-
公开(公告)号：JPH0887495A
公开(公告)日：1996-04-02
申请号：JP22154394
申请日：1994-09-16
Applicant: IBM JAPAN
Inventor： YAMASHITA AKIO , HIRAYAMA TADAKI
Abstract: PURPOSE: To paste text display data of a text editing program, which are tabulated and not formatted, to a table area of a spread sheet program as normal cell data. CONSTITUTION: The titled purpose is attained by providing a spread sheet type data converting means which processes data sent to a clip board or transferred to the spread sheet program by DDE on the side of the text editing program where the data are cut and pasted. The sperad sheet type data converting means converts the data into a format wherein table strucyture such as CSV is reflected by detecting ruled line information of the table of the text editing program, spaces present between columns, etc.
-
公开(公告)号：JPH0668300A
公开(公告)日：1994-03-11
申请号：JP33377891
申请日：1991-12-18
Applicant: IBM
Inventor： YAMASHITA AKIO , TOYOKAWA KAZUHARU
Abstract: PURPOSE: To provide a method for extracting a tree structure by using the result of the picture analysis of an actual picture, and preparing a general layout model. CONSTITUTION: After a tree structure is automatically extracted based on the picture analysis of a document, a user operates graphical correction so that the tree structure or a layout model can be newly prepared. That is, an image inputted document picture 51 is physically analyzed, a separator which is the most likely to separate the constituting elements of a document is automatically extracted, and a document picture 51 is divided into plural areas 51A-51G according to the information of the separator. Then, this area division is displayed on a display device 13 with the document picture 51, and the user adds necessary correction by an interactive input so that the desired tree structure can be defined. Moreover, a parameter is set to each node 61A-61G of this tree structure so that a general layout model 80 can be completed.
-
公开(公告)号：JPH04278634A
公开(公告)日：1992-10-05
申请号：JP5554391
申请日：1991-02-28
Applicant: IBM
Inventor： AMANO TOMIO , YAMASHITA AKIO , TAKAHASHI HIROYASU
Abstract: PURPOSE: To enable a user to easily describe a tree and also to generate a data structure showing the tree in a memory by showing the attributes to be given to the nodes of the tree in every type of them. CONSTITUTION: A table contains the columns of the levels (depth), the node type names and plural attributes, and a user writes the necessary information into these columns. For example, 0 is written in the level column for a node A that is positioned at the highest part of a tree together with 1 written for nodes B to E as the children of A and the sequence relation of the children reflected on the upper-lower relation respectively. The number of nodes, for example, is written in the attribute column, and a tree shown by this table includes the nodes belonging to the types B to E as the children of the node A as long as both maximum and minimum value are set to 1. In other words, the pointer data can be generated through such interpretation as mentioned above and the user can easily input the pointer data by means of a text editor.
-
公开(公告)号：JPS6472294A
公开(公告)日：1989-03-17
申请号：JP22423387
申请日：1987-09-09
Applicant: IBM
Inventor： TAKAHASHI HIROYASU , YAMASHITA AKIO
Abstract: PURPOSE: To improve the efficiency at recognition time by generating a recognition dictionary in tree structure which considers redundancy by using not binary feature quantities such as whether a pixel is black or white, but multi-valued feature quantities having maximum and minimum values. CONSTITUTION: A system consists of a personal computer 1, a bit map display 2, a scanner 3 and an auxiliary storage device 4, and a recognition part 5, a postprocessing part 6, and a user interface part 7 which are encircled with a broken line are actualized as software. Then the tree-structured dictionary is searched by using a multi-valued feature quantity extracted from an unknown input character pattern, an upper-limit or lower-limit value is assigned to at least one branch in the tree-structured dictionary, and the search of the tree- structured dictionary is quit when the multi-valued feature quantity extracted from the unknown input character pattern for the branching deviates from the upper-limit or lower-limit value. Consequently, the tree-structured dictionary need not be searched unnecessarily and a character can be recognized fast.
