-
公开(公告)号：KR1020120112293A
公开(公告)日：2012-10-11
申请号：KR1020120092705
申请日：2012-08-24
Applicant: 전자부품연구원
Abstract: PURPOSE: An apparatus and a method for judging positions of a moving body are provided to reduce image storage capacity and computational load for comparing images because images comparing with photographed images are selected by a predetermined standard. CONSTITUTION: An apparatus for judging positions of a moving body is as follows. A position judging device stores images after stratifying multiple reference images according to the position of an object(S14). The reference images with similarity over a predetermined value are grouped in a similar image and the rest images are classified into an identifying images(S15). The position judging device compares photographed images and the reference images to sort a matching reference image with the similarity over the predetermined value(S16). The position judging device reflects weight according as the matching reference images belongs to the similar image or the identifying image to calculate the probability that moving body exists at the position of the object corresponding to the matching reference image. The position judging device judges the position of the moving body according to the calculated results(S17,S18,S19,S20). [Reference numerals] (S11) Taking a photograph; (S12) Extracting feature points; (S13) Calculating a number of the feature points; (S14) Selecting and storing reference images; (S15) Grouping similar images and classifying identifying images; (S16) Sorting matching reference images corresponding to photographed images; (S17) Similar images?; (S18) Calculating position probability with reflecting the weighted value of the similar images; (S19) Calculating the position probability with reflecting the weighted value of the identifying images; (S20) Determining a position
Abstract translation: 目的：提供一种用于判断移动体的位置的装置和方法，以减少用于比较图像的图像存储容量和计算负荷，因为与拍摄图像进行比较的图像是以预定标准选择的。 构成：用于判断移动体的位置的装置如下。 位置判断装置根据物体的位置对多个参考图像进行分层后存储图像（S14）。 具有预定值的相似度的参考图像被分组在类似的图像中，并且其余的图像被分类为识别图像（S15）。 位置判断装置将拍摄图像和参考图像进行比较，对具有预定值的相似度的匹配参考图像进行排序（S16）。 位置判断装置根据匹配的参考图像属于相似图像或识别图像来反映权重，以计算移动体存在于与匹配参考图像相对应的对象的位置处的概率。 位置判断装置根据计算结果来判断移动体的位置（S17，S18，S19，S20）。 （附图标记）（S11）拍摄照片; （S12）提取特征点; （S13）计算特征点的数量; （S14）选择并存储参考图像; （S15）对相似图像进行分组并对识别图像进行分类; （S16）对与拍摄图像对应的匹配参考图像进行排序; （S17）相似图片 （S18）反映相似图像的加权值计算位置概率; （S19）通过反映识别图像的加权值来计算位置概率; （S20）确定位置
-
公开(公告)号：KR1020120104044A
公开(公告)日：2012-09-20
申请号：KR1020110022115
申请日：2011-03-11
Applicant: 전자부품연구원
Abstract: PURPOSE: A rope cart for a gondola robot system and a method for estimating the movement of a rope thereof are provided to exactly estimate the movement of a rope by analyzing an image of the rope. CONSTITUTION: A method for estimating the movement of a rope of a rope cart(10) for a gondola robot system(100) is as follows. The rope cart obtains the shape information of the rope(18). The rope cart monitors the movement of the rope through a camera unit(17). The rope cart films the rope when the rope moves while monitoring. The rope cart estimates a movement distance and speed based on a captured image.
Abstract translation: 目的：提供一种用于吊车机器人系统的绳车和用于估计其绳索的运动的方法，以通过分析绳索的图像来精确估计绳索的运动。 构成：用于估计吊篮机器人系统（100）的绳车（10）的绳索的运动的方法如下。 绳索车获得绳索（18）的形状信息。 绳车通过相机单元（17）监测绳索的运动。 当绳子在监控时移动时，绳索车将电缆绳绳。 绳子车推测出基于拍摄图像的移动距离和速度。
-
公开(公告)号：KR101170594B1
公开(公告)日：2012-08-02
申请号：KR1020110023596
申请日：2011-03-16
Abstract: PURPOSE: An apparatus for measuring a force of a minimally invasive surgery tool and a system including the same are provided to measure a force generated in the end of the minimally invasive surgery tool by mounting a detachable insulation sensor device between a trocar and a single path guide. CONSTITUTION: A trocar(100) includes a cannula(150), a trocar housing(120), and a housing hole(140). The cannula is inserted into a body via the skin of a cut part. A flange(220) is extended from the end of a body unit(210) to the outside of the body unit. A pressure sensor(230) generates pressure sensor information corresponding force measurement information. A signal line transmits the pressure sensor information to an external computer.
