-
-
公开(公告)号:KR101694300B1
公开(公告)日:2017-01-09
申请号:KR1020140025530
申请日:2014-03-04
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00302 , B33Y80/00 , G05B19/40931 , G05B2219/2636 , G05B2219/49023 , G06K9/00 , G06K9/00214 , G06K9/00268 , G06T19/20 , Y02P90/265
Abstract: 본발명은이기종센서들로부터획득한얼굴데이터로부터얼굴영역및 부위별특징을검출하고전역/지역적변형최적화를통해 3D 프린팅에적합한자동 3D 개인피규어생성장치및 그방법에관한것이다. 3D 개인피규어생성장치는복원대상에해당하는사용자의얼굴데이터를획득하는단계, 얼굴데이터에서부위별특징점을추출하고, 추출한특징점을기반으로얼굴의 3D 고유모델을복원하는단계, 3D 고유모델과이전에저장된얼굴표정모델, 바디/치장모델을기반으로 3D 피규어모델을생성하는단계및 3D 피규어모델을실제 3D 프린팅에해당하는구조및 형태인지검증하고, 검증한결과를토대로수정및 편집하여 3D 프린팅에해당하는 3D 피규어모델을출력하는단계를수행한다.
Abstract translation: 本文公开了一种用于通过从由异质传感器获取的面部数据和优化全局/局部变换来检测来自相应区域的面部区域和特征来自动创建适于3D打印的3D个性化图案的装置和方法。 3D个性化图形生成装置获取与重建对象对应的用户的面部数据; 从面部数据中提取各个区域的特征点,并基于提取的特征点重建用户脸部的独特3D模型; 基于独特的3D模型和先前存储的面部表情模型和身体/装饰模型创建3D图形模型; 并验证每个3D图形模型是否具有与实际3D打印相对应的结构和形状,基于验证结果校正和编辑3D图形模型,并输出与3D打印相对应的3D图形模型。
-
公开(公告)号:KR1020160073727A
公开(公告)日:2016-06-27
申请号:KR1020140182386
申请日:2014-12-17
Applicant: 한국전자통신연구원
CPC classification number: B33Y50/02 , G05B19/4099 , G05B2219/35134 , G06T19/20 , G06T2219/2021 , Y02P90/265 , B33Y99/00 , Y10S715/964
Abstract: 3차원객체저작장치가개시된다. 이장치는상기예제객체들별로다수의디자인속성을설정하고, 상기다수의디자인속성이설정된상기예제객체들을특징벡터공간의좌표값으로표현가능하도록상기디자인속성에특징값을할당하고, 상기특징값을할당받은예제객체들을초기예제객체들로서생성하는초기예제객체생성부와, 비선형변형(Nonlinear Deformation) 기법에따라상기각 디자인속성별로할당된특징값을세분화하여확장하고, 상기세분화된특징값을조합하여상기특징벡터공간에서표현가능한확장예제객체를생성하는확장예제객체생성부및 상기특징벡터공간내에서, 사용자입력에따라입력된특징값이속한공간영역의꼭지점에해당하는확장예제개체를선택하고, 상기선택된확장예제객체들의가중값을계산하고, 상기계산된가중값이반영된상기선택된확장예제객체의메쉬정점(mesh vertices)을선형중첩하여, 사용자가원하는디자인속성을부여한사용자객체를생성하는사용자객체생성부를포함한다.
Abstract translation: 公开了一种三维对象创作装置。 本发明的目的是提供一种能够通过在非线性特征空间中应用变换方法来实时地创建用于三维打印的三维对象的三维对象创作装置及其方法。 该设备包括:初始示例对象生成单元,用于为每个示例对象设置多个设计属性,以向设计属性分配特征值以表示其中设计属性被设置为特征向量的坐标值的示例对象 空间,并产生接收特征值作为初始示例对象的示例对象; 扩展示例对象生成单元,用于根据非线性变形方法细分并扩展由每个设计属性分配的特征值,以组合细分特征值,并产生能够在特征向量空间中表达的扩展示例对象 ; 以及用户对象生成单元,用于选择对应于由特征向量空间内的用户输入输入的特征值的空间区域的顶点对应的扩展示例对象,以计算所选择的扩展示例对象的加权值,以线性重叠 所选择的扩展示例对象的网格顶点,其中反映了所计算的加权值,并且产生其中给出了用户的期望设计属性的用户对象。
-
公开(公告)号:KR1020130026854A
公开(公告)日:2013-03-14
申请号:KR1020110090282
申请日:2011-09-06
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00275 , G06K9/00993 , G06T17/00
Abstract: PURPOSE: A face data management system and a method thereof are provided to confirm searched face data through a display device by searching the location and direction of a desired face by using an interface device. CONSTITUTION: A face data storage unit(200) stores global face data, local face data, and global and local face matching data generated by a face data generation device(100). A face data search device(300) receives user input by using a user interface device. The face data search device outputs a face image on a display unit(340). [Reference numerals] (110a) Structured light generating unit; (110b) Camera; (110c) Storage media; (112) Global face data generating unit; (114) Global face posture extracting unit; (120) Local face photographing unit; (122) Local photographing device posture extracting unit; (124) Local face data generating unit; (130) Global and local face matching unit; (210) Global face data storage unit; (220) Local face data storage unit; (230) Matched data storage unit; (310) Data selecting unit; (320) Rendering unit; (330) Converted image generating unit; (340) Display unit
