-
公开(公告)号：KR1020040035390A
公开(公告)日：2004-04-29
申请号：KR1020020064520
申请日：2002-10-22
Applicant: 한국전자통신연구원
IPC: H04N19/139 , H04N19/51
CPC classification number: H04N5/145 , H04N19/533 , H04N19/57
Abstract: PURPOSE: A motion estimation device using an adaptive pattern in a video image compression system is provided to determine a pattern used for searching based on a previous searching result, and to adaptively determine a horizontal search pattern and a vertical search pattern, thereby improving a calculation amount and calculation speed as improving accuracy in motion estimation. CONSTITUTION: A motion estimation portion(807) transmits a driving signal to a pattern determining portion(806). The pattern determining portion(806) applies the driving signal to a pattern memory portion(805). The pattern memory portion(805) receives a search pattern. The search pattern transmitted to the pattern memory portion(805) is retransmitted to the motion estimation portion(807) for performing a searching process. The motion estimation portion(807) transmits a search pattern to be used in a next step to the pattern determining portion(806) and the pattern memory portion(805) through the driving signal according to a searched result.
Abstract translation: 目的：提供一种在视频图像压缩系统中使用自适应模式的运动估计装置，用于基于先前的搜索结果确定用于搜索的模式，并且自适应地确定水平搜索模式和垂直搜索模式，从而改进计算 数量和计算速度提高了运动估计的精度。 构成：运动估计部分（807）将驱动信号发送到模式确定部分（806）。 图案确定部分（806）将驱动信号施加到图案存储部分（805）。 图案存储部（805）接收搜索图案。 发送到模式存储部分（805）的搜索模式被重新发送到用于执行搜索处理的运动估计部分（807）。 运动估计部分（807）根据搜索结果通过驱动信号将下一步骤中使用的搜索模式发送到模式确定部分（806）和模式存储部分（805）。
-
公开(公告)号：KR1020040006611A
公开(公告)日：2004-01-24
申请号：KR1020020040932
申请日：2002-07-13
Applicant: 한국전자통신연구원
CPC classification number: G06K9/00187 , G06F17/30943 , G06K9/00181
Abstract: PURPOSE: A device and a method for certifying an online signature using mapping between extreme points are provided to verify identification of a subscriber by detecting a maximum point and a minimum point of a horizontal/vertical direction expressed on a locus of the inputted signature, detecting a matched sector through the mapping between the extreme points matched with each other, and judging agreement of the sector. CONSTITUTION: A signature data input part(10) reads the locus of the user's signature as a sequence of the points sampled by a same time interval. A characteristic extracting part(20) extracts the extreme points from the sequence. A matching relation generating part(30) generates a mapping relation between an extreme point list of the registered signature and the extracted signature. A similarity calculating part(40) calculates similarity between the matched sectors of the registered signature and the inputted signature. A judging part(50) judges the agreement of the registered signature and the inputted signature based on the similarity.
Abstract translation: 目的：提供使用极点之间的映射来验证在线签名的设备和方法，以通过检测在输入的签名的轨迹上表示的水平/垂直方向的最大点和最小点来检验用户的识别，检测 通过匹配的极值点之间的映射，并判断该部门的协议，匹配的部门。 构成：签名数据输入部分（10）将用户签名的轨迹作为以相同时间间隔采样的点的序列读取。 特征提取部分（20）从序列中提取极值点。 匹配关系生成部（30）生成登记签名的极点列表与提取的签名之间的映射关系。 相似度计算部（40）计算登记签名的匹配扇区与输入的签名的相似度。 判断部（50）基于相似度来判断登记的签名的一致性和输入的签名。
-
公开(公告)号：KR1020030050617A
公开(公告)日：2003-06-25
申请号：KR1020010081102
申请日：2001-12-19
Applicant: 한국전자통신연구원
IPC: G06T7/00
Abstract: PURPOSE: A system for extracting a face color area using edge and color information, and a method thereof are provided to improve the accuracy of face authentication by preventing a face area from merging with a background area. CONSTITUTION: A color area dividing unit(10) detects edge information from an input color image and divides areas from an entire image by the edge information. An isolated area forming unit(11) makes a background area and a face area binary for forming each isolated area. A skin color area selecting unit(12) extracts the isolated area adjacent to a center position of an image in the isolated areas for selecting an area satisfying conditions of mean gray level, size information, and color distribution as a representative face candidate area. An image merging unit(13) extracts the areas satisfying the conditions of the color and position information from the adjacent isolated areas based on the mean color distribution and position information of the selected representative face candidate area for merging images, so that the image merging unit extracts a face color area.
