公开(公告)号：US6002492A

公开(公告)日：1999-12-14

申请号：US748658

申请日：1996-11-14

申请人： Koichi Kamon ,  Hideo Kumashiro ,  Hiroshi Murakami

发明人： Koichi Kamon ,  Hideo Kumashiro ,  Hiroshi Murakami

IPC分类号： G03B27/62 ,  G03G15/00 ,  G03G15/36 ,  H04N1/00 ,  H04N1/04 ,  H04N1/12 ,  H04N1/193 ,  H04N1/387

CPC分类号： H04N1/04 ,  H04N1/3876 ,  H04N1/12 ,  H04N1/193 ,  H04N2201/0414

摘要： The present invention provides an apparatus which separately reads and combines a plurality of partial images of a large document so as to produce simply and accurately produce a composite image of excellent precision at the joint boundary of each partial image without complex positioning operation.

摘要翻译： 本发明提供了一种单独读取并组合大型文档的多个部分图像的装置，以便在每个部分图像的联合边界处简单且准确地产生具有优异精度的复合图像，而无需复杂的定位操作。

公开(公告)号：US6148118A

公开(公告)日：2000-11-14

申请号：US675485

申请日：1996-07-03

申请人： Hiroshi Murakami ,  Hideo Kumashiro ,  Koichi Kamon

发明人： Hiroshi Murakami ,  Hideo Kumashiro ,  Koichi Kamon

IPC分类号： G06T11/00 ,  G06K9/36

CPC分类号： H04N1/3876 ,  G06T11/00

摘要： In a method of joining image data divided into a plurality of parts each having an overlap region with the other at a border line portion into a single image for output, an image processing apparatus joins the plurality of partial images having overlap regions L at border line portions. The image processing apparatus extracts a joining line connecting pixels with small density difference between partial images from the overlap region, and joins the partial images with each other along the joining line. The image processing apparatus extracts a line connecting pixels with small density difference with adjacent pixels from the overlap region of the partial images as a joining line upon extracting the joining line. The density difference at the joining border is minimized in order to make the joining line look more natural, and make offset of images in the overlap region less recognizable.

摘要翻译： 在图像处理装置中，将分割成具有与边界线部分的重叠区域的多个部分的图像数据连接成单个图像以输出的方法，图像处理装置将具有重叠区域L的多个部分图像连接到边界线 部分。 图像处理装置提取从重叠区域连接部分图像之间具有小密度差的像素的连接线，并且沿着接合线将部分图像相互连接。 图像处理装置在提取接合线时，提取从部分图像的重叠区域连接具有小密度差的像素作为接合线的线。 连接边界处的密度差最小化，以使接合线看起来更自然，并使得重叠区域中的图像的偏移不太可识别。

公开(公告)号：US5446802A

公开(公告)日：1995-08-29

申请号：US262392

申请日：1994-06-20

申请人： Kaoru Tada ,  Hideo Kumashiro ,  Hiroya Sugawa ,  Shigeru Moriya ,  Hiroshi Murakami

发明人： Kaoru Tada ,  Hideo Kumashiro ,  Hiroya Sugawa ,  Shigeru Moriya ,  Hiroshi Murakami

IPC分类号： G06T5/00 ,  H04N1/40 ,  G06K9/00

CPC分类号： G06T7/0081 ,  H04N1/40062 ,  G06T2207/10008 ,  G06T2207/30176

摘要： In an image processing apparatus, first peak density pixels of one dimension each having a peak density in one-dimension direction larger than a predetermined density are detected from image data located in a predetermined image area of a document image, and second peak density pixels of two dimensions each having peak densities in two-dimension directions larger than the predetermined density are detected from the image data. Further, there are counted a number of the detected first peak density pixels of one dimension and a number of the detected second peak density pixels of two dimensions, and then, it is judged whether or not the image data located in the predetermined image area are image data of a dot photograph image, a photograph image or a character image based on the counted numbers.

