公开(公告)号：KR1020090021527A

公开(公告)日：2009-03-04

申请号：KR1020070086105

申请日：2007-08-27

Applicant: 삼성전자주식회사 ,  재단법인서울대학교산학협력재단

Inventor： 김수균 ,  문재원 ,  조성대 ,  오상욱 ,  손병준 ,  홍태화 ,  정희원 ,  조남익 ,  이상화 ,  하성종

IPC: H04N5/262

CPC classification number: G06T7/20 ,  G03B37/04 ,  H04N5/232 ,  H04N5/23238 ,  H04N5/2624

Abstract: A method for photographing a panorama picture is provided to correct and adopt an image obtained by a camera by considering the relation with a real image of a subject so that the movement of a camera can be recognized accurately. When photographing each image which configures a panorama picture, a projected image is obtained by projecting an image, which is continuously inputted at a predetermined time interval, onto a virtual plane(200-215). A motion of a corresponding photographing device is detected by comparing a current image with a previous image based on a motion estimation technique by which exposure calibration is applied to the projected image(220). The photographing point for each image is judged by confirming, according to the movement of the detected motion of the corresponding photographing device, that the motion depending on photographing direction reaches a threshold value(230,240). Each image is inputted manually or is photographed at each photographing point of each image(250).

Abstract translation: 提供拍摄全景图像的方法，以通过考虑与被摄体的实际图像的关系来校正和采用通过照相机获得的图像，使得可以准确地识别照相机的移动。 当拍摄配置全景图像的每个图像时，通过将以预定时间间隔连续输入的图像投影到虚拟平面（200-215）上来获得投影图像。 通过基于将投射图像应用曝光校准的运动估计技术（220）来比较当前图像与先前图像来检测相应拍摄装置的运动。 通过根据相应的拍摄装置的检测到的运动的移动确定根据拍摄方向的运动达到阈值（230,240）来判断每个图像的拍摄点。 每个图像被手动输入或在每个图像的每个拍摄点拍摄（250）。

公开(公告)号：KR100796849B1

公开(公告)日：2008-01-22

申请号：KR1020060084880

申请日：2006-09-04

Applicant: 삼성전자주식회사 ,  재단법인서울대학교산학협력재단

Inventor： 박경주 ,  조성대 ,  김수균 ,  문재원 ,  조남익 ,  이상화 ,  구형일 ,  하성종

IPC: H04N5/262

CPC classification number: G06K9/36 ,  G06K2009/2045 ,  G06T3/4038

Abstract: A method for capturing a panorama mosaic image in a mobile terminal is provided to perform mosaic configuration with a relatively simple calculation by using limited hardware resources. A portion of a previously captured image is displayed as a guide image on an LCD(Liquid Crystal Display) viewer of a currently captured image and a plurality of images constituting a mosaic image is captured according to user manipulation(30). When the captured image is obtained in a panorama mode, projection transformation can be approximated to a rotational transformation matrix and each image is re-projected to a common cylindrical curved surface(32). Overlapped images of the re-projected images are locally aligned to determine an optimal position by using a hexagonal search scheme(34). Image stitching is performed on the overlapped regions by using a dynamic programming scheme to determine a boundary for connecting the overlapped images most naturally(36). A blending process is performed by removing a sudden change of brightness and the color sense based on the boundary region of the images(38).

Abstract translation: 提供一种用于在移动终端中捕获全景马赛克图像的方法，以通过使用有限的硬件资源以相对简单的计算来执行马赛克配置。 先前拍摄的图像的一部分作为引导图像显示在当前拍摄图像的LCD（液晶显示器）观察器上，并且根据用户操纵（30）捕获构成马赛克图像的多个图像。 当以全景模式获得拍摄图像时，可以将投影变换近似为旋转变换矩阵，并且将每个图像重新投影到公共圆柱形曲面（32）。 重新投影的图像的重叠图像被局部对齐以通过使用六边形搜索方案来确定最佳位置（34）。 通过使用动态规划方案在重叠区域上执行图像拼接，以确定用于最重要地连接重叠图像的边界（36）。 通过基于图像的边界区域去除亮度和颜色感觉的突然变化来执行混合处理。

公开(公告)号：KR100788643B1

公开(公告)日：2007-12-26

申请号：KR1020010001130

申请日：2001-01-09

Applicant: 삼성전자주식회사

Inventor： 최양림 ,  조남익 ,  곽정원

IPC: G06F17/30

CPC classification number: G06F17/30262 ,  G06F17/3025 ,  Y10S707/99936

Abstract: 색 특징 및 질감 특징을 적절히 조합함으로써 보다 검색 성능이 향상된 영상 검색 방법이 개시된다. 본 발명에 따른 영상 검색 방법은 (a) 쿼리 영상과 데이터 영상 사이의 색 거리와 질감 거리를 구하는 단계와, (b) 구한 색 거리와 질감 거리에 소정의 가중치를 사용하여 가중시키는 단계, (c) 인간의 시감적 특성을 고려하여 가중된 색거리와 질감 거리를 조합한 거리를 구하는 단계, 및 (d) 조합된 특징 거리를 사용하여 쿼리 영상과 유사한 영상을 결정하는 단계를 포함하는 것을 특징으로 한다. 본 발명에 따르면 영역 영상에서 추출된 색과 질감 특징을 이용하여 이들 정보를 동시에 조합함으로써 보다 사람 시각적인 감각에 적합한 검색 결과를 얻을 수 있다. 특히, 영역별 검색으로 한 영상에 있는 많은 물체(object)와 정보를 적은 계산에 의하여 보다 정밀하게 검색하는 것이 가능하다.

公开(公告)号：KR1020070087455A

公开(公告)日：2007-08-28

申请号：KR1020060017881

申请日：2006-02-23

Applicant: 삼성전자주식회사 ,  재단법인서울대학교산학협력재단

Inventor： 조남익 ,  박찬율

IPC: G06T3/40

CPC classification number: G06T3/4084 ,  H04N19/117 ,  H04N19/18 ,  H04N19/48 ,  H04N19/59 ,  H04N19/60

Abstract: An image resizing method using DCT(Discrete Cosine Transform) and an apparatus therefor are provided to easily resize an input image by performing coarse resizing for inputted image information and performing fine resizing. A receiving unit(810) receives a DCT coefficient of an input image. A transformation matrix determining unit(820) determines a transformation matrix for transforming the input image at a ratio approximate to a resizing ratio. A coarse resizing unit(830) performs the coarse resizing of the input image in a DCT region by using the transformation matrix. An IDCT(Inverse DCT) unit(880) performs IDCT in the coarse-resized region and acquires a spatial image. A fine resizing unit(890) performs the fine resizing of the spatial image in a space region and forms an output image.

Abstract translation: 提供了使用DCT（离散余弦变换）及其装置的图像尺寸调整方法，以便通过对输入的图像信息进行粗调整并执行精细调整大小来容易地调整输入图像的大小。 接收单元（810）接收输入图像的DCT系数。 变换矩阵确定单元（820）确定用于以接近于调整大小比率的比率变换输入图像的变换矩阵。 粗调整单元（830）通过使用变换矩阵在DCT区域中执行输入图像的粗调整。 IDCT（逆DCT）单元（880）在粗调大小区域中执行IDCT并获取空间图像。 精细调整单元（890）在空间区域中执行空间图像的精细调整，并形成输出图像。