Abstract translation: 目的：提供使用边缘和颜色信息提取面部颜色区域的系统及其方法，其方法是通过防止面部区域与背景区域合并来提高面部认证的准确性。 构成：颜色区域划分单元（10）检测来自输入彩色图像的边缘信息，并且将来自整个图像的区域除以边缘信息。 隔离区域形成单元（11）使得形成每个隔离区域的背景区域和面部区域成为二进制形式。 皮肤颜色区域选择单元（12）提取与分离区域中的图像的中心位置相邻的隔离区域，用于选择满足平均灰度级别，大小信息和颜色分布条件的区域作为代表面部候选区域。 图像融合单元（13）基于用于合并图像的所选择的代表面部候选区域的平均颜色分布和位置信息，提取满足来自相邻隔离区域的颜色和位置信息的条件的区域，使得图像合成单元 提取面部颜色区域。
-
公开(公告)号：KR100353842B1
公开(公告)日：2002-09-26
申请号：KR1019990061217
申请日：1999-12-23
Applicant: 한국전자통신연구원
IPC: G06T1/00
Abstract: 본발명은쿼시(Quasi) 3차마르코프랜덤필드(MRF)를이용한영역분할영상후처리방법및 상기방법을실현시키기위한프로그램을기록한컴퓨터로읽을수 있는기록매체에관한것으로서, 영상처리시스템에서쿼시(Quasi) 3차마르코프랜덤필드(MRF)의통계학적접근을통해영역분할(Segmentation)의오분류(Misclassification) 에러를줄이기위한영역분할영상후처리방법및 상기방법을실현시키기위한프로그램을기록한컴퓨터로읽을수 있는기록매체를제공하기위하여, 영상처리시스템에서영역분할된영상의후처리방법에있어서, 영역분할된영상에서보다많은주변픽셀들의영역레벨정보를얻기위하여, 쿼시(Quasi) 3차마르코프랜덤필드(MRF)를사용하여작은윈도우로구성된이웃시스템을정의하는제 1 단계; 상기쿼시 3차마르코프랜덤필드(MRF)에의해정의된상기이웃시스템의중심픽셀과주변픽셀들과의관계를각기다른영역레벨의조건하에서각각의조건확률함수로표현하고, 그중 가장높은확률을제공하는영역레벨을해당윈도우의중심픽셀의영역레벨로보정하는제 2 단계; 및각 윈도우를서로중첩시켜에러를포함하고있는영역분할된전체영상을보정하는제 3 단계를포함하되, 상기이웃시스템은 2차 MRF에해당하는 8개의픽셀과 3차 MRF 이웃시스템에해당하는 5×5 윈도우의외곽픽셀 16개중 중심픽셀로부터수평, 수직그리고 2개의대각선방향에위치하여있는 8개의픽셀로구성되며, 영상분할및 분류처리등에이용됨.
-
公开(公告)号：KR100349673B1
公开(公告)日：2002-08-22
申请号：KR1019990061254
申请日：1999-12-23
Applicant: 한국전자통신연구원
IPC: G06F17/30
Abstract: 본발명은고차원영상특징량의대표값을이용한영상검색방법에관한것으로, 고차원의특징량벡터로부터그 특징량을대표할수 있는대표값을추출하여, 검색대상이되는데이터베이스의특징량범위를한정함으로써, 효율적인검색속도및 검색효율을제공할수 있는고차원영상특징량의대표값을이용한영상검색방법을제공하기위하여, 특징추출부가질의영상특징량을추출하는제 1 단계; 검색범위한정부가상기질의영상특징량을수신하여적어도하나이상의대표값을추출하는제 2 단계; 상기검색범위한정부가상기질의영상과유사도계산을위해상기대표값으로부터소정범위내의대표값을적어도하나이상의후보집합을선정하는제 3 단계; 상기검색범위한정부가상기적어도하나이상의후보집합을비트배열의형태로구성하는제 4 단계; 상기검색범위한정부가복수의비트배열로구성된적어도하나이상의후보집합에비트와이즈앤드연산을수행하여, 최종검색후보집합을구성하는제 5 단계; 및유사도평가부가상기질의영상특징량과최조검색후보집합의특징량의유사도를평가하여질의영상에대응되는영상을검색하는제 6 단계를포함하며, 영상검색시스템등에이용됨.
-
Patent Agency Ranking
公开(公告)号：KR1020000039197A
公开(公告)日：2000-07-05
申请号：KR1019980054452
申请日：1998-12-11
Applicant: 한국전자통신연구원
IPC: H04N7/087
CPC classification number: H04N7/0255 , G06K9/3258
Abstract: PURPOSE: An apparatus for separating text region using geographic text characteristics information and a method therefor are provided to detect characteristic points of texts and collect the detected points. CONSTITUTION: An apparatus for separating text region using geographic text characteristics information includes a geographic text characteristics separator/candidate region separator(110) and a candidate region prover(120). The geographic text characteristics separator/candidate region separator(110) detects the geographic characteristics of text of an input image from outside and defines the candidate region. The candidate region prover(120) increases the size of the candidate region in the geographic text characteristics separator/candidate region separator(110) proves whether the region is a real text region.
Abstract translation: 目的：提供一种使用地理文本特征信息分离文本区域的方法及其方法，用于检测文本的特征点并收集检测到的点。 构成：使用地理文本特征信息分离文本区域的装置包括地理文本特征分隔符/候选区域分隔符（110）和候选区域证明器（120）。 地理文本特征分离器/候选区域分离器（110）从外部检测输入图像的文本的地理特征并定义候选区域。 候选区域识别器（120）增加地理文本特征分离器/候选区域分离器（110）中的候选区域的大小，证明该区域是否是真实文本区域。
-
-
-
-
-
-
Search Results
183
records
(took 0.02 seconds)
