-
-
公开(公告)号:KR101805154B1
公开(公告)日:2017-12-06
申请号:KR1020160020710
申请日:2016-02-22
Applicant: 한국전자통신연구원
Abstract: 객체추적장치및 그방법이개시된다. 일실시예에따른객체추적장치는, 입력이미지에서타원객체를포함하는객체탐색영역을획득하는탐색부와, 객체탐색영역이미지를분석하여조명의현재상태를판단하고판단된조명의현재상태에따라객체위치결정을위한데이터를획득하는분석부와, 분석부를통해획득된데이터를이용하여전경및 배경을분리하고분리된전경이미지에서객체위치를결정하는결정부를포함한다.
-
公开(公告)号:KR101706200B1
公开(公告)日:2017-02-16
申请号:KR1020110124582
申请日:2011-11-25
Applicant: 한국전자통신연구원
CPC classification number: G06T7/2033 , A63B37/0022 , A63B69/3658 , G06T7/246
Abstract: 회전체의회전율, 회전축, 회전각과같은회전특성을측정하기위한방법및 장치가개시된다. 회전특성측정방법은, 회전체의영상으로부터회전체영역을추출하는단계, 추출한회전체영역의명도에기초하여상기회전체의표면패턴을추출하는단계및 추출한표면패턴의변화를기반으로상기회전체의회전특성을획득하는단계를포함한다. 따라서, 조명의조도, 카메라의감도및 카메라의노출시간에따른회전체표면영역의명도값 변화에관계없이회전체의회전율을정확하게측정할수 있다.
Abstract translation: 公开了一种用于测量旋转体的旋转速度,旋转轴线和旋转角度等旋转特性的方法和装置。 测量旋转特性的方法包括从旋转体的图像中提取旋转体区域; 基于所提取的旋转体区域的亮度值提取旋转体的表面图案,并且基于提取的表面图案的变化来获取旋转体的旋转特性。 因此,根据灯的照明,照相机的灵敏度和照相机的曝光时间,可以精确地测量旋转特性,而不管旋转体的表面区域的亮度值的变化如何。
-
公开(公告)号:KR1020160098543A
公开(公告)日:2016-08-19
申请号:KR1020150019206
申请日:2015-02-09
Applicant: 한국전자통신연구원
CPC classification number: H04N13/282 , H04N13/243 , H04N13/246 , H04N17/002 , H04N13/128 , G06T13/20 , H04N13/122 , H04N13/167
Abstract: 본발명은넓은공간에서다중영상입력장치의보정파라미터(Intrinsic/extrinsic parameter) 를별도의보정툴 없이간편하게자동으로얻을수 있는다중입력영상보정장치및 방법을제공한다.
Abstract translation: 本发明提供了一种用于校正多输入图像的装置和方法,其可以在没有附加校正工具的情况下,在宽的空间中简单且自动地获得多图像输入装置的校正参数(固有/外在参数)。 多输入图像校正装置包括:图像输入同步单元; 图像获取单元; 特征点检测单元; 相应的点组生成单元; 和校正参数计算单元。
-
公开(公告)号:KR1020160096473A
公开(公告)日:2016-08-16
申请号:KR1020150018189
申请日:2015-02-05
Applicant: 한국전자통신연구원
IPC: G06K9/00 , A63F13/213 , G06T13/40
CPC classification number: G06K9/00342 , G06T7/251 , G06T2207/10016 , G06T2207/30196 , G06T2207/30241 , A63F13/213 , G06T13/40
Abstract: 동작평가시스템및 방법이개시된다. 동작평가시스템은사용자를촬영하여사용자의관절데이터를획득하는복수의센서장치, 복수의센서장치로부터입력된관절데이터의캘리브레이션및 동기화를수행하여사용자의스켈레톤을생성하고, 신체부위별로산출하는자세벡터를이용하여사용자의자세와전문가의자세를비교하는분석장치및 스켈레톤을이용하여사용자동작에따라움직이는사용자캐릭터를전문가캐릭터와함께출력하고, 비교에대한결과를출력하는디스플레이장치를포함한다.
Abstract translation: 公开了一种用于评估可以直接比较用户与具有不同身体尺寸的其他专家的运动的运动的系统和方法,从而学习舞蹈,运动等的操作。 运动评估系统包括:多个传感器装置,用于通过拍摄用户来获取用户的联合数据; 一种分析装置,用于通过校准并同步由多个传感器装置输入的关节数据,并通过使用由每个身体部位计算出的姿势向量来比较用户的姿势与专家的姿势; 以及显示装置,用于以专家字符输出通过用户使用骨架的运动移动的用户角色,并输出比较结果。
-
公开(公告)号:KR1020150061549A
公开(公告)日:2015-06-04
申请号:KR1020140083916
申请日:2014-07-04
Applicant: 한국전자통신연구원
CPC classification number: G06T7/579 , G06K9/6289 , G06T7/251 , G06T7/55 , G06T2207/10016 , G06T2207/10028 , G06T2207/20221
Abstract: 본발명은사람의동작을획득하기위해사용되는동작분석시스템에서, 하이브리드카메라에서획득되는고해상도의영상데이터와저해상도의깊이데이터를이용하여피사체의동작을추적하는장치및 방법에관한것으로, 데이터수집부, 데이터결합부, 데이터분할부, 대응점추적부, 관절추적부를포함할수 있으며, 이를통해하이브리드카메라에서획득되는고해상도영상데이터와저해상도깊이데이터를결합하여피사체의동작을정밀하게추적하는데그 목적이있다.
