公开(公告)号：KR1020050024104A

公开(公告)日：2005-03-10

申请号：KR1020030061915

申请日：2003-09-04

Applicant: 한국전자통신연구원

Inventor： 김욱중 ,  기명석 ,  김규헌 ,  김진웅

IPC: H04N13/02

CPC classification number: H04N19/597

Abstract: PURPOSE: A system and a method for serving multi-visual point video based on the object are provided to supply services of resolution and image quality differentiated by a visual point to a user by dividing a panoramic image into several parts and then objectifying them and compressively coding them. CONSTITUTION: An image generation part(100) generates a panoramic image from a plurality of partial images. An image objectification part(200) objectifies the panoramic image generated by the image generation part and then compressively codes it by each object image. The image generation part includes an image obtaining part(110) for obtaining the partial images that correspond to parts of an object, respectively. The image generation part further includes a mosaic process part(120) for coupling the partial images to each other and thereby generating the panoramic image.

Abstract translation: 目的：提供一种基于该对象服务多视点视频的系统和方法，用于通过将全景图像划分成若干部分，然后对其进行客观化和压缩来向用户提供由视觉点区分的分辨率和图像质量的服务 编码它们。 构成：图像生成部（100）从多个部分图像生成全景图像。 图像对象部分（200）将由图像生成部生成的全景图像对象化，然后对每个对象图像进行压缩编码。 图像生成部分包括用于分别获得对应于对象部分的部分图像的图像获取部分（110）。 图像生成部还包括用于将部分图像彼此耦合并从而​​生成全景图像的马赛克处理部分（120）。

公开(公告)号：KR100466587B1

公开(公告)日：2005-01-24

申请号：KR1020020070773

申请日：2002-11-14

Applicant: 한국전자통신연구원

Inventor： 정세윤 ,  김규헌

IPC: G06T1/00

Abstract: PURPOSE: A method for extracting camera information for a synthesized image contents producing tool is provided to extract camera information from previously photographed motion pictures and easily produce synthesized contents. CONSTITUTION: A motion picture from which camera information will be extracted is opened(301). The first frame of the motion picture is used as the first image required for obtaining initial camera information(303), and a user directly selects the second image required for obtaining the initial camera information from the motion picture(304). The camera information is extracted using coordinates of base characteristic points with respect to the first and second images(306). The camera information is extracted using projection points of corner points of virtual objects with respect to the first and second images(308). The extracted camera information is stored(309). Varied coordinates of characteristic points and virtual object projection points are tracked for other frames of the opened motion picture to extract camera information.

Abstract translation: 目的：提供一种用于提取合成图像内容产生工具的相机信息的方法，以从先前拍摄的运动图像中提取相机信息并容易地产生合成内容。 构成：打开从其中提取相机信息的运动图像（301）。 运动图像的第一帧被用作获得初始摄像机信息所需的第一图像（303），并且用户直接从运动图像（304）中选择获得初始摄像机信息所需的第二图像。 使用相对于第一和第二图像的基本特征点的坐标提取相机信息（306）。 使用虚拟对象的角点相对于第一和第二图像的投影点提取相机信息（308）。 所提取的相机信息被存储（309）。 跟踪打开的运动图像的其他帧的特征点和虚拟对象投影点的变化的坐标以提取相机信息。

公开(公告)号：KR1020040052330A

公开(公告)日：2004-06-23

申请号：KR1020020080194

申请日：2002-12-16

Applicant: 한국전자통신연구원

Inventor： 김현철 ,  김규헌 ,  김진웅

IPC: H04N21/60 ,  H04N21/2343

Abstract: PURPOSE: A transmission/reception system of an interactive contents for performing a conversation with a low delay and the method thereof are provided to store an object for storing a buffer and use the stored object on demand, thereby capable of minimizing a possibility of delay in a conversation with a user. CONSTITUTION: The interactive contents transmission apparatus(10) comprises an interactive contents input unit(11) for receiving interactive contents inputted from the outside. An interactive contents processing unit(12) separates interactive contents inputted from the interactive contents input unit(11) into object/screen information and each object and analyzes the interactive contents. A storage object discriminating unit(13) determines objects which will be previously stored to a buffer of a contents receiving apparatus(20) among the objects constructing the interactive contents. A transmission/reception unit(14) transmits object/scene information and each object constructing the interactive contents to a receiving apparatus and receives a object requisition message from the receiving apparatus by a user or time information.

