-
公开(公告)号:KR1020140109185A
公开(公告)日:2014-09-15
申请号:KR1020130023578
申请日:2013-03-05
Applicant: 삼성전자주식회사 , 유니버시티 칼리지 런던
Abstract: An image processing apparatus is provided. The image processing apparatus comprises: a determination unit which determines a rendering parameter for rendering an inputted 3D model to a set degree of reality by using cost function including the relationship information between at least a portion of characteristic parameters of a rendered 3D model, a rendering method, rendering parameters of the rendering method, and a degree of reality of the rendered result image; and a render unit which renders the inputted 3D model by using the determined rendering parameter.
Abstract translation: 提供一种图像处理装置。 图像处理装置包括:确定单元,其通过使用成本函数来确定用于将输入的3D模型渲染到设定的实际程度的渲染参数,所述成本函数包括呈现的3D模型的特征参数的至少一部分之间的关系信息, 方法,渲染方法的渲染参数以及渲染结果图像的现实程度; 以及渲染单元,其通过使用所确定的渲染参数来呈现输入的3D模型。
-
公开(公告)号:KR1020130060158A
公开(公告)日:2013-06-07
申请号:KR1020120137122
申请日:2012-11-29
Applicant: 삼성전자주식회사
IPC: G06T15/50
CPC classification number: G06T15/506 , G06T19/006 , G06T2215/16 , H04N5/2258 , H04N13/239 , H04N13/243 , H04N13/25 , H04N13/271 , H04N13/275
Abstract: PURPOSE: An image processing device and method are provided to obtain accurate lamp environment information without considering kinds or distance of individual lamps included in lamp environment of a surrounding object, thereby reflecting the obtained information in rendering. CONSTITUTION: A light capture unit(110) photographs lamp environment of a surrounding object and generates lamp environment information. An object capture unit(120) generates object information by photographing the object. A modeling unit(130) generates a 3D model related to the object by reflecting the lamp environment information and the object information. A rendering unit(140) renders a 3D image by combining information for the 3D model with separately inputted virtual object information. [Reference numerals] (110) Light capture unit; (120) Object capture unit; (130) Modeling unit; (140) Rendering unit
Abstract translation: 目的:提供一种图像处理装置和方法,以便在不考虑周围物体的灯环境中包括的各个灯的种类或距离的情况下获得准确的灯环境信息,从而将获得的信息反映在渲染中。 构成:光捕获单元(110)照射周围物体的灯环境并产生灯环境信息。 对象捕获单元(120)通过拍摄对象来生成对象信息。 建模单元(130)通过反射灯环境信息和对象信息来生成与对象相关的3D模型。 渲染单元(140)通过将3D模型的信息与单独输入的虚拟对象信息组合来呈现3D图像。 (附图标记)(110)光捕获单元; (120)对象捕获单元; (130)建模单元; (140)渲染单元
-
公开(公告)号:KR1020120115063A
公开(公告)日:2012-10-17
申请号:KR1020110077389
申请日:2011-08-03
Applicant: 삼성전자주식회사 , 더 텍사스 에이 앤드 엠 유니버시티 시스템
CPC classification number: G06T7/285 , G06F17/30247 , G06T7/50
Abstract: PURPOSE: A device for estimating motion using a depth image and a method thereof are provided to precisely estimate motion under complicated illumination or complicated backgrounds. CONSTITUTION: A pose parameter determining unit(120) determines an initial pose parameter corresponding to an input depth image(110). A pose generating unit(140) applies the pose parameter to an operating model. The pose generating unit generates a first result pose(150) of the operating model. The pose generating unit generates a second result pose of the operating model based on a difference of the input depth image. [Reference numerals] (110) Input depth image; (120) Pose parameter determining unit; (130) Database; (140) Pose generating unit; (150) Result pose
