-
公开(公告)号：KR100965720B1
公开(公告)日：2010-06-24
申请号：KR1020080017305
申请日：2008-02-26
Applicant: 삼성전자주식회사
CPC classification number: G06T3/4038 , G06T11/00 , G06T2200/32
Abstract: 본 발명에 따른 모자이크 영상 생성방법은 영상에 대해 모자이크 처리를 수행하여 모자이크 영상을 생성하는 방법에 있어서, 입력받은 영상으로부터, 모자이크에 사용될 객체(object)를 획득하는 과정과, 획득된 상기 객체를 저장매체에 저장하는 과정과, 모자이크 처리할 영상을 입력받고, 모자이크 블록의 크기를 설정하는 과정과, 저장매체에 저장된 객체로부터 모자이크 블록에 적용될 적어도 하나의 객체를 선택하는 과정과, 설정된 모자이크 블록의 크기에 대응하여 모자이크 처리할 영상을 복수의 영역으로 분할하는 과정과, 분할된 각 영역의 색상을 상기 객체에 반영하여, 모자이크 블록으로 설정하는 과정을 포함한다.
모자이크, 객체, 히스토그램, 최대값, 대표값-
公开(公告)号：KR1020100052185A
公开(公告)日：2010-05-19
申请号：KR1020080111098
申请日：2008-11-10
Applicant: 삼성전자주식회사
CPC classification number: H04N5/3575 , H04N5/367 , H04N9/045 , H04N2209/046
Abstract: PURPOSE: A method of reducing image noise is provided, which reduces a noise by color channel sensitivity by calculating weighted mean value based on a color channel. CONSTITUTION: A sampler samples a pixel value within window, of which a preset size is preset at an image of image sensor(200). A mean value calculator(120) calculates the average of the pixel values. An edge discriminator(130) distinguishes a target pixel within an edge area. A light/darkness discriminator(140) distinguishes the target pixel within a bright area or a dark zone. A weight mean filter(160) applies a weight values to a neighbor pixel values. The weight mean filter calculates the average of weighted pixel values.
Abstract translation: 目的：提供一种降低图像噪声的方法，通过基于颜色通道计算加权平均值，减少颜色通道灵敏度的噪声。 构成：采样器对窗口内的像素值进行采样，其中预先设定的尺寸在图像传感器（200）的图像上。 平均值计算器（120）计算像素值的平均值。 边缘鉴别器（130）区分边缘区域内的目标像素。 亮/暗鉴别器（140）将目标像素区分在明亮区域或暗区域内。 权重均值滤波器（160）对相邻像素值应用权重值。 权重均值滤波器计算加权像素值的平均值。
-
公开(公告)号：KR1020100013106A
公开(公告)日：2010-02-09
申请号：KR1020080074623
申请日：2008-07-30
Applicant: 삼성전자주식회사
CPC classification number: H04N5/45 , G06F3/0481 , G06F3/0488 , G06F2203/04805 , G09G5/14 , G09G5/397 , G09G2320/103 , G09G2340/0407 , G09G2340/045 , G09G2340/0464 , G09G2340/12 , H04N5/44591 , H04N21/4316 , H04N21/440263 , H04N21/4728 , H04N2005/44595
Abstract: PURPOSE: An apparatus and a method for displaying the zoomed-displaying a target area of an reproduceding image are provided to enable a user to freely adjust a window on which a target object is displayed, thereby improving user convenience. CONSTITUTION: Generation Generated information about a window is set(305). The window is overlapped on with a reproduced image and the overlapped image is displayed. A target area is designated in the reproduced image(310). A window is generated using the generation generated information of the window(315). An image of the designated target area is displayed on the generated window.
Abstract translation: 目的：提供一种用于显示缩放显示再现图像的目标区域的装置和方法，以使用户能够自由地调整显示目标对象的窗口，从而提高用户便利性。 构成：生成有关窗口的生成信息（305）。 窗口与再现的图像重叠，并且显示重叠的图像。 在再现图像（310）中指定目标区域。 使用窗口的生成生成信息生成窗口（315）。 指定的目标区域的图像显示在生成的窗口上。
-
公开(公告)号：KR1020090097057A
公开(公告)日：2009-09-15
申请号：KR1020080022223
申请日：2008-03-10
Applicant: 삼성전자주식회사 , 홍익대학교 산학협력단
Abstract: A method for encoding and decoding connectivity information of 3D data implemented as a mesh model is provided to efficiently reduce a size of encoding data of 3D data expressed as the mesh model. A gate including the first vertex and the second vertex is selected based on information about a vertex included in geometric data(120). At least one point positioned at surroundings of the first vertex of the gate is set up as a candidate point(130). At least one candidate point is arranged based on a distance between a point included in the gate and the candidate point. Information about distance relation between the point included in the gate and the candidate point is obtained(150). Connection information of a plurality of points included in a polygon is encoded(160).
