JPH02136956A - Extracting method for layout information - Google Patents

Abstract

PURPOSE:To improve the labor saving effect by controlling the document reading results at the head position of a line after dividing them into blocks and handling the output character strings of the reading results at the side of a user device as a document where the line heads are arranged. CONSTITUTION:The characteristic symbols or symbol strings, e.g., G1 - G6 groups 1 - 6 are stored in a dictionary table to be added to the head, the middle and the end of a document respectively. Then the (candidate) characters or character strings of the head or the middle of lines, i.e., the reading results are compared with a dictionary so that an itemized part, for example, can be detected out of a document. Therefore the document reading results can be divided into blocks for each line and the head, middle and end positions of lines are controlled for each block. Thus the output character strings, i.e., the reading results can be treated at the user device side as a document where the head, middle and end of lines are arranged. As a result, a user can omit the layout connection jobs an the labor saving effect is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method for extracting layout information using a character reading device.

[Conventional technology]

Ordinary printed documents have layouts such as broken paragraphs or bullet points to make it easier for readers to understand the meaning and content, and such layout information is useful to users. Therefore, when inputting a printed document using a character reading device, for example, a blank area at the beginning of a sentence is converted into text as a continuous blank by the character reading device, and the number of characters is calculated by dividing the width W of the blank area by the character width. I'm trying to convert it.

[Problem to be solved by the invention] The above method has the following drawbacks.

■The width of the blank area at the beginning of a sentence is not necessarily specified in terms of the number of characters during phototypesetting, but is sometimes specified in mm units, but the result of this division is generally not divisible.

■For example, the width of the number "1" is smaller than that of the kanji "yomi". Therefore, in the case of a horizontally written document, when "1" comes at the beginning of a sentence, the width of the blank area becomes larger than when "yomi" comes at the beginning of the sentence.

■Since the estimated value of character width is calculated for each line, there will be some differences between them.

■For example, in a list, there are cases where you want the middle character string positions to be aligned.

For these reasons, the number of blank characters at the beginning of a sentence is not constant even within a paragraph, and even though the number of blank characters at the beginning of a sentence and the end of one line in a line are uniform, the number of blank characters at the beginning of a sentence is not constant for the user of a character reading device. This results in the output of a document in which the end of one line is not aligned. For this reason, even if there is no error in reading the document itself, the user has to edit the document that has been read, and it is difficult to say that the labor-saving effect of the character reading device is fully demonstrated. Word processors and civil processing programs generally do not cover all the functions of phototypesetting machines that create printed documents.

Currently, only limited functions are supported especially regarding layout information and font information.

In summary, the purpose of character reading devices is to save labor during the initial input of documents and when reusing existing documents. First, it can be said that it is converted into reusable computer information.

However, when a conventional character reading device processes spaces at the beginning of a line or in the middle of a line, even if a document in which the characters at the beginning of a sentence or in a line are aligned is input, a document with the characters at the beginning of a line or in a line not aligned is output. Therefore, even if there is no error in the reading of the characters, the user has to edit the read document, resulting in the problem that the labor-saving effect is not fully utilized.

Therefore, an object of the present invention is to enable layout information to be extracted from a character reading device.

[Means to solve the problem]

Examples of cases in which it is necessary to align the beginning of a line or the end of each line of characters within the same paragraph include bulleted lists. In such a case, for example (1), (2), (3)...
Symbols or symbol strings such as ..., ■, ■, ■, etc. are often used. Therefore, such characteristic symbols or symbol strings are stored in a dictionary, and when these symbols or symbol strings are detected at the beginning of a line or in a line of a manuscript, it is indicated at the beginning of the line or in the line. Insert information, for example a tab symbol. Then, these tab symbols are replaced with a certain number of spaces so that the beginning or middle of a line is aligned when the document is output. In addition, in lists etc.
(blank), "...", etc. are characteristic symbols or symbol strings.

[Effect]

For example, (1), (2), (3)...
By detecting characteristic symbols or symbol strings at the beginning of sentences, such as (a), (b), (C), etc., it is possible to control the position of the beginning of a line by dividing the document reading result into blocks line by line. Therefore, the output character string of the reading result can be treated as a document with the beginnings of the lines aligned on the side of the device that uses it, and the user can omit the work of modifying the layout, increasing the labor-saving effect.

In addition, in the case of a directory, by using "" (blank) as a characteristic symbol or symbol string, the character positions in the lines can be adjusted, making it possible to input the data in the same layout as the original. This allows items such as name, place of birth, date of birth, and address to be extracted at the same time, making it easier to create a database.

〔Example〕

The figure is an explanatory diagram for explaining the present invention in detail. This shows an example of a characteristic symbol or symbol string (hereinafter also referred to as an initial character string) that appears at the beginning of a line in a bulleted document, and is stored as a dictionary in a table in a character reading device or computer. In the figure, group Gl is an example of numbers in parentheses (1), (2), (3), etc., and group G2 is an example of numbers in circles: ■, ■, ■, ■...・
For example, group G3 has real child letters a, b, c, d...
For example, group G4 consists of uppercase letters A, B, C, D, etc.
. . , groups G5 and G6 are examples of the symbols "■" and ".", respectively.

