-
公开(公告)号：KR101587962B1
公开(公告)日：2016-01-28
申请号：KR1020100132882
申请日：2010-12-22
Applicant: 한국전자통신연구원
CPC classification number: G06T7/251 , G06T7/13 , G06T2207/20044
Abstract: 모션캡처장치및 방법이개시된다. 본발명에따른모션캡처장치는 T+1 프레임에서개체를촬영한다시점영상을이용하여상기개체의 3차원외형을복원하는 3차원외형복원부; 상기 3차원외형및 T 프레임의모션을가진상기개체의모델을이용하여 T+1 프레임의 3차원뼈대를생성하는 3차원뼈대화부; 및상기 T+1 프레임의 3차원뼈대에기반하여상기모델의모션을변경하여상기모델의 T+1 프레임의모션을복원하는모션복원부를포함한다.
Abstract translation: 这里公开了一种运动捕捉装置和方法。 运动捕捉装置包括三维（3D）外观重建单元，用于使用通过在第T + 1帧捕获实体而获得的多视图图像来重构实体的3D外观。 3D骨骼生成单元使用3D外观和具有第T帧的运动的实体的模型来生成第T + 1帧的3D骨架。 运动重建单元通过基于第T + 1帧的3D骨架改变模型的运动来重建模型的第T + 1帧的运动。
-
公开(公告)号：KR1020140077752A
公开(公告)日：2014-06-24
申请号：KR1020120146866
申请日：2012-12-14
Applicant: 한국전자통신연구원
IPC: G06T17/00
Abstract: A method for recreating the same effects as an actual makeup on an image is disclosed. The method for recreating makeup effects on an image comprises the steps of generating a cosmetics pattern, based on a difference between an image before makeup and an image after makeup; generating a makeup tool pattern with reference to the characteristics of a makeup tool; generating an applying pattern synthesized by resizing the cosmetics pattern and the makeup tool pattern; and applying the applying pattern to an original image by using the applying pattern as an alpha channel to recreate a makeup effect. Therefore, the skin can be recreated realistically as if cosmetics are actually applied to the skin.
Abstract translation: 公开了一种用于重建与图像上的实际化妆相同效果的方法。 用于重建对图像的化妆效果的方法包括基于化妆前的图像与化妆后的图像之间的差异来生成化妆品图案的步骤; 参考化妆工具的特点产生化妆工具图案; 生成通过调整化妆品图案和化妆工具图案的大小而合成的施加图案; 以及通过使用所述应用模式作为阿尔法通道将所述应用模式应用于原始图像以重新形成化妆效果。 因此，皮肤可以被真实地再现，就好像化妆品实际应用于皮肤一样。
-
公开(公告)号：KR1020130047194A
公开(公告)日：2013-05-08
申请号：KR1020110112068
申请日：2011-10-31
Applicant: 한국전자통신연구원
CPC classification number: G06T13/40 , G06T17/30 , G06T2200/08
Abstract: PURPOSE: An automatic skinning and appearance restoring device of a dynamic object and a method thereof are provided to automatically generate an intrinsic mesh model without geometric correction by using multiview image information including only appearance information of the dynamic object. CONSTITUTION: An image capturing unit(200) generates a multiview image of a dynamic object, multiview silhouette information, and a primary globally fitted standard mesh model based on a dynamic image object and a standard mesh model. A 3D image restoring unit(300) generates an NURBS-based intrinsic mesh model in the dynamic object about the primary globally fitted standard mesh. A data output unit(400) generates and outputs animation data and a final intrinsic mesh model. [Reference numerals] (100) Camera; (200) Image capturing unit; (300) 3D image restoring unit; (400) Skinning and skin data output unit; (AA) Standard mesh data; (BB) Final intrinsic mesh model; (CC) Animation data
Abstract translation: 目的：提供一种动态对象的自动皮肤和外观恢复装置及其方法，通过使用仅包括动态对象的外观信息的多视点图像信息，自动生成不进行几何校正的内在网格模型。 构成：图像捕获单元（200）基于动态图像对象和标准网格模型生成动态对象的多视图图像，多视角剪影信息和主要全局拟合的标准网格模型。 3D图像恢复单元（300）在关于主要全局拟合的标准网格的动态对象中生成基于NURBS的内在网格模型。 数据输出单元（400）生成并输出动画数据和最终的本征网格模型。 （附图标记）（100）相机; （200）图像捕获单元; （300）3D图像恢复单元; （400）皮肤和皮肤数据输出单元; （AA）标准网格数据; （BB）最终内在网格模型; （CC）动画数据
