公开(公告)号：KR1020130074383A

公开(公告)日：2013-07-04

申请号：KR1020110142431

申请日：2011-12-26

Applicant: 삼성전자주식회사

Inventor： 백아론

IPC: G06T15/00

CPC classification number: G06T15/205 ,  H04N13/111 ,  H04N13/122 ,  H04N13/128 ,  H04N13/261 ,  H04N13/349 ,  H04N19/597

Abstract: PURPOSE: A method and apparatus for processing an image using a multi-layer representation are provided to generate the multi-layer representation of a three-dimensional scene from multi-images and depth maps corresponding to the multi-images. CONSTITUTION: A method for processing an image includes: a step (110) of generating a multi-layer representation for a referential view having information about a scene for processing the image using at least one input view; a step (120) of expanding at least one layer of the representation; and a step (S130) of generating the output view of an output view point using the representation. The input views are respectively the views of certain view points for the scene, have images and depth maps related to the images. The representation has at least one layer generated on the basis of the image and the depth map. [Reference numerals] (110) Multi-layer representation generation; (120) Layer expansion; (130) Output view generation; (AA) Start; (BB) End

Abstract translation: 目的：提供一种使用多层表示处理图像的方法和装置，用于从对应于多图像的多图像和深度图生成三维场景的多层表示。 一种用于处理图像的方法包括：步骤（110），用于生成具有关于使用至少一个输入视图处理图像的场景的信息的参考视图的多层表示; 扩展所述表示的至少一层的步骤（120）; 以及使用该表示生成输出视点的输出视图的步骤（S130）。 输入视图分别是场景的某些视点的视图，具有与图像相关的图像和深度图。 该表示具有至少一个基于图像和深度图生成的层。 （附图标记）（110）多层表示生成; （120）层膨胀; （130）输出视图生成; （AA）开始; （BB）结束

公开(公告)号：KR1020130056808A

公开(公告)日：2013-05-30

申请号：KR1020120043506

申请日：2012-04-25

Applicant: 삼성전자주식회사

Inventor： 오관정 ,  이재준 ,  박두식 ,  백아론 ,  이진영

IPC: H04N7/24 ,  H04N13/00

CPC classification number: H04N19/597 ,  H04N19/102 ,  H04N19/46 ,  H04N19/503 ,  H04N19/85

Abstract: PURPOSE: A device for encoding and decoding additional information of a 3D image and a method thereof are provided to efficiently encode and decode the additional information by encoding and decoding only updated additional information. CONSTITUTION: A camera information receiving unit(210) receives camera information. A depth range receiving unit(220) receives depth range information. An encoding determining unit(240) determines an updated parameter from among parameters included in the camera information and the depth range information as a parameter to be encoded and generates update information including information about updated parameters and non-updated parameters. An encoding unit(270) encodes the update information and the parameter to be encoded. [Reference numerals] (210) Camera information receiving unit; (220) Depth range receiving unit; (230) Focusing depth receiving unit; (240) Encoding determining unit; (250) SEI encoding determining unit; (260) Floating point converting unit; (270) Encoding unit; (AA) Encoded additional information

Abstract translation: 目的：提供用于对3D图像的附加信息进行编码和解码的装置及其方法，以通过仅编码和解码仅更新的附加信息来有效地编码和解码附加信息。 构成：相机信息接收单元（210）接收相机信息。 深度范围接收单元（220）接收深度范围信息。 编码确定单元（240）从包含在相机信息中的参数和深度范围信息中确定更新参数作为要编码的参数，并生成包括关于更新参数和未更新参数的信息的更新信息。 编码单元（270）对更新信息和要编码的参数进行编码。 （附图标记）（210）相机信息接收单元; （220）深度范围接收单元; （230）聚焦深度接收单元; （240）编码确定单元; （250）SEI编码确定单元; （260）浮点转换单元; （270）编码单元; （AA）编码的附加信息

