公开(公告)号：JP2005129077A

公开(公告)日：2005-05-19

申请号：JP2004368135

申请日：2004-12-20

Applicant: Hewlett Packard Co ,  ヒューレット・パッカード・カンパニーHewlett−Packard Company

Inventor： POLLARD STEPHEN BERNARD ,  KAHN RICHARD OLIVER

IPC: G06T1/00 ,  G06T3/00 ,  H04N1/047 ,  H04N1/107 ,  H04N1/40

CPC classification number: H04N1/047 ,  H04N1/107 ,  H04N2201/02425 ,  H04N2201/02439 ,  H04N2201/03162 ,  H04N2201/0414 ,  H04N2201/0471 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04737 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791 ,  H04N2201/04794

Abstract: PROBLEM TO BE SOLVED: To provide a method capable of accurately restoring an original image even though the original image is captured in a distorted state. SOLUTION: A scanning device with a navigation sensor sensing a sensor position corresponding to the original image scans the original image, and captures image data and position data. Pixel grids for an image to be restored are defined, and linear image data is mapped into the restored image. One zone is defined by mapping a pair of linear image data. Regarding each pixel in the zone, each intensity value is determined. The intensity value is calculated on the basis of the image data obtained at positions of a plurality of sensors which straddles the pixel in which its intensity is determined. Non-distorted image can be obtained by the determined intensity value. COPYRIGHT: (C)2005,JPO&NCIPI

Abstract translation: 要解决的问题：即使在畸变状态下捕获原始图像，也提供能够准确地恢复原始图像的方法。 解决方案：具有感测与原始图像对应的传感器位置的导航传感器的扫描装置扫描原始图像，并且捕获图像数据和位置数据。 定义要恢复的图像的像素网格，并将线性图像数据映射到恢复的图像中。 通过映射一对线性图像数据来定义一个区域。 对于区域中的每个像素，确定每个强度值。 基于在跨越其强度确定的像素的多个传感器的位置处获得的图像数据来计算强度值。 可以通过确定的强度值获得非失真图像。 版权所有（C）2005，JPO＆NCIPI

公开(公告)号：JP3376077B2

公开(公告)日：2003-02-10

申请号：JP4509494

申请日：1994-02-18

Applicant: 株式会社リコー

Inventor： ジェー ウォルフ グレゴリー ,  ジー ストーク ディビッド

IPC: H04N1/387 ,  G06T3/00 ,  H04N1/047 ,  H04N1/10 ,  H04N1/193

CPC classification number: G06T3/0031 ,  H04N1/047 ,  H04N1/1013 ,  H04N1/193 ,  H04N2201/02425 ,  H04N2201/03162 ,  H04N2201/0434 ,  H04N2201/04703 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04743 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791

Abstract: An improved method and apparatus for correcting for splay is provided. A document distorted by the curvature of a page of text away from a platen is converted to a digital image. The digital image is the manipulated to remove the distortion by fitting the lines of text in an unsplayed portion to a skew line, which represents the deviation of lines of text in the digital image from horizontal. Then the splay is determined for each line of text. Once the skew and the splay are determined, an inverse transformation is done to straighten the lines of text. A horizontal stretching is also applied to the text to correct for the projection angle of the original document.

公开(公告)号：JPS61105954A

公开(公告)日：1986-05-24

申请号：JP22851084

申请日：1984-10-30

Applicant: Toshiba Corp

Inventor： ISHIGAKI TOSHINORI

IPC: G06K9/74 ,  G06K7/10 ,  G06K9/22 ,  G06K9/32 ,  G06T1/00 ,  H04N1/04 ,  H04N1/047 ,  H04N1/107 ,  H04N1/191

CPC classification number: H04N1/1072 ,  H04N1/047 ,  H04N1/107 ,  H04N1/1911 ,  H04N2201/02425 ,  H04N2201/02439 ,  H04N2201/03162 ,  H04N2201/0414 ,  H04N2201/0422 ,  H04N2201/0426 ,  H04N2201/04703 ,  H04N2201/0471 ,  H04N2201/0472 ,  H04N2201/04727 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04732 ,  H04N2201/04734 ,  H04N2201/04749 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791 ,  H04N2201/04794