摘要翻译： 在图像处理装置中，从位于原稿图像的预定图像区域的图像数据中检测出具有大于预定浓度的一维方向上的峰值密度的一维的第一峰值密度像素，第二峰值密度像素 从图像数据中检测出具有大于预定浓度的二维方向的峰值密度的两个维度。 此外，对所检测到的一维的第一峰值密度数和检测到的二维的第二峰值密度像素的数量进行计数，然后判断位于预定图像区域中的图像数据是否为 基于所计数的点数的点照片图像，照片图像或字符图像的图像数据。

公开(公告)号：US20070195992A1

公开(公告)日：2007-08-23

申请号：US11786611

申请日：2007-04-12

申请人： Yoshikazu Ikenoue ,  Hideo Kumashiro ,  Munehiro Nakatani

发明人： Yoshikazu Ikenoue ,  Hideo Kumashiro ,  Munehiro Nakatani

IPC分类号： G06K9/00

CPC分类号： H04N1/0084 ,  G06T1/0021 ,  G06T1/0028 ,  G06T2201/0051 ,  G06T2201/0061 ,  G06T2201/0064 ,  H04N1/00846 ,  H04N1/00854 ,  H04N1/00867 ,  H04N1/0087 ,  H04N1/00875 ,  H04N1/00877 ,  H04N1/32149 ,  H04N1/32203 ,  H04N1/32208 ,  H04N1/32213 ,  H04N1/32229 ,  H04N1/32251 ,  H04N1/32261 ,  H04N1/32288 ,  H04N2201/3205 ,  H04N2201/3214 ,  H04N2201/3215 ,  H04N2201/3226 ,  H04N2201/3232 ,  H04N2201/3242 ,  H04N2201/3246 ,  H04N2201/327 ,  H04N2201/3274 ,  H04N2201/3283 ,  Y10S283/901 ,  Y10S283/902

摘要： An additional data useful for the management of copies of a document for an image forming apparatus is embedded in a hard copy of the document. The additional data is embedded in an inherent image of a document as discrete dots arranged in a prescribed format. Preferably, the size of each of said discrete dots is not so large as to be recognized easily with naked eyes. The additional data can be embedded in unit of blocks. Further, the density of the discrete dots is different from the density adjacent to the discrete dots. The image data for embedding an additional data can be obtained by reading a document, or by receiving data sent from a computer or read from a floppy disk. An additional data embedded in a hard copy can be extracted from the image data on the hard copy. Further, the production of a copy according to the image data can be controlled according to the extracted additional data. For example, if the additional data means that the source of the hard copy is a secret document, the copying is allowed only for a legitimate user. An additional data is generated for each copy according to the extracted additional data for embedding it in a hard copy to identify each copy. The additional data generated for each copy are compiled and they are used to manage copies of various sources systematically and illegal copies can be traced easily. If the additional data received from a printer is not consistent with the additional data already stored, the printing may be prohibited. A transmitter for transmitting an additional data may be provided in-a document such as a book.

摘要翻译： 用于管理图像形成装置的文档的副本的附加数据被嵌入文档的硬拷贝中。 附加数据嵌入文档的固有图像中，作为以规定格式排列的离散点。 优选地，每个所述离散点的尺寸不是很大，以致容易用肉眼识别。 附加数据可以以块为单位嵌入。 此外，离散点的密度与离散点相邻的密度不同。 用于嵌入附加数据的图像数据可以通过读取文档或通过接收从计算机发送的数据或从软盘读取来获得。 可以从硬拷贝上的图像数据中提取嵌入在硬拷贝中的附加数据。 此外，可以根据所提取的附加数据来控制根据图像数据的副本的制作。 例如，如果附加数据意味着硬拷贝的来源是秘密文件，则仅允许合法用户复制。 根据提取的附加数据为每个副本生成附加数据，以将其嵌入硬拷贝中以识别每个副本。 为每个副本生成的附加数据进行编译，并用于系统地管理各种源的副本，并且可以轻松地跟踪非法副本。 如果从打印机接收到的附加数据与已经存储的附加数据不一致，则可能禁止打印。 用于发送附加数据的发送器可以在诸如书籍的文档中提供。

公开(公告)号：US06369912B1

公开(公告)日：2002-04-09

申请号：US09234513

申请日：1999-01-21

申请人： Hideo Kumashiro

发明人： Hideo Kumashiro

IPC分类号： H04N1405

CPC分类号： H04N1/4053

摘要： An image processing apparatus capable of reducing a nonuniform pitch in the sub scanning direction of an image, caused such as by unstable paper feed of a printer, for example, is formed as described below. An adder adds the density value of one pixel of an input multi-valued image and a diffused error output from an error weighting filter. A comparator compares an output from the adder and a corresponding threshold of a variable threshold matrix, and outputs the result as a binary image. The variable threshold matrix is formed of two sub matrices with a size of 3×3 connected in a main scanning direction, and an average value of thresholds in one sub matrix is larger than another. Accordingly, a line component vertical to the main scanning direction is added in an output image. This line component masks a nonuniform pitch in the sub scanning direction.