Abstract translation: 目的：提供用于执行低延迟的对话的交互内容的发送/接收系统及其方法，用于存储用于存储缓冲器的对象并且按需使用所存储的对象，从而能够最小化延迟的可能性 与用户的对话。 交互式内容发送装置（10）包括：交互式内容输入单元（11），用于接收从外部输入的交互内容。 交互式内容处理单元（12）将从交互内容输入单元（11）输入的交互内容分离成对象/屏幕信息和每个对象，并分析交互内容。 存储对象识别单元（13）将构成交互内容的对象中预先存储在内容接收装置（20）的缓冲器中的对象确定。 发送/接收单元（14）将对象/场景信息和构成交互内容的每个对象发送到接收设备，并由用户或时间信息从接收设备接收对象请求消息。

公开(公告)号：KR1020040041979A

公开(公告)日：2004-05-20

申请号：KR1020020070092

申请日：2002-11-12

Applicant: 한국전자통신연구원

Inventor： 정원식 ,  유영재 ,  김현철 ,  기명석 ,  정예선 ,  이인재 ,  김규헌 ,  김진웅 ,  정세윤

IPC: G11B27/031 ,  H04N21/431

Abstract: PURPOSE: A GUI(Graphic User Interface) is provided to perform editing and making by considering respective materials existing in a moving image, an audio, and a still image as one object, reedit an edited/made result, and control the object interactively depending on user's request. CONSTITUTION: The first interface(110) inserts each object, and displays/edits an attribute of each object. The second interface(120) displays/edits a time/spatial layout of each object. The third interface(130) displays/edits the user interactive. The fourth interface(140) displays/edits a logical value and a field value of a node forming the currently made/edited object. A data access application program interface exchanges the internal database structure storing the editing/making information of the object, and the making information between the first to the fourth interfaces.

Abstract translation: 目的：提供GUI（图形用户界面），通过将运动图像，音频和静止图像中存在的各种材料作为一个对象进行编辑和制作，重新编辑编辑/制作的结果，并根据 根据用户的要求 构成：第一个界面（110）插入每个对象，并显示/编辑每个对象的属性。 第二接口（120）显示/编辑每个对象的时间/空间布局。 第三接口（130）显示/编辑用户交互。 第四接口（140）显示/编辑形成当前制作/编辑对象的节点的逻辑值和字段值。 数据访问应用程序接口交换存储对象的编辑/制作信息的内部数据库结构以及第一至第四接口之间的制作信息。

公开(公告)号：KR1020030056276A

公开(公告)日：2003-07-04

申请号：KR1020010086473

申请日：2001-12-28

Applicant: 한국전자통신연구원

Inventor： 김현철 ,  정원식 ,  김규헌 ,  김진웅

IPC: H04N19/68

CPC classification number: H04N19/68 ,  H04N19/91

Abstract: PURPOSE: A video encoding system and method are provided to accurately detect an error generated in the event of encoding to enable video communication with high picture quality. CONSTITUTION: An image applied to a video encoding system is split into a plurality of blocks and a number is given to each of the blocks(S202). A specific rule for inserting a resynchronization code into the split blocks is designated according to encoding information of the image. The resynchronization code is inserted into the blocks according to the designated rule to encode the blocks(S207). The encoding information includes the bit rate and frame rate of the image, the size of the image, and a video packet size.

