-
公开(公告)号：KR1020170082078A
公开(公告)日：2017-07-13
申请号：KR1020160001249
申请日：2016-01-05
Applicant: 한국전자통신연구원
Abstract: 신원인식장치및 방법이개시된다. 본발명의일실시예에따른신원인식장치는카메라를이용하여촬영되는영상내부에서사람의전신을추출하고, 상기전신과데이터베이스에존재하는의복정보를비교한결과를이용하여, 상기의복의인식을수행하는의복인식부; 상기영상에서추출된상기사람의얼굴과상기데이터베이스에존재하는얼굴정보를비교한결과를이용하여, 상기얼굴의인식을수행하는얼굴인식부; 및상기얼굴의인식결과에대응하는값의크기에기반하여, 상기의복의인식결과및 상기얼굴의인식결과중 어느하나이상을이용할것인지판단하고, 상기판단의결과에기반하여상기사람의신원을인식하는신원인식부를포함한다.
Abstract translation: 公开了一种用于识别身份的装置和方法。 根据本发明实施例的用于识别身份的设备从使用照相机拍摄的图像中提取人的身体，并且使用比较存在于数据库中的身体信息和服装信息的结果， 服装识别单元执行; 脸部识别单元，用于使用将从图像中提取的人物的脸部与存在于数据库中的脸部信息进行比较的结果来识别脸部; 和基于对应于所述衣服的脸部识别结果的识别结果的值的大小，并确定是否基于所述判断结果使用一个或多个脸部的识别结果，识别人的身份 等等。
-
公开(公告)号：KR1020150039252A
公开(公告)日：2015-04-10
申请号：KR1020130117373
申请日：2013-10-01
Applicant: 한국전자통신연구원
IPC: G06T7/20
CPC classification number: G06K9/00342 , G06K9/6269 , G06K2009/4666
Abstract: 본발명의행동인식기반의응용서비스제공장치는, 카메라로부터제공되는깊이영상을수집하는영상입력블록과, 수집된깊이영상으로부터인체를검출하는인체검출블록과, 상기인체로부터추출한 3차원액션볼륨과기 학습된행동모델에의거하여상기인체의행동을인식하는행동인식블록을포함할수 있다.
Abstract translation: 基于动作识别的应用服务提供装置的本发明包括：图像输入块，其收集由相机提供的深度图像; 人体检测块，其从所收集的深度图像中检测人体; 以及动作识别块，其根据从人体提取的3维动作体积识别人体的动作，以及预先获知的动作模型。
-
公开(公告)号：KR101484043B1
公开(公告)日：2015-01-19
申请号：KR1020100122186
申请日：2010-12-02
Applicant: 한국전자통신연구원
Abstract: 본 발명은 자동차 번호판의 문자열/숫자열에 기초하여 차량 식별자로 자동차 번호판을 인식하는 시스템 및 그 방법에 관한 것이다. 본 발명은 다양한 형태의 자동차 번호판에 대한 선행 학습을 통해 얻은 결과를 기초로 문자열/숫자열과 관련한 관심영역 후보군을 검출하고, 미리 정해진 5가지 조건 중 적어도 하나의 조건을 이용하여 검출된 관심영역 후보군을 검증하며, ROI 영역의 높이와 너비의 비를 고려해서 검증된 관심영역 후보군으로부터 선택된 ROI 영역 내의 번호판 MBR 영역을 결정하여 자동차 번호판을 인식한다. 본 발명에 따르면, 각 나라마다 정의된 다양한 형태의 번호판 규격과 관계없이 번호판의 위치를 자동으로 검출할 수 있다.
-
公开(公告)号：KR1020140137893A
公开(公告)日：2014-12-03
申请号：KR1020130059117
申请日：2013-05-24
Applicant: 한국전자통신연구원
IPC: G06T7/20
CPC classification number: G06K9/32 , G06K2009/3291 , G06T7/246 , G06T2207/10024 , G06T2207/10028 , G06T2207/30196 , G06T2207/30232
Abstract: 객체 추적 방법 및 장치가 제공되며, 영상 획득 장치에 의해 촬영된 영상에 대응하는 영상 프레임을 입력받는 단계, 영상 프레임으로부터, 추적 객체, 추적 객체와 깊이가 유사한 깊이 유사 장애물 및, 추적 객체와 형상이 유사한 유형 유사 장애물을 검출하는 단계, 검출된 추적 객체, 깊이 유사 장애물 및 유형 유사 장애물을 추적하는 단계, 검출된 추적 객체가 깊이 유사 장애물과 겹치는 경우, 추적 객체와 깊이 유사 장애물의 추적 점수 변화량을 비교하는 단계, 추적 객체의 추적 점수 변화량이 깊이 유사 장애물의 추적 점수보다 낮은 경우 추적 객체를 계속 추적하고, 추적 객체의 추적 점수 변화량이 깊이 유사 장애물의 추적 점수 변화량 보다 높은 경우, 다음 영상 프레임으로 진행하는 단계를 포함한다.