摘要翻译： 如下所述，可以形成能够减少图像的副扫描方向上的不均匀间距的图像处理装置，例如由于打印机的送纸不稳定引起的。 加法器将输入多值图像的一个像素的密度值和来自误差加权滤波器的扩散误差输出相加。 比较器比较来自加法器的输出和可变阈值矩阵的对应阈值，并将结果输出为二进制图像。 可变阈值矩阵由在主扫描方向上连接的尺寸为3×3的两个子矩阵形成，并且一个子矩阵中的阈值的平均值大于另一个。 因此，在输出图像中附加与主扫描方向垂直的线分量。 该线分量在副扫描方向上屏蔽不均匀的间距。

公开(公告)号：US5987127A

公开(公告)日：1999-11-16

申请号：US893141

申请日：1997-07-15

申请人： Yoshikazu Ikenoue ,  Hideo Kumashiro

发明人： Yoshikazu Ikenoue ,  Hideo Kumashiro

IPC分类号： G06T1/00 ,  H04N1/00 ,  H04N1/32 ,  H04N7/16 ,  H04L9/00

CPC分类号： H04N1/0084 ,  G06T1/0028 ,  H04N1/00854 ,  H04N1/00859 ,  H04N1/00867 ,  H04N1/0087 ,  H04N1/00875 ,  G06T2201/0051 ,  G06T2201/0064 ,  H04N2201/3205 ,  H04N2201/327

摘要： Additional data useful for the management of copies of a document for an image forming apparatus is embedded in a hard copy of the document. The additional data is embedded in an inherent image of a document as pixels arranged in a prescribed format, and, preferably, the size of the pixels is such as to not be easily recognized with the naked eye. The image data for embedding the additional data can be obtained by reading a document, or by receiving data sent from a computer or read from a floppy disk. Additional data embedded in a hard copy can be extracted from the image data on the hard copy, and the production of a copy according to the image data can be controlled according to the extracted additional data. For example, if the additional data means that the source of the hard copy is a secret document, copying thereof is allowed only for a legitimate user. Additional data generated for each copy is compiled and is used to manage copies of various sources systematically.

摘要翻译： 用于管理图像形成装置的文档的副本的附加数据嵌入在文档的硬拷贝中。 附加数据被嵌入作为以规定格式布置的像素的文档的固有图像，并且优选地，像素的大小不能用肉眼容易地识别。 用于嵌入附加数据的图像数据可以通过读取文档或者通过接收从计算机发送的数据或从软盘读取来获得。 可以从硬拷贝上的图像数据中提取嵌入在硬拷贝中的附加数据，并且可以根据提取的附加数据来控制根据图像数据的副本的生成。 例如，如果附加数据意味着硬拷贝的来源是秘密文件，则只允许其复制是为了合法用户。 为每个副本生成的附加数据被编译，并用于系统地管理各种源的副本。

公开(公告)号：US5341227A

公开(公告)日：1994-08-23

申请号：US34900

申请日：1993-03-22

申请人： Hideo Kumashiro

发明人： Hideo Kumashiro

IPC分类号： H04N1/40 ,  H04N1/46

CPC分类号： H04N1/40062

摘要： In a dot image discrimination circuit, a dot image in an area composed of many multi-tone pixels including a central pixel is discriminated. A pixel (change point) at which the density of the image data is prominent in a first direction is detected in a plurality of pixels. Then, it is inspected if change points exist continuously in a second direction perpendicular to the first direction. The change points are counted in an area of a prescribed size around the central pixel, but some change points existing continuously in the second direction are not counted. Thus, a pixel in an image made of regular dots can be discriminated from a pixel in another type of image such as a character image. The discrimination of a dot image can be performed fast and precisely.