Abstract translation: 提供了一种用于编码和解码实现为网格模型的3D数据的连通性信息的方法，以有效地减小表示为网格模型的3D数据的编码数据的大小。 基于关于包括在几何数据（120）中的顶点的信息来选择包括第一顶点和第二顶点的门。 位于门的第一顶点周围的至少一个点被设置为候选点（130）。 基于门中包含的点与候选点之间的距离来布置至少一个候选点。 获得关于包括在门中的点与候选点之间的距离关系的信息（150）。 包括在多边形中的多个点的连接信息被编码（160）。
-
公开(公告)号：KR1020090092035A
公开(公告)日：2009-08-31
申请号：KR1020080017305
申请日：2008-02-26
Applicant: 삼성전자주식회사
CPC classification number: G06T3/4038 , G06T11/00 , G06T2200/32
Abstract: A method for producing a mosaic and an apparatus thereof capable of producing the mosaic image and processing a mosaic process about the image are provided to produce the mosaic image which actively reflects the demand of a user by producing the mosaic image by using an object. An object used for mosaic is obtained from an image which is input(100). The obtained object is stored in a storage media. The size of the mosaic block is set up. One or more object used as the mosaic block is selected from the stored object. The image mosaic is divided to a plurality of domain frames(400). The color of the divided angular region is reflected in object.
Abstract translation: 提供一种用于制造马赛克的方法及其制造马赛克图像的装置以及处理关于图像的拼接处理，以通过使用对象产生马赛克图像来产生主动反映用户需求的马赛克图像。 从输入（100）的图像获得用于马赛克的物体。 获得的对象被存储在存储介质中。 马赛克块的大小设置。 从存储的对象中选择用作镶嵌块的一个或多个对象。 图像拼接被划分成多个域帧（400）。 分割角区域的颜色反映在物体中。
-
Patent Agency Ranking
公开(公告)号：KR1020090087384A
公开(公告)日：2009-08-17
申请号：KR1020080012784
申请日：2008-02-12
Applicant: 삼성전자주식회사 , 전남대학교산학협력단
IPC: H04B1/40
CPC classification number: G06K9/38 , G06K9/00449 , G06K9/346
Abstract: A method for recognizing the automatic accompaniment music in a portable terminal is provided to remove the staff of a photographed image of a sheet of music without separately performing inclination compensation needing the much amount of calculation. In consideration of an area including the staff and an area including a note mark, an area in which an accompaniment code exists is detected from an image including a sheet of music(200,300). From the sheet of music inputted from the outside, the staff is extracted and removed. And then, the note mark and accompaniment code are extracted and recognized(400,500).
Abstract translation: 提供用于识别便携式终端中的自动伴奏音乐的方法，以便不需要单独执行需要很多计算量的倾斜补偿来移除一张乐曲的拍摄图像的人员。 考虑到包括工作人员在内的区域和包括音符标记的区域，从包括一张音乐的图像（200,300）中检测出存在伴奏码的区域。 从从外部输入的乐谱中，人员被抽取和移除。 然后，提取并识别音符和伴奏码（400,500）。
-
公开(公告)号：KR1020090061929A
公开(公告)日：2009-06-17
申请号：KR1020070128943
申请日：2007-12-12
Applicant: 삼성전자주식회사
CPC classification number: H04N1/41 , G09G3/2048 , G09G2320/0242 , H04N1/405 , H04N1/58
Abstract: A dithering method and an apparatus performing the method are provided to distribute an arbitrary random number used in a dithering process in the whole image, thereby minimizing the color distortion of the whole image and reducing image degradation. An image frame including plural pixels is received from the outside(51). A difference is checked between color difference signal values of a specific pixel and a pixel adjacent to the specific pixel within an inputted image(53). The specific pixel is designated as a flat region or a non-flat region in consideration of the confirmation result(55). Dithering is performed for pixels designated as the non-flat region(57,60).
