-
公开(公告)号:KR1020100020069A
公开(公告)日:2010-02-22
申请号:KR1020080078714
申请日:2008-08-12
Applicant: 전자부품연구원
IPC: H04N5/262
CPC classification number: G06K9/00255 , H04N5/225 , H04N5/335
Abstract: PURPOSE: A method for recognizing head pose using a thermal camera is provided to recognize the head pose regardless of the change in an adjacent region by recognizing the head pose through a thermal image near a face. CONSTITUTION: A thermal image camera(100) photographs the heat of a human body, and a face divider(120) divides the thermal image of the face in the terminal image. A thermal region extraction unit(130) extracts a thermal region, and a weight value extraction unit(140) extracts the coordinates of a center in the thermal. An end coordinate detector(150) detects X and Y axes of pixels in the horizontal and vertical ends of the thermal image. A head pose judgment unit(160) judges the head pose through the X and Y axes and the coordinates of the center.
Abstract translation: 目的:提供一种使用热像仪识别头部姿势的方法,通过通过靠近脸部的热图像识别头部姿势来识别头部姿势,而不管相邻区域的变化如何。 构成:热像摄像机(100)拍摄人体的热量,脸部分割器(120)将终端图像中的脸部的热图像分割开。 热区域提取单元(130)提取热区域,并且权重值提取单元(140)提取热中心的坐标。 端部坐标检测器(150)检测热图像的水平和垂直端中的像素的X轴和Y轴。 头姿势判断单元(160)通过X轴和Y轴以及中心的坐标来判断头姿势。
-
公开(公告)号:KR100930626B1
公开(公告)日:2009-12-09
申请号:KR1020070084856
申请日:2007-08-23
Applicant: 전자부품연구원
IPC: G05D1/02
Abstract: 본 발명은 스테레오 카메라를 구비한 로봇의 물체 자세 인식 방법에 관한 것으로서, 인식 대상 물체의 실효성 있는 특징점(Constrained Feature)들로 이루어지는 3차원 점군 데이터(3D Point Cloud)를 획득하여 데이터 베이스에 저장하고, 스테레오 카메라로 상기 인식 대상 물체를 촬영한 영상의 실효성 있는 특징점(Constrained Feature)들로 이루어지는 3차원 점군 데이터를 획득한 후, 두 개의 3차원 점군 데이터를 정합하여 촬영한 인식 대상 물체의 자세를 인식하는 것을 특징으로 한다.
본 발명에 의하면, ICP(Iterative Closest Point) 알고리즘을 사용하여 두 영상을 정합함으로써, 물체를 바라보는 위치에 관계없이 인식이 가능하고 또한 물체가 부분적으로 가려지더라도 정확한 포즈 인식이 가능하며, 동일한 물체에 대한 두 개의 영상에서 동일한 특징을 찾아 정합함으로써, ICP(Iterative Closest Point) 알고리즘 수행시 작업 시간을 줄일 수 있다
물체 인식, 로봇, 스테레오 카메라, ICP 알고리즘, 영상 정합, 포즈Abstract translation: 提供一种包含立体相机的机器人的物体姿态识别方法,以减少3D点云数据,并通过在相同物体的两个图像中搜索相同特征并匹配相同特征来处理ICP算法时减少操作时间。 包括立体相机的机器人的物体姿态识别方法如下。 获取识别对象整体的3D点云数据(3D Point Cloud)并将其存储在数据库(S100)。 3D点云数据是通过使用立体照相机从识别对象拍摄照片之后拍摄的照片的图像获得的(S110)。 用拍摄照片的图像获取3D点群组数据的步骤。 通过使存储在数据库中的对象整体的3D点群组数据与拍摄照片的图像的3D点群组数据匹配来识别拍摄照片的识别对象素材的姿态(S120)。
-
公开(公告)号:KR100894423B1
公开(公告)日:2009-04-22
申请号:KR1020070120026
申请日:2007-11-23
Applicant: 전자부품연구원
Abstract: A mobile platform is provided to move a device of specific purpose efficiently while preventing the distortion of the wheel axle by lessening the vertical impact acting on the wheel and the circumferential impact on the axle. A mobile platform(100) comprises a base plate(110), a pair of left and right drive wheel modules(140,150) installed on the base plate opposite to each other, a control drive(145) which is mounted on the base plate and controls the left and right drive wheel modules, a drive wheel module buffer(160) buffering the external force acting on the left and right drive wheel modules, a front wheel caster(170) which is pivotally installed in the base plate tip on the center line of the left and right drive wheel modules via a caster supporting plate(171), and a mounting plate(120) which is supported above the top of the base plate with a plurality of profiles(130). The drive wheel module buffer includes a plurality of linear bushes(161) installed on the left and right drive wheel modules, a linear bush shaft(165), and a pressure spring(167) inserted into the linear bush shaft.