Abstract translation: 目的:提供一种脸部数据管理系统及其方法,通过使用界面装置搜索所需面部的位置和方向,通过显示装置确认搜索到的面部数据。 构成:面部数据存储单元(200)存储面部数据生成装置(100)生成的全局面部数据,局部脸部数据以及全局和局部脸部对照数据。 面部数据搜索装置(300)通过使用用户界面装置接收用户输入。 面部数据搜索装置在显示单元(340)上输出面部图像。 (附图标记)(110a)结构光产生单元; (110b)相机; (110c)存储介质; (112)全局人脸数据生成单元; (114)全球面部姿势提取单元; (120)本地面部摄影单元; (122)本地拍摄装置姿势提取单元; (124)局部面部数据生成单元; (130)全球和本地面部匹配单位; (210)全球人脸数据存储单元; (220)本地面部数据存储单元; (230)匹配数据存储单元; (310)数据选择单元; (320)渲染单元; (330)转换图像生成单元; (340)显示单元
-
公开(公告)号:KR101208427B1
公开(公告)日:2012-12-05
申请号:KR1020080131664
申请日:2008-12-22
Applicant: 한국전자통신연구원
CPC classification number: H04N5/232 , H04N5/23203 , H04N5/247 , H04N5/77 , H04N7/173 , H04N9/8227 , H04N21/21805 , H04N21/231 , H04N21/242 , H04N21/4223
Abstract: 본발명은여러대의 HD 카메라를제어하여동기화된영상을얻고그것을실시간무손실로저장하며각 카메라의팬(pan), 틸트(tilt), 줌(zoom) 등의동작을제어하기위한동기화된실시간다시점영상관리장치및 방법에관한것이다. 본발명은, 다중카메라로부터동기화된영상을무손실로취득하는영상취득부와, 상기취득한영상을저장하고상기취득한영상을기반으로상기카메라의팬, 틸트, 줌동작을제어하는하나이상의인제스트에이전트, 및상기인제스트에이전트에제어명령을전송하고상기저장된영상을수신하여통합저장하는중앙서버를포함하는카메라제어및 영상저장장치를제공한다. 본발명에의하면, 카메라제어, 영상취득, 동기화등의다양한기능을소프트웨어적으로수행함으로써시스템이단순해지고업그레이드가간편해지며, 다중영상을실시간무손실로저장할수 있게되며각종영상처리를쉽게구현하고영상전송효율성을높일수 있게되는효과가있다.
-
公开(公告)号:KR1020120068470A
公开(公告)日:2012-06-27
申请号:KR1020100130107
申请日:2010-12-17
Applicant: 한국전자통신연구원
Inventor: 황본우
CPC classification number: G06T7/0075 , G06T7/593 , G06T2207/10012 , H04N13/106 , H04N13/207 , H04N2013/405
Abstract: PURPOSE: A stereo image matching apparatus and a method thereof are provided to determine the displacement of a stereo image by calculating a pair of windows having maximum similarity sum of images in a local support area. CONSTITUTION: A local supporting area determining unit(50) extracts a similarity mask and a local support area mask according to similarity difference of window images. A similarity extracting unit(60) calculates local support area similarity of the local support area mask. A displacement selecting unit(70) selects a pair of window images having great local support area similarity. The displacement selecting unit determines the displacement of a three-dimensional image.
Abstract translation: 目的:提供一种立体图像匹配装置及其方法,用于通过计算具有本地支持区域中的图像的最大相似和的一对窗口来确定立体图像的位移。 构成:本地支持区域确定单元(50)根据窗口图像的相似度差异提取相似度掩码和本地支持区域掩码。 相似度提取单元(60)计算本地支持区域掩码的本地支持区域相似度。 位移选择单元(70)选择具有大的局部支持区域相似度的一对窗口图像。 位移选择单元确定三维图像的位移。
-
公开(公告)号:KR101078641B1
公开(公告)日:2011-11-01
申请号:KR1020080126032
申请日:2008-12-11
Applicant: 한국전자통신연구원 , 명지대학교 산학협력단
CPC classification number: H04N5/765 , G06F3/016 , H04N5/85 , H04N9/8205 , H04N21/41 , H04N21/42201 , H04N21/485 , H04N21/812 , H04N21/8133 , H04N21/8166 , H04N21/8186
Abstract: 본발명은감각재생장치에관계된메타데이터를이용한멀티미디어응용기술에관한것으로, 동영상콘텐츠시청에있어동영상콘텐츠를구성하는내용에따라디스플레이기기의색감및 주변조명등 주변감각재생장치를제어하는멀티미디어응용시스템에대한메타데이터를구성하는과정에있어서, 감각재생장치를효과적으로제어하기위한감각효과정보(SEI) 메타데이터를활용하는방법을포함하고있으며, 메타데이터에기반을둔 콘텐츠활용방법적도구들을포함함으로써, 동영상콘텐츠제작에서최종소비에이르기까지감각재생공정간에제작자의의도에따른소비자중심의고품질멀티미디어서비스제공을특징으로한다. 본발명에의하면, 동영상콘텐츠의내용에따라주변감각재생장치를제어하기위한메타데이터를활용하여감각재생시 제작자의의도에따른소비자중심의고품질멀티미디어서비스제공을가능하게할 수있다.
