-
公开(公告)号：KR1020080051013A
公开(公告)日：2008-06-10
申请号：KR1020070094739
申请日：2007-09-18
Applicant: 한국전자통신연구원
CPC classification number: G06T7/215
Abstract: An apparatus and a method for recognizing person behaviors using a motion vector are provided to cope with an emergency situation effectively that may occur with an observation target person. A recognition table storage unit(11) stores behavior configuration information corresponding to an average motion vector. An image collecting unit(12) collects images inputted from the exterior. A motion vector calculating unit(13) recognizes an object from an image collected by the image collecting unit and calculates motion vectors of the object. An average motion vector calculating unit(14) calculates an average motion vector with respect to the motion vectors calculated by the motion vector calculating unit. A behavior recognizing unit(15) searches the recognition table storage unit with the average motion vector calculated by the average motion vector calculating unit to recognize a corresponding behavior.
Abstract translation: 提供一种使用运动矢量识别人物行为的装置和方法，以便有效地处理可能与观察目标人物发生的紧急情况。 识别表存储单元（11）存储对应于平均运动矢量的行为配置信息。 图像采集单元（12）收集从外部输入的图像。 运动矢量计算单元（13）从由图像采集单元收集的图像识别对象，并计算对象的运动矢量。 平均运动矢量计算单元（14）针对由运动矢量计算单元计算的运动矢量计算平均运动矢量。 行为识别单元（15）利用由平均运动矢量计算单元计算的平均运动矢量来识别识别表存储单元，以识别对应的行为。
-
公开(公告)号：KR100813008B1
公开(公告)日：2008-03-13
申请号：KR1020060123331
申请日：2006-12-06
Applicant: 한국전자통신연구원
IPC: G01N33/48
Abstract: An apparatus and a method for predicting a gene module are provided to predict the gene module(a group of genes) controlled by the same transcription factor group and performing the same function in a cell from gene expression data and transcription factor binding information. An apparatus for predicting a gene module(103) comprises: a gene expression similarity matrix generation means(104) which generates a gene expression similarity matrix using gene expression date from the outside gene expression data; a gene expression similarity graph generation means(105) which generates a gene expression similarity graph using the gene expression similarity matrix generated from the gene expression similarity matrix generation means; a crowd portion graph generation means(106) which applies a crowd portion graph algorithm to the gene expression similarity graph to generate a crowd portion graph; and a significant gene module generation means(107) which extracts the transcription factor binding information from the outside transcription factor binding database using the crowd portion graph to generate a significant gene module, wherein the crowd portion graph generation means generates the crowd portion graph accepting the replication of the gene.
Abstract translation: 提供一种用于预测基因模块的装置和方法，以预测由相同转录因子组控制的基因模块（一组基因），并从基因表达数据和转录因子结合信息在细胞中起相同的功能。 一种用于预测基因模块（103）的装置，包括：使用基因表达日期从外部基因表达数据生成基因表达相似性矩阵的基因表达相似性矩阵生成装置（104） 使用从基因表达相似矩阵生成单元生成的基因表达相似性矩阵生成基因表达相似图的基因表达相似图生成单元（105） 人群部分图生成装置，将群体部分图算法应用于基因表达相似图，生成人群部分图; 以及使用人群部分图从外部转录因子结合数据库中提取转录因子结合信息以产生重要基因模块的显着基因模块产生装置（107），其中，群体部分图生成装置生成接受 基因的复制。
-
-
公开(公告)号：KR100550329B1
公开(公告)日：2006-02-08
申请号：KR1020030080817
申请日：2003-11-15
Applicant: 한국전자통신연구원
IPC: G06F17/30
Abstract: 본 발명은 단백질 3차원 구조공간상에서 원자 또는 펩티드 결합관계에 의해 각 에지패턴의 분포인 3차원 에지 히스토그램을 생성하고 그 히스토그램의 유사도를 통해 사용자의 질의 단백질과 구조적으로 유사한 단백질들을 검출 제공하는 단백질 구조 비교검색 장치 및 방법에 관한 것이다.
본 발명은 사용자로부터 질의 단백질을 입력받아 단백질 구조 검색서버에 유사 단백질의 검색을 요청하고 그 검색 결과를 출력하는 검색 클라어언트; 각종 단백질들에 대한 3D 에지 히스토그램을 생성하고 이들을 데이터베이스화하는 3D 에지 히스토그램 추출/저장장치; 및 질의 단백질에 대한 3D 에지 히스토그램을 생성한 후 이를 상기 데이터베이스화된 각종 단백질의 3D 에지 히스토그램들과 상호 비교하여 유사도를 산출하고 소정 이상의 유사도를 갖는 단백질들을 검색 제공하는 단백질 구조 검색서버;로 구성된다.
