公开(公告)号：KR1020120072291A

公开(公告)日：2012-07-03

申请号：KR1020110007484

申请日：2011-01-25

Applicant: 전남대학교산학협력단

Inventor： 양형정 ,  김선희 ,  캄스위

IPC: G06F19/00 ,  A61B5/0476

CPC classification number: G06F17/16 ,  A61B5/0476

Abstract: PURPOSE: A method for replacing missing values which uses an EM(Expectation Maximization) PCA is provided to maximize data accuracy and reduce computation complexity by measuring incremental EM PCA. CONSTITUTION: An input vector is received. If the input vector includes a value, the value is initialized as average of data matrix. A PCA(Principle Component Analysis) is performed about the data matrix. The data matrix is re-configured. A mixing value is replaced with the missing value of the reconfigured data matrix. An error rate about the data matrix is performed until error rate is less than a critical value.

Abstract translation: 目的：提供一种用于替换使用EM（期望最大化）PCA的缺失值的方法，以通过测量增量EM PCA来最大化数据精度并降低计算复杂度。 构成：接收输入向量。 如果输入向量包含一个值，则该值被初始化为数据矩阵的平均值。 对数据矩阵进行PCA（原理分量分析）。 重新配置数据矩阵。 混合值被替换为重新配置的数据矩阵的缺失值。 执行关于数据矩阵的错误率，直到错误率小于临界值。

公开(公告)号：KR1020110028034A

公开(公告)日：2011-03-17

申请号：KR1020090085937

申请日：2009-09-11

Applicant: 삼성전자주식회사 ,  전남대학교산학협력단

Inventor： 황성택 ,  안희범 ,  이귀상 ,  김수형 ,  양형정 ,  박종현 ,  임준식 ,  김정식 ,  이명훈 ,  정종문

IPC: G06T7/40 ,  G06F17/30

CPC classification number: G06K9/3258 ,  G06K2209/01

Abstract: PURPOSE: A label search method and a device thereof are provided to search corresponding wind label by using a wine label image which is obtained through a camera which is included in a portable terminal. CONSTITUTION: A label database(50) includes a label and information about label. A character recognition(20) outputs an recognition result by recognizing an extracted character. A research unit(10) comprises a preparatory label candidate including detected label. An image comparison unit(30) comprises a final label candidate. The character recognition u nit includes a character area detecting module, a character division module, and a character recognition module.

Abstract translation: 目的：提供标签搜索方法及其装置，通过使用通过便携式终端中包括的相机获得的酒标图像来搜索对应的风标。 规定：标签数据库（50）包括标签和关于标签的信息。 字符识别（20）通过识别提取的字符来输出识别结果。 研究单元（10）包括包括检测到的标签的预备标签候选。 图像比较单元（30）包括最终标签候选。 字符识别包括字符区域检测模块，字符分割模块和字符识别模块。

公开(公告)号：KR1020090130651A

公开(公告)日：2009-12-24

申请号：KR1020080056371

申请日：2008-06-16

Applicant: 전남대학교산학협력단

Inventor： 양형정 ,  정종문 ,  오진식

IPC: G06F3/03 ,  G06F3/02 ,  G06F3/14 ,  G06F1/00

CPC classification number: G06F3/011 ,  G06F1/1601 ,  G06F3/016 ,  G06F3/0304

Abstract: PURPOSE: An AR(Augmented Reality) providing system of a table top structural is provided to extend a range of an application of a table top system by using various physical input devices. CONSTITUTION: A haptic unit(100) includes a housing having a shape, which can be reeled from a user. The haptic unit radiates an optical position detection signal to a bottom surface of the housing. The haptic unit includes a vibrator for realizing AR based on a displacement vector of the optical signal. A display panel of a display control unit(200) displays video information. The display control unit generates collision information between the video information and vector information for the displacement vector. The display control unit computes a collision vector of the haptic unit based on the collision information. The display control unit transmits collision vector data to the vibrator.

Abstract translation: 目的：提供一种提供桌面结构的AR（增强现实）提供系统，通过使用各种物理输入设备来扩展桌面系统的应用范围。 构成：触觉单元（100）包括具有能够从使用者卷取的形状的壳体。 触觉单元将光学位置检测信号发射到壳体的底表面。 触觉单元包括用于基于光信号的位移矢量实现AR的振动器。 显示控制单元（200）的显示面板显示视频信息。 显示控制单元在视频信息和位移矢量的向量信息之间产生冲突信息。 显示控制单元基于碰撞信息来计算触觉单元的碰撞矢量。 显示控制单元将碰撞矢量数据发送到振动器。

公开(公告)号：KR1020090075287A

公开(公告)日：2009-07-08

申请号：KR1020080001077

申请日：2008-01-04

Applicant: 전남대학교산학협력단

Inventor： 양형정 ,  오진식 ,  정종문 ,  김정식

IPC: G06F3/03 ,  G06F3/041

CPC classification number: G06F3/0416 ,  G06F3/0304 ,  G06F3/0412 ,  G06F3/042

Abstract: An LED(Light Emitting Diode) base object control system for a table top display for supplying the improved accuracy of input signal recognition through an ultraviolet ray emitting intensity measurement is provided to sense the input signal from the outside without additional work by including an infrared ray emitting diode and a switch. A radiated intensity measuring unit(220) generates radiated intensity information by measuring the transition change of the radiated ratio by unit stereo angle. If the value of the radiant intensity information is changed, a switch(240) authorizes the radiant intensity information to a controller(250). According to the radiant intensity information applied from the switch, the controller produces the external input information.

