-
公开(公告)号：KR1020090042503A
公开(公告)日：2009-04-30
申请号：KR1020070108292
申请日：2007-10-26
Applicant: 아주대학교산학협력단
IPC: G06Q50/22 , G05D1/02 , H04N21/231 , H04N21/236 , G06T7/00
CPC classification number: G06Q50/22 , G05D1/0278 , G06T7/344 , H04N21/231 , H04N21/236
Abstract: A system for human activity recognition and location tracking is provided to prevent children or elderly people with dementia from kid napping or getting lost by delivering the behavior state in a real time basis to external device and tracking the location through GPS. A smart pendant(200) comprises the followings. An action recognition analysis unit(210) recognizes and analyzes clearly an action of a user from the image inputted from outside and gets a grip on a current situation of the user. An automatic capturing unit(220) captures an image inputted through a camera installed inside the action recognition analysis unit. A recording unit(230) records a movement situation of the user as a video. A location analysis unit(240) receives a string signal value from GPS in order to chase the location of the user. The action recognition analysis unit analyses the images depends on walk, stop, left turn, and right turn of the user by using a camera installed inside the action recognition analysis unit.
Abstract translation: 提供人体活动识别和位置跟踪系统，以防止儿童或老年痴呆症患者通过将实时行为状态实时传递给外部设备并通过GPS跟踪位置，从而引起孩童的睡眠或迷失。 智能吊坠（200）包括以下内容。 动作识别分析单元（210）从从外部输入的图像中清楚地识别和分析用户的动作，并且掌握用户的当前状况。 自动拍摄单元（220）捕获通过安装在动作识别分析单元内的照相机输入的图像。 记录单元（230）将用户的移动状况记录为视频。 位置分析单元（240）从GPS接收字符串信号值，以便追踪用户的位置。 动作识别分析单元通过使用安装在动作识别分析单元内的摄像机来分析图像取决于用户的行走，停止，左转和右转。
-
公开(公告)号：KR1020090021613A
公开(公告)日：2009-03-04
申请号：KR1020070086253
申请日：2007-08-27
Applicant: 아주대학교산학협력단
CPC classification number: G06K9/00771 , G06T7/292 , G06T2207/10016
Abstract: An apparatus and a method for inferring the configuration of an image acquisition device network by tracking movement are provided to clarify the phase and the distance between image acquisition devices in a network which consists of multiple image acquisition devices so that the movable object of the image acquisition network can be accurately detected. An object extractors(131-1 to 131-N) extracts at least one movable object from an image acquired according to plural image acquisition devices. An emerging data generator(150) determines an image acquisition device at which a movable object appears as an emerging image device and sets a time at which the movable object appears as an emerging time to generate emerging data. In addition, the emerging data generator creates the disappearing time of a disappearing image device at which the movable object disappears. An emerging database(170) stores the emerging time and the disappearing time, and a phase inferring unit(190) infers a phase of a network.
Abstract translation: 提供了一种用于通过跟踪移动来推断图像采集设备网络的配置的装置和方法，以阐明由多个图像采集设备组成的网络中的图像采集设备之间的相位和距离，使得图像采集的可移动对象 可以准确检测网络。 物体提取器（131-1至131-N）从根据多个图像获取装置获取的图像中提取至少一个可移动物体。 新出现的数据发生器（150）确定图像获取装置，在该图像采集装置中可移动对象出现为新出现的图像装置，并设置可移动对象出现的时间作为新兴时间以产生新出现的数据。 另外，新兴的数据发生器产生消失的图像设备消失的时间，在该消失的图像设备可移动物体消失。 新兴数据库（170）存储正在出现的时间和消失时间，而相位推断单元（190）推断网络的相位。
-
公开(公告)号：KR1020080022934A
公开(公告)日：2008-03-12
申请号：KR1020060086794
申请日：2006-09-08
Applicant: 아주대학교산학협력단
IPC: G06Q50/10
Abstract: An application framework system for distributing and migrating an application in a ubiquitous environment is provided to support a platform independent application to users, share computing resources by using a distribution MVC(Model, View and Controller) model, and guarantee migration to the users in the unique environment. An MVC/UAML(Ubiquitous Application Markup Language) manager(210) manages a model, a view, and a controller, and parses/manages UMAL defining the model, the view, and the controller by including a model manager(212), a view manager(214) and a controller manager(216). A service manager(220) operates and manages an automatic update service and migration service of the model, the view, and the controller. A runtime manager(230) manages a life cycle related to loading, executing, and terminating an application, and stores/manages information of the model, the view, and the controller. An event manager transfers a message and data among the model, the view, and the controller. A network manager(250) transfers the message generated from the application to other frameworks and devices.