-
公开(公告)号：KR1020130037996A
公开(公告)日：2013-04-17
申请号：KR1020110102601
申请日：2011-10-07
Applicant: 한국전자통신연구원
CPC classification number: G06T7/0081 , G06K9/00664 , G06K9/00697 , G06T7/11 , G06T7/136 , G06T7/194 , G06T2200/04 , G06T2207/10028 , G06T2207/20021 , G06T5/00 , G06T17/00 , H04N5/21
Abstract: PURPOSE: A space dividing method of a 3D point cloud is provided to easily divide buildings, trees, or vehicles from points of an object automatically divided into a floor and other objects. CONSTITUTION: A space dividing unit divides a space of a 3D point cloud into a plurality of grid cells(S10). The space dividing unit sets a base surface as a floor on the space(S20). An accumulator accumulates a point of all gird cells on the base surface in a grid cell of the base surface(S30). A grouping unit divides the grid cell into a floor and objects according to the number of accumulated points(S70). [Reference numerals] (AA) Start; (BB) End; (S10) Equally dividing into a plurality of grid cells; (S20) Setting a base surface; (S30) Accumulating a point of a grid cell vertical to the base surface; (S40) Setting a group; (S50) Setting a boundary bounding box; (S60) Determining an object group bounding box; (S70) Determining a floor and objects;
Abstract translation: 目的：提供3D点云的空间分割方法，以容易地将建筑物，树木或车辆从物体的点自动分为地板和其他物体。 构成：空间分割单元将3D点云的空间划分为多个网格单元（S10）。 空间分割单元将空间的底面设为底面（S20）。 累加器在基面的网格单元中累积基底表面上的所有网格单元的点（S30）。 分组单元根据累积点的数量将网格单元划分为楼层和对象（S70）。 （附图标记）（AA）开始; （BB）结束; （S10）同样地分成多个网格单元; （S20）设置基面; （S30）堆积垂直于基面的网格单元的点; （S40）设定组; （S50）设定边界边界框; （S60）确定对象组边界框; （S70）确定楼层和物体;
-
公开(公告)号：KR1020130026853A
公开(公告)日：2013-03-14
申请号：KR1020110090281
申请日：2011-09-06
Applicant: 한국전자통신연구원
IPC: G06T15/04
CPC classification number: G06T15/04 , G06T17/00 , G06T2210/56
Abstract: PURPOSE: A rendering method of point cloud using a voxel grid and a device thereof are provided to maintain fixed rendering speed by determining the rendering speed for the number of voxel grids. CONSTITUTION: A user interface(100) sets parameters which are necessary for a rendering process. A voxel grid generation unit(110) generates a bounding box which is able to include a point cloud mode and divides the bounding box with a voxel grid. A texture plane generation unit(120) assigns two or more texture planes to each voxel grid. A texture image generation unit(130) generates a texture image. A rendering unit(140) executes a rendering process by using the location information of a camera and the texture image stored in a memory. [Reference numerals] (100) User interface; (110) Voxel grid generation unit; (120) Texture plane generation unit; (130) Texture image generation unit; (140) Rendering unit