Abstract translation: 目的：提供一种用于测量微创外科手术工具的力的装置和包括该微创外科手术工具的系统的装置，以通过在套管针和单个路径之间安装可拆卸的绝缘传感器装置来测量在微创外科手术工具的端部中产生的力 指南。 构成：套管针（100）包括套管（150），套管针壳体（120）和容纳孔（140）。 插管通过切割部分的皮肤插入体内。 凸缘（220）从主体单元（210）的端部延伸到主体单元的外部。 压力传感器（230）产生对应的力测量信息的压力传感器信息。 信号线将压力传感器信息传送到外部计算机。
-
公开(公告)号：KR1020120033414A
公开(公告)日：2012-04-09
申请号：KR1020100094914
申请日：2010-09-30
Applicant: 전자부품연구원
Abstract: PURPOSE: A robot with a self-location recognizing function and an operating method thereof are provided to accurately perform various operations established on a robot based accurate self-location recognition by performing the self-location recognition of the robot without errors. CONSTITUTION: A robot(100) with a self-location recognizing function comprises a storage unit(150), a camera(130), and a control unit(160). The storage unit stores a topological map and front and ceiling reference images registered for each spot of the topological map. The camera takes ceiling images and front images at the time of the self-location recognition. The control unit compares the ceiling based images with the ceiling images to detect at least one ceiling based image matched the ceiling image.
Abstract translation: 目的：提供具有自身位置识别功能的机器人及其操作方法，以通过无错误地执行机器人的自位置识别来精确地执行基于机器人的精确自定位识别所建立的各种操作。 构成：具有自位置识别功能的机器人（100）包括存储单元（150），相机（130）和控制单元（160）。 存储单元存储为拓扑图的每个点注册的拓扑图和前后参考图像。 摄像机在自定位识别时拍摄天花板图像和正面图像。 控制单元将基于天花板的图像与天花板图像进行比较，以检测与天花板图像匹配的至少一个基于天花板的图像。
-
公开(公告)号：KR1020120004132A
公开(公告)日：2012-01-12
申请号：KR1020100064826
申请日：2010-07-06
Applicant: 전자부품연구원
CPC classification number: E02F9/205 , B60Y2200/412 , E02F9/24 , E02F9/26 , G08C17/02
Abstract: PURPOSE: An excavator automation system and an integral management method thereof are provided to collectively manage a drilling work plan and an excavator when controlling the excavator according to the drilling work plan from a distance. CONSTITUTION: An excavator automation system comprises a remote controlled excavator(100), a remote control station(200), a TPS(Task Planning System) server(300), and a 3D modeling system(400). The remote control excavator includes a communication modem, implements excavation according to commands inputted through a communication link, senses work environments, and transmits the sensing information through the communication link. The remote control station generates a command for controlling the remote controlled excavator according to an excavation work plan produced by the TPS server and transmits the command through the communication link to the remote controlled excavator. The TPS server produces the excavation work plan based on 3D digital terrain information of the work area. The 3D modeling system obtains 3D digital terrain information of the whole work area.
Abstract translation: 目的：提供一种挖掘机自动化系统及其综合管理方法，以便在距离范围内根据钻井作业计划对挖掘机进行控制时，共同管理钻井工作计划和挖掘机。 构成：挖掘机自动化系统包括遥控挖掘机（100），远程控制站（200），TPS（任务计划系统）服务器（300）和3D建模系统（400）。 遥控挖掘机包括通信调制解调器，根据通过通信链路输入的命令实现挖掘，感测工作环境，并通过通信链路发送感测信息。 远程控制站根据由TPS服务器生成的挖掘工作计划生成用于控制遥控挖掘机的命令，并且通过通信链路将命令发送到遥控挖掘机。 TPS服务器根据工作区域的3D数字地形信息生成挖掘工作计划。 3D建模系统获得整个工作区域的3D数字地形信息。
-
Patent Agency Ranking
公开(公告)号：KR101088883B1
公开(公告)日：2011-12-07
申请号：KR1020090132255
申请日：2009-12-28
Applicant: 전자부품연구원
IPC: G06K9/00
Abstract: 본 발명은 이종 센서 융합 기반 사용자 인식 방법 및 장치에 관한 것로, 이러한 본 발명은 적어도 2개의 서로 다른 모달리티 센서를 통해 인식 대상의 모달리티를 센싱하는 모달리티 센서부; 및 상기 모달리티 센서가 감지한 모달리티와 미리 생성한 모달리티의 불확실성 모델의 기 설정한 신뢰도 범위 내에 속하는지 여부에 따라 상기 인식 대상을 인증하는 식별부;를 포함하는 사용자 인식 장치와 이에 따른 인식 방법을 제공한다.
