-
公开(公告)号：KR102233427B1
公开(公告)日：2021-03-29
申请号：KR1020140044423A
申请日：2014-04-14
CPC classification number: A61B8/5261 , A61B5/055 , A61B5/1128 , A61B5/113 , A61B5/7289 , A61B6/032 , A61B6/4417 , A61B6/486 , A61B6/5247 , A61B6/5288 , A61B8/463 , G06T7/33 , A61B2505/05 , A61B2576/02 , A61B6/037 , G01R33/5608 , G06T2207/10072 , G06T2207/10116 , G06T2207/10132 , G16H30/40
Abstract: 개시된 실시 예들은 의료 영상 정합 방법 및 장치에 관한 것으로, 실시 예에 따른 의료 영상 정합 방법은 료 시술 이전에 호흡 변형 정보가 반영된 비 실시간 의료영상들을 생성하고, 의료 시술 중에는 실시간 의료영상과 생성된 비 실시간 의료 실시간 의료 영상을 강체 정합하여 호흡 변형 정보가 고려된 정확도 높은 정합 영상을 얻을 수 있다.
-
公开(公告)号：KR100419777B1
公开(公告)日：2004-02-21
申请号：KR1020010062085
申请日：2001-10-09
Applicant: 한국과학기술원
IPC: G06T7/20
Abstract: PURPOSE: A method and a system for recognizing successive sign language based on computer vision are provided to efficiently recognize had gesture and output the meaning of the sign language into voice signal. CONSTITUTION: A data obtaining part obtains color image data of a person talking with the hands(S510). A pre-processor processes hand video data for obtaining hand locus data and hand gesture data(S511-S513). A hand motion dividing part divides the hand locus data into individual sign language sentences and again divides into sign language(S514-S515). A hand motion classifying part classifies the least unit of sign language by using the individual sign language and the hand gesture data(S516-S517). A sign language interpreting part recognizes sign language words by combining the classified unit of sign language(S518). The sign language interpreting part interprets the sign language words into the sign language sentences by considering a sign language grammar(S519-S520). A sound generating part outputs the interpreted sign language sentences in a voice(S521).
Abstract translation: 目的：提供一种基于计算机视觉的连续手语识别方法和系统，以有效识别手势并将手语的含义输出为语音信号。 组成：数据获得部分获得与手谈话的人的彩色图像数据（S510）。 预处理器处理手部视频数据以获得手部轨迹数据和手势数据（S511-S513）。 手部运动分割部分将手部轨迹数据分成单独的手语句子并再次分成手语（S514-S515）。 手部运动分类部件通过使用个别手语和手势数据来分类手语的最小单位（S516-S517）。 手语翻译部分通过组合手语的分类单元来识别手语词（S518）。 手语翻译部分通过考虑手语语法将手语翻译为手语语句（S519-S520）。 声音生成部分以语音输出解释的手语语句（S521）。
-
公开(公告)号：KR1020030077348A
公开(公告)日：2003-10-01
申请号：KR1020020016422
申请日：2002-03-26
Applicant: 한국과학기술원
IPC: H04N7/08
CPC classification number: G09B21/009 , G06T13/20 , G06T15/04
Abstract: PURPOSE: A system for generating three-dimensional finger language animation based on TV subtitle signals is provided to help a deaf and dumb person to understand subtitle broadcasting. CONSTITUTION: A subtitle decoder(101) extracts subtitle signals included in TV signals. A pre-processor(102) processes the extracted subtitle signals into subtitle signals appropriate for the finger language expression. A morpheme analyzing unit(103) analyzes the subtitle signal obtained from the pre-processor into morphemes. A three-dimensional finger language animation database(104) provides appropriate three-dimensional finger language animation data according to the result of the morpheme analysis. A three-dimensional finger language speaker modeling unit realizes three-dimensional finger language animation by using the three-dimensional finger language animation data. A three-dimensional finger language animation information display unit displays the three-dimensional finger language speaker modeling unit and related three-dimensional information.
