公开(公告)号：US6055336A

公开(公告)日：2000-04-25

申请号：US751755

申请日：1996-11-18

Applicant: Toru Niki

Inventor： Toru Niki

IPC: G06K9/20 ,  G06K9/38 ,  H04N1/403 ,  H04N1/41 ,  H04N1/411 ,  G06K9/36

CPC classification number: G06K9/00456 ,  G06K9/38 ,  H04N1/403 ,  H04N1/41 ,  H04N1/4115 ,  G06K2209/01

Abstract: An optical character recognition system recognizes character images in a document image comprised of character areas and non-character areas. The system includes a step of obtaining multi-value image data representing the document image, the multi-value image data having a first resolution and comprising plural pixels each having a pixel density value. The system binarizes the multi-value image data to produce binary image data having a second resolution for the document image, the second resolution being greater than the first resolution. The target pixel in the multi-value image data is binarized based on pixels adjacent to the target pixel. The system further includes steps for performing block selection on the binary image data for the document image in order to identify character areas of the document image and non-character areas of the document image, and performing optical character recognition processing on binary image data for the character areas of the document image to obtain computer codes for recognized character images in the character areas.

Abstract translation: 光学字符识别系统识别由字符区域和非字符区域组成的文档图像中的字符图像。 该系统包括获得表示文档图像的多值图像数据的步骤，该多值图像数据具有第一分辨率并且包括多个具有像素密度值的像素。 该系统对多值图像数据进行二值化以产生具有用于该文件图像的第二分辨率的二值图像数据，该第二分辨率大于该第一分辨率。 基于与目标像素相邻的像素，多值图像数据中的目标像素被二值化。 该系统还包括用于对文档图像的二进制图像数据执行块选择以识别文档图像的文字图像和文档图像的非字符区域的字符区域的步骤，以及对二进制图像数据执行光学字符识别处理 文字图像的字符区域，以获得字符区域中识别的字符图像的计算机代码。

公开(公告)号：US06512848B2

公开(公告)日：2003-01-28

申请号：US08751677

申请日：1996-11-18

Applicant: Shin-Ywan Wang ,  Toru Niki

Inventor： Shin-Ywan Wang ,  Toru Niki

IPC: G06K934

CPC classification number: G06K9/00456

Abstract: A method for increasing the accuracy of image data classification in a page analysis system for analyzing image data of a document page. The method includes inputting image data of a document page as pixel data, analyzing the pixel data in order to locate all connected pixels, rectangularizing connected pixel data into blocks, analyzing each of the blocks of pixel data in order to determine the type of image data contained in the block, outputting an attribute corresponding to the type of image data determined in the analyzing step, and performing optical character recognition to attempt to recognize a character of the block of image data in the case that the analyzing step cannot determine the type of image data contained in the block.

Abstract translation: 一种用于在用于分析文档页面的图像数据的页面分析系统中提高图像数据分类的精度的方法。 该方法包括输入文档页面的图像数据作为像素数据，分析像素数据以定位所有连接的像素，将连接的像素数据矩形化为块，分析像素数据块中的每一个以确定图像数据的类型 输出与分析步骤中确定的图像数据的类型对应的属性，并且在分析步骤不能确定图像数据的类型的情况下，执行光学字符识别以尝试识别图像数据块的字符 包含在块中的图像数据。

公开(公告)号：US6125204A

公开(公告)日：2000-09-26

申请号：US742222

申请日：1996-10-23

Applicant: Tadanori Nakatsuka ,  Toru Niki ,  Kazuyuki Saito ,  Akihiro Matsumoto

Inventor： Tadanori Nakatsuka ,  Toru Niki ,  Kazuyuki Saito ,  Akihiro Matsumoto

IPC: G06K9/20 ,  G06K9/32 ,  G06K9/34

CPC classification number: G06K9/00456 ,  G06K9/3208 ,  G06K2209/01

Abstract: Image processing method and apparatus are provided. Image information is inputted. The input image information is divided into a plurality of areas. Radial line segments are extended in upper, lower, left, and right directions from a point in an arbitrary notice area in the divided areas and connection information between the line segments and the input image is detected. Characteristics of the notice area are discriminated in accordance with the connection information detected. The connection information of the areas is a relative position of each area. The image information is dot information. The characteristics to be discriminated are such that the image information of the area is a headline.

