURPOSE: A post filter, a loop filter, a video data encoding or decoding device and an operation method thereof are provided to improve the quality of a depth image in an image pressure system. CONSTITUTION: A post filter(100) inserted into a video data encoding device includes an input unit(110) and a processing unit(120). The input unit receives a pressure image for a depth image inputted to the video data encoding device. The pressure image includes a plurality of slices. The processing unit determines optical filter parameters for executing an intension of boundary line for an image profile. [Reference numerals] (101) Intra prediction unit; (102) Movement estimation and compensation unit; (104) Conversion and quantization unit; (105) Inverse quantization and inverse conversion unit; (107) Entropy coding unit; (108) Picture server; (110) Input unit; (120) Processing unit; (130) Output unit; (AA) Input images; (BB) Bit-stream; (CC) Compressed images; (DD) Filter parameters
Abstract translation: 目的：提供后置滤波器，环路滤波器，视频数据编码或解码装置及其操作方法，以提高图像压力系统中的深度图像的质量。 构成：插入到视频数据编码装置中的后置滤波器（100）包括输入单元（110）和处理单元（120）。 输入单元接收用于输入到视频数据编码装置的深度图像的压力图像。 压力图像包括多个切片。 处理单元确定用于执行图像轮廓的边界线的强度的光学滤波器参数。 （附图标记）（101）帧内预测单元; （102）运动估算和赔偿单位; （104）转换和量化单元; （105）逆量化和逆转换单元; （107）熵编码单元; （108）图片服务器; （110）输入单元; （120）处理单元; （130）输出单元; （AA）输入图像; （BB）比特流; （CC）压缩图像; （DD）过滤参数
-
公开(公告)号：KR1020130008435A
公开(公告)日：2013-01-22
申请号：KR1020110102374
申请日：2011-10-07
Applicant: 삼성전자주식회사
CPC classification number: H04N13/111 , H04N13/128 , H04N13/351 , H04N2213/003
Abstract: PURPOSE: A multi-view image processing method using hole rendering and a device thereof are provided to grant priority to pixels in a hole area and restore the pixels in a hole area in descending order of the priority. CONSTITUTION: A multi view image processing device generates an output view and parallax information corresponding to the output view through one or more input views and parallax information corresponding to the input views(1110). The device generates priority for each hole pixel indicating a hole in the output view by using the parallax information corresponding to the output view(1120~1130). The device applies hole rendering to each hole pixel in descending order of the priority(1140~1170). [Reference numerals] (1110) Generating an output view; (1120) Determining a background adjacent hole boundary; (1122) Structure priority operation; (1124) Confidence priority operation; (1126) Parallax allocation; (1128) Parallax priority operation; (1130) Priority combination; (1140) Priority arrangement; (1150) Generating a target patch; (1160) Searching an optimal source patch; (1170) Copying the patch; (AA) Start; (BB) End
Abstract translation: 目的：提供使用孔渲染的多视图图像处理方法及其装置，以优先考虑孔区域中的像素，并以优先级的降序恢复孔区域中的像素。 构成：多视点图像处理装置通过与输入视图对应的一个或多个输入视图和视差信息生成与输出视图对应的输出视图和视差信息（1110）。 通过使用与输出视图对应的视差信息（1120〜1130），设备对于表示输出视图中的孔的每个孔像素产生优先级。 该设备按照优先级的降序对每个孔像素应用孔渲染（1140〜1170）。 （参考数字）（1110）产生输出视图; （1120）确定背景相邻孔边界; （1122）结构优先操作; （1124）信心优先操作; （1126）视差分配; （1128）视差优先操作; （1130）优先组合; （1140）优先安排; （1150）生成目标补丁; （1160）搜索最佳源补丁; （1170）复制补丁; （AA）开始; （BB）结束
-
公开(公告)号：KR101213205B1
公开(公告)日：2012-12-17
申请号：KR1020060116566
申请日：2006-11-23
Applicant: 삼성전자주식회사
IPC: H04N9/64
CPC classification number: G09G5/02 , G01J3/462 , G01J3/465 , G09G2340/0407 , G09G2340/06 , G09G2360/16 , H04N1/6052 , H04N1/6072 , H04N5/57 , H04N9/3182 , H04N9/3188 , H04N9/67 , H04N9/68 , H04N9/735 , H04N21/4318 , H04N21/44008 , H04N21/440263
Abstract: 본발명은컬러재현시스템및 방법에관한것으로서, 본발명에따른컬러재현시스템은입력된이미지를구성하는픽셀의색상분포를분석하는이미지컨텐츠분석유닛과, 상기픽셀에관한속성정보와상기색상분포및 이미지크기정보를기초로하여스케일링파라미터를예측하는파라미터예측유닛및 상기예측된파라미터를이용하여상기픽셀의최종밝기와채도를결정하는밝기-채도결정유닛을포함한다.
-
公开(公告)号：KR1020120083209A
公开(公告)日：2012-07-25
申请号：KR1020110141112
申请日：2011-12-23
Applicant: 삼성전자주식회사
CPC classification number: H04N19/597 , H04N19/103 , H04N19/139 , H04N19/14 , H04N19/159 , H04N19/176 , H04N19/567 , H04N19/86 , H04N2213/005
Abstract: PURPOSE: A depth image coding/decoding apparatus and a method thereof are provided to arrange and encode the priority of each coding mode according to the property of a depth image. CONSTITUTION: An intra mode prediction unit(210) predicts the directionality of a current macro block. A coding mode determination unit(218) determines a coding mode corresponding to smallest RDO(Rate-Distortion Optimization) cost as a coding mode of the current macro block in consideration of the priority of each coding mode included intra candidate coding modes and inter candidate coding modes. An intra prediction mode performing intra coding is included in the intra candidate coding modes.
Abstract translation: 目的：提供深度图像编码/解码装置及其方法，以根据深度图像的属性对每个编码模式的优先级进行排列和编码。 构成：帧内模式预测单元（210）预测当前宏块的方向性。 考虑到包括帧内候选编码模式和帧间候选编码模式的每个编码模式的优先级，编码模式确定单元（218）确定与最小RDO（速率失真优化）成本相对应的编码模式作为当前宏块的编码模式 模式。 执行帧内编码的帧内预测模式包括在帧内候选编码模式中。
-
-
-
-
-
-
-
-
-
Search Results
415
records
(took 0.034 seconds)