Abstract translation: 目的：提供使用体素网格及其装置的点云的渲染方法，以通过确定体素网格数目的渲染速度来保持固定的渲染速度。 构成：用户界面（100）设置渲染过程所必需的参数。 体素网格生成单元（110）生成能够包括点云模式的边界框，并且使用体素网格划分边界框。 纹理平面生成单元（120）将两个或多个纹理平面分配给每个体素网格。 纹理图像生成单元（130）生成纹理图像。 渲染单元（140）通过使用摄像机的位置信息和存储在存储器中的纹理图像来执行渲染处理。 （附图标记）（100）用户界面; （110）体素网格生成单元; （120）纹理平面生成单元; （130）纹理图像生成单元; （140）渲染单元
-
Patent Agency Ranking
公开(公告)号：KR1020130022041A
公开(公告)日：2013-03-06
申请号：KR1020110084703
申请日：2011-08-24
Applicant: 한국전자통신연구원
CPC classification number: G06T17/00 , G06T2200/08
Abstract: PURPOSE: A local multi-resolution 3 dimensional face intrinsic model generating device, a generating method thereof, and a facial skin management system are provided to conveniently generate a local multi-resolution 3 dimensional face intrinsic model by automatically compositing an image photographed from a set position with the 3 dimensional face intrinsic model. CONSTITUTION: A 3 dimensional face intrinsic model generating part(210) generates a 3 dimensional face intrinsic model by compositing at least one 3 dimensional face model. A local photographing camera(230) photographs a local region of a face. A control part(220) controls the position of the local photographing camera at the 3 dimensional face intrinsic model. A local multi-resolution 3 dimensional face intrinsic model generating part(240) generates a local multi-resolution face intrinsic model by compositing the 3 dimensional face intrinsic model with an image secured from the local photographing camera. [Reference numerals] (200-1) 3 dimensional face model generating camera 1; (200-2) 3 dimensional face model generating camera 2; (200-N) 3 dimensional face model generating camera N; (210) 3 dimensional face intrinsic model generating part; (220) Control part; (230) Local photographing camera; (240) Local multi-resolution 3 dimensional face intrinsic model generating part; (250) Display part
Abstract translation: 目的：提供一种局部多分辨率三维面内在模型生成装置，其生成方法和面部皮肤管理系统，以便通过自动合成从组中拍摄的图像来方便地生成局部多分辨率三维面部内在模型 位置与3维面内在模型。 构成：通过合成至少一个三维面部模型，三维面部内在模型生成部分（210）产生三维面部内在模型。 本地拍摄照相机（230）拍摄面部的局部区域。 控制部件（220）在3维面部固有模型中控制本地拍摄相机的位置。 本地多分辨率三维人脸本征模型生成部分（240）通过将三维人脸本征模型与从本地拍摄照相机确保的图像合成来产生局部多分辨率面内在模型。 （200-1）三维人脸模型生成摄像机1; （200-2）三维面部模型生成摄像机2; （200-N）三维人脸模型生成相机N; （210）3维面内在模型生成部分; （220）控制部分; （230）本地摄像机; （240）本地多分辨率3维面内在模型生成部分; （250）显示部分
-
公开(公告)号：KR1020130009603A
公开(公告)日：2013-01-23
申请号：KR1020120038466
申请日：2012-04-13
Applicant: 한국전자통신연구원
CPC classification number: G06T7/85 , G06T2207/30244 , H04N5/247
Abstract: PURPOSE: A camera network revision method with a small revision pattern and a device thereof are provided to easily revise a camera of a camera network. CONSTITUTION: A projection matrix generated from each camera is inputted to plural revision patterns. A part of the projection matrix is generated as a sub projection matrix. The sub projection matrix is connected to generate a sub measurement matrix. SVD(Singular Value Decomposition) is performed to the sub measurement matrix. The sub measurement matrix is converted into a matrix having a 3-rank. The SVD is performed to the converted sub measurement matrix. The network revision of the camera is performed.