Abstract translation: 提供图像处理方法和装置。 输入图像信息。 输入图像信息被分成多个区域。 径向线段从分割区域中的任意通知区域的点向上，下，左，右方向延伸，并且检测线段与输入图像之间的连接信息。 根据检测到的连接信息来辨别通知区域的特征。 区域的连接信息是每个区域的相对位置。 图像信息是点信息。 要区分的特征使得该区域的图像信息是标题。

公开(公告)号：US5774584A

公开(公告)日：1998-06-30

申请号：US685779

申请日：1996-07-24

Applicant: Akihiro Matsumoto ,  Toru Niki ,  Tadanori Nakatsuka ,  Kazuyuki Saito

Inventor： Akihiro Matsumoto ,  Toru Niki ,  Tadanori Nakatsuka ,  Kazuyuki Saito

IPC: G06T7/60 ,  G06K9/20 ,  G06K9/46 ,  G06K9/66

CPC classification number: G06K9/00449

Abstract: Image processing method and apparatus are provided. A histogram of image information is obtained. A peak position is detected from the histogram obtained. A small area is decided by using the detected peak position as a reference. A histogram of the decided small area of the image information is obtained. A specific point is determined from the peak position of the histogram of the small area obtained. The size of small area is decided in accordance with a peak width of the histogram of the image information. The image information is inputted by a scanner. A table can be used as image information and the specific point which is decided is an intersection of ruled lines of the table. A cell area of the table is judged from the intersection decided.

Abstract translation: 提供图像处理方法和装置。 获得图像信息的直方图。 从获得的直方图中检测出峰值位置。 通过使用检测到的峰值位置作为参考来决定小区域。 获得所确定的图像信息的小区域的直方图。 从获得的小区域的直方图的峰值位置确定特定点。 小区域的大小根据图像信息的直方图的峰值宽度来决定。 图像信息由扫描器输入。 一张桌子可以作为图像信息，决定的具体点是表格的一条线。 从交会决定表格的单元格区域。

公开(公告)号：US5703962A

公开(公告)日：1997-12-30

申请号：US934313

申请日：1992-08-25

Applicant: Toru Niki ,  Nasami Kugai ,  Tadanori Nakatsuka

Inventor： Toru Niki ,  Nasami Kugai ,  Tadanori Nakatsuka

IPC: G06F3/0485 ,  G06F17/21 ,  G06K9/03 ,  G06K9/20 ,  G06K9/00

CPC classification number: G06F17/21 ,  G06F17/212 ,  G06F3/0485 ,  G06K9/033 ,  G06K9/2054 ,  G06K2209/01

Abstract: In a character recognition method and apparatus, image information is input, the input image information is displayed, a desired area in the input image information is assigned, the number of characters contained in the assigned area is calculated, and the size of a picture frame for displaying a result of character recognition of the image information is determined according to the calculated number of characters.

Abstract translation: 在字符识别方法和装置中，输入图像信息，显示输入图像信息，分配输入图像信息中的期望区域，计算包含在分配区域中的字符数，并且图像帧的大小 根据计算出的字符数确定用于显示图像信息的字符识别结果的显示。

公开(公告)号：US5671293A

公开(公告)日：1997-09-23

申请号：US700828

申请日：1996-08-21

Applicant: Toru Niki

Inventor： Toru Niki

IPC: G06K9/62

CPC classification number: G06K9/6234 ,  G06K9/6218 ,  G06K9/6235

Abstract: A pattern recognizing apparatus comprises a feature vector extractor to extract a feature vector of an input pattern; a converting unit to convert the feature vector extracted by the feature vector extractor into the feature vector which is effective for selection between categories; a classification processing unit for calculating an inner product with a predetermined basic vector for the converted vector after the conversion was performed by the converting unit; and a category selecting unit for selecting and outputting a category group to which the input pattern belongs with reference to a predetermined category table on the basis of the result of the calculation by the classification processing unit.

Abstract translation: 图案识别装置包括：特征向量提取器，用于提取输入图案的特征向量; 转换单元，用于将由特征向量提取器提取的特征向量转换成对于类别之间的选择有效的特征向量; 分类处理单元，用于通过转换单元执行转换后的转换矢量的具有预定基本矢量的内积; 以及类别选择单元，用于基于分类处理单元的计算结果，参考预定类别表来选择和输出输入模式所属的类别组。