-
公开(公告)号：KR101749100B1
公开(公告)日：2017-07-03
申请号：KR1020100134081
申请日：2010-12-23
Applicant: 한국전자통신연구원
CPC classification number: G08C23/00 , G06F3/017 , G06F3/038 , G06F2203/0381 , G10L15/22
Abstract: 본발명에따른제스처/음향융합인식시스템은, 영상정보로부터사용자명령에해당하는제스처특징정보를추출하고, 상기제스처특징정보로부터제스처인식정보를획득하는제스처인식부; 음향정보로부터미리구축된배경음모델을이용하여배경음정보를획득하는배경음인식부; 상기음향정보로부터사용자명령에해당하는음향특징정보를추출하되상기배경음정보를바탕으로상기음향특징정보를추출하고, 상기음향특징정보로부터음향인식정보를획득하는음향인식부; 및상기제스처인식정보와상기음향인식정보를융합하여융합인식정보를생성하는융합인식부를포함하는것을특징으로한다.
Abstract translation: 根据本发明的手势/声音融合识别系统包括：手势识别单元，用于从图像信息中提取与用户命令对应的手势特征信息，并从手势特征信息中获取手势识别信息; 背景声音识别单元，用于通过使用从声音信息预先构建的背景声音模型来获取背景声音信息; 声学识别单元，从声学信息中提取与用户命令对应的声学特征信息，基于背景声音信息提取声学特征信息，并从声学特征信息中获取声学识别信息; 以及融合识别单元，用于通过融合手势识别信息和声音识别信息来生成融合识别信息。
-
公开(公告)号：KR1020120072244A
公开(公告)日：2012-07-03
申请号：KR1020100134081
申请日：2010-12-23
Applicant: 한국전자통신연구원
CPC classification number: G08C23/00 , G06F3/017 , G06F3/038 , G06F2203/0381 , G10L15/22
Abstract: PURPOSE: A gesture/sound mixing recognizing system for device control and a method thereof are provided to combine a gesture of a user with a sound of the user and recognize the combined gesture and sound under noises. CONSTITUTION: A gesture recognizing unit(210) obtains gesture recognizing information from gesture feature information. A sound recognizing unit(230) extracts sound feature information from a sound section selected using the gesture recognizing information. The sound recognizing unit obtains sound recognizing information from the sound feature information. A mixture recognizing unit(250) mixes the gesture recognizing information with the sound recognizing information. The mixture recognizing unit generates mixture recognizing information.
Abstract translation: 目的：提供一种用于设备控制的手势/声音混合识别系统及其方法，用于将用户的手势与用户的声音相结合，并在噪音下识别组合的手势和声音。 构成：手势识别单元（210）从手势特征信息获取手势识别信息。 声音识别单元（230）从使用手势识别信息选择的声音部分中提取声音特征信息。 声音识别单元从声音特征信息获得声音识别信息。 混合识别单元（250）将手势识别信息与声音识别信息混合。 混合识别单元产生混合识别信息。
-
公开(公告)号：KR1020120062495A
公开(公告)日：2012-06-14
申请号：KR1020100123782
申请日：2010-12-06
Applicant: 한국전자통신연구원
Inventor： 정영규
Abstract: PURPOSE: A vehicle door protection apparatus is provided to prevent the damage of the vehicle next to a vehicle due to carelessness while opening and closing a vehicle door. CONSTITUTION: A vehicle door protection apparatus comprises a distance measuring sensor(110), an opening and closing control unit(120), and an alarm sound output unit(130). The distance measuring sensor measures the distance between vehicles. The opening and closing control unit compares the measured distance and a reference distance and calculates the opening angle of a vehicle door. The opening and closing control unit controls the opening of a vehicle door according to the calculated opening angle of a vehicle door. The alarm sound output unit outputs alarm sound through receiving the control signals from the opening and closing control unit.