Abstract translation: 目的：提供一种基于电视字幕信号产生三维手指语言动画的系统，以帮助聋哑人了解字幕广播。 构成：字幕解码器（101）提取包括在TV信号中的字幕信号。 预处理器（102）将提取的字幕信号处理成适合于手指语言表达的字幕信号。 语素分析单元（103）将从预处理器获得的字幕信号分析成语素。 三维手指语言动画数据库（104）根据语素分析的结果提供适当的三维手指语言动画数据。 三维手指语言扬声器建模单元通过使用三维手指语言动画数据实现三维手指语言动画。 三维手指语言动画信息显示单元显示三维手指语言说话者建模单元和相关的三维信息。
-
公开(公告)号：KR1020030030232A
公开(公告)日：2003-04-18
申请号：KR1020010062085
申请日：2001-10-09
Applicant: 한국과학기술원
IPC: G06T7/20
Abstract: PURPOSE: A method and a system for recognizing successive sign language based on computer vision are provided to efficiently recognize had gesture and output the meaning of the sign language into voice signal. CONSTITUTION: A data obtaining part obtains color image data of a person talking with the hands(S510). A pre-processor processes hand video data for obtaining hand locus data and hand gesture data(S511-S513). A hand motion dividing part divides the hand locus data into individual sign language sentences and again divides into sign language(S514-S515). A hand motion classifying part classifies the least unit of sign language by using the individual sign language and the hand gesture data(S516-S517). A sign language interpreting part recognizes sign language words by combining the classified unit of sign language(S518). The sign language interpreting part interprets the sign language words into the sign language sentences by considering a sign language grammar(S519-S520). A sound generating part outputs the interpreted sign language sentences in a voice(S521).
Abstract translation: 目的：提供一种基于计算机视觉识别连续手语的方法和系统，以有效识别手势并将手语的意义输出为语音信号。 构成：数据获取部件获取与人交谈的人的彩色图像数据（S510）。 预处理器处理手视频数据以获得手轨迹数据和手势数据（S511-S513）。 手动分割部将手轨迹数据划分为单独的手语句，并再次分为手语（S514-S515）。 手动分类部分通过使用个体手语和手势数据对手语的最小单位进行分类（S516-S517）。 手语翻译部分通过组合手语的分类单位来识别手语单词（S518）。 手语解释部分通过考虑手语语法（S519-S520）将手语语言解释为手语语句。 声音产生部分输出语音中解释的手语语句（S521）。
-
公开(公告)号：KR101932721B1
公开(公告)日：2018-12-26
申请号：KR1020120099548
申请日：2012-09-07
Abstract: 복수개의 의료 영상들을 정합하는 방법이 개시된다. 본 발명의 일 실시예에 따른 의료 영상 정합 방법은, 의료 시술 이전에 촬영된 제1 의료 영상 및 의료 시술 중 실시간으로 촬영된 제2 의료 영상을 획득하고, 환자의 관심 장기와 인접하는 다수의 해부학적 개체들 중에서 상기 해상도가 낮은 제2 의료 영상에서 식별 가능한 적어도 두 개의 인접 개체들 각각에 대한 특징점들을 상기 제1 의료 영상과 상기 제2 의료 영상에서 추출하고, 상기 제1 의료 영상의 특징점들이 나타내는 상기 인접 개체들 간의 기하학적 관계와 상기 제2 의료 영상의 특징점들이 나타내는 상기 인접 개체들 간의 기하학적 관계에 기초하여 상기 제1 의료 영상과 상기 제2 의료 영상을 정합한다.
