-
公开(公告)号：KR1020150077392A
公开(公告)日：2015-07-07
申请号：KR1020150086731
申请日：2015-06-18
Applicant: 삼성전자주식회사
IPC: H04N19/117 , H04N19/119 , H04N19/89 , H04N19/122 , H04N19/172
CPC classification number: H04N19/117 , H04N19/119 , H04N19/172 , H04N19/174 , H04N19/176 , H04N19/46 , H04N19/59 , H04N19/82 , H04N19/86
Abstract: 영역기반의부호화장치및 복호화장치가개시된다. 본발명의일실시예에따른부호화장치는영상프레임의분할된영역과관련된영역분할모드들중 최적의영역분할모드를선택하고, 상기최적의영역분할모드에따라분할된영역들각각의최적의영상필터링방법과최적의필터계수를최적의영역분할모드와함께복호화장치에전송할수 있다.
Abstract translation: 公开了一种基于区域的编码和解码装置。 根据本发明的实施例的编码装置在与图像帧的分割区域相关的区域分割模式中选择最佳区域划分模式; 并且向解码装置发送用最佳区域划分模式划分的每个区域的最优滤波方法和具有最佳区域划分模式的最佳滤波器系数。
-
公开(公告)号：KR1020130037843A
公开(公告)日：2013-04-17
申请号：KR1020110102323
申请日：2011-10-07
Applicant: 삼성전자주식회사
IPC: H04N19/59
CPC classification number: H04N19/159 , H04N19/182 , H04N19/597
Abstract: PURPOSE: A prediction pixel generation device and an operating method thereof are provided to improve a compression ratio of a depth image by generating a prediction pixel using one or both of a look-up table and a boundary line. CONSTITUTION: A processing unit(110) detects an LUT(Look Up Table) corresponding to an intra prediction direction of a prediction mode for intra prediction among preset LUTs. A generating unit(120) generates a prediction pixel using a depth value of a reference pixel related to the prediction pixel and a prediction coefficient recorded in the detected LUT. [Reference numerals] (101) Deblocking filter unit; (102) Movement estimation and compensation unit; (104) Conversion and quantization unit; (105) De-quantization and deconversion unit; (107) Entropy coding unit; (108) Picture buffer; (110) Processing unit; (120) Generating unit; (AA) Image input; (BB) Bit stream;
Abstract translation: 目的：提供预测像素生成装置及其操作方法，以通过使用查找表和边界线中的一个或两者生成预测像素来提高深度图像的压缩率。 构成：处理单元（110）检测与预设LUT之间的帧内预测的预测模式的帧内预测方向对应的LUT（查找表）。 生成单元（120）使用与预测像素相关的参考像素的深度值和记录在检测到的LUT中的预测系数来生成预测像素。 （附图标记）（101）去块滤波器单元; （102）运动估算和赔偿单位; （104）转换和量化单元; （105）去量化和解转换单元; （107）熵编码单元; （108）图片缓冲; （110）处理单元; （120）发电机组; （AA）图像输入; （BB）比特流;
-
公开(公告)号：KR1020120083825A
公开(公告)日：2012-07-26
申请号：KR1020110041902
申请日：2011-05-03
Applicant: 삼성전자주식회사
IPC: H04N19/117 , H04N19/85 , H04N13/00
CPC classification number: H04N19/117 , H04N19/13 , H04N19/182 , H04N19/59 , H04N19/96
Abstract: PURPOSE: An apparatus and a method for processing a depth image are provided to divide a depth image into a plurality of areas and to process a plurality of the areas. CONSTITUTION: An area division unit(110) divides a compressed depth image into a plurality of areas. The area division unit computes flatness according to the area division unit. The area division unit classifies a plurality of areas into a plurality of classes based on the flatness. A determination unit(120) determines filter parameter value corresponding to each class. An image filtering unit(130) executes an image filtering according to each class based on the filter parameter value.
Abstract translation: 目的：提供一种用于处理深度图像的装置和方法，以将深度图像分割成多个区域并处理多个区域。 构成：区域划分单元（110）将压缩深度图像分割成多个区域。 区域分割单元根据区域划分单位来计算平坦度。 区域分割单元基于平坦度将多个区域分类为多个类别。 确定单元（120）确定与每个类对应的过滤器参数值。 图像滤波单元（130）根据滤波器参数值根据每个类执行图像滤波。
-
公开(公告)号：KR1020110090366A
公开(公告)日：2011-08-10
申请号：KR1020100010096
申请日：2010-02-03
Applicant: 삼성전자주식회사
IPC: H04N13/02
CPC classification number: G06T15/005 , G06K9/00 , H04N13/218
Abstract: PURPOSE: A method and device for processing an image of high quality are provided to realize an output image without resolution loss by processing an image by using an image data maintaining original resolution. CONSTITUTION: A video input unit(110) receives an original left image, original center image and original right image. An image processing unit(120) generates a processing left image and processing right image by processing an original left image and an original right image. A second frame image is comprised of a first frame, processing left image and processing right image which is comprised of the original center image. An image generating unit(130) generates a 3D image data including the second frame image.