단백질 구조 검색, 결합선 분포, 3차원 에지 히스토그램, 생물정보학, PDB, PCA-
公开(公告)号：KR100501413B1
公开(公告)日：2005-07-18
申请号：KR1020030074429
申请日：2003-10-23
Applicant: 한국전자통신연구원
CPC classification number: G06Q50/22 , G06F19/00 , Y10S707/99933 , Y10S707/99943
Abstract: 본 발명은 유엠엘에스(UMLS : United Medical Language System)를 기반으로 생물학 문헌으로부터 생물학적 개체명을 인식하는 장치 및 그 방법에 관한 것이다. 본 발명의 장치 및 방법은 유엠엘에스에서 메타시소러스를 제공받아 개체명 인식에 사용될 언어자원인 개념어, 단일어 및 범주키텀 데이터베이스를 각각 구축하고, 상기 개념어 데이터베이스에 저장된 각 개념어를 입력받아 상기 단일어 및 범주키텀 데이터베이스에 저장된 자료를 이용하여 각 개념어에 대한 자질을 추출하며, 상기 추출된 결과를 이용하여 개체명을 인식하기 위한 규칙 생성 및 규칙 필터링 과정을 거쳐 규칙 데이터베이스를 구축하며, 생물학 문서를 입력받아 개체명 후보가 되는 명사 및 명사구를 추출하여 상기 규칙 데이터베이스에 저장된 규칙을 상기 명사 및 명사구에 적용하여 개체명 인식을 수행한다. 이렇게 함으로써, 입력문서에서 중요 정보 개체로 활용될 수 있는 생물학적 개체명들을 효과적으로 추출할 수 있다.
-
Patent Agency Ranking
公开(公告)号：KR1020050046960A
公开(公告)日：2005-05-19
申请号：KR1020030080817
申请日：2003-11-15
Applicant: 한국전자통신연구원
IPC: G06F17/30
Abstract: 본 발명은 단백질 3차원 구조공간상에서 원자 또는 펩티드 결합관계에 의해 각 에지패턴의 분포인 3차원 에지 히스토그램을 생성하고 그 히스토그램의 유사도를 통해 사용자의 질의 단백질과 구조적으로 유사한 단백질들을 검출 제공하는 단백질 구조 비교검색 장치 및 방법에 관한 것이다.
본 발명은 사용자로부터 질의 단백질을 입력받아 단백질 구조 검색서버에 유사 단백질의 검색을 요청하고 그 검색 결과를 출력하는 검색 클라어언트; 각종 단백질들에 대한 3D 에지 히스토그램을 생성하고 이들을 데이터베이스화하는 3D 에지 히스토그램 추출/저장장치; 및 질의 단백질에 대한 3D 에지 히스토그램을 생성한 후 이를 상기 데이터베이스화된 각종 단백질의 3D 에지 히스토그램들과 상호 비교하여 유사도를 산출하고 소정 이상의 유사도를 갖는 단백질들을 검색 제공하는 단백질 구조 검색서버;로 구성된다.-
公开(公告)号：KR1020050039067A
公开(公告)日：2005-04-29
申请号：KR1020030074429
申请日：2003-10-23
Applicant: 한국전자통신연구원
CPC classification number: G06Q50/22 , G06F19/00 , Y10S707/99933 , Y10S707/99943
Abstract: The present invention relates to an apparatus and method for recognizing biological named entity from biological literature based on united medical language system (UMLS). The apparatus and the method receives metathesaurus from the UMLS, constructs a concept name database, a single name database and a category keyterm database, which are language resources to be used recognize a named entity, receives each concept name stored in the concept name database, extracts features of each of the concept names by using data stored in the single name database and the category keyterm database, constructs a rule database by creating rules used to recognize the named entity and filtering the rules by using the extracted features, receives a biological literature, extracts nouns and noun phrases that are candidate named entities, applies the rules stored in the rule database to the nouns and the noun phrases, and recognizes the named entities. In the present invention, the biological named entities can be effectively extracted which can be used as important information individual in input literature.
-
公开(公告)号：KR100456627B1
公开(公告)日：2004-11-10
申请号：KR1020020078187
申请日：2002-12-10
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A system and a method for predicting a function of a macromolecular based on a 3D(Dimensional) structure are provided to predict the macromolecular fast and accurately by automatically processing a process for expressing the macromolecular as a descriptor, searching a similar structure, modeling a combination structure of the macromolecular, and predicting the function. CONSTITUTION: The first database(100) stores the data of the macromolecular that the function and the structure are defined. A 3D expressing/searching part(200) expresses the 3D structure of the macromolecular data stored in the first database by the selected descriptor and expressing method, classifies the 3D descriptors into each group, and searches the similar structure. The second database(300) stores the 3D descriptors of the 3D expressing/searching part, stores the 3D descriptors of the first database after classifying into each similar structure group, and stores a folder structure characteristic of the macromolecular. A 3D modeling/function predicting part(400) models the 3D structure and analyzes the folder structure characteristic.