Abstract translation: 提供了一种用于通过紫外线发射强度测量提供输入信号识别的改进精度的用于桌面显示器的LED（发光二极管）基础对象控制系统，用于通过包括红外线来检测来自外部的输入信号而无需额外的工作 发光二极管和开关。 辐射强度测量单元（220）通过以单位立体角测量辐射比的转变变化来产生辐射强度信息。 如果辐射强度信息的值改变，则开关（240）授权辐射强度信息给控制器（250）。 根据从开关施加的辐射强度信息，控制器产生外部输入信息。

公开(公告)号：KR100882275B1

公开(公告)日：2009-02-06

申请号：KR1020070050761

申请日：2007-05-25

Applicant: 전남대학교산학협력단

Inventor： 양형정 ,  김정식 ,  이형재

IPC: G06K9/46 ,  G06K9/80 ,  G06K9/00 ,  G06F19/28

Abstract: 본 발명은 얼굴영상을 이용한 질병분류시스템에 관한 것이다. 본 발명에 따른 얼굴영상을 이용한 질병분류시스템은, 얼굴사진을 로딩하는 얼굴이미지 로딩모듈과, 로딩된 얼굴사진의 각 픽셀별 색차신호값들의 비교를 통해 얼굴이미지를 추출하는 얼굴영역 추출모듈과, 추출된 얼굴이미지를 소정 개수의 타일로 분할하는 얼굴영역 분할모듈과, 분할된 얼굴이미지별 휘도값, 색차신호값, 색상값, 채도값 및 명도값들의 평균과 표준편차들을 해당 얼굴이미지들의 특징값으로 추출하는 특징추출모듈과, 추출된 얼굴이미지별 특징값을 정규화하는 정규화모듈 및 정규화모듈로부터 정규화된 얼굴이미지를 인가받고, 질병정보DB를 참조하여 다수의 질병정보 특징값을 각기 SVM에 인가하여 사용자의 질병을 분류하는 질병분류모듈을 포함한다.
얼굴영역, 얼굴이미지, 분할, SVM, 질병분류

公开(公告)号：KR1020080074627A

公开(公告)日：2008-08-13

申请号：KR1020070013993

申请日：2007-02-09

Applicant: 전남대학교산학협력단

Inventor： 양형정 ,  김경윤 ,  이형재

IPC: G06Q50/04 ,  G06F17/50 ,  G06F17/30 ,  G06F3/14

CPC classification number: G06Q50/04 ,  G06F3/14 ,  G06F17/30244 ,  G06F17/5081

Abstract: A system and a method for reusing product design information by using a 2D(Dimensional) image of a product are provided to recognize the 2D image of the product in a part unit, and use a small storage space and offer a search request quickly by storing image information in a 2D mode. An image receiving module(100) receives a 2D image. An image dividing module(200) divides the 2D image and extracts an image of parts by merging the divided images. An ARG(Attributed Rational Graph) generating module(300) generates an ARG by using information for the part images and matches the ARG with product design information. A storing module(400) stores the 2D image, the divided part image, and the ARG. An image searching module(500) searches similar images by calculating similarity between the generated ARG and the stored ARG of the images dividing the 2D image of the desired product, and detects a solution for misconnection of the part image or incorrect part information. A display module(600) outputs a result searched by comparison of the image searching module.

Abstract translation: 提供了通过使用产品的2D（维度）图像来重新使用产品设计信息的系统和方法，以识别部件单元中的产品的2D图像，并且使用小的存储空间并通过存储快速提供搜索请求 2D模式下的图像信息。 图像接收模块（100）接收2D图像。 图像分割模块（200）通过合并分割图像来划分2D图像并提取部分的图像。 ARG（Attributed Rational Graph）生成模块（300）通过使用部件图像的信息并将ARG与产品设计信息相匹配来生成ARG。 存储模块（400）存储2D图像，分割部分图像和ARG。 图像搜索模块（500）通过计算生成的ARG和分割所需产品的2D图像的图像的存储的ARG之间的相似度来搜索相似的图像，并且检测部分图像的误连接或不正确的部分信息的解决方案。 显示模块（600）通过图像搜索模块的比较输出搜索到的结果。