Abstract translation: 目的：提供一种车门保护装置，用于在打开和关闭车门时，防止由于粗心而导致车辆旁边车辆的损坏。 构成：一种车门保护装置，包括距离测量传感器（110），开闭控制单元（120）和报警声输出单元（130）。 距离测量传感器测量车辆之间的距离。 打开和关闭控制单元将测量的距离和参考距离进行比较，并计算车门的打开角度。 打开和关闭控制单元根据计算出的车门的开度来控制车门的打开。 报警声输出单元通过接收来自开闭控制单元的控制信号输出报警声。
-
公开(公告)号：KR1020120003588A
公开(公告)日：2012-01-11
申请号：KR1020100064269
申请日：2010-07-05
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A health care system and method thereof equipped with a metaverse interaction function are provided to supply an optimized exercise guideline by supplying an exercise environment and display an exercise state on a virtual space. CONSTITUTION: A metaverse server(210) manages a coordinate of an object for indicating the user within a metaverse space corresponding to the motion of a user. The metaverse server offers real effect information corresponding to the coordinate. A metaverse client(220) displays a virtual world through a graphic. A reality providing device(230) offers reality effect corresponding to the control of the metaverse client.
Abstract translation: 目的：提供一种配备有虚拟空间相互作用功能的保健系统及其方法，以通过提供运动环境并在虚拟空间上显示锻炼状态来提供优化的锻炼指南。 构成：虚拟服务器（210）管理对象的坐标，用于在对应于用户的运动的虚拟空间内指示用户。 虚拟服务器提供与坐标相对应的实际效果信息。 虚拟客户端（220）通过图形显示虚拟世界。 现实提供设备（230）提供与虚拟客户端的控制相对应的现实效果。
-
公开(公告)号：KR1020110066587A
公开(公告)日：2011-06-17
申请号：KR1020090123303
申请日：2009-12-11
Applicant: 한국전자통신연구원
CPC classification number: G08B27/00 , F21W2111/02 , H05B37/0227
Abstract: PURPOSE: A guard lamp for transmitting an accident alarm, providing a vehicle parking guide broadcast, and performing an emergency call-out vehicle guide function are provided to rapidly evacuate residents by transmitting a warning call for informing an accident. CONSTITUTION: A guard lamp(100) includes a network connection unit, a speaker drive unit, a warning lamp drive unit, and a control unit. The network connection unit is connected to a guard lamp management server(200). The speaker drive unit transmits an accident alarm through a speaker and provides a vehicle parking guide broadcast. A warning lamp drive unit drives a warning lamp for guiding an emergency vehicle to an accident spot. The control unit controls operations of the speaker driving unit and the warning lamp drive unit according to a control command of the guard lamp management server.
Abstract translation: 目的：提供一种用于传输事故报警器，提供车辆停车指南广播和执行紧急呼出车辆引导功能的防护灯，通过发送用于通知事故的警告呼叫来快速疏散居民。 构成：保护灯（100）包括网络连接单元，扬声器驱动单元，警告灯驱动单元和控制单元。 网络连接单元连接到防护灯管理服务器（200）。 扬声器驱动单元通过扬声器发送事故警报并提供车辆停车指南广播。 警告灯驱动单元驱动用于将紧急车辆引导到事故点的警告灯。 控制单元根据防护灯管理服务器的控制命令来控制扬声器驱动单元和警告灯驱动单元的操作。
-
Patent Agency Ranking
公开(公告)号：KR1020110037721A
公开(公告)日：2011-04-13
申请号：KR1020090095272
申请日：2009-10-07
Applicant: 한국전자통신연구원
CPC classification number: G06Q10/087 , G06F17/30241 , G06K17/0029 , G06Q50/28 , H04N7/18
Abstract: PURPOSE: A product map generating apparatus and method thereof within a product warehouse are provided to reduce the time to search for a product due to easily searching for a location of a product in a product warehouse by using a product map. CONSTITUTION: A palette(100) attaches a light emitting tag which emits identification light. A recognizer(300) recognizes the location of the palette by using light which is photographed through a camera. A map generator(400) generates a product map based on the recognized location of the palette. The light emitting tag includes one or more light source. The light is emitted from one light source.
Abstract translation: 目的：提供产品仓库内的产品地图生成装置及其方法，以便通过使用产品图来容易地在产品仓库中搜索产品的位置，从而减少搜索产品的时间。 构成：调色板（100）附着发出识别光的发光标签。 识别器（300）通过使用通过相机拍摄的光来识别调色板的位置。 地图生成器（400）基于所识别的调色板的位置生成商品地图。 发光标签包括一个或多个光源。 光从一个光源发射。
-
公开(公告)号：KR100999664B1
公开(公告)日：2010-12-08
申请号：KR1020080109683
申请日：2008-11-06
Applicant: 한국전자통신연구원
CPC classification number: G08G1/20
Abstract: 어린이의 위치 및 주변환경 정보를 어린이의 부모에게 제공하도록 한 어린이 보호 알림 장치 및 방법이 제시된다. 제시된 어린이 보호 알림 장치는, 사용자 단말기로부터 송출되는 사용자 정보를 수신하는 통신부; 각각의 사용자 정보에 해당하는 단말기 정보를 저장하고 있는 알림정보 저장부; 및 통신부가 수신한 사용자 정보에 해당하는 단말기 정보를 알림정보 저장부로부터 검출하고, 통신부를 통해 검출한 단말기 정보에 해당하는 단말기로 사용자 단말기의 위치정보를 전송하는 알림 제어부를 포함한다.