Abstract translation: 提供了一种用于在无所不在的环境中分发和迁移应用程序的应用程序框架系统，以支持与用户的平台独立应用程序，通过使用分发MVC（模型，视图和控制器）模型共享计算资源，并保证向用户提供迁移 独特的环境。 MVC / UAML（无处不在的应用标记语言）管理器（210）管理模型，视图和控制器，并且通过包括模型管理器（212），解析/管理定义模型，视图和控制器的UMAL 视图管理器（214）和控制器管理器（216）。 服务管理器（220）操作和管理模型，视图和控制器的自动更新服务和迁移服务。 运行时管理器（230）管理与加载，执行和终止应用有关的生命周期，并存储/管理模型，视图和控制器的信息。 事件管理器在模型，视图和控制器之间传输消息和数据。 网络管理器（250）将从应用产生的消息传送到其他框架和设备。
-
公开(公告)号：KR100659566B1
公开(公告)日：2006-12-22
申请号：KR1020050056335
申请日：2005-06-28
Applicant: 아주대학교산학협력단
IPC: A61B5/11
Abstract: A method for recognizing a sleep state is provided to select one of sleep, sleepy and awake states and give a proper service by collecting a situation context or an activity of a human and guessing an intention of the human. A method for recognizing a sleep state includes the steps of: a)classifying a sleeping state of a human into sleep, sleepy and awake states, and determining a normal state probability based on three elements including a time, a position and a movement of a human; and b)repeatedly observing a personal life pattern in view of the time, position and movement based on the normal state probability defined at the step a), determining that the normal state probability is applied to each person correctly, and finally determining the state probability matched to the personal life pattern by continuously correcting the personal life pattern in case that the personal life pattern is not matched with the normal state probability.
Abstract translation: 提供一种用于识别睡眠状态的方法来选择睡眠状态，睡眠状态和清醒状态中的一个，并通过收集人的情境背景或活动并猜测人的意图来提供适当的服务。 一种用于识别睡眠状态的方法，包括以下步骤：a）将人的睡眠状态分类为睡眠状态，睡眠状态和清醒状态，以及基于包括时间，位置和移动的三个元素来确定正常状态概率 人类; 以及b）基于步骤a）中定义的正常状态概率，根据时间，位置和移动反复观察个人生活模式，确定正常状态概率正确地应用于每个人，并且最终确定状态概率 在个人生活模式与正常状态概率不匹配的情况下，通过不断纠正个人生活模式与个人生活模式相匹配。
-
-
Patent Agency Ranking
-
公开(公告)号：KR101797013B1
公开(公告)日：2017-11-13
申请号：KR1020120122928
申请日：2012-11-01
Applicant: 아주대학교산학협력단
Abstract: 본발명은텍스트편집시 한/영모드자동변환시스템및 방법에관한것으로, 보다상세하게는사용자가별도의한/영변환키를눌러서한/영모드를변환시키는것이아니라마우스의클릭또는드래그방법으로인식한글자를분석함으로써상황에맞는한/영모드를자동변환하여제공하는시스템및 방법에관한것이다. 이를위해, 본발명에따른한/영모드자동변환시스템은인식부, 연산부, 한/영모드설정부및 제공부를포함한다. 상기인식부는텍스트편집이가능한영역에서마우스클릭이벤트가발생한경우, 상기마우스클릭이벤트를감지하여클릭된위치의앞뒤글자를인식한다. 상기연산부는상기인식부에서마우스클릭이벤트로인식한앞뒤글자가아스키코드값에따라구분되어있는언어범위중 어떤언어범위에속하는지를판단하여, 상기앞뒤글자각각에대해언어범위판단값을도출한다. 상기한/영모드설정부는상기연산부에서도출한앞뒤글자의언어범위판단값을이용하여한/영모드의설정값을도출한다. 상기제공부는상기한/영모드설정부에서도출한설정값을운영체제로제공하여한/영모드를변환한다.