-
Patent Agency Ranking
公开(公告)号：KR1020140032810A
公开(公告)日：2014-03-17
申请号：KR1020120099548
申请日：2012-09-07
CPC classification number: G06T7/0016 , G06T7/33 , G06T2207/10081 , G06T2207/10084 , G06T2207/10088 , G06T2207/10104 , G06T2207/10136 , G06T2207/30056 , G06T2207/30084 , G06T2207/30101
Abstract: A method for matching a plurality of medical images is disclosed. The medical image matching method according to an embodiment of the present invention comprises: obtaining a first medical image taken before a medical procedure and a second medical image taken in real-time during the medical procedure; extracting feature points for each of at least two adjacent entities, which are identifiable in the second medical image having a low resolution, from the first medical image and the second medical image among a plurality of anatomical entities adjacent to an organ of interest of a patient; and matching the first medical image and the second medical image based on a geometrical relationship between the adjacent objects displayed by the feature points of the first medical image and a geometrical relationship between the adjacent objects displayed by the feature points of the second medical image. [Reference numerals] (210) First medical image storage unit; (220) Second medical image storage unit; (230) Feature point extracting unit; (231) Adjacent objects extracting unit; (232) Coordinate extracting unit; (240) Matching unit; (241) Vector calculating unit; (242) Matrix calculating unit; (243) Basic matching unit; (244) Boundary area selecting unit; (245) Matching image correcting unit; (AA) First medical image; (BB) Second medical image; (CC) Image process processor
Abstract translation: 公开了一种用于匹配多个医学图像的方法。 根据本发明的实施例的医学图像匹配方法包括：获得在医疗过程之前拍摄的第一医疗图像和在医疗过程中实时拍摄的第二医疗图像; 从与患者感兴趣的器官相邻的多个解剖实体中的第一医学图像和第二医疗图像提取至少两个相邻实体中的每一个的特征点，其可以在具有低分辨率的第二医学图像中被识别 ; 并且基于由第一医学图像的特征点显示的相邻对象之间的几何关系和由第二医学图像的特征点显示的相邻对象之间的几何关系，匹配第一医疗图像和第二医疗图像。 （附图标记）（210）第一医用图像存储单元; （220）第二医用图像存储单元; （230）特征点提取单元; （231）相邻对象提取单元; （232）坐标提取单元; （240）配套单元; （241）向量计算单元; （242）矩阵计算单元; （243）基本匹配单位; （244）边界区选择单位; （245）匹配图像校正单元; （AA）第一医学图像; （BB）第二医学图像; （CC）图像处理处理器
-
公开(公告)号：KR100443365B1
公开(公告)日：2004-08-09
申请号：KR1020020016422
申请日：2002-03-26
Applicant: 한국과학기술원
IPC: H04N7/08
Abstract: PURPOSE: A system for generating three-dimensional finger language animation based on TV subtitle signals is provided to help a deaf and dumb person to understand subtitle broadcasting. CONSTITUTION: A subtitle decoder(101) extracts subtitle signals included in TV signals. A pre-processor(102) processes the extracted subtitle signals into subtitle signals appropriate for the finger language expression. A morpheme analyzing unit(103) analyzes the subtitle signal obtained from the pre-processor into morphemes. A three-dimensional finger language animation database(104) provides appropriate three-dimensional finger language animation data according to the result of the morpheme analysis. A three-dimensional finger language speaker modeling unit realizes three-dimensional finger language animation by using the three-dimensional finger language animation data. A three-dimensional finger language animation information display unit displays the three-dimensional finger language speaker modeling unit and related three-dimensional information.
Abstract translation: 目的：提供一种基于电视字幕信号生成三维手指语言动画的系统，以帮助聋哑人理解字幕广播。 构成：小标题解码器（101）提取包含在电视信号中的小标题信号。 预处理器（102）将提取的字幕信号处理成适合手指语言表达的字幕信号。 语素分析单元（103）将从预处理器获得的字幕信号分析为语素。 根据词素分析的结果，三维手指语言动画数据库（104）提供适当的三维手指语言动画数据。 三维手指语言说话人建模单元通过使用三维手指语言动画数据来实现三维手指语言动画。 三维手指语言动画信息显示单元显示三维手指语言说话者建模单元和相关的三维信息。
-
公开(公告)号：KR100457928B1
公开(公告)日：2004-11-18
申请号：KR1020020070120
申请日：2002-11-12
Applicant: 한국과학기술원
IPC: G06F19/00
Abstract: PURPOSE: A method for recognizing a hand signal by a small group based classification and a recording medium recording the same are provided to recognize the hand signal according to a hand pose of a user by extracting a characteristic from a queried hand image, classifying the characteristic into a small group, and recognizing the hand pose within the classified small group. CONSTITUTION: If the queried hand image is inputted, the first characteristic value is extracted from the queried hand image. The queried hand image is compared with a representative hand pose model image of each small group. The small group including the queried hand image is recognized by using the first characteristic value and an image comparing result value. The second characteristic value of the queried hand image is extracted. A hand pose model matched with the queried hand image is recognized by comparing the second characteristic value with the hand pose models included in the small group, and the hand signal assigned to the hand pose model is recognized.