Abstract translation: 提供移动平台以有效地移动具有特定目的的装置,同时通过减小作用在车轮上的垂直冲击和对车轴的周向冲击来防止轮轴的变形。 移动平台(100)包括:基板(110),安装在彼此相对的基板上的一对左右驱动轮模块(140,150);安装在基板上的控制驱动器(145) 控制左右驱动轮模块,缓冲作用在左右驱动轮模块上的外力的驱动轮模块缓冲器(160),前轮脚轮(170),该前轮脚轮可枢转地安装在中心的底板尖端 通过脚轮支撑板(171)的左右驱动轮模块的线路,以及安装板(120),其被支撑在具有多个轮廓(130)的基板的顶部上方。 驱动轮模块缓冲器包括安装在左驱动轮模块和右驱动轮模块上的多个线性衬套(161),线性衬套轴(165)和插入到线性衬套轴中的压力弹簧(167)。
-
公开(公告)号:KR100882139B1
公开(公告)日:2009-02-06
申请号:KR1020070092417
申请日:2007-09-12
CPC classification number: G06K9/4604 , G06K9/00221 , G06T5/20 , G06T7/12 , G06T7/155 , G06T7/66 , G06T7/90 , G06T2207/20036 , G06T2207/30201
Abstract: A method for detecting a face at improved speed is provided to improve the speed for detecting the face in a captured image by finding central coordinates of a leg part from the captured image, finding the centroid of a detected skin color area, and detecting a face area having the centroid having the closest value to the central coordinates of the leg part. An edge image is generated by extracting an edge from a captured image(S110). An outline of a target person is extracted by removing background from the edge image based on distance information to the target person(S120). Skin color areas of the target person are extracted from the captured image(S130). An area is labeled after noises are removed by using a morphology filter(S140). Coordinates are found at a central position of both legs after a lag part is extracted from the image extracting the outline(S150). A face area is detected by finding centroid of the skin color areas and comparing the centroid with the coordinates corresponding to the middle point of both legs(S160).
Abstract translation: 提供一种用于以改善的速度检测面部的方法,以通过从所捕获的图像中找到腿部的中心坐标,找到检测到的皮肤颜色区域的重心并且检测面部来提高用于检测拍摄图像中的脸部的速度 具有与腿部的中心坐标最接近的重心的区域。 通过从拍摄图像中提取边缘来生成边缘图像(S110)。 通过基于与目标人物的距离信息从边缘图像中移除背景来提取目标人物的轮廓(S120)。 从拍摄图像中提取目标人物的皮肤颜色区域(S130)。 通过使用形态滤波器去除噪声后,标记区域(S140)。 在从提取轮廓的图像中提取滞后部分之后,在两条腿的中心位置处找到坐标(S150)。 通过找到皮肤颜色区域的质心并将质心与对应于两条腿的中点的坐标进行比较来检测面部区域(S160)。
-
公开(公告)号:KR100792852B1
公开(公告)日:2008-01-14
申请号:KR1020060104191
申请日:2006-10-25
Applicant: 전자부품연구원
Abstract: A method for extracting a distance of a landmark of a mobile robot is provided to reduce loads of a main processor by avoiding use of an algorithm for processing a signal. A method for extracting a distance of a landmark of a mobile robot, includes a step of calculating the horizontal distance(D1) between the point where the landmark is projected into an image plane of a camera of the mobile robot to the perpendicular bisector of the image plane, a step of moving the robot along a focal axis of the camera from the position to a target point, and photographing the landmark, a step of calculating the horizontal distance(D2) between the point where the landmark photographed at the target point is projected into an image plane of the camera to the perpendicular bisector of the image plane, and a step of calculating the distance(D) between the camera of the robot at the target point to the point where the focal axis of the camera and the plane including the landmark form a perpendicularity by using the distances(D1,D2).