Abstract translation: 更具体地说，本发明涉及用于在文本编辑期间自动转换文本/语音模式的系统和方法，并且更具体地涉及用于自动将文本/ 本发明还提供了一种通过分析识别出的一个字符根据情况自动转换一对一模式的系统和方法。 为此，根据本发明的自动/自动模式自动转换系统包括识别单元，操作单元，单向模式设置单元和提供单元。 当在文本可编辑区域中发生鼠标点击事件时，识别单元检测到鼠标点击事件并识别点击位置之前和之后的字符。 算术单元根据ASCII码值确定识别单元将前向和后向字符识别为鼠标点击事件的语言范围中的哪一个语言范围根据ASCII码值进行分类，并且为前向和后向字符中的每一个导出语言范围确定值。 一归零模式设置单元使用从操作单元导出的前后字符的语言范围确定值来导出一归零模式的设置值。 提供单元向操作系统提供从一/零模式设置单元导出的设置值以转换该一/零模式。
-
公开(公告)号：KR1020170123948A
公开(公告)日：2017-11-09
申请号：KR1020160053314
申请日：2016-04-29
Applicant: 아주대학교산학협력단
CPC classification number: Y02D70/10
Abstract: 본발명의일 실시예에따른클라우드기반의저전력 GPS 데이터공유방법은 복수의인원각각으로부터스마트폰의 GPS 데이터를수신하는단계; 상기수신된 GPS 데이터에기초하여, 상기복수의인원중에서소정의거리, 이동방향및 이동속도에대한조건을충족하는인원으로구성되는그룹을생성하는단계; 스마트폰의배터리잔량에기초하여, 상기그룹의그룹원중에서하나의헤더를선정하는단계; 및상기헤더로부터스마트폰의배터리잔량에기초하여적응적으로 GPS데이터를수집하는단계를포함한다.
Abstract translation: 根据本发明实施例的基于云的低功率GPS数据共享方法包括：从多个人员中的每个人员接收智能手机的GPS数据; 确定，基于所接收的GPS数据，并创建由所述多个满足条件的预定距离，移动方向和移动速度的人员的人员的一个基团; 基于智能手机的电池剩余量在组的组成员中选择一个头部; 并根据标题中智能手机的电池剩余量自适应收集GPS数据。
-
公开(公告)号：KR101783183B1
公开(公告)日：2017-10-23
申请号：KR1020160019040
申请日：2016-02-18
Applicant: 아주대학교산학협력단
Abstract: 스마트기기사용자의감정상태분류방법을제공한다. 본발명의일 실시예에따른스마트기기사용자의감정상태분류방법은스마트기기로부터복수의애플리케이션종류별사용정보를포함하는제1 앱사용정보를수집하는단계; 상기스마트기기의사용자로부터소정시간구간동안의감정상태에대한정보를포함하는감정정보를복수회수집하는단계; 및상기소정시간구간별상기감정정보및 상기제1 앱사용정보에기초하여상기스마트기기사용자의애플리케이션종류별사용정보와감정상태간의관계를나타내는감정상태결정모델을생성하는단계;를포함한다.
-
公开(公告)号：KR1020160136906A
公开(公告)日：2016-11-30
申请号：KR1020150071130
申请日：2015-05-21
Applicant: 아주대학교산학협력단
Abstract: 본발명은관심지점을이용한이동패턴분석시스템및 방법에관한것으로, 보다상세하게는비콘신호를이용하여사용자의이동패턴을분석하는기술에관한것이다. 이러한목적을달성하기위하여본 발명의일 실시예에따른관심지점을이용한이동패턴분석시스템은시간에따라수집된사용자의위치정보를기반으로상기사용자의생활패턴을추출하는제1 분석부, 및상기분석된사용자의생활패턴을기반으로관심지점(POI, Point of Interest)을산출하고, 상기산출된관심지점을기반으로 POI 시퀀스를모델링하는제2 분석부를포함할수 있다. 이러한본 발명은비콘신호를이용하여사용자의배회, 혹은평소다니는경로의이탈여부를감지할수 있으며, 특히사용자의평소이동패턴을기반으로비콘신호를이용해사용자의이상여부를파악할수 있는효과가있다.
-
-
-
-
-
-
-
-
-
Search Results
219
records
(took 0.027 seconds)