公开(公告)号：KR1020130056684A

公开(公告)日：2013-05-30

申请号：KR1020110122409

申请日：2011-11-22

Applicant: 삼성전자주식회사

Inventor： 오관정 ,  백아론 ,  이진영 ,  박두식 ,  이재준

IPC: H04N19/46 ,  H04N19/105 ,  H04N13/00

CPC classification number: H04N19/463 ,  H04N19/597

Abstract: PURPOSE: A device for encoding/decoding additional information about color images and depth images and a method thereof are provided to supply information about the encoding and decoding of the color images and the depth images by providing information about a view, an image type, a reference image, and focus depth. CONSTITUTION: A reference image confirming unit(140) confirms a color image and a depth image. A reference information generating unit(150) generates reference information including the number of color images, a list of the color images, the number of depth images and a list of the depth image. An encoding unit(180) encodes additional information including the reference information. [Reference numerals] (110) View confirming unit; (120) Image type determining unit; (130) View information generating unit; (140) Reference image confirming unit; (150) Reference information generating unit; (160) Focus depth receiving unit; (170) Encoding determining unit; (180) Encoding unit; (AA) Encoded additional information

Abstract translation: 目的：提供一种用于编码/解码关于彩色图像和深度图像的附加信息的装置及其方法，用于提供关于彩色图像和深度图像的编码和解码的信息，通过提供关于视图，图像类型， 参考图像和焦点深度。 构成：参考图像确认单元（140）确认彩色图像和深度图像。 基准信息生成部（150）生成包括彩色图像的数量，彩色图像的列表，深度图像的数量和深度图像的列表的参考信息。 编码单元（180）对包括参考信息的附加信息进行编码。 （附图标记）（110）查看确认单元; （120）图像类型确定单元; （130）查看信息生成单元; （140）参考图像确认单元; （150）参考信息生成单元; （160）聚焦深度接收单元; （170）编码确定单元; （180）编码单元; （AA）编码的附加信息

公开(公告)号：KR1020110090488A

公开(公告)日：2011-08-10

申请号：KR1020100010300

申请日：2010-02-04

Applicant: 삼성전자주식회사

Inventor： 백아론 ,  정용주 ,  정영주 ,  김지원

IPC: H04N13/02

CPC classification number: H04N13/271 ,  H04N13/243 ,  H04N13/246 ,  H04N13/279 ,  H04N13/282

Abstract: PURPOSE: A system and method generating time point by using a time difference surface rendering are provided to preferentially process foreground object than background setting. CONSTITUTION: An input unit(710) receives two color images and two disparity maps which is estimated by two color images. Two line expresses two time different function about to time difference map. A distance calculation unit(720) respectively calculates a distance between two lines and at least one spot. A time surface configuring unit(730) configures a time difference surface is a combination of a coordinate having a minimal cost based on a cost according to the calculated distance. A time location determination unit(740) determines a location of a new time point based on the time difference surface.

Abstract translation: 目的：提供通过使用时差表面渲染产生时间点的系统和方法来优先处理前景对象而不是背景设置。 构成：输入单元（710）接收由两个彩色图像估计的两个彩色图像和两个视差图。 两行表示两个时间不同的函数关于时差图。 距离计算单元（720）分别计算两条线和至少一个点之间的距离。 时间表面配置单元（730）构成时差表面是根据计算出的距离的成本，具有最小成本的坐标的组合。 时间位置确定单元（740）基于时差表面确定新时间点的位置。

公开(公告)号：KR1020100135007A

公开(公告)日：2010-12-24

申请号：KR1020090053417

申请日：2009-06-16

Applicant: 삼성전자주식회사

Inventor： 백아론 ,  정용주

IPC: H04N13/00 ,  G02F1/1335 ,  G02B27/22

CPC classification number: G02B27/2214 ,  H04N5/23219 ,  H04N5/23293 ,  H04N13/305 ,  H04N13/351 ,  H04N13/366 ,  H04N13/368 ,  H04N13/111 ,  G02B27/22 ,  G02F1/1335

Abstract: PURPOSE: A multi-view image display device and a method thereof are provided to track a location of a viewer and locally display a multi-view image at the tracked location, thereby preventing a dead zone and supplying the same image without respect to the location of the viewer. CONSTITUTION: If a location of a head is changed, a variation measuring unit(514) measures a location variation of the head based on a location tracking result of a location tracking unit(511). A determining unit(515) compares the measured location variation with a threshold value. The determining unit determines whether the measured location variation exceeds the threshold value. If the measured location variation exceeds the threshold value, a transmitting unit(516) transmits the location tracking result to a display unit(512).