Abstract translation: 目的:提供一种用于使用深度图像估计运动的装置及其方法,以精确地估计复杂照明或复杂背景下的运动。 构成:姿势参数确定单元(120)确定对应于输入深度图像(110)的初始姿态参数。 姿势生成单元(140)将姿势参数应用于操作模型。 姿势生成单元生成操作模型的第一结果姿势(150)。 姿势生成单元基于输入深度图像的差异生成操作模型的第二结果姿势。 (附图标记)(110)输入深度图像; (120)姿势参数确定单元; (130)数据库; (140)姿势发生单元; (150)结果姿势
-
公开(公告)号:KR1020120041086A
公开(公告)日:2012-04-30
申请号:KR1020100109294
申请日:2010-11-04
Applicant: 삼성전자주식회사 , 더 텍사스 에이 앤드 엠 유니버시티 시스템
CPC classification number: G06T17/00 , G06N3/006 , G06T7/285 , G06T19/003 , G06T2207/20044
Abstract: PURPOSE: An avatar generation processing device and a method thereof are provided to generate an accurate avatar operation considering the variation of a target user body shape and a sensor attachment position. CONSTITUTION: A skeleton database stores skeleton sizes on reference poses of a plurality of users. A difference function shows the difference between a forward kinematics function and a sensor position attached to a target user. A function determining unit(120) determines a skeleton prior function based on the statistics of skeleton sizes on a reference pose of users. A calculating unit(130) calculates a local coordinate corresponding to each joint skeleton size and a sensor(50) by using a difference function and the skeleton prior function.
Abstract translation: 目的:提供一种化身生成处理装置及其方法,用于考虑目标用户身体形状和传感器安装位置的变化来生成准确的化身操作。 构成:骨架数据库在多个用户的参考姿势上存储骨架尺寸。 差分函数显示正向运动学函数和附加到目标用户的传感器位置之间的差异。 功能确定单元(120)基于用户的参考姿势上的骨架尺寸的统计来确定骨架先验功能。 计算单元(130)通过使用差分函数和骨架先验函数来计算与每个关节骨架尺寸相对应的局部坐标和传感器(50)。
-
公开(公告)号:KR1020110032366A
公开(公告)日:2011-03-30
申请号:KR1020090089809
申请日:2009-09-22
Applicant: 삼성전자주식회사
IPC: G06T15/00
Abstract: PURPOSE: An image processing apparatus and a method thereof for minimizing aliasing are provided to increase rendering process speed while maintaining the quality of the image processing apparatus. CONSTITUTION: A mesh showing geometry information of an object of 3D model is projected on a plane of a first point of view. A first calculating unit(110) calculates the projected plane. A sampling unit(120) adaptively assigns the sampling point about mesh. A second calculating unit(130) writes out a shadow map about the sampling point.
Abstract translation: 目的:提供一种用于最小化混叠的图像处理装置及其方法,以在保持图像处理装置的质量的同时提高渲染处理速度。 构成:显示3D模型对象的几何信息的网格投影在第一视角的平面上。 第一计算单元(110)计算投影平面。 采样单元(120)自适应地分配关于网格的采样点。 第二计算单元(130)写出关于采样点的阴影图。
-
公开(公告)号:KR1020110014795A
公开(公告)日:2011-02-14
申请号:KR1020090072327
申请日:2009-08-06
Applicant: 삼성전자주식회사
Inventor: 하인우
IPC: G06T15/50
CPC classification number: G06T15/60
Abstract: PURPOSE: An image processing apparatus and a method thereof are provided to reduce the rendering time and conspicuously reduce amount of the rendering calculation without decreasing the quality of the stereoscopic image. CONSTITUTION: A first calculation unit(110) renders a first image of a first viewpoint about a 3D model including one or more objects. In case the distance between a first object and a first point is greater than a first critical value, a second calculation unit(120) renders a second image of a second point with the reuse of the pixel value. An image composition unit(140) generates a stereoscopic image.