Abstract translation: 目的：提供一种视频编码系统和方法，用于准确检测在编码时产​​生的错误，以实现具有高图像质量的视频通信。 构成：将应用于视频编码系统的图像分割成多个块，向每个块赋予数（S202）。 根据图像的编码信息指定用于将再同步码插入到分割块中的具体规则。 根据指定的规则将重新同步码插入到块中以对块进行编码（S207）。 编码信息包括图像的比特率和帧速率，图像的大小和视频分组大小。

公开(公告)号：KR1020030043612A

公开(公告)日：2003-06-02

申请号：KR1020020053957

申请日：2002-09-06

Applicant: 한국전자통신연구원 ,  학교법인경희학원

Inventor： 정원식 ,  김규헌 ,  박광훈 ,  이윤진 ,  김진웅 ,  안치득

IPC: H04N19/154 ,  H04N19/167

CPC classification number: H04N19/34 ,  H04N19/112 ,  H04N19/126 ,  H04N19/129 ,  H04N19/154 ,  H04N19/176 ,  H04N19/184 ,  H04N19/46 ,  H04N19/61

Abstract: PURPOSE: A smart water ring scanning system and method using a picture quality improving factor and a video coding/decoding system and method using the scanning system and method are provided to offer images suitable for the human visual system. CONSTITUTION: A water ring scanning system applied to a video encoding/decoding system includes a water ring generation position determining unit(61), a water ring position determining unit(69), and a water ring quality factor determining unit(63). The water ring generation position determining unit determines a water ring generation position in a video frame. The water ring position determining unit decides the i-th water ring position surrounding the water ring generation position. The water ring picture factor determining unit determines a water ring quality factor according to image region significance of the water ring generation position and the water ring position.

Abstract translation: 目的：提供使用图像质量改进因子的智能水环扫描系统和方法以及使用扫描系统和方法的视频编码/解码系统和方法来提供适合于人类视觉系统的图像。 构成：应用于视频编码/解码系统的水环扫描系统包括水环产生位置确定单元（61），水环位置确定单元（69）和水环质量因子确定单元（63）。 水环产生位置确定单元确定视频帧中的水环产生位置。 水环位置确定单元决定围绕水环产生位置的第i个水环位置。 水环图像因子确定单元根据水环产生位置和水环位置的图像区域有效性来确定水环质量因子。

公开(公告)号：KR100385864B1

公开(公告)日：2003-06-02

申请号：KR1019990025033

申请日：1999-06-28

Applicant: 한국전자통신연구원

Inventor： 김규헌 ,  정세윤 ,  전병태 ,  이재연 ,  배영래

IPC: G06F17/30

Abstract: PURPOSE: A texture-based searching method with co-occurrence matrix and normalized Euclidean distance is provided to search a moving image effectively based on co-occurrence matrix and normalized Euclidean distance. CONSTITUTION: A texture feature vector is obtained through co-occurrence matrix of a selected distance of a sample image. The input image is analyzed based on a window. A texture feature is obtained by the co-occurrence matrix of the analyzed window-based image. A normalized Euclidean distance between the texture feature vectors of the selected distance and the window is obtained. The normalized Euclidean distance is compared with a specific value. Finally, the current texture is recognized as a different one if the normalized Euclidean distance is larger or the same as the specific value. Otherwise, the texture is the same.

Abstract translation: 目的：提出一种基于共生矩阵和归一化欧几里德距离的基于纹理的搜索方法，以有效地搜索运动图像。 构成：纹理特征向量是通过样本图像的选定距离的共现矩阵获得的。 输入图像基于窗口进行分析。 通过分析的基于窗口的图像的共现矩阵获得纹理特征。 获得所选距离的纹理特征向量与窗口之间的归一化欧几里德距离。 将归一化的欧几里德距离与特定值进行比较。 最后，如果归一化的欧几里德距离大于或等于特定值，则当前纹理被识别为不同的纹理。 否则，纹理是相同的。