-
公开(公告)号:KR1020110066609A
公开(公告)日:2011-06-17
申请号:KR1020090123341
申请日:2009-12-11
Applicant: 한국전자통신연구원
CPC classification number: G06K9/38 , G06T7/11 , G06T7/143 , G06T7/194 , G06T2207/20081
Abstract: PURPOSE: A device for dividing a panorama and background based on a code book in a multi view image and method thereof are provided to minimize a division performance degradation of a panorama and background. CONSTITUTION: A background model generating unit(100) extracts a codebook from a multi-view background image. The background model generating unit generates a codeword mapping table interlinking with the codebook. A panorama and background separation executing unit(150) divides a multi-viewpoint image into panorama and a background. A codebook generating unit(102) extracts a codeword index from each pixel information of the multi-view background image.
Abstract translation: 目的:提供一种基于多视图图像中的码本划分全景和背景的装置及其方法,以最小化全景和背景的分割性能下降。 构成:背景模型生成单元(100)从多视点背景图像中提取码本。 背景模型生成单元生成与码本互连的码字映射表。 全景和背景分离执行单元(150)将多视点图像划分为全景图像和背景图像。 码本生成部(102)从多视点背景图像的各像素信息中提取码字索引。
-
公开(公告)号:KR1020100073174A
公开(公告)日:2010-07-01
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: PURPOSE: A method and an apparatus for deforming a shape of a human body model are provided to automatically and naturally deform the shape of a human body model by forming the shape of the human body according to the motion capture data. CONSTITUTION: A human body generator(200) reconfigures a Nurbs curved surface model based on a joint skeleton structure, and a statistical transformation information generator(300) generates the statistical deformation information for a control parameter of the Nurbs curved surface model based on a parameter of a joint and key section curve in various specific motions. A human body shape deformation unit(400) deforms the human body model shape according to the statistical deformation information and the Nurbs curved surface model.
Abstract translation: 目的:提供一种用于使人体模型的形状变形的方法和装置,以通过根据运动捕获数据形成人体的形状来自动且自然地变形人体模型的形状。 构成:人体发生器(200)基于联合骨架结构重新配置Nurbs曲面模型,并且统计变换信息生成器(300)基于参数生成Nurbs曲面模型的控制参数的统计变形信息 在各种具体运动中的关节和关键曲线曲线。 根据统计变形信息和Nurbs曲面模型,人体形状变形单元(400)使人体模型形状变形。
-
公开(公告)号:KR1020100012013A
公开(公告)日:2010-02-04
申请号:KR1020080126032
申请日:2008-12-11
Applicant: 한국전자통신연구원 , 명지대학교 산학협력단
CPC classification number: H04N5/765 , G06F3/016 , H04N5/85 , H04N9/8205 , H04N21/41 , H04N21/42201 , H04N21/485 , H04N21/812 , H04N21/8133 , H04N21/8166 , H04N21/8186
Abstract: PURPOSE: A multimedia application system using metadata for a sensory device and a method thereof are provided to add a sensory function about moving picture color reproduction made by an intention of a producer from moving picture content production to final consumption, thereby providing a consumer customized high quality multimedia service. CONSTITUTION: A sensory device engine unit(108) receives SEI(Sensory Effect Information) according to moving picture contents. The sensory device engine unit receives UPI(User Preference Information) of the sensory device(114~120). When DCI(Device Capability Information) is inputted, the sensory device engine unit generates SDC(Sensory Device Commands) for controlling the sensory device based on the input information. A sensory device controller(110) controls each sensory device.
Abstract translation: 目的:提供一种使用用于感觉装置的元数据的多媒体应用系统及其方法,用于将由生产者的意图产生的运动图像色彩再现从运动图像内容生成添加到最终消费中,从而为消费者定制高 高质量的多媒体服务。 构成:感觉装置引擎单元(108)根据运动图像内容接收SEI(感觉效果信息)。 感觉装置引擎单元接收感觉装置(114〜120)的UPI(用户偏好信息)。 当输入DCI(设备能力信息)时,感觉设备引擎单元根据输入信息产生用于控制感觉设备的SDC(感觉设备命令)。 感觉装置控制器(110)控制每个感觉装置。
-
-
-
-
-
-
-
-
-