Abstract: PURPOSE: To make constitution of the mechanics simple and to make image inputting of high quality by simple operation by adding a mark having fixed regularity to a medium to be read, reading above-mentioned mark at every scanning, and making correction of overlapping of inputted data, correction of deviation of starting position of scanning, etc.
CONSTITUTION: A black mark 23 is printed on an underlay 22 for setting an inputting paper 21. The mark 23 is tapered upward, and the width of the mark along the direction of scanning narrows gradually upward. Straight lines (guide lines) 24 are provided at intervals equal to the width of reading of a sensor to use as parallelism guide at the time of input scanning. Thus, by adding a mark having fixed regularity on a medium to be read on which images of characters, graphics, etc. that become objects of inputting, and reading above- mentioned mark at every scanning, correction of overlapping of inputted data, correction of deviation of starting position of scanning etc. can be executed.
COPYRIGHT: (C)1986,JPO&Japio

Abstract translation: 目的：使机械结构简单，通过在待读取的介质上添加具有固定规则性的标记，通过简单的操作进行高质量的图像输入，在每次扫描时读取上述标记，并对输入的重叠进行校正 数据，扫描起始位置的偏差校正等。构成：黑色标记23印刷在底层22上，用于设置输入纸21。标记23向上逐渐变细，并且标记沿着扫描方向的宽度 逐渐变窄。 直线（引导线）24以等于传感器的读取宽度的间隔设置，以在输入扫描时用作平行引导件。 因此，通过在要读取的介质上添加具有固定规则性的标记，在每次扫描时成为对象的字符，图形等的图像，以及在每次扫描时读取上述标记，校正输入数据的重叠，校正 可以执行扫描起始位置的偏差等。

公开(公告)号：EP0873003B1

公开(公告)日：2007-03-21

申请号：EP97302519.0

申请日：1997-04-14

Applicant: Hewlett-Packard Company A Delaware Corporation

Inventor： Kahn, Richard Oliver ,  Pollard, Stephen Bernard

IPC: H04N1/047

CPC classification number: H04N1/047 ,  H04N1/107 ,  H04N2201/02425 ,  H04N2201/02439 ,  H04N2201/03162 ,  H04N2201/0414 ,  H04N2201/0471 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04737 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791 ,  H04N2201/04794

Abstract: The present invention relates to a method of reconstructing an image from data captured by a sensor, and is particularly applicable to the case of sequential capture in relative movement between a scanning device and the original image. The scanning device comprises navigation means for detecting the position of the sensor relative to the original image. A pixel grid for the reconstructed image is determined, and correspondence between the pixel grid and the sensor data is identified using sensor position detection data. The intensity of pixels is determined from sensor data selected as relevant to the pixel under consideration.

公开(公告)号：EP0946043A3

公开(公告)日：2002-08-28

申请号：EP99200855.7

申请日：1999-03-19

Applicant: Agfa Corporation

Inventor： Larsen, David B. ,  Nolan John F. ,  Coppeta, David ,  Rolfe, Norman F.

IPC: H04N1/191

CPC classification number: B41J2/435 ,  G06K15/1238 ,  G06K2215/111 ,  H04N1/053 ,  H04N1/1135 ,  H04N1/12 ,  H04N1/1903 ,  H04N1/192 ,  H04N1/193 ,  H04N1/1934 ,  H04N2201/03162 ,  H04N2201/0471 ,  H04N2201/04712 ,  H04N2201/04732 ,  H04N2201/04744 ,  H04N2201/04755 ,  H04N2201/04787

Abstract: The present invention eliminates artifacts in an image formed using a plurality of imaging sources. Visible seams in the image are eliminated by randomizing the stitch point (18) between the scan lines (44, 46) produced by each imaging source. The randomization may be optimized by additionally applying a method for relocating the random stitch point (18) based on the data content of the scan line, adjacent scan lines, and other criteria. The present invention further compensates for in-scan and cross-scan errors caused by thermally induced errors, spinner synchronization errors, mechanical misalignment, and other factors associated with the use of a plurality of imaging systems (30). A photodetector system (120, 122, 124), comprising a mask (124) having a pair of triangular openings (128,128'), provides measurements of the in-scan and cross-scan errors.

Abstract translation: 本发明消除了使用的成像源形成的多个的图像中的伪影。 在图像中可见的接缝是由随机化由每个成像源产生的扫描线（44,46）之间的落针位置（18）消除。 随机化可通过附加地施加搬迁基于所述扫描线，相邻的扫描线，和其他标准的数据内容的随机落针位置（18）的方法进行优化。 本发明进一步补偿了在扫描和由热引起的误差，旋转同步误差，机械失准，并且与使用的成像系统（30）中的多个相关联的其他因素引起的交叉扫描误差。 光电检测器系统（120，122，124）包括一个面罩（124），其具有一对三角形的开口（128,128“）的提供在扫描和交叉扫描误差的测量。

公开(公告)号：EP0987873A3

公开(公告)日：2001-02-28

申请号：EP99307023.4

申请日：1999-09-03

Applicant: Hewlett-Packard Company

Inventor： Beausoleil, Raymond G. Jr. ,  Allen, Ross R.