-
公开(公告)号：JPH06176188A
公开(公告)日：1994-06-24
申请号：JP32394492
申请日：1992-12-03
Applicant: IBM JAPAN
Inventor： HIRAYAMA TADAKI , AMANO TOMIO , YAMASHITA AKIO
Abstract: PURPOSE:To draw recognized characters by outline font with the same size of characters and locations or the like as a mentioned text. CONSTITUTION:The device is provided with a recognition means providing information on recognized characters having text code information and character layout information, outline font table 34 holding the outline font data of characters, character box expansion means 31 expanding a rectangular circumscribed to the obtained recognition characters by referring to it by the ratio between the character box of the outline font and a character section area and correcting information with the use of the expanded circumscribed rectangular as character box data, means discriminating the section in the recognized character string, and a means correcting the character box data consisting of the circumscribed rectangular expanded by the width which is obtained by equally allocating the width of the character box by the number of characters between sections. When a character is written by outfont based on the information, it can be prevented from becoming small and the sectioned character string can be treated in a batch.
-
Patent Agency Ranking
公开(公告)号：JPH04248687A
公开(公告)日：1992-09-04
申请号：JP2147191
申请日：1991-01-23
Applicant: IBM
Inventor： YAMASHITA AKIO
Abstract: PURPOSE: To accurately detect constitutional elements(CEs) from a document picture in document picture layout analysis using a model even when information to be applied to the model is a little and to allow the model itself to sufficiently correspond to variation the positions or size of the CEs. CONSTITUTION: The model includes the positional relation data of respective CEs, data indicating the existence of respective CEs on paper, data indicating the number of character strings included in each CE located on each leaf of tree structure, and data indicating the number of direct lower CEs of each CE located on a node of the tree structure. In addition to character string areas 1 to 7, a separator for delimiting respective CBs and subseparators to be delimitation candidates are extracted from the document picture. Labels expressing the names of respective CEs are put on the areas 1 to 7 based upon the separator and the subseparators so as not to generate contradiction to the model.
-
公开(公告)号：CA2157972A1
公开(公告)日：1996-03-17
申请号：CA2157972
申请日：1995-09-11
Applicant: IBM
Inventor： YAMASHITA AKIO , HIRAYAMA YUKI
Abstract: A technique to make it possible to paste text display data in a text-editing program, which is arranged in table format and not formed into a geometrical format, onto a table region of a spreadsheet program as normal cell data by providing a table-format converte r for processing data that is sent to a clipboard or transferred to a spreadsheet program by DDE. The table-format data converter converts data to a format reflecting a table structure such as a C SV format by detecting the ruled-line data or spaces between columns of a table of the text-editing program.
-
公开(公告)号：DE3851867D1
公开(公告)日：1994-11-24
申请号：DE3851867
申请日：1988-08-25
Applicant: IBM
Inventor： TAKAHASHI HIROYASU GOH HIGASHI , YAMASHITA AKIO
-
公开(公告)号：CA1299292C
公开(公告)日：1992-04-21
申请号：CA575692
申请日：1988-08-25
Applicant: IBM
Inventor： TAKAHASHI HIROYASU , YAMASHITA AKIO
Abstract: CHARACTER RECOGNITION APPARATUS A recognition dictionary of a tree structure is created by using multi-value feature with maximum and minimum values such as a white run length, rather than a binary feature value such as black or white pels, so that the dictionary can be created through very simple operations with fewer samples. In addition, recognition error due to erroneous branching can be detected at the time of searching of the dictionary tree by using a ternary tree instead of a binary tree, to improve the efficiency in the recognition. Combination of the recognition technique with the tree structure dictionary and the pattern matching technique also provides a feature where the former can efficiently narrow down the candidate character categories, and then the latter can order the candidates. Thus, because high speed and high recognition ratio can be realized, and the processing is simple, a practical system can be implemented on a general purpose personal computer in software only, without using special hardware. Additionally, it is easy to improve the recognition ratio through adaption because addition or modification of the recognition dictionary can be very easily performed. JA9-82-002
-
公开(公告)号：DE3851867T2
公开(公告)日：1995-04-20
申请号：DE3851867
申请日：1988-08-25
Applicant: IBM
Inventor： TAKAHASHI HIROYASU GOH HIGASHI , YAMASHITA AKIO
-
-
-
-
-
-
-
-
-
Search Results
10
records
(took 0.019 seconds)