화상, 열화상, 확률 분포-
公开(公告)号：KR101087058B1
公开(公告)日：2011-11-28
申请号：KR1020090132256
申请日：2009-12-28
Applicant: 전자부품연구원
Abstract: 본 발명은 전방향 카메라 장치 및 이의 구동 방법에 관한 것으로, 이러한 본 발명은, 영상 촬영을 위한 카메라 모듈을 구비한 복수의 카메라 장치부; 상기 복수의 카메라 장치 각각의 카메라 모듈이 360도 전방향을 구분하여 바깥으로 대향하도록 설치되는 받침부; 및 상기 각각의 카메라 모듈이 촬영한 화상 신호를 합성하여 전방향의 영상을 출력하는 화상 처리부;를 포함하는 것을 특징으로 하는 전방향 카메라 장치 및 이의 구동 방법을 제공한다.
로봇, 카메라, 해상도, 전방향, 해상도-
公开(公告)号：KR1020110075730A
公开(公告)日：2011-07-06
申请号：KR1020090132256
申请日：2009-12-28
Applicant: 전자부품연구원
CPC classification number: H04N7/181 , H04N5/2251 , H04N5/23206 , H04N5/23238 , H04N5/23296
Abstract: PURPOSE: An omni-directional camera apparatus and driving method thereof are provided to increase processing speed by photographing an object after partitioning an omni-direction. CONSTITUTION: A plurality of camera device units(100) includes a camera module(110) for image photographing. A plurality of camera device units is installed in a support unit(200). An image processing unit(300) synthesizes an image signal of each camera module. The image processor outputs an omniidirectional image.
Abstract translation: 目的：提供一种全方位相机装置及其驱动方法，通过在分割全方向后拍摄对象来提高处理速度。 构成：多个相机装置单元（100）包括用于图像拍摄的相机模块（110）。 多个照相机装置单元安装在支撑单元（200）中。 图像处理单元（300）合成每个相机模块的图像信号。 图像处理器输出全向图像。
-
公开(公告)号：KR1020110071885A
公开(公告)日：2011-06-29
申请号：KR1020090128577
申请日：2009-12-22
Applicant: 전자부품연구원
Abstract: PURPOSE: Sensor signal processing system and method for a robot are provided to easily upgrade the sensor processing ability of the robot since a robot controller only remains and only a sensor related device is replaced. CONSTITUTION: A robot system(1) comprises a robot controller(200), a sensor group(10) and a sensor signal processing system(100). The robot controller controls the movement of a robot mechanism. The sensor group comprises one or more sensors. The one or more sensors are arranged on at least one of the robot mechanism and a robot mechanism environment. The sensor signal processing system is arranged between the robot controller and the sensor group. After the sensor signal processing system, in common, processes or does not process the sensor signals of sensors of the sensor group, it transfer signals to the robot controller.
Abstract translation: 目的：提供用于机器人的传感器信号处理系统和方法，以便轻松升级机器人的传感器处理能力，因为仅保留机器人控制器，仅替换传感器相关设备。 构成：机器人系统（1）包括机器人控制器（200），传感器组（10）和传感器信号处理系统（100）。 机器人控制器控制机器人机构的运动。 传感器组包括一个或多个传感器。 一个或多个传感器布置在机器人机构和机器人机构环境中的至少一个上。 传感器信号处理系统设置在机器人控制器和传感器组之间。 在传感器信号处理系统的共同之处，处理或不处理传感器组的传感器的传感器信号，它将信号传送到机器人控制器。
-
公开(公告)号：KR1020090050567A
公开(公告)日：2009-05-20
申请号：KR1020070117082
申请日：2007-11-16
Applicant: 전자부품연구원
IPC: H04N13/00
CPC classification number: G06T5/002 , G06T7/13 , G06T17/00 , G06T2207/20024 , H03K19/20 , H04N13/204
Abstract: 본 발명은 TOF(Time of Flight) 카메라를 이용한 로봇의 물체 자세 인식방법에 관한 것으로서, TOF 카메라를 이용하여 인식 대상 물체의 거리 정보 이미지를 획득하고, 일반 영상 카메라를 통해 촬영한 영상 이미지에서 에지(Edge) 이미지를 검출한 후, 거리 정보 이미지와 에지 이미지를 AND 연산하여 인식 대상 물체의 실효성 있는 특징점(Constrained Feature)들로 이루어진 3차원 점군 데이터(3D Point Cloud)를 획득하고, 이를 데이터 베이스에 저장된 인식 대상 물체 전체의 3차원 점군 데이터와 정합함으로써, 인식 대상 물체의 자세를 인식하는 것을 특징으로 한다.
본 발명에 의하면, 유리나 투명 아크릴판과 같은 비 시인성 물체들의 경우에도 양질의 3차원 점군 데이터를 얻을 수 있고, 실효성 있는 특징점(Constrained Feature)들로 이루어진 3차원 점군 데이터(3D Point Cloud)를 사용함으로써, ICP(Iterative Closest Point) 알고리즘 수행시 작업 시간을 획기적으로 줄일 수 있다.
로봇, 물체 자세 인식, TOF 카메라, 3D Point Cloud, 특징점
-
-
-
-
-
-
-
-
-
Search Results
162
records
(took 0.057 seconds)
