-
公开(公告)号：KR1020140143310A
公开(公告)日：2014-12-16
申请号：KR1020130131658
申请日：2013-10-31
Applicant: 삼성전자주식회사 , 임페리얼 이노베이션스 리미티드
CPC classification number: G06T7/20 , G06K9/00382 , G06K9/00389 , G06K9/6219 , G06K9/6256 , G06K9/6259 , G06K9/627 , G06K9/6282 , G06K2209/05 , G06T7/11 , G06T7/143 , G06T7/593 , G06T7/75 , G06T2207/10028 , G06T2207/20081
Abstract: A present invention relates to an estimator learning method and a pose estimation method using a depth image. The estimator learning method according to an embodiment of the present invention allows a user to learn an estimator for estimating the pose of an object based on correlation between synthesis data and actual data. The pose estimation method is to estimate the pose of the object by using the learned estimator.
Abstract translation: 本发明涉及使用深度图像的估计器学习方法和姿态估计方法。 根据本发明的实施例的估计器学习方法允许用户基于合成数据和实际数据之间的相关性来学习估计对象的姿态的估计器。 姿态估计方法是通过使用学习估计器来估计对象的姿态。
-
公开(公告)号：KR1020140056992A
公开(公告)日：2014-05-12
申请号：KR1020120123557
申请日：2012-11-02
Applicant: 삼성전자주식회사
CPC classification number: G06T7/285 , G06K9/6247 , G06T7/251 , G06T2207/10028
Abstract: Disclosed is a motion tracking method using a depth image. According to an embodiment, the present invention includes a step of obtaining continuous depth images of at least two frames; a step of calculating first parameters related to a depth flow between adjacent frames based on the depth images; a step of obtaining second parameters related to motion priors; and a step of estimating a pose change rate between the at least two continuous frames based on the first parameters and the second parameters.
Abstract translation: 公开了使用深度图像的运动跟踪方法。 根据实施例，本发明包括获得至少两帧的连续深度图像的步骤; 基于深度图像计算与相邻帧之间的深度流相关的第一参数的步骤; 获得与运动先验相关的第二参数的步骤; 以及基于所述第一参数和所述第二参数来估计所述至少两个连续帧之间的姿态变化率的步骤。
-
公开(公告)号：KR1020140010500A
公开(公告)日：2014-01-27
申请号：KR1020120076099
申请日：2012-07-12
Applicant: 삼성전자주식회사
CPC classification number: G06K9/00389 , G06F3/017 , G06K9/46
Abstract: A pose recognition apparatus and an operating method for the same are disclosed. According to an embodiment of the present invention, a pose recognition apparatus for recognizing the pose of a hand using a learned random forest comprises: an input unit for receiving an image; an extraction unit for extracting, from the image, a region of interest related to the hand; a computation unit for computing a center point for the region of interest; a generation unit for generating a feature vector based on a feature vector generator and the center point used when generating the learned random forest; and a processing unit for inputting the feature vector to the learned random forest, collecting the recognition result from each of a plurality of trees included in the learned random forest, and selecting, as the pose of the hand, a recognition result for which a majority votes among the collected recognition results. [Reference numerals] (100) Pose recognition apparatus; (120) Acquire the pose of a hand
Abstract translation: 公开了一种姿势识别装置及其操作方法。 根据本发明的实施例，使用学习随机森林识别手的姿态的姿势识别装置包括：用于接收图像的输入单元; 提取单元，用于从图像中提取与手相关的感兴趣区域; 用于计算感兴趣区域的中心点的计算单元; 用于生成基于特征向量生成器的特征向量的生成单元和生成学习随机森林时使用的中心点; 以及处理单元，用于将特征向量输入到所学习的随机森林中，从包括在所学习的随机森林中的多个树中的每一棵树中收集识别结果，并且选择作为手的姿势的多数的识别结果 在收集的认可结果中投票。 （附图标记）（100）姿态识别装置; （120）获取一只手的姿势
-
公开(公告)号：KR1020140009913A
公开(公告)日：2014-01-23
申请号：KR1020130017404
申请日：2013-02-19
Applicant: 삼성전자주식회사
CPC classification number: A63F13/65 , H04N13/20 , H04N21/42202 , H04N21/4223 , H04N21/4316 , H04N21/4725
Abstract: A virtual world processing apparatus and a method thereof are disclosed. According to embodiments, an interaction between a real world and a virtual world can be realized by delivering sensed information on the photographed image of the real world to the virtual world by using the characteristics of an image sensor which is information about the characteristics of the image sensor. [Reference numerals] (321) Reception unit; (322) Processing unit; (323) Transmission unit