Abstract translation: 目的：提供一种基于3D（Dimensional）结构预测大分子功能的系统和方法，通过自动处理表达大分子的过程作为描述符，快速准确地预测大分子，寻找类似的结构，建模 大分子的组合结构，并预测其功能。 构成：第一个数据库（100）存储定义了功能和结构的大分子数据。 3D表现/搜索部分（200）通过所选择的描述符和表达方法来表达存储在第一数据库中的大分子数据的3D结构，将3D描述符分类为每个组，并且搜索相似的结构。 第二数据库（300）存储3D表达/搜索部分的3D描述符，将分类后的第一数据库的3D描述符存储到每个相似结构组中，并存储大分子的文件夹结构特征。 3D建模/功能预测部件（400）对3D结构进行建模并分析文件夹结构特性。
-
公开(公告)号：KR1020040050372A
公开(公告)日：2004-06-16
申请号：KR1020020078187
申请日：2002-12-10
Applicant: 한국전자통신연구원
Abstract: PURPOSE: A system and a method for predicting a function of a macromolecular based on a 3D(Dimensional) structure are provided to predict the macromolecular fast and accurately by automatically processing a process for expressing the macromolecular as a descriptor, searching a similar structure, modeling a combination structure of the macromolecular, and predicting the function. CONSTITUTION: The first database(100) stores the data of the macromolecular that the function and the structure are defined. A 3D expressing/searching part(200) expresses the 3D structure of the macromolecular data stored in the first database by the selected descriptor and expressing method, classifies the 3D descriptors into each group, and searches the similar structure. The second database(300) stores the 3D descriptors of the 3D expressing/searching part, stores the 3D descriptors of the first database after classifying into each similar structure group, and stores a folder structure characteristic of the macromolecular. A 3D modeling/function predicting part(400) models the 3D structure and analyzes the folder structure characteristic.
Abstract translation: 目的：提供一种基于3D（Dimensional）结构预测大分子功能的系统和方法，通过自动处理表达大分子作为描述符的方法，快速准确地预测大分子，搜索类似结构，建模 大分子的组合结构，并预测功能。 构成：第一个数据库（100）存储大分子的功能和结构的定义数据。 3D表示/搜索部分（200）通过所选择的描述符和表达方法表示存储在第一数据库中的大分子数据的3D结构，将3D描述符分类到每个组中，并搜索类似的结构。 第二数据库（300）存储3D表达/搜索部分的3D描述符，在分类到每个相似结构组之后存储第一数据库的3D描述符，并存储大分子的文件夹结构特征。 3D建模/功能预测部分（400）模拟3D结构并分析文件夹结构特征。
-
公开(公告)号：KR1020020053968A
公开(公告)日：2002-07-06
申请号：KR1020000082179
申请日：2000-12-26
Applicant: 한국전자통신연구원
IPC: G06F17/30
CPC classification number: G06F17/30259 , G06F17/3025
Abstract: PURPOSE: An image data color and shape searching system and method is provided to register a fuzzy representation value on a color and a shape with a natural language, and to enable a user to input a natural language for searching a color or a shape. CONSTITUTION: The system comprises a query sentence input module(11), a dictionary storage(12), a natural language processor(13), a color/shape word recognizer(14), a color/shape threshold database(15), a color/shape threshold database constructor(16), a search module(17), and a search result output module(18). The query sentence input module(11) enables a user to input a natural language sentence describing a desired color or shape to be searched for. The dictionary storage(12) stores dictionary data for processing a sentence. The natural language processor(13) analyzes the natural language sentence to be input via the query sentence input module(11) by using the dictionary data stored at the dictionary storage(14). The color/shape threshold database(15) stores a color histogram value mapped with a word in relation with a color, and edge data corresponding to a word in relation with a shape. The search module(17) searches the database(15) for the color and the shape requested by the color/shape word recognizer(14). The search result output module(18) offers the searched color and shape to the user. The natural language processor(13) includes a tokenizer and a parser. The tokenizer separates the input sentence into words, and the parser searches for morphemes corresponding to each word output by the tokenizer, and analyzes the structure of the sentence.
Abstract translation: 目的：提供一种图像数据颜色和形状搜索系统和方法，以自然语言对颜色和形状注册模糊表示值，并使用户能够输入搜索颜色或形状的自然语言。 构成：系统包括查询语句输入模块（11），字典存储器（12），自然语言处理器（13），颜色/形状字识别器（14），颜色/形状阈值数据库（15）， 颜色/形状阈值数据库构造器（16），搜索模块（17）和搜索结果输出模块（18）。 查询语句输入模块（11）使得用户能够输入描述要搜索的期望颜色或形状的自然语言句子。 字典存储器（12）存储用于处理句子的字典数据。 自然语言处理器（13）通过使用存储在字典存储器（14）中的字典数据，经由查询语句输入模块（11）分析要输入的自然语言句子。 颜色/形状阈值数据库（15）存储与颜色相关的字映射的颜色直方图值，以及与形状相关联的字的边缘数据。 搜索模块（17）在数据库（15）中搜索由颜色/形状字识别器（14）请求的颜色和形状。 搜索结果输出模块（18）向用户提供所搜索的颜色和形状。 自然语言处理器（13）包括标记器和解析器。 标记器将输入的句子分成单词，并且解析器搜索与分词器输出的每个单词对应的语素，并分析句子的结构。
-
-
-
-
-
-
-
-
-
Search Results
133
records
(took 0.033 seconds)