Abstract translation: 目的：提供用于处理高品质图像的方法和装置，以通过使用保持原始分辨率的图像数据处理图像来实现没有分辨率损失的输出图像。 构成：视频输入单元（110）接收原始左图像，原始中心图像和原始右图像。 图像处理单元（120）通过处理原始左图像和原始右图像来生成处理左图像和处理右图像。 第二帧图像包括由原始中心图像组成的第一帧，处理左图像和处理右图像。 图像生成单元（130）生成包括第二帧图像的3D图像数据。
-
公开(公告)号：KR1020110086521A
公开(公告)日：2011-07-28
申请号：KR1020110006393
申请日：2011-01-21
Applicant: 삼성전자주식회사
IPC: H04N19/119 , H04N19/46 , H04N19/117 , H04N19/174 , H04N19/82 , H04N19/86 , H04N19/59
CPC classification number: H04N19/117 , H04N19/119 , H04N19/172 , H04N19/174 , H04N19/176 , H04N19/46 , H04N19/59 , H04N19/82 , H04N19/86
Abstract: PURPOSE: A region-based encoding/decoding device and method thereof are provided to enable an encoding device to select an optimized division mode among an area division modes related to the divided area of an image frame and to transmit optimized image filtering method and an optimized filter coefficient to a decoding device with the optimized division mode. CONSTITUTION: A encoding device(101) includes as follows. A division mode selecting unit selects an optimized region-dividing mode among an area division modes about an image frame. A division mode transmitting unit transmits the selected area division mode to a decoding mode(102). A filter information transmitting unit transmits an optimized image filtering method and optimized filter coefficient to the decoding device.
Abstract translation: 目的：提供一种基于区域的编码/解码装置及其方法，使得编码装置能够在与图像帧的分割区域相关的区域分割模式中选择优化的分割模式，并且发送优化的图像滤波方法和优化的图像滤波方法 滤波器系数到具有优化分频模式的解码装置。 构成：编码装置（101）包括如下。 分割模式选择单元在关于图像帧的区域划分模式中选择优化的分区模式。 分割模式发送单元将所选择的区域划分模式发送到解码模式（102）。 滤波器信息发送单元向解码装置发送优化图像滤波方法和优化的滤波器系数。
-
Patent Agency Ranking
公开(公告)号：KR1020110049079A
公开(公告)日：2011-05-12
申请号：KR1020090105922
申请日：2009-11-04
Applicant: 삼성전자주식회사
IPC: H04N19/46 , H04N19/176 , H04N19/134 , H04N19/117 , H04N19/59 , H04N19/177 , H04N19/103 , H04N7/24
CPC classification number: H04N19/44 , H04N19/117 , H04N19/134 , H04N19/14 , H04N19/176 , H04N19/177 , H04N19/46 , H04N19/587 , H04N19/59 , H04N7/24 , H04N19/103 , H04N19/80
Abstract: PURPOSE: An apparatus and method for compressing and restoration image are provided to reduce the deterioration of image quality in a reduce bandwidth using filter information. CONSTITUTION: An image reduction unit(201) generates a resolution-reduced reduction image by performing sampling for an input image. A filter information determination unit(202) determines filter information through the decoded reduction image and the input image. A bit stream generator(203) generates bit stream by coding the information and the reduced image. An image division unit(204) divides the decoded reduced images into a plurality of blocks.
Abstract translation: 目的：提供一种用于压缩和恢复图像的装置和方法，以使用滤波器信息在减少带宽中减少图像质量的劣化。 构成：图像缩小单元（201）通过对输入图像进行采样来生成分辨率降低的缩小图像。 滤波器信息确定单元（202）通过解码的缩小图像和输入图像来确定滤波器信息。 比特流生成器（203）通过对信息进行编码和缩小图像来生成比特流。 图像分割单元（204）将解码的缩小图像划分为多个块。
-
公开(公告)号：KR1020110027231A
公开(公告)日：2011-03-16
申请号：KR1020090085227
申请日：2009-09-10
Applicant: 삼성전자주식회사
IPC: H04N19/597 , H04N13/00
CPC classification number: H04N19/507 , H04N19/597
Abstract: PURPOSE: A three-dimensional image compression processing method and an apparatus thereof are provided to reduce the amount of data by sharing the motion vector of a multiple-viewpoint image and to reduce the amount of bits which are generated after compressing a three-dimensional image. CONSTITUTION: A motion estimating unit(110) generates a color image motion vector. A depth estimating unit(120) generates a depth image of a multiple-viewpoint image. An encoding unit(130) generates a compression image after compressing the color image motion vector. A motion determining unit determines whether there is image motion between two frames.