Abstract translation: 目的:提供一种图像处理装置及其方法,以减少渲染时间并显着减少渲染计算量,而不降低立体图像的质量。 构成:第一计算单元(110)呈现关于包括一个或多个对象的3D模型的第一视点的第一图像。 在第一对象和第一点之间的距离大于第一临界值的情况下,第二计算单元(120)利用像素值的重用来呈现第二点的第二图像。 图像合成单元(140)生成立体图像。
-
公开(公告)号:KR1020090092618A
公开(公告)日:2009-09-01
申请号:KR1020080017958
申请日:2008-02-27
Applicant: 삼성전자주식회사
Inventor: 하인우
CPC classification number: G06T15/06 , G06T2210/52
Abstract: A 3D image processor and a processing method thereof are provided to reduce the calculation quantity and performing the calculation parallel in case of using a ray tracing technique. A space divider divides virtual object spaces in which virtual 3D objects are positioned into a plurality of sub spaces(110,120,130,140). A plurality of processors are matched with the divided sub spaces and calculates the locus of rays incident from the sub spaces. When rays within the first sub space are incident to the second sub space, the first processor corresponding to the first sub space stores locus information of the ray. The second processor corresponding to the second sub space calculates the locus of the ray inside the second sub space.
Abstract translation: 提供3D图像处理器及其处理方法以减少计算量并且在使用光线跟踪技术的情况下并行执行计算。 空间分隔器将虚拟对象空间(虚拟对象空间)分割成多个子空间(110,120,130,140)。 多个处理器与划分的子空间匹配,并且计算从子空间入射的光线的轨迹。 当第一子空间内的射线入射到第二子空间时,对应于第一子空间的第一处理器存储射线的轨迹信息。 对应于第二子空间的第二处理器计算第二子空间内的光线的轨迹。
-
-
公开(公告)号:KR101845231B1
公开(公告)日:2018-04-04
申请号:KR1020110057471
申请日:2011-06-14
Applicant: 삼성전자주식회사
CPC classification number: G06T19/006
Abstract: 영상처리장치가제공된다. 모델링부는입력칼라영상및 상기입력칼라영상에대응하는입력깊이영상으로부터실제오브젝트에대응하는 3D 모델을생성할수 있다. 그리고계산부는, 입력조명환경정보를이용하여입력가상오브젝트및 상기 3D 모델에대해포톤기반렌더링을수행하여, 상기가상오브젝트가상기 3D 모델에삽입됨으로써발생되는칼라왜곡정보를포함하는차분영상을생성하고, 렌더링부는상기차분영상을상기입력칼라영상과합성하여상기가상오브젝트를포함하는결과영상을생성할수 있다.
Abstract translation: 提供了一种图像处理装置。 建模单元可以从输入彩色图像和与输入彩色图像对应的输入深度图像生成与实际对象相对应的3D模型。 然后,计算单元使用输入的照明环境信息对输入的虚拟对象和3D模型执行基于光子的渲染,以产生包括通过将虚拟对象插入到3D模型中而生成的颜色失真信息的差异图像 渲染单元可以将差分图像与输入彩色图像合成以生成包括虚拟对象的结果图像。
-
公开(公告)号:KR1020170045809A
公开(公告)日:2017-04-28
申请号:KR1020150145737
申请日:2015-10-20
Applicant: 삼성전자주식회사
CPC classification number: G06T15/06 , G06T15/04 , G06T15/506 , G06T15/80
Abstract: 광원으로부터방출된광선이제1 객체를구성하는입자들을통과하여제2 객체와만나는교차점들의위치를산출하고, 교차점들의위치에커스틱텍스쳐를적용한결과에기초한커스틱맵을이용하여제1 객체를표현하는, 커스틱을표현하는방법및 장치를제공할수 있다.
Abstract translation: 通过使用心理声学地图基于通过构成光束从光源计算与所述第二对象的交叉路口相遇的位置所发出现在第一对象,应用心理声学纹理到交叉点的位置的颗粒的结果表示所述第一对象 可以提供用于表示光标的方法和设备。
-
-
-
-
-
-
-
-
-