公开(公告)号：KR1020020003249A

公开(公告)日：2002-01-12

申请号：KR1019990056216

申请日：1999-12-09

Applicant: 한국전자통신연구원

Inventor： 정세윤 ,  김규헌 ,  전병태 ,  이재연 ,  배영래

IPC: G06T7/00

Abstract: PURPOSE: A color image search system using wavelet conversion and a method for the same are provided to improve a search speed of a color image by using a wavelet conversion method. CONSTITUTION: A graphic user interface(100) is used for controlling a total flow chart of a system. A preprocessor(102) performs a normalization and color space conversion process for an input image and an inquiry image. A characteristic extractor(104) performs a wavelet conversion process and calculates an average and a standard deviation of the first band and a centroid of gravity of neighboring bands. A searcher(106) receives characteristic data of the inquiry image and performs a search process. An average and standard deviation characteristic data manager(108) stores the average and the standard deviation data. A gravity centroid characteristic data manager(110) stores the gravity centroid characteristic data. An image database(112) stores the input image, receives a result signal of the searcher(106), and transmits the corresponding image to the graphic user interface(100). A display device(114) displays an index processing procedure and a search processing procedure received from the graphic user interface(100).

Abstract translation: 目的：提供使用小波变换的彩色图像搜索系统及其方法，以通过使用小波变换方法来提高彩色图像的搜索速度。 规定：图形用户界面（100）用于控制系统的总流程图。 预处理器（102）对输入图像和查询图像执行归一化和颜色空间转换处理。 特征提取器（104）执行小波变换处理，并计算第一频带的平均值和标准偏差以及相邻频带的重心质心。 搜索器（106）接收查询图像的特征数据并进行搜索处理。 平均和标准差特征数据管理器（108）存储平均值和标准偏差数据。 重力中心特征数据管理器（110）存储重力中心特征数据。 图像数据库（112）存储输入图像，接收搜索器（106）的结果信号，并将对应的图像发送到图形用户界面（100）。 显示装置（114）显示从图形用户界面（100）接收的索引处理过程和搜索处理过程。

公开(公告)号：KR100315426B1

公开(公告)日：2001-11-29

申请号：KR1019990061895

申请日：1999-12-24

Applicant: 한국전자통신연구원

Inventor： 전병태 ,  정세윤 ,  김규헌 ,  이재연 ,  배영래

IPC: H04N5/445 ,  H04N7/08

Abstract: 본발명은블록매칭기법을이용한자막영역의원영상복원장치및 그방법과기록매체에관한것으로, 영상처리시스템에서내용기반영상처리를위해정지영상, 동영상등과같은영상에서자막영역의원영상을복원하기위한블록매칭기법을이용한자막영역의원영상복원장치및 그방법과기록매체를제공하기위하여, 영상으로부터영상장면전환정보를추출하는제 1 단계; 영상으로부터복원될자막정보(자막의시작프레임, 끝프레임, 자막문자의구성요소단위)를추출하는제 2 단계; 추출된영상장면전환정보와추출된자막정보의자막의시작프레임및 끝프레임을이용하여복원의방향성을결정하는제 3 단계; 및결정된복원의방향성에따라, 각프레임마다추출된자막정보의문자의구성요소단위로원영상과블록매칭을수행하여자막영역을원영상으로복원하는제 4 단계를포함하며, 멀티미디어영상처리분야등에이용됨.

公开(公告)号：KR101878218B1

公开(公告)日：2018-07-16

申请号：KR1020120024671

申请日：2012-03-09

Applicant: 한국전자통신연구원 ,  경희대학교 산학협력단

Inventor： 임현정 ,  권형진 ,  정광희 ,  이광순 ,  윤국진 ,  정원식 ,  허남호 ,  김병철 ,  박종환 ,  김규헌 ,  서덕영 ,  박광훈

IPC: H04N21/43 ,  H04N13/00 ,  H04L29/06

Abstract: 본발명은 MPEG-2 TS 기반방송서비스환경에서전달시간이다른콘텐츠를연동하여 3D 서비스를제공하기위한시그널링기술방법및 동기화방법으로서, 본발명에따른실시간, 비실시간연동형서비스기술에의하면기존방송과저장된파일의연동형미디어를도출함으로써, 몰입형멀티미디어의접근이쉬워지며, 다양한저장매체를활용하는서비스의기반기술로서활용될수 있으며, 본발명에서개발할 NRT저장콘텐츠와실시간방송을연동하여 3D 서비스를제공하기위한방법은향후, 두개의기술규격뿐만아니라, 다양한저장매체를활용하는연동형서비스기술개발의방향성을제시할것으로기대할수 있다.