摘要翻译： 在点图像识别电路中，区分由包括中心像素的许多多色调像素组成的区域中的点图像。 在多个像素中检测图像数据的密度在第一方向上突出的像素（变化点）。 然后，如果在垂直于第一方向的第二方向上连续存在变化点，则检查它。 将变化点计数在中心像素周围的规定尺寸的区域中，但是不计数在第二方向上连续存在的一些变化点。 因此，可以从诸如字符图像的另一类型图像中的像素辨别由常规点形成的图像中的像素。 可以快速，精确地执行点图像的辨别。

公开(公告)号：US6144777A

公开(公告)日：2000-11-07

申请号：US810240

申请日：1997-03-03

申请人： Kaoru Tada ,  Hiroya Sugawa ,  Atsushi Ishikawa ,  Akio Nakajima ,  Hideo Kumashiro

发明人： Kaoru Tada ,  Hiroya Sugawa ,  Atsushi Ishikawa ,  Akio Nakajima ,  Hideo Kumashiro

IPC分类号： H04N1/387

CPC分类号： H04N1/3877

摘要： Originals are fed onto a platen glass one by one by an automatic original feeder to be read out by a reading system. The originals are converted into digital image data. The image data is transferred to a memory unit. The memory unit is controlled by a CPU. The memory unit is configured of an image memory, a sentence direction recognizing portion, an RAM producing a management table, an original direction (top and bottom direction) recognizing portion, a rotation processing portion, a sign memory and the like. Based on top and bottom directions of original images and directions in which sentences run included in the images, arrangement of images in a 2in1 or 4in1 mode is determined.

摘要翻译： 原稿通过自动原稿进给器一个接一个地送入平板玻璃板，由读取系统读出。 原件被转换为数字图像数据。 图像数据被传送到存储器单元。 存储单元由CPU控制。 存储单元由图像存储器，句子方向识别部分，产生管理表的RAM，原始方向（上下方向）识别部分，旋转处理部分，符号存储器等构成。 基于图像的原始图像的顶部和底部方向以及图像中包括的句子的方向，确定2in1或4in1模式中的图像的排列。

公开(公告)号：US5898796A

公开(公告)日：1999-04-27

申请号：US784319

申请日：1997-01-16

申请人： Hideo Kumashiro

发明人： Hideo Kumashiro

IPC分类号： H04N1/40 ,  H04N1/405 ,  G06K9/36 ,  G06K9/38 ,  G06K9/46 ,  G06K9/66

CPC分类号： H04N1/4052 ,  H04N1/40087

摘要： An input signal represented by values of many levels obtained by reading out an original image is applied to a pixel characteristic extracting unit for extraction of a characteristic of a pixel. Based on the extracted pixel characteristic, a gain calculating process is effected based on the pixel characteristic in a first gain calculating unit. An attribute flag showing an image characteristic extracted by the pixel characteristic extracting unit is added to the gain calculated error data which is stored. An area characteristic is determined in an area characteristic extracting unit based on the flag. Based on the determination result, error correction with high precision is carried out in a second gain calculating unit.

摘要翻译： 通过读出原始图像获得的许多级别的值所表示的输入信号被施加到用于提取像素的特性的像素特征提取单元。 基于所提取的像素特性，基于第一增益计算单元中的像素特性来实现增益计算处理。 示出由像素特征提取单元提取的图像特征的属性标志被添加到存储的增益计算出的错误数据中。 基于该标志在区域特征提取单元中确定区域特性。 基于确定结果，在第二增益计算单元中执行高精度的纠错。

公开(公告)号：US5870503A

公开(公告)日：1999-02-09

申请号：US545412

申请日：1995-10-19

申请人： Hideo Kumashiro

发明人： Hideo Kumashiro

IPC分类号： G06K9/36

CPC分类号： G06K9/36

摘要： An 8-bit input image data value has its range adjusted by an image range conversion lookup table and is converted into a 4-bit dither image data value f2 with a multi-value dither circuit. The converted 4-bit dither image data value f2 is error-corrected by an error correction unit, and then binarized with a predetermined binarization threshold value of Th=8. An error between binarized data value g and error-correction image data value f2' is calculated by the error calculation unit to be stored in an error storage line memory. Error data e' is provided with respect to a pixel of interest by a peripheral error weighting filter, whereby error is propagated to peripheral pixels of the pixel of interest. As a result, an image processing apparatus of high picture quality, high speed, and low cost is provided.

摘要翻译： 8位输入图像数据值的范围由图像范围转换查找表进行调整，并用多值抖动电路转换为4位抖动图像数据值f2。 转换的4位抖动图像数据值f2由误差校正单元进行纠错，然后用Th = 8的预定二值化阈值二值化。 由误差计算单元计算二值化数据值g与误差校正图像数据值f2'之间的误差，以存储在错误存储行存储器中。 通过周边误差加权滤波器相对于感兴趣的像素提供误差数据e'，由此将误差传播到感兴趣像素的周边像素。 结果，提供了高画质，高速，低成本的图像处理装置。