Abstract translation: 本发明涉及一种用于在用于获得人的运动的运动分析系统中使用高分辨率图像数据和从混合照相机获取的低分辨率深度数据跟踪被摄体的运动的装置和方法。 该装置可以包括:数据收集单元; 数据组合单元; 数据分割单元; 相应的点跟踪单元; 和联合跟踪单元。 因此,本发明的目的是通过组合高分辨率图像数据和从混合照相机获得的低分辨率深度数据来精确地跟踪对象的运动。
-
公开(公告)号:KR1020130068613A
公开(公告)日:2013-06-26
申请号:KR1020110135907
申请日:2011-12-15
Applicant: 한국전자통신연구원
IPC: A63F13/573 , A63F13/816 , A63F13/812 , G06F9/455
CPC classification number: G06F17/5009 , A63F13/10 , A63F2300/643 , A63F2300/646 , A63F2300/8011 , G06F2217/16 , G06T13/20 , G06T2210/21 , A63F13/573 , A63F13/812 , A63F13/816 , G06F9/455
Abstract: PURPOSE: A method for simulating a collision of a rotating ball is provided to increase realism in a sport game using a ball by accurately simulating a collision trace of the ball and providing various trace changes of the ball in a real filed. CONSTITUTION: A collision of a rotating ball is detected and a speed after the collision is calculated by considering friction between the ball and a ground surface(S200). A central point of the ball is calculated and the rotation of the ball is stopped(S400). The collision of the ball is determined by checking a collision between bottom vectors of the ball and the ground surface in a current position. [Reference numerals] (AA,CC,EE,GG) No; (BB,DD,FF,HH) Yes; (S100) Flight; (S110,S310) Collision?; (S200) Bounce; (S210) Height?; (S300) Cloud; (S320) Distance? Speed?; (S400) Stop
Abstract translation: 目的:提供一种用于模拟旋转球的碰撞的方法,以通过精确地模拟球的碰撞轨迹并在真实场景中提供球的各种迹线变化来增加使用球的运动游戏中的现实性。 构成:通过考虑球与地面之间的摩擦力,可以检测出旋转球的碰撞,碰撞后的速度(S200)。 计算球的中心点并停止球的旋转(S400)。 通过在当前位置检查球的底部矢量与地面之间的碰撞来确定球的碰撞。 (附图标记)(AA,CC,EE,GG)否; (BB,DD,FF,HH)是; (S100)飞行; (S110,S310)碰撞? (S200)反弹; (S210)高度? (S300)云; (S320)距离? 速度?; (S400)停止
-
公开(公告)号:KR1020120138376A
公开(公告)日:2012-12-26
申请号:KR1020110057806
申请日:2011-06-15
Applicant: 한국전자통신연구원
CPC classification number: G01P3/38 , G06T7/248 , G06T7/262 , H04N5/2353
Abstract: PURPOSE: An automatic high speed moving object detection method utilizing a multi-exposure camera image is provided to automatically detect a moving object, which is a target of speed measurement, at high speed by automatically labeling mutually linked moving object pixels among moving object pixels at high speed. CONSTITUTION: The original version of a multi-exposure image data is obtained from a multi-exposure camera(S110). A multi-exposure image on a first grid is generated(S120). Probability distribution of the pixel values on the horizontal and vertical scan lines of the multi-exposure image on the first grid is calculated(S130). Pixels having a difference of more than constant multiple from the standard deviation are determined as moving object pixels, and generate a binary image(S160). A binary image on a second grid is generated(S170). A label image on the second grid is generated(S180). A label image of a moving object generated on the first grid is generated from the label image generated on the second grid(S200). A moving object is automatically detected from the multi-exposure image on the original grid at high speed(S210). [Reference numerals] (AA) Start; (BB) Finish; (S110) Obtaining multiple expose image data; (S120) Sampling multiple expose image (original grid -> first grid); (S130) Calculating horizontal and vertical probability distribution; (S140) Calculating horizontal and vertical accumulation distribution; (S150) Generating horizontal and vertical background image; (S160) Generating binary image; (S170) Sampling binary image (first grid -> second grid); (S180) Leveling connection element; (S190) Area filtering; (S200) Level image warping (second grid -> original grid); (S210) Detecting a movable object