IPC: H04N1/047

CPC classification number: H04N1/047 ,  H04N1/107 ,  H04N2201/02425 ,  H04N2201/03162 ,  H04N2201/04712 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04737 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791

Abstract: A method and apparatus are provided for efficiently obtaining navigational position of a sensor array (22) based on a correlation between images captured by the sensor array at successive points in time. This is achieved by first acquiring a reference frame through said sensor array (22), successively acquiring a sample frame through said sensor array. Correlation data representing a relationship between the sample frame and the reference frame is then generated and a displacement of a reference feature contained in both the reference frame and the sample frame is determined based on the correlation data. The correlation surface of the correlation data is modelled as a general two-dimensional Taylor series expansion. The method and apparatus can be employed in a scanning device (10) to provide efficient scheme for obtaining navigational information based on a comparison of image frames captured by a sensor array of the scanning device at successive points in time.

公开(公告)号：EP0869659A3

公开(公告)日：2000-05-03

申请号：EP98105835.7

申请日：1998-03-31

Applicant: KONICA CORPORATION

Inventor： Washio, Koji ,  Ichihara, Yoshiyuki

IPC: H04N1/047 ,  H04N1/387 ,  G06T3/60

CPC classification number: H04N1/00721 ,  H04N1/00681 ,  H04N1/00734 ,  H04N1/00745 ,  H04N1/00748 ,  H04N1/00753 ,  H04N1/00758 ,  H04N1/00766 ,  H04N1/047 ,  H04N1/12 ,  H04N1/193 ,  H04N2201/02425 ,  H04N2201/03162 ,  H04N2201/0471 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04732 ,  H04N2201/04737 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791

Abstract: A correction method of a document inclination angle includes the steps of: first detecting an edge of a document; second detecting an inclination amount of the document corresponding to an inclination angle according to a detected result of the edge of the document. The inclination angle represents an inclination angle of the edge of the document with respect to a first reference direction and a second reference direction perpendicular to the first reference direction. The correction method further includes correcting the inclination angle of image data of the document with respect to the first and second reference directions. The step of the correcting includes a shift processing conducting a first shift processing in which the image data are shifted to one of the first and second reference directions according to the inclination amount and conducting a second shift processing in which the image data are shifted to the other of the first and second reference directions according to the inclination amount.

公开(公告)号：EP0987873A2

公开(公告)日：2000-03-22

申请号：EP99307023.4

申请日：1999-09-03

Applicant: Hewlett-Packard Company

Inventor： Beausoleil, Raymond G. Jr. ,  Allen, Ross R.

IPC: H04N1/047

CPC classification number: H04N1/047 ,  H04N1/107 ,  H04N2201/02425 ,  H04N2201/03162 ,  H04N2201/04712 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04737 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791

Abstract: A method and apparatus are provided for efficiently obtaining navigational position of a sensor array (22) based on a correlation between images captured by the sensor array at successive points in time. This is achieved by first acquiring a reference frame through said sensor array (22), successively acquiring a sample frame through said sensor array. Correlation data representing a relationship between the sample frame and the reference frame is then generated and a displacement of a reference feature contained in both the reference frame and the sample frame is determined based on the correlation data. The correlation surface of the correlation data is modelled as a general two-dimensional Taylor series expansion. The method and apparatus can be employed in a scanning device (10) to provide efficient scheme for obtaining navigational information based on a comparison of image frames captured by a sensor array of the scanning device at successive points in time.

Abstract translation: 提供了一种方法和装置，用于基于传感器阵列在连续时间点捕获的图像之间的相关性来有效地获得传感器阵列（22）的导航位置。 这通过首先通过所述传感器阵列（22）获取参考帧，通过所述传感器阵列连续获取采样帧来实现。 然后生成表示采样帧和参考帧之间的关系的相关数据，并且基于相关数据来确定包含在参考帧和采样帧中的参考特征的位移。 相关数据的相关表面被建模为一般的二维泰勒级数展开。 可以在扫描装置（10）中采用该方法和装置，以提供基于在连续时间点上由扫描装置的传感器阵列捕获的图像帧的比较来获得导航信息的有效方案。