Abstract translation: 目的：提供一种通过基于小组的分类来识别手势的方法和记录该手势的记录介质，以通过从查询的手图像中提取特征来识别手势信号，根据用户的手势来对手势进行识别， 分成一小组，并识别分类小组内的手势。 构成：如果输入查询的手图像，则从查询的手图像中提取第一特征值。 将查询的手部图像与每个小组的代表性手部姿势模型图像进行比较。 通过使用第一特征值和图像比较结果值来识别包括查询的手图像的小组。 提取查询到的手部图像的第二特征值。 通过将第二特征值与包括在小组中的手部姿势模型进行比较来识别与查询的手部图像匹配的手部姿势模型，并且识别分配给手部姿势模型的手部信号。
-
公开(公告)号：KR1020040042002A
公开(公告)日：2004-05-20
申请号：KR1020020070120
申请日：2002-11-12
Applicant: 한국과학기술원
IPC: G06F19/00
Abstract: PURPOSE: A method for recognizing a hand signal by a small group based classification and a recording medium recording the same are provided to recognize the hand signal according to a hand pose of a user by extracting a characteristic from a queried hand image, classifying the characteristic into a small group, and recognizing the hand pose within the classified small group. CONSTITUTION: If the queried hand image is inputted, the first characteristic value is extracted from the queried hand image. The queried hand image is compared with a representative hand pose model image of each small group. The small group including the queried hand image is recognized by using the first characteristic value and an image comparing result value. The second characteristic value of the queried hand image is extracted. A hand pose model matched with the queried hand image is recognized by comparing the second characteristic value with the hand pose models included in the small group, and the hand signal assigned to the hand pose model is recognized.
Abstract translation: 目的：提供一种通过基于小组的分类来识别手信号的方法和记录该信号的记录介质的方法，用于通过从查询的手图像中提取特征来根据用户的手姿态来识别手信号， 进入一个小组，并认可在分类的小组内的手姿势。 构成：如果输入查询的手图像，则从查询的手图像中提取第一特征值。 将查询的手图像与每个小组的代表性手姿态模型图像进行比较。 通过使用第一特征值和图像比较结果值来识别包括查询的手图像的小组。 提取查询手图像的第二个特征值。 通过将第二特征值与包括在小组中的手姿态模型进行比较来识别与查询的手图像相匹配的手姿势模型，并且识别分配给手姿势模型的手信号。
-
公开(公告)号：KR1020030037692A
公开(公告)日：2003-05-16
申请号：KR1020010068528
申请日：2001-11-05
Applicant: 한국과학기술원
IPC: H04Q9/00
Abstract: PURPOSE: A remote control system and method using hand indication recognition are provided to drive electronic devices by tracking/recognizing human's hand gesture by using cameras. CONSTITUTION: A data obtaining module(110) includes plural cameras(250a,250b,250c) for recognizing a hand direction of a user. Each camera includes a pan/tilt module. An image processing board(240) and a main computer(220) process color image data obtained by the cameras. A pan/tilt controller(230) drives the pan/tilt module upon receiving a driving command from the main computer(220). The main computer(220) performs calculations made from a pre-processing module(120) and a hand indication recognition module(140). An electronic device control module(150) drives electronic devices(200a,200b) indicated by the user through a remote controller board(210).
Abstract translation: 目的：通过使用相机来跟踪/识别人的手势，提供使用手指识别的遥控系统和方法来驱动电子设备。 构成：数据获取模块（110）包括用于识别用户手指方向的多个照相机（250a，250b，250c）。 每个相机包括一个云台模组。 图像处理板（240）和主计算机（220）处理由相机获得的彩色图像数据。 在从主计算机（220）接收到驾驶命令时，摇摄/倾斜控制器（230）驱动摇摄/倾斜模块。 主计算机（220）执行由预处理模块（120）和手指示识别模块（140）进行的计算。 电子设备控制模块（150）通过遥控器板（210）驱动由用户指示的电子设备（200a，200b）。
-
-
-
-
-
-
-
-
-
Search Results
13
records
(took 0.03 seconds)