Abstract translation: 目的：提供三维图像压缩处理方法及其装置，以通过共享多视点图像的运动矢量来减少数据量，并且减少在压缩三维图像之后产生的位数量 。 构成：运动估计单元（110）生成彩色图像运动矢量。 深度估计单元（120）生成多视点图像的深度图像。 编码单元（130）在压缩彩色图像运动矢量之后生成压缩图像。 运动确定单元确定两帧之间是否存在图像运动。
-
公开(公告)号：KR1020100122999A
公开(公告)日：2010-11-24
申请号：KR1020090041973
申请日：2009-05-14
Applicant: 삼성전자주식회사
CPC classification number: H04N19/593 , H04N19/105 , H04N19/159 , H04N19/597
Abstract: PURPOSE: A multiview image compression device and a method thereof for performing the compression of a multiview image and designing a space prediction structure by using a geometric structure of a camera are provided to improve compression efficiency and maximize the space correlation by designing the space prediction structure. CONSTITUTION: A geometric information extractor(110) extracts geometry information from a plurality of cameras. A space prediction structure determiner(120) determines the space prediction structure by using the geometry information. A multiview image compressor(130) compresses the camera image according to the space prediction structure. The geometry information includes location information and posture information of a plurality of cameras.
Abstract translation: 目的：提供一种多视图图像压缩装置及其方法，用于通过使用相机的几何结构来执行多视图图像的压缩和设计空间预测结构，以通过设计空间预测结构来提高压缩效率和最大化空间相关性 。 构成：几何信息提取器（110）从多个摄像机提取几何信息。 空间预测结构确定器（120）通过使用几何信息来确定空间预测结构。 多视点图像压缩器（130）根据空间预测结构压缩摄像机图像。 几何信息包括多个照相机的位置信息和姿势信息。
-
公开(公告)号：KR1020100088774A
公开(公告)日：2010-08-11
申请号：KR1020090007875
申请日：2009-02-02
Applicant: 삼성전자주식회사
CPC classification number: H04N13/111 , H04N13/00
Abstract: PURPOSE: An apparatus and a method for generating an intermediate view image are provided to prevent the hole or blurring which may happen in an intermediate view image by generating an intermediate view image through the right and left view images. CONSTITUTION: In consideration of a disparity map, a generation unit generates a middle view image from one image among right or left view image. A block setting unit(231) sets up a reference block of the size determined for a region adjacent to a covered region of the intermediate view image. A search unit(232) searches a reference block and the reference block for another image. In consideration of another image, a region supplement unit(233) compensates the hidden region.
Abstract translation: 目的：提供一种用于产生中间视图图像的装置和方法，以通过通过左视图图像和左视图图像生成中间视图图像来防止可能在中间视图图像中发生的孔或模糊。 构成：考虑到视差图，生成单元从右图像或左视图图像中的一个图像生成中间视图图像。 块设置单元（231）设置为与中间视图图像的被覆盖区域相邻的区域确定的尺寸的参考块。 搜索单元（232）搜索参考块和另一图像的参考块。 考虑到另一图像，区域补充单元（233）补偿隐藏区域。
-
公开(公告)号：KR1020100073724A
公开(公告)日：2010-07-01
申请号：KR1020080132475
申请日：2008-12-23
Applicant: 삼성전자주식회사
IPC: G06T7/00
CPC classification number: G06T3/4023 , G06F17/17 , H04N1/417
Abstract: PURPOSE: An apparatus and a method for extrapolating an image through multi-fitting are provided to increase the visual satisfaction of a viewer by increasing output resolution based on the estimation of an external signal of an image with utmost signals. CONSTITUTION: A boundary-adjacent input signal extractor(110) extracts an input signal for an adjacent boundary of the input image by using an extension coefficient and an input image. A monotonicity decision unit(120) analyzes the input signal of extracted boundary vicinity, and judges the monotonicity through the analysis of the input signal. According to the determined monotonicity, a monotonicity preservation signal generator(130) generates the monotonicity preservation signal.
Abstract translation: 目的：提供一种通过多重拟合来推断图像的装置和方法，以通过基于对具有最大信号的图像的外部信号的估计来增加输出分辨率来增加观看者的视觉满意度。 构成：边界相邻输入信号提取器（110）通过使用扩展系数和输入图像提取输入图像的相邻边界的输入信号。 单调性判定单元（120）分析提取的边界附近的输入信号，并通过对输入信号的分析来判断单调性。 根据确定的单调性，单调性保持信号发生器（130）产生单调性保持信号。
-
-
-
-
-
-
-
-
-
Search Results
56
records
(took 0.032 seconds)