Abstract translation: 公开了一种虚拟世界处理装置及其方法。 根据实施例，可以通过使用作为关于图像的特征的信息的图像传感器的特性将感知到的真实世界的拍摄图像上的信息传递到虚拟世界来实现真实世界和虚拟世界之间的交互 传感器。 （附图标记）（321）接收单元; （322）处理单元; （323）传输单元
-
公开(公告)号：KR1020130083997A
公开(公告)日：2013-07-24
申请号：KR1020120004653
申请日：2012-01-16
Applicant: 삼성전자주식회사
CPC classification number: G06T17/00 , G06K9/00369 , G06K9/6272
Abstract: PURPOSE: A body posture group formation apparatus and a method thereof use imbalance measure to make the number of postures uniform. CONSTITUTION: A data receiver receives at least one posture data. A characteristic extractor (120) extracts a characteristic vector. A cluster classifier classifies similar postures into each cluster. A cluster controller (140) makes the number of postures uniform. Each cluster produces a posture group. [Reference numerals] (110) Data receiver; (120) Characteristic extractor; (130) Cluster classifier; (140) Cluster controller; (150) Posture group generation unit
Abstract translation: 目的：一种身体姿势组形成装置及其使用方法，使姿势数量均匀。 构成：数据接收器接收至少一个姿势数据。 特征提取器（120）提取特征向量。 集群分类器将类似的姿势分类到每个集群中。 集群控制器（140）使得姿势数量一致。 每个集群都会生成一个姿势组。 （附图标记）（110）数据接收机; （120）特征提取器; （130）集群分类器; （140）集群控制器; （150）姿势组生成单元
-
Patent Agency Ranking
公开(公告)号：KR1020120132337A
公开(公告)日：2012-12-05
申请号：KR1020120047215
申请日：2012-05-04
Applicant: 삼성전자주식회사
Abstract: PURPOSE: An UI(user interface) control apparatus and method thereof are provided to clearly separate a start point of voice recognition from an end point of the voice recognition by determining the start and the end point and by detecting a location of a hand from an image. CONSTITUTION: A reception unit(110) receives an image for a user from a sensor. A detection unit(120) detects a location of a hand and a face from the image. A processing unit(130) calculates a difference between the location of the hand and the face. When the difference is under a threshold value, a control unit(140) controls an UI(user interface) based on the voice recognition by recognizing the voice from the user. The detection unit detects the posture of the hand from the image. When the calculated difference is under the threshold value and the posture of the hand is matched with a start posture, the control unit recognizes the voice. [Reference numerals] (110) Reception unit; (120) Detection unit; (130) Processing unit; (140) Control unit
Abstract translation: 目的：提供一种用户界面（用户界面）控制装置及其方法，用于通过确定开始和结束点来清楚地将语音识别的起始点与语音识别的终点分开，并通过从 图片。 构成：接收单元（110）从传感器接收用户的图像。 检测单元（120）从图像中检测手和脸部的位置。 处理单元（130）计算手和脸部的位置之间的差异。 当差值低于阈值时，控制单元（140）通过识别来自用户的声音来基于语音识别来控制UI（用户界面）。 检测单元从图像检测手的姿势。 当计算出的差值低于阈值并且手的姿势与起始姿势相匹配时，控制单元识别出声音。 （附图标记）（110）接收单元; （120）检测单元; （130）处理单元; （140）控制单元
-
公开(公告)号：KR1020120020045A
公开(公告)日：2012-03-07
申请号：KR1020110046174
申请日：2011-05-17
Applicant: 삼성전자주식회사
IPC: G06F3/03 , G06F3/0346 , G06F3/0481 , G06F3/14
Abstract: PURPOSE: An interface control apparatus and method thereof are provided to offer an UI(User Interface) including a high unity sense by controlling a two and a three dimensional UI object based on information for hand movements and gaze. CONSTITUTION: A reception unit(201) receives image information including a depth image related to a user from a sensor. A processing unit(202) creates gaze information for gaze and motion information for hand movements based on the image information. A control unit(203) controls a two dimensional image or a two dimensional GUI(Graphic User Interface) based on the motion and the gaze information. The processing unit creates the motion information by calculating ROI(Region Of Interest) based on the image information. The motion information includes the information for the hand movements.