公开(公告)号：EP0698991A2

公开(公告)日：1996-02-28

申请号：EP95305810.4

申请日：1995-08-21

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor： Edgar, Albert Durr ,  Penn, Steven Craig

IPC: H04N1/047 ,  H04N1/40

CPC classification number: H04N1/40068 ,  H04N1/047 ,  H04N1/12 ,  H04N1/193 ,  H04N2201/02425 ,  H04N2201/02439 ,  H04N2201/03162 ,  H04N2201/0408 ,  H04N2201/04705 ,  H04N2201/04708 ,  H04N2201/0472 ,  H04N2201/04729 ,  H04N2201/04731 ,  H04N2201/04732 ,  H04N2201/04734 ,  H04N2201/04749 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791 ,  H04N2201/04794

Abstract: A scanner converts an image, for example from a film, paper, or slide, into a set of digital pixel values. A reference pattern 464, 466 is disposed to one side of the image. The scanning means, e.g., a CCD array sensor 39 simultaneously scans lines of the image and the reference pattern as the image is moved relative to the scanning means. The scanner contains software and/or hardware for interpreting the sensed portion of the reference pattern as a position, which can then be associated with the raw scan line of the image which was sensed at the same time as the sensed portion of the reference pattern. This position information can be used to scale the pixels from their raw values onto a specified output grid. Typically, the raw scan lines are prescaled between a raw resolution between scan lines as they are received from the sensor to an intermediate resolution between scan lines. Next the prescaled lines are scaled cross-wise (ie parallel to the scan line), and then length-wise, to locate the pixels onto the chosen output grid. The scaling may be performed in the linear lumen domain using a sinc filter. Typically the raw image data is oversampled, and at a higher sampling frequency than the specified output grid.

Abstract translation: 扫描仪将图像（例如，电影，纸张或幻灯片）转换成一组数字像素值。 参考图案464,466设置在图像的一侧。 当图像相对于扫描装置移动时，扫描装置，例如CCD阵列传感器39同时扫描图像线和参考图案。 扫描仪包含用于将感测到的参考图案的部分解释为位置的软件和/或硬件，该位置然后可以与在参考图案的感测部分同时感测到的图像的原始扫描线相关联。 该位置信息可用于将像素从其原始值缩放到指定的输出网格上。 通常，原始扫描线在从传感器接收到扫描线之间的中间分辨率之间的扫描线之间的原始分辨率之间预分频。 接下来，预分线是横向缩放的（即平行于扫描线），然后纵向地将像素定位到所选择的输出栅格上。 缩放可以使用sinc滤光片在线性管腔域中进行。 通常，原始图像数据被过采样，并且处于比指定的输出网格更高的采样频率。 $$

公开(公告)号：EP0605131A1

公开(公告)日：1994-07-06

申请号：EP93310053.9

申请日：1993-12-14

Applicant: XEROX CORPORATION

Inventor： Curry, Douglas M.

IPC: H04N1/04

CPC classification number: B41J2/471 ,  G03G15/0435 ,  G03G15/326 ,  G03G2215/00075 ,  G06K15/12 ,  G06K15/1219 ,  G06K15/128 ,  H04N1/047 ,  H04N1/0473 ,  H04N1/1135 ,  H04N1/12 ,  H04N1/193 ,  H04N1/40037 ,  H04N2201/02425 ,  H04N2201/02441 ,  H04N2201/03162 ,  H04N2201/0458 ,  H04N2201/0471 ,  H04N2201/04725 ,  H04N2201/04731 ,  H04N2201/04734 ,  H04N2201/04787 ,  H04N2201/04789 ,  H04N2201/04791 ,  H04N2201/04793 ,  H04N2201/04796

Abstract: A method and apparatus is described for eliminating banding, misregistration and bowing by controlling a composite light intensity profile and phase shifting of a spatial location at which the composite light intensity profile crosses a xerographic threshold in a two dimensional high addressability printer operating in an overscan mode. The contrast, pitch grating and phase shifting are dependent on at least one of a predetermined amount of overlap, an exposure profile and an intensity value of each of the plurality of laser beam spots formed on the photosensitive medium of a two dimensional high addressability printer.

Abstract translation: 的方法和装置描述了用于消除带，配准不良和通过控制复合光强度分布和空间位置的相移在其中复合光强度分布交叉在过扫描模式下操作在两台维高寻址能力打印机一个静电复印阈弯曲 ， 对比度，音调光栅和相移依赖于重叠的预定量中的至少一个在暴露轮廓和每个形成的二维高寻址能力的打印机的感光介质上的激光束光斑的多个强度值的。