Abstract translation: 提供了一种用于提取移动机器人的地标的距离的方法,以通过避免使用用于处理信号的算法来减少主处理器的负载。 一种用于提取移动机器人的地标的距离的方法包括以下步骤:计算在所述地标被投影到所述移动机器人的照相机的像平面中的点与所述移动机器人的垂直平分线之间的水平距离(D1) 图像平面,使所述机器人沿所述摄像机的焦点从所述位置移动到目标点的步骤,以及拍摄所述地标;步骤,计算在所述目标点拍摄的地标的点之间的水平距离(D2) 被投影到相机的图像平面到图像平面的垂直二等分线,以及计算机器人在目标点处的相机与相机的焦点和相机的点之间的距离(D)的步骤 包括地标的平面通过使用距离(D1,D2)形成垂直度。
-
公开(公告)号:KR100756223B1
公开(公告)日:2007-09-06
申请号:KR1020060056381
申请日:2006-06-22
Applicant: 전자부품연구원
IPC: H02N2/10
CPC classification number: H02N2/123 , H01L41/0536 , H02N2/006 , H02N2/0065
Abstract: A rotor actuating device for controlling a micro displacement is provided to obtain a strong rotation force by using a compressive piezoelectric element as an actuator. A rotor actuating device for controlling a micro displacement includes a base plate(100), a rotor(110), a friction element(120), an actuator(140), a compressing unit(150), and a fixing block(130). The base plate(100) includes a lower part of the rotor(110) actuating device and supports other components. The rotor(110) has a disk shape and is rotatably installed on the base plate(100). The refractor has a bar shape. An end(120a) of the friction element(120) has a predetermined gap from an exterior of the rotor(110). An other gap(120b) of the friction element(120) is rotatably installed on the base plate(100). The actuator(140) actuates the friction element(120). The compressing unit(150) is installed on the fixing block(130) for elastically compressing the friction element(120) to the actuator(140). The fixing block(130) supports the friction element(120), the actuator(140), and the compressing unit(150), and is fixed on the base plate(100).
Abstract translation: 提供用于控制微位移的转子致动装置,以通过使用压电压电元件作为致动器来获得强的旋转力。 用于控制微位移的转子致动装置包括基板(100),转子(110),摩擦元件(120),致动器(140),压缩单元(150)和固定块(130) 。 基板(100)包括转子(110)致动装置的下部并支撑其它部件。 转子(110)具有圆盘形状并可旋转地安装在基板(100)上。 折射器具有条形。 摩擦元件(120)的端部(120a)与转子(110)的外部具有预定的间隙。 摩擦元件(120)的另一间隙(120b)可旋转地安装在基板(100)上。 致动器(140)致动摩擦元件(120)。 压缩单元(150)安装在用于将摩擦元件(120)弹性压缩到致动器(140)的固定块(130)上。 固定块(130)支撑摩擦元件(120),致动器(140)和压缩单元(150),并且固定在基板(100)上。
-
-
公开(公告)号:KR1020170134838A
公开(公告)日:2017-12-07
申请号:KR1020160065037
申请日:2016-05-26
Applicant: 전자부품연구원
Abstract: 본발명은물류운송로봇의충전시스템에관한것으로, 더욱상세하게는물류센터내에서충전시설이차지하는공간을최소화할수 있는물류운송로봇의충전장치및 이를이용한충전시스템에관한것이다. 본발명에따른물류운송로봇용충전장치는일측면이개방되며내부에물류운송로봇을수용하기위한공간이형성된충전본체, 충전본체의하부면에구비되어물류운송로봇의제1 충전단자와접촉되어물류운송로봇에전원을공급하는전원공급부를포함한다.