Abstract translation: 目的：提供一种具有小修订版本的相机网络修订方法及其设备，以便轻松修改相机网络的相机。 构成：从每个摄像机产生的投影矩阵被输入到多个修订图案。 投影矩阵的一部分被生成为子投影矩阵。 子投影矩阵被连接以产生子测量矩阵。 对子测量矩阵执行SVD（奇异值分解）。 子测量矩阵被转换为具有3级的矩阵。 对转换的子测量矩阵执行SVD。 执行相机的网络版本。
-
公开(公告)号：KR101221451B1
公开(公告)日：2013-01-11
申请号：KR1020080131769
申请日：2008-12-22
Applicant: 한국전자통신연구원
CPC classification number: G06T13/40
Abstract: 본 발명은 다시점(multi-view) 카메라를 이용하여 동적 개체의 동적 모션을 캡쳐한 다시점 비디오 영상으로부터 대상체의 표준모델을 이용하여 프레임 별로 대상체의 고유 외형 변형 정보와 관절 모션을 자동으로 복원하고 복원된 정보들을 이용하여 동적 개체의 디지털 클론을 생성하는 방법에 관한 것이다. 외형과 모션의 복원을 위해 각 프레임의 다시점 영상에서 얻어지는 실루엣 정보와 동적 개체의 투영영역의 칼라값, 그리고 동적 개체 골격계의 운동학적(Kinematic) 제약조건을 이용한다. 생성된 동적 개체의 디지털 클론은 표준모델의 메쉬 구조(Mesh Topology)를 따르므로 외형이나 모션의 제어에 대한 시공간적 일치성(spatio-temporal coherency)을 확보할 수 있으며 사용자가 원하는 형상변형이나 모션제어를 일관성 있게 적용할 수 있다.
삼차원 모델 생성, 골격계, 다시점 영상, 파라미트릭 형상 변형, 관절 제어, 삼차원 복원-
公开(公告)号：KR101194604B1
公开(公告)日：2012-10-25
申请号：KR1020080131768
申请日：2008-12-22
Applicant: 한국전자통신연구원
IPC: G06T17/00
CPC classification number: G06T17/30 , G06T13/40 , G06T19/20 , G06T2219/2021
Abstract: 본 발명은 조인트-스켈레톤(joint-skeleton) 구조의 파라메트릭 제어 넙스(nurbs)곡면 기반 인체 모델을 이용하여 연기자의 모션을 캡쳐한 모션 캡쳐 데이터에 따라 자동으로 인체 형상이 자연스럽게 변형되는 인체 모델 형상 변형 기술에 관한 것으로, 3차원 폴리곤 메쉬 인체 모델을 입력받아 스켈레톤 구조를 갖는 계층구조의 조인트, 조인트와 조인트 사이의 키 단면 곡선, 키 단면 곡선들을 근사화하여 생성한 넙스 파라메트릭 제어 곡면, 넙스 파라메트릭 제어 곡면과 스킨 버텍스 사이의 종속관계 설정으로 구성되며, 각 관절부위의 특징적인 특정 모션(포즈)에 대한 학습(training) 데이터로부터의 넙스 곡면의 제어 파라미터에 대한 통계적 변형 정보를 이용하여 연기자의 자연스러운 여러 모션(동작)을 인체 모델의 형상 변형으로 실시간 자동 재현할 수 있 다.
3차원 인체 모델, 넙스(Nurbs) 기반 파라메트릭(parametric) 제어곡면, 키 단면 곡선, 통계적 변형 정보-
公开(公告)号：KR1020120071299A
公开(公告)日：2012-07-02
申请号：KR1020110065053
申请日：2011-06-30
Applicant: 한국전자통신연구원
CPC classification number: G06T15/205 , G06T19/20
Abstract: PURPOSE: A shape transition method of a three-dimensional model and an apparatus thereof are provided to generate a three-dimensional model for expressing a motion of the object. CONSTITUTION: A multi-view camera unit obtains a plurality of two-dimensional images in a first frame(S200). A shape restoring unit of a shape transition controller generates a first volume model(S210). The shape transition controller generates a first shape transition model(S220). The multi-view camera unit obtains a plurality of two-dimensional images in a second frame(S230). The shape restoring unit generates a second volume model(S240).
Abstract translation: 目的：提供三维模型的形状转换方法及其装置，以产生用于表示物体运动的三维模型。 构成：多视点摄像机单元在第一帧中获得多个二维图像（S200）。 形状转换控制器的形状恢复单元生成第一体积模型（S210）。 形状转变控制器产生第一形状转变模型（S220）。 多视点相机单元在第二帧中获得多个二维图像（S230）。 形状恢复单元生成第二体积模型（S240）。
-
-
-
-
-
-
-
-
-
Search Results
148
records
(took 0.031 seconds)