Abstract translation: 提供了一种跟踪对象的方法和装置。 该方法包括：从摄像装置接收与拍摄图像相对应的图像帧; 检测要跟踪的物体，具有与要跟踪的物体的深度类似的深度的类似深度障碍物以及类似于要跟踪物体的形状的形状障碍物; 跟踪检测到的要跟踪的物体，类似的深度障碍物和类似的形状障碍物; 比较当要跟踪的对象与相似的深度障碍物重叠时，要跟踪的对象的跟踪得分变化量和类似的深度障碍物; 并且当跟踪对象被跟踪的跟踪得分变化量小于相似深度障碍物的跟踪分数或进行到下一个图像帧时，跟踪待追踪的对象， 被跟踪的大于相似深度障碍物的跟踪分数。
-
公开(公告)号：KR1020140076964A
公开(公告)日：2014-06-23
申请号：KR1020120145578
申请日：2012-12-13
Applicant: 한국전자통신연구원
Abstract: A multiple intelligence test apparatus according to an embodiment of the present invention includes an image sensing device which receives an image for a multiple intelligence test from a user; a multiple intelligence measurement model unit which receives information of the image from the image sensing device and performs the multiple intelligence test in the way of selecting any one of a first reaction and a second reaction; and a content unit which receives an evaluated multiple intelligence result from the multiple intelligence measurement model unit and generates a personal portfolio based on the received result, wherein the multiple intelligence measurement model unit selects any one of the first and second reactions based on a reference reaction according to the emotion and behavior pattern of the user.
Abstract translation: 根据本发明的实施例的多智能测试装置包括从用户接收用于多智能测试的图像的图像感测装置; 多智能测量模型单元，其从所述图像感测装置接收图像的信息，并且以选择第一反应和第二反应中的任何一个的方式进行所述多重智能测试; 以及内容单元，其从多个智能测量模型单元接收评估的多个智能结果，并且基于接收结果生成个人投资组合，其中多个智能测量模型单元基于参考反应选择第一和第二反应中的任何一个 根据用户的情感和行为模式。
-
Patent Agency Ranking
公开(公告)号：KR1020140025814A
公开(公告)日：2014-03-05
申请号：KR1020120091970
申请日：2012-08-22
Applicant: 한국전자통신연구원
CPC classification number: G06F13/14 , B25J9/1656 , B25J13/02 , G06F9/451 , G06F17/2217 , G06F17/30023
Abstract: Service robot developers and application programmers have produced a service application in a manner of directly controlling hardware, and this causes limitations on the reusability of the service application and the expandability of a service robot market as wells as the overall change in the service application whenever the hardware is changed. According to an embodiment of the present invention, provided is an apparatus for a human-robot interaction service, which can be reused without wasting resources, such as time, money, and manpower, by integrating various sensors and hardware (actuators) in a service robot and maintaining compatibility such that service application developers can develop an application in unified standards. Furthermore, the present invention provides an apparatus for a human-robot interaction service, which can promote the reusability of components and improve the efficiency of service application and component development by proposing a unified interface between the components of a service robot technique and the application program, for the service robot market. [Reference numerals] (210) HRI engine 1; (212) Face recognition unit; (214) Wheel control unit; (220) HRI engine 2; (222) RFID tag detection unit; (224) Leg control unit; (AA) Service application 1; (BB) Service application 1'; (CC) Compatibility
Abstract translation: 服务机器人开发人员和应用程序员已经以直接控制硬件的方式生产了服务应用程序，这导致服务应用程序的可重用性和服务机器人市场的可扩展性的限制作为服务应用程序的整体变化 硬件更改。 根据本发明的实施例，提供了一种人机交互服务的装置，其能够通过将各种传感器和硬件（执行器）集成在服务中而不浪费资源（诸如时间，金钱和人力）而被重用 机器人和维护兼容性，使服务应用程序开发人员可以以统一标准开发应用程序。 此外，本发明提供了一种人机交互服务的装置，其可以通过提出服务机器人技术的组件与应用程序之间的统一接口来提高组件的可重用性并提高服务应用和组件开发的效率 ，为服务机器人市场。 （附图标记）（210）HRI发动机1; （212）面部识别单元; （214）车轮控制单元; （220）HRI发动机2; （222）RFID标签检测单元; （224）腿部控制单元; （AA）服务申请1; （BB）服务申请1'; （CC）兼容性
-
公开(公告)号：KR101081972B1
公开(公告)日：2011-11-09
申请号：KR1020080128695
申请日：2008-12-17
Applicant: 한국전자통신연구원
Abstract: 본발명은하이브리드특징벡터를이용하여화자인식의정확성을향상시키는방법에관한것으로서, 특징벡터 AS-MFCC(Autocorrelation Sequence - Mel-Frequency Cepstral Coefficient)와 RAS-MFCC(Relative AS-MFCC)를동시에사용하는멀티스트리밍방법과서브밴드특징벡터재결합방법을혼합하여이용하며, 본발명에의하면잡음환경에서정확하고신뢰도높은인식결과를제공할수 있다.