Abstract translation: 目的:提供一种利用多曝光摄像机图像的自动高速移动物体检测方法,通过自动识别运动物体像素之间的相互关联的运动物体像素,以高速自动检测作为速度测量目标的运动物体 高速。 构成:从多曝光照相机获得多曝光图像数据的原始版本(S110)。 生成第一格栅上的多曝光图像(S120)。 计算第一格子上的多曝光图像的水平和垂直扫描线上的像素值的概率分布(S130)。 将与标准偏差大于常数倍的像素确定为移动物体像素,生成二值图像(S160)。 生成第二格子上的二进制图像(S170)。 生成第二格子上的标签图像(S180)。 从在第二格栅上生成的标签图像生成在第一格栅上生成的移动物体的标签图像(S200)。 从原始格栅上的多曝光图像以高速自动检测移动物体(S210)。 (附图标记)(AA)开始; (BB)完成; (S110)获取多个曝光图像数据; (S120)采样多曝光图像(原网格 - >第一格); (S130)计算水平和垂直概率分布; (S140)计算水平和垂直积累分布; (S150)生成水平和垂直背景图像; (S160)生成二进制图像; (S170)采样二值图像(第一格 - >第二格); (S180)调平连接元件; (S190)区域滤波; (S200)水平图像翘曲(第二格 - >原网格); (S210)检测可移动物体
-
公开(公告)号:KR1020110066610A
公开(公告)日:2011-06-17
申请号:KR1020090123342
申请日:2009-12-11
Applicant: 한국전자통신연구원
CPC classification number: G06T7/73 , G06T2207/10016
Abstract: PURPOSE: A pause search device of motion capture data and method thereof are provided to automatically find a frame with a similar pose by analyzing motion capture data for a character animation. CONSTITUTION: A motion which has incorrect character size in a motion data is detected. A scale adjusting unit(106) performs a size adjustment. An ICP algorithm performing unit(108) finds matching transformation matrix by using ICP algorithm. The ICP algorithm performing unit determines a pose with the most similar frame having the smallest value with an input pose.
Abstract translation: 目的:提供运动捕捉数据的暂停搜索装置及其方法,通过分析用于角色动画的运动捕获数据来自动找到具有相似姿势的帧。 构成:检测到运动数据中具有不正确字符尺寸的运动。 刻度调节单元(106)进行尺寸调整。 ICP算法执行单元(108)通过使用ICP算法找到匹配变换矩阵。 ICP算法执行单元使用具有输入姿势的具有最小值的最相似的帧来确定姿势。
-
公开(公告)号:KR1020080053124A
公开(公告)日:2008-06-12
申请号:KR1020060125123
申请日:2006-12-08
Applicant: 한국전자통신연구원
CPC classification number: H04W28/18 , H04B17/318 , H04W24/00 , H04W64/00 , H04W84/10 , H04W92/18 , H04W88/18
Abstract: An automatic network setting apparatus between adjacent wireless terminals and a method therefor are provided to measure proximity using signal intensity between wireless terminals and perform a network connection automatically between the wireless terminals adjacent to each other within a predetermined distance on the basis of the measured proximity, thereby automating complicated network connection settings only with an action of making a wireless terminal approach the other wireless terminal within the predetermined distance anywhere. A wireless terminal comprises a RF(Radio Frequency) unit(21), a control unit(22), a storage unit(23), and an input/output interface unit(24). The RF unit monitors the intensity of a wireless signal received from an adjacent wireless terminal periodically. The control unit controls the RF unit to calculate a distance corresponding to the intensity of the signal transferred from the RF unit using signal intensity or a distance correction constant in a current wireless environment and perform the automatic network connection of the wireless terminal when the calculated distance is less than a threshold distance. The storage unit stores the signal intensity or a threshold value for the distance for determining proximity with the other wireless terminal, the correction constant for the current wireless environment, and network setting parameters for automatic network settings.
Abstract translation: 提供相邻无线终端之间的自动网络设置装置及其方法,用于测量无线终端之间的信号强度的接近度,并且基于测量的接近度在预定距离内彼此相邻的无线终端之间自动执行网络连接, 从而仅通过使无线终端在任何地方的预定距离内接近另一无线终端的动作自动化复杂的网络连接设置。 无线终端包括RF(射频)单元(21),控制单元(22),存储单元(23)和输入/输出接口单元(24)。 RF单元周期性地监视从相邻无线终端接收的无线信号的强度。 控制单元控制RF单元,以使用当前无线环境中的信号强度或距离校正常数来计算与从RF单元传送的信号的强度相对应的距离,并且当计算出的距离 小于阈值距离。 存储单元存储用于确定与另一无线终端的接近度的距离的信号强度或阈值,当前无线环境的校正常数和用于自动网络设置的网络设置参数。
-
-
-
-
-
-
-
-
-