Abstract translation: 提供抖动方法和执行该方法的装置，以在整个图像中分布在抖动处理中使用的任意随机数，从而最小化整个图像的颜色失真并减少图像劣化。 从外部（51）接收包括多个像素的图像帧。 在特定像素的色差信号值与与输入图像（53）内的特定像素相邻的像素之间检查差异。 考虑到确认结果，特定像素被指定为平坦区域或非平坦区域（55）。 针对指定为非平坦区域的像素进行抖动（57,60）。
-
公开(公告)号：KR100871693B1
公开(公告)日：2008-12-05
申请号：KR1020060120062
申请日：2006-11-30
Applicant: 삼성전자주식회사
Inventor： 조성대
IPC: H01L21/56 , H01L21/301
CPC classification number: H01L21/78 , H01L23/3185 , H01L23/49575 , H01L24/03 , H01L24/05 , H01L24/06 , H01L24/48 , H01L24/85 , H01L2224/04042 , H01L2224/05554 , H01L2224/05556 , H01L2224/05599 , H01L2224/45144 , H01L2224/48091 , H01L2224/48227 , H01L2224/48247 , H01L2224/48465 , H01L2224/48472 , H01L2224/48599 , H01L2224/4899 , H01L2224/85399 , H01L2924/00014 , H01L2924/01005 , H01L2924/01006 , H01L2924/01023 , H01L2924/01033 , H01L2924/01047 , H01L2924/01082 , H01L2924/10157 , H01L2924/10158 , H01L2224/45099 , H01L2924/00
Abstract: 본 발명은 반도체 칩 및 그의 제조 방법에 관한 것으로서, 더욱 구체적으로는 반도체 다이(die)의 가장자리의 적어도 일부분에 대하여, 모서리의 적어도 일부분이 절연성 물질로 이루어진 반도체 칩 및 그의 제조 방법에 관한 것이다.
본 발명의 반도체 칩 및 제조 방법에 따르면, 본딩 와이어가 반도체 칩의 표면 및/또는 모서리와 접촉함으로써 발생할 수 있는 누설 전류, 반도체 칩의 오작동 및/또는 회로의 단락과 같은 부작용을 제거하거나 줄일 수 있고, 반도체 칩을 적층시켜 구성되는 반도체 패키지의 두께를 줄일 수 있는 효과가 있다.
스크라이브 레인, 오목부, 절연성 물질, 모서리-
公开(公告)号：KR100866230B1
公开(公告)日：2008-10-30
申请号：KR1020070036056
申请日：2007-04-12
Applicant: 삼성전자주식회사 , 재단법인서울대학교산학협력재단
CPC classification number: H04N5/23238 , G03B37/04 , G06K9/6201 , G06T7/20 , H04N5/23293 , H04N5/76
Abstract: 본 발명은 파노라마 사진 촬영 방법에 있어서, 파노라마 사진을 구성하는 각 화상을 촬영시에, 노출 보정을 적용한 모션 추정 기법을 사용하여 실시간적으로 현재 입력되는 영상과 이전 영상을 비교함으로써 해당 촬영 장치의 움직임을 확인하는 과정과, 확인한 해당 촬영 장치의 움직임에 따라 촬영 방향에 따른 움직임이 미리 설정된 임계값에 도달하였음을 확인함으로써 각 화상의 촬영 시점을 판단하는 과정과, 각 화상의 촬영 시점에서 각 화상을 수동 입력에 따라, 또는 자동으로 촬영하는 과정을 포함한다.
파노라마, 움직임, 모션벡터Abstract translation: 公开了一种拍摄全景图像的方法，包括以下步骤：通过用曝光补偿通过运动估计机构比较当前实时输入图像与先前图像来识别相应拍摄装置的移动，确定拍摄每个下一张图像的时间 通过确定拍摄方向上的移动是否达到预设的阈值，并且在所确定的时间通过手动或自动操作来拍摄每个下一张照片。
-
公开(公告)号：KR1020080092578A
公开(公告)日：2008-10-16
申请号：KR1020070036056
申请日：2007-04-12
Applicant: 삼성전자주식회사 , 재단법인서울대학교산학협력재단
CPC classification number: H04N5/23238 , G03B37/04 , G06K9/6201 , G06T7/20 , H04N5/23293 , H04N5/76
Abstract: A method for photographing a panorama picture is provided to check motion information about a terminal having a camera through an image processing operation in photographing the panorama picture and determine a photographing visual point automatically, thereby enabling a user to photograph the panorama picture easily and accurately without a separate device or an additional hardware. A method for photographing a panorama picture comprises the following steps of: setting various photographing environments for photographing the panorama picture(200); starting the panorama picture photographing(210); checking a motion of a camera by using a current input image and a previous input image(220); displaying a direction in which the camera moves on a display unit(230); checking whether the camera is currently positioned in each visual point for image photographing according to the motion of the camera(240); and photographing a corresponding image(250).
Abstract translation: 提供一种用于拍摄全景图像的方法，用于通过拍摄全景图像中的图像处理操作来检查关于具有相机的终端的运动信息，并且自动确定拍摄视点，从而使用户能够容易且准确地拍摄全景图像而无需 单独的设备或附加硬件。 一种用于拍摄全景图像的方法包括以下步骤：设置用于拍摄全景图像的各种拍摄环境（200）; 开始全景图片拍摄（210）; 通过使用当前输入图像和先前输入图像来检查相机的运动（220）; 显示相机在显示单元（230）上移动的方向; 根据相机的运动，检查相机当前是否位于每个可视点以进行图像拍摄; 并拍摄对应的图像（250）。
-
-
-
-
-
-
-
-
-
Search Results
202
records
(took 0.029 seconds)