Abstract translation: 目的：提供多视图图像显示装置及其方法来跟踪观看者的位置并在被跟踪位置处本地显示多视点图像，从而防止死区并且不依赖于位置提供相同的图像 的观众。 构成：如果头部的位置改变，则变化测量单元（514）基于位置跟踪单元（511）的位置跟踪结果来测量头部的位置变化。 确定单元（515）将测量的位置变化与阈值进行比较。 确定单元确定测量的位置变化是否超过阈值。 如果测量的位置变化超过阈值，则发送单元（516）将位置跟踪结果发送到显示单元（512）。

公开(公告)号：KR1020100096644A

公开(公告)日：2010-09-02

申请号：KR1020090015626

申请日：2009-02-25

Applicant: 삼성전자주식회사

Inventor： 백아론 ,  정용주

IPC: G06T15/00

CPC classification number: G06T15/10 ,  G06T15/04 ,  G06T15/205 ,  G06T17/00

Abstract: PURPOSE: A device and a method for rendering are provided to prevent that occurrence of hole or wrong depth information included in output image. CONSTITUTION: A modeling unit(111) generates 3D model about the depth map based on depth value stored in a depth map(120). By projecting an input image(130) to the 3D modeul, a generating unit map a texture about the input image to the 3D model. By projecting the 3D model which the texture is mapped on an image plane, the generating unit generates an output image(140).

Abstract translation: 目的：提供一种用于渲染的设备和方法，以防止输出图像中包含的孔或错误深度信息的发生。 构成：建模单元（111）基于深度图（120）中存储的深度值生成关于深度图的3D模型。 通过将输入图像（130）投影到3D模式，生成单元将关于输入图像的纹理映射到3D模型。 通过将纹理映射到图像平面上的3D模型投影，生成单元生成输出图像（140）。

公开(公告)号：KR100809354B1

公开(公告)日：2008-03-05

申请号：KR1020070011197

申请日：2007-02-02

Applicant: 삼성전자주식회사

Inventor： 백아론 ,  박두식

IPC: H04N19/132 ,  H04N19/587

CPC classification number: H04N7/014 ,  H04N7/0127 ,  H04N19/513 ,  H04N19/587 ,  H04N19/61

Abstract: An apparatus and a method for up-converting a frame rate of restored frames are provided to up-convert a frame rate quickly and effectively and be suitably used for devices having limited memory resource and low-power conditions. A motion vector resampling unit(300) re-samples a motion vector for an interpolation frame by filtering a motion vector of a current block by using the motion vector of the current block and motion vectors of neighbor blocks of the current block from a restored frame. A motion compensating unit(400) generates the interpolation frame by performing motion compensation on the restored frame by the resampled motion vector. A switch(500) selects one of the restored frame and the interpolation frame according to the temporal order of the frames and outputs the same.

Abstract translation: 提供一种用于上变频恢复帧速率的装置和方法，用于快速有效地对帧速率进行上变频，并适用于具有有限的存储资源和低功率条件的设备。 运动矢量重采样单元（300）通过使用当前块的运动矢量和来自恢复帧的当前块的相邻块的运动矢量对当前块的运动矢量进行滤波，重新采样内插帧的运动矢量 。 运动补偿单元（400）通过对采样的运动矢量对恢复的帧执行运动补偿来生成内插帧。 开关（500）根据帧的时间顺序选择恢复帧和插值帧中的一个，并输出。