Here, in the case of bullet points, both "■" and r (1) J are used, but they are never used together. However, r(1)1 and r(2)1 are used together. That is, r(1), (2), (3), (4
)...", "■,■.

"■, ■..." each form a group. on the other hand,
"・" and "■" form a group by themselves. Therefore,
Numbers are assigned to groups so that they can be distinguished from each other, and numbers are assigned if there is a hierarchy within the group. m), (2), (3), (4)...''
Then, the order of (1) is 1, and the order of (2) is 2. [(
In a), (b), (c), (d) -J, the ranking of (a) is 1 and the ranking of (b) is 2. The order of ■ is “■, ■, ■.

It is 2 for "■..." and 1 for "■, ◎, 0.0...".

Also, as shown in group G1, symbol (column) r(1)
” is r(1)” or “(” “l)”
' (J 'I
J') It can also be recognized as a combination of the three letters J. If there are multiple combinations of character strings like this, remember all of them. However, select one typical combination method and make it easy to understand, and place this at the beginning of the combination (for example. Then, when cutting out lines from a document, the width of the blank area at the beginning of the line can be converted into a physical quantity. It is stored in units (for example, mm, point, dat) in a table called a blank table at the beginning of a line.

After creating such a table, (
Compare the character string (candidate) with the initial character string dictionary, and if there is a match, store the line number, the group number of the initial character string, and the order within the group. The range in which initial letter strings of the same group number appear in order in a document or page is examined and is called a block. A collection of lines that do not belong to any group is called a root block. A block consists of multiple consecutive lines. rA, B, C...'' etc. may appear at the beginning of a line even if they are not used as an initial character string, so if they do not follow the order or are on a single line, they are not considered a block. Therefore, such a line belongs to the root block even if the initial string is detected. Note that the root block may be composed of discrete rows. Then, the widths of blank areas in the same block are retrieved from the line head blank table, averaged, and the line head blank table is rewritten. Do the same for the root block. Additionally, a tab symbol is inserted at the beginning of the line of the text of the recognition result.

Thereafter, when the document is output, the tab symbol at the beginning of each line is replaced so that the beginnings of the lines are aligned on a device that uses text. In other words, if it is a word processing program that can specify the position of the beginning of a line in millimeters, it will convert tabs according to that format, and if it is a word processor that can specify the position of the beginning of a line only by the number of spaces, it will convert the document or page by referring to the beginning of line blank table. Converts a constant character width to a constant number of spaces. Lines within a block without an initial string are lower in rank than lines with an initial string. The number of spaces inserted for this purpose is made to match the number of characters of the representative combination in the initial character string dictionary.

The above is the case at the beginning of a line; for example, in a directory, ['' (blank) is the focus symbol, and tabs are added to arrange the list. It goes without saying that the same process as above can be performed at the end of the second line if necessary.

〔Effect of the invention〕

Characteristic symbols or symbol strings at the beginning of a line or at the end of three lines of a document, such as (1), (2), (3)... (a)
, (b), (c)...... Store symbols or symbol strings such as "" (blank) in a dictionary table, and select (candidate) characters or character strings at the beginning or in the line of the reading result. For example, bullet points in a document can be detected by comparing them with a dictionary. As a result, the reading results of a document can be divided into blocks line by line, and the line start position, middle line position, and line end position can be controlled for each block. .

It becomes possible to handle the document as a document with the same endings for all three lines.

As a result, the user can save the work of modifying the layout, and the advantage is that the labor-saving effect is increased.

[Brief explanation of the drawing]

The figure is an explanatory diagram for explaining the method according to the present invention. Code explanation 01-G5...Symbol or symbol string group. Agent Patent Attorney Akio Namiki Agent Patent Attorney Kiyoshi Matsuzaki Procedural Amendment (Method) 1. Display of the case Mr. Shida Bunki Patent Application No. 289963 of 1989 2. Method for extracting invention title layout information 3. Amendment Relationship with the case of a person who does
3) Fuji Electric Co., Ltd. 4, Agent 8105 Telephone 03 (580)
95135, date of amendment order 6, subject of amendment (one other person) December 20, 1985 Specification and drawings 7, contents of amendment (1) Specification, page 6, line 2 (detailed description of the invention) 10th line of the specification (brief explanation column), the phrase ``The drawing is'' was corrected to ``Fig. Figure 1 has been corrected to ``. (2) Amend the drawing as shown in the attached sheet (that is, amend the conventional drawing to a drawing with Figure 1 and figure numbers attached throughout the drawing. Figure 1)

Claims (1)

[Claims] When a predefined characteristic symbol or symbol string is memorized and the memorized symbol or symbol string is detected in a document to be read, a predetermined symbol indicating this is displayed at the corresponding position. A method for extracting layout information, characterized in that the information is generated and made available as information for document layout.