-
公开(公告)号：KR1020110003811A
公开(公告)日：2011-01-13
申请号：KR1020090061263
申请日：2009-07-06
Applicant: 한국전자통신연구원
Abstract: PURPOSE: An interacting robot is provided to improve a processing speed and process efficiency by generally controlling the reaction of the human body recognized by a recognition module. CONSTITUTION: An interacting robot comprises a recognition part(110), a service provider(120) and an interaction part(130). The recognition part figures out the person's intention by using sensing information about the human. The service provider offers the service corresponding to the intention and senses the reaction of the human body.
Abstract translation: 目的：提供一种相互作用的机器人，以通过一般地控制由识别模块识别的人体的反应来提高处理速度和处理效率。 构成：相互作用的机器人包括识别部分（110），服务提供者（120）和交互部分（130）。 识别部分通过使用关于人的感知信息来了解人的意图。 服务提供者提供对应于意图的服务，并感知人体的反应。
-
公开(公告)号：KR1020110003146A
公开(公告)日：2011-01-11
申请号：KR1020090060771
申请日：2009-07-03
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00355 , G06T7/20 , G06T7/00 , G06T7/40
Abstract: PURPOSE: A gesture recognition apparatus, and a robot system having the same are provided to recognize four gestures such as waving, calling, raising, stopping gesture for remote distance interactivity. CONSTITUTION: A human detection unit(120) detects the face region of a user from an input image. A gesture area setting unit(130) establishes a gesture domain based on the detected face region. An arm detection unit(150) detects arm region of the user within the gesture domain. A gesture decision unit(160) distinguishes a target gesture of the user within the gesture domain with the location of the arm domain, and the analysis of the type information and movement directivity.
Abstract translation: 目的：提供一种手势识别装置和具有该手势识别装置的机器人系统以识别诸如挥动，呼叫，提升，停止用于远距离交互的手势的四种手势。 构成：人体检测单元（120）从输入图像检测用户的脸部区域。 手势区域设置单元（130）基于检测到的脸部区域建立手势域。 手臂检测单元（150）检测手势域内的用户的手臂区域。 手势判定单元（160）将手势域内的用户的目标手势与手臂域的位置，以及类型信息和移动方向性的分析区分开。
-
公开(公告)号：KR100883519B1
公开(公告)日：2009-02-13
申请号：KR1020070086101
申请日：2007-08-27
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00275 , G06T1/0014 , G06T7/33 , G06T2207/20084
Abstract: A face recognition result analysis system of an image recognition robot and a method thereof are provided automatically to analyze a failure cause and inform the cause to a user when a face recognition failure of an image recognition robot occurs. A feature vector extraction unit(100) extracts a feature vector necessary for the discrimination of face recognition by using the motion information of an image obtained through an image recognition robot. A differential binary image conversion unit(102) finds out a differential image of two continued images, that is to say, differential value between two pixels of two images. A connection component analysis unit(104) carries out connected component analysis as to the binary image converted in the differential binary image conversion unit to detect a region which is changed spatially. A user region detection unit merges regions, which are adjacent to each other, among the regions detected in a connection component analyzer. A feature vector calculation unit(108) calculates feature amount vectors based on the user regions merged in the user region detection unit.
Abstract translation: 提供图像识别机器人的面部识别结果分析系统及其方法，当图像识别机器人出现脸部识别失败时，自动分析故障原因并向用户通知原因。 特征向量提取单元（100）通过使用通过图像识别机器人获得的图像的运动信息来提取用于识别面部识别所必需的特征向量。 差分二进制图像转换单元（102）找出两个连续图像的差分图像，也就是说，两个图像的两个像素之间的差分值。 连接分量分析单元（104）对在差分二进制图像转换单元中转换的二进制图像执行连接分量分析，以检测在空间上变化的区域。 用户区域检测单元将在连接分量分析器中检测到的区域中彼此相邻的区域合并。 特征矢量计算单元（108）基于在用户区域检测单元中合并的用户区域来计算特征量向量。
-
-
-
-
-
-
-
-
-
Search Results
130
records
(took 0.03 seconds)