Abstract translation: 目的：提供一种接口控制装置及其方法，用于通过基于用于手部移动和注视的信息来控制两个和三维UI对象来提供包括高统一感的UI（用户界面）。 构成：接收单元（201）从传感器接收包括与用户相关的深度图像的图像信息。 处理单元（202）基于图像信息创建用于手部移动的注视和运动信息的注视信息。 控制单元（203）基于运动和注视信息来控制二维图像或二维GUI（图形用户界面）。 处理单元通过基于图像信息计算ROI（感兴趣区域）来创建运动信息。 运动信息包括用于手动的信息。
-
公开(公告)号：KR1020110114979A
公开(公告)日：2011-10-20
申请号：KR1020100034399
申请日：2010-04-14
Applicant: 삼성전자주식회사
IPC: G06F3/01 , G06N3/00 , G06T19/00 , G06F3/0487 , H04N1/00
CPC classification number: G06F3/016 , G06T19/006 , A63F2300/6045 , G06F3/017 , G06F3/0487 , G06N3/006 , G06T2215/16 , H04N1/00381
Abstract: 가상 세계 처리 장치 및 방법이 개시된다. 본 발명의 실시예들에 따르면 센싱한 감지 정보에 대응하는 가상 세계의 가상 객체에 대한 촉각 정보를 추출하여 촉각 피드백 장치로 전송함으로써, 현실 세계와 가상 세계의 상호 작용을 구현할 수 있고, 또한, 가상 객체의 유형에 따라 데이터 구조를 달리함으로써, 가상 객체에 대한 촉각 정보의 처리 속도를 향상 시킬 수 있다.
Abstract translation: 公开了一种虚拟世界处理设备和方法。 通过根据本发明的触觉提取在虚拟世界中的虚拟对象的触觉信息，对应于发送到反馈系统所感测的检测信息的实施例，能够实现在现实世界和虚拟世界和虚拟之间的相互作用 通过根据对象的类型改变数据结构，可以提高用于虚拟对象的触觉信息的处理速度。
-
公开(公告)号：KR1020110114793A
公开(公告)日：2011-10-20
申请号：KR1020100034125
申请日：2010-04-14
Applicant: 삼성전자주식회사
CPC classification number: G06F3/0487 , A63F13/21 , A63F13/211 , A63F13/235 , A63F13/327 , A63F13/42 , A63F2300/6045 , G06N3/006 , G06T15/00 , G06T2215/16 , G06F3/017 , G06K9/00335 , G06T19/006 , H04N1/00381
Abstract: 가상 세계 처리 장치 및 방법이 개시된다. 실시예들은 센서의 특성에 관한 정보인 센서 특성을 이용하여 현실 세계로부터 측정한 정보를 가상 세계로 전달하여 현실 세계와 가상 세계의 상호 작용을 구현할 수 있다. 상기 측정된 정보 중 이전에 측정된 정보와 상이한 것만을 선택적으로 전달하는 가상 세계 처리 장치 및 방법이 개시된다. 상기 측정된 정보가 이전에 측정된 정보와 많은 차이가 있을 경우 상기 측정된 정보 전체를 전달하고, 차이가 많지 않을 경우 상기 측정된 정보 중 상기 이전에 측정된 정보와 상이한 것만을 선택적으로 전달하는 가상 세계 처리 장치 및 방법이 개시된다.
Abstract translation: 公开了一种用于处理虚拟世界的设备和方法。 根据本发明的实施例，使用传感器的特性从真实世界测量的信息被传送到虚拟世界，从而实现现实世界和虚拟世界之间的交互。 所公开的用于处理虚拟世界的设备和方法涉及从测量信息中选择性地传送与先前测量的信息不同的信息。 所公开的用于处理虚拟世界的设备和方法涉及在测量信息与先前测量的信息显着不同的情况下传送整个测量信息，并且用于从测量信息中选择性地传送与测量信息不同的信息， 事件的差异不显着。
-
公开(公告)号：KR1020110103571A
公开(公告)日：2011-09-21
申请号：KR1020100022696
申请日：2010-03-15
Applicant: 삼성전자주식회사
CPC classification number: H04N5/232 , G01C3/08 , G01S11/12 , G06T5/003 , G06T7/529 , H04N5/225 , G01C3/02
Abstract: 디스플레이 패널의 각 픽셀들의 개구부를 통해 촬영하는 촬영 장치 및 방법이 제공된다. 일 측면에 촬영 장치는 카메라부를 디스플레이 패널 뒤에 위치시키고 디스플레이 패널의 각 픽셀들의 개구부를 통해 촬영하도록 해서 카메라부의 위치와 디스플레이부의 위치를 일치시키고, 디스플레이 패널의 개구부를 통해 촬영할 때 발생하는 Blur 효과를 이용해서 물체와의 거리를 측정하고 Blur 효과가 반영된 영상을 복원한다.
Abstract translation: 提供了一种通过显示面板的每个像素的至少一个孔径拍摄对象的成像设备，方法和计算机可读介质。 成像装置包括位于显示面板背面的相机单元，并通过显示面板的每个像素的至少一个孔眼拍摄该物体，以使相机单元的位置与显示器的位置相同 使用在通过显示面板的至少一个孔径拍摄时发生的模糊效果来测量到物体的距离，并增强包括模糊效果的图像。
-
-
-
-
-
-
-
-
-
Search Results
197
records
(took 0.037 seconds)