-
公开(公告)号：KR1020100070100A
公开(公告)日：2010-06-25
申请号：KR1020080128698
申请日：2008-12-17
Applicant: 한국전자통신연구원
CPC classification number: H04L41/082 , H04W84/18
Abstract: PURPOSE: A node updating method based on a database and a target node operating method thereof are provided to update an image of a sensor node by communicating only delta information consisting of difference between a present execution image and an image to update of the sensor node. CONSTITUTION: A database storing a node identifier and an execution image about each node is constructed. The execution image of a target node from the database is extracted by using an identifier of the target node(S330). The similarity of data between the execution image and an updating image is compared with reference to an address field. The data includes a record of the updating image which is not similar. Delta information is generated(S350).
Abstract translation: 目的：提供一种基于数据库和其目标节点操作方法的节点更新方法，用于通过仅传达包括当前执行图像与图像之间的差异的增量信息来更新传感器节点的图像以更新传感器节点。 构成：构建存储有关每个节点的节点标识符和执行映像的数据库。 通过使用目标节点的标识符来提取来自数据库的目标节点的执行图像（S330）。 将执行图像和更新图像之间的数据的相似性与参考地址字段进行比较。 该数据包括不相似的更新图像的记录。 生成增量信息（S350）。
-
公开(公告)号：KR1020100066352A
公开(公告)日：2010-06-17
申请号：KR1020090104128
申请日：2009-10-30
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A context recognition apparatus and a situation method using the same capable of recognizing the emotion status of a user are provided to efficiently recognize the emotion state of the user by utilizing a recognition result of multiple sensors. CONSTITUTION: A voice-based recognizing unit(110) recognizes the emotion state of a user based on voice signal. A movement-based recognizing unit(120) recognizes the emotion state of the user based on motion. A location recognizing unit(130) recognizes the place where the user belongs. A merged recognizing unit(150) analyzes recognition results and recognizes the situation of the user.
Abstract translation: 目的：通过利用多个传感器的识别结果，提供能够识别用户的情感状态的上下文识别装置和情况方法，以有效地识别用户的情感状态。 构成：基于语音识别单元（110）基于语音信号识别用户的情绪状态。 基于运动的识别单元（120）基于运动识别用户的情感状态。 位置识别单元（130）识别用户所属的位置。 合并识别单元（150）分析识别结果并识别用户的情况。
-
公开(公告)号：KR100919227B1
公开(公告)日：2009-09-28
申请号：KR1020070088918
申请日：2007-09-03
Applicant: 한국전자통신연구원
IPC: G10L15/02 , G10L15/04 , G10L15/06 , G08G1/0968
Abstract: 본 발명은 음성 인식 장치 및 방법에 관한 것으로, 적은 리소스를 사용하여 음성 입력된 단어를 인식하기 위한 장치 및 방법을 제공한다.
이를 위해 본 발명에 따른 네비게이션을 이용한 음성 인식 방법은, (a) 음성을 통하여 입력된 단어의 특징 데이터를 추출하는 단계; (b) 상기 추출된 단어 특징 데이터와 미리 등록된 단어 특징 데이터를 비교함으로써 상기 단어를 인식하는 단계; (c) 상기 추출된 단어 특징 데이터와 일치하는 상기 미리 등록된 단어 특징 데이터를 갖는 등록 단어가 없다면, 상기 입력된 음성의 초성 특징 데이터를 추출하는 단계; 및 (d) 상기 추출된 초성 특징 데이터를 미리 등록된 초성 특징 데이터와 비교함으로써 상기 단어를 인식하는 단계를 포함함으로써, 적은 리소스를 사용하여 음성 입력된 단어를 인식할 수 있는 이점이 있다.
-
-
-
-
-
-
-
-
-
Search Results
44
records
(took 0.021 seconds)