Abstract translation: 本发明涉及一种充电装置和使用本发明的充电系统中的充电系统涉及一种机器人运输物流,并且更具体地,物流机器人,可以通过最小化在仓库中的充电设施所占据的空间。 被打开根据本发明的用于物流运输机器人的充电装置被设置在填充体的下表面一侧,所述填充体的空间被形成为内部容纳物流机器人与物流运输机器人的第一充电端子接触 以及用于向物流运输机器人供电的供电单元。
-
公开(公告)号:KR1020170134837A
公开(公告)日:2017-12-07
申请号:KR1020160065035
申请日:2016-05-26
Applicant: 전자부품연구원
Abstract: 본발명은자동으로팔레트를보관하고, 공급할수 있는팔레트공급장치및 이를이용한물류자동화시스템에관한것이다. 본발명에따른팔레트공급장치는물품분류대로부터팔레트를전달받는이동레일부, 이동레일부의상부에구비되며, 이동레일부에안착된팔레트를파지하는그리퍼, 그리퍼의상부에구비되며, 그리퍼를상승또는하강시켜팔레트를보관하는상하이동부, 이동레일부를통해물품분류대로부터팔레트를전달받고, 이동레일부를통해전달받은팔레트를그리퍼를통해파지한상태로상하이동부를통해상승시켜보관하거나, 하강시킨상태에서그리퍼의파지를해제하여물류운송로봇으로공급하도록제어하는공급제어부를포함한다.
Abstract translation: 托盘供给装置以及使用该托盘供给装置的物流自动化系统技术领域本发明涉及能够自动储存托盘的托盘供给装置以及使用该托盘供给装置的物流管理系统。 在项目分类描述托盘供给装置设置在上移动轨道部设置,所述可动轨道部从所述AS接收托盘的转移,被设置在夹持器的顶部,用于夹紧托盘夹持器搁置在移动轨道,提高抓具 或降低由上海东告知从托盘中,如,铁路部分将货物分类存放的调色板,通过导轨部件交付移动调色板,而它是通过上海东方升起握hansangtae通过抓取器来存储或下跌 以及供给控制单元,用于释放夹持器的夹持并控制对物流运输机器人的供应。
-
公开(公告)号:KR1020160132149A
公开(公告)日:2016-11-17
申请号:KR1020150062855
申请日:2015-05-06
Applicant: 전자부품연구원
Abstract: 본발명은관절구동모듈에관한것으로, 위치맞춤갭을사용하지않으면서작업자의숙련도에크게의존하지않고절대엔코더의조립을쉽게수행하기위한것이다. 본발명에따른관절구동모듈은모듈고정부, 모듈출력부, 출력축샤프트및 절대엔코더를포함한다. 모듈고정부는양쪽으로중공형의모터축이돌출되게설치된모터를구비한다. 모듈출력부는모터의일측으로돌출된모터축부분에결합된감속기와, 감속기에결합되어모터축방향으로발생되는토크를축정하는 1축토크센서를구비한다. 출력축샤프트는모터축에삽입되어설치되되, 일측은모듈출력부에고정되고, 타측은모듈고정부밖으로돌출되며, 모듈고정부밖으로돌출된타단부에숫나사산이형성되어있다. 그리고절대엔코더는출력축샤프트에나사결합방식으로결합된다.
-
-
-
-
-
-
-
-
-