JP2001216296A - Character retrieval device, character retrieval method, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information retrieval device, an information retrieval method, ad storage medium capable of easily retrieving character information whose read is unknown, relating to a character information retrieval device, a character information retrieval method and a recording medium for retrieving the character information of a character whose read is unknown. SOLUTION: By inputting similar Kanji (Chinese character) containing the components of retrieval Kanji, the retrieval Kanji is retrieved.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character information retrieval apparatus, a character information retrieval method, and a recording medium, and more particularly to a character information retrieval apparatus, a character information retrieval method, and a recording medium for retrieving character information of characters that cannot be read. About.

2. Description of the Related Art Kana-kanji conversion is performed in kanji input in a computer. In the kana-kanji conversion, a kanji corresponding to the input reading is searched by inputting the reading of the kanji, and the conversion is performed.

In the kana-kanji conversion, the conversion cannot be performed unless the kanji to be input is read. Therefore,
There is a need for a kanji search device that can efficiently input kanji that cannot be read.

[0004]

2. Description of the Related Art Generally, a computer inputs characters such as kanji using a kana-kanji conversion program.
The kana-kanji conversion program searches for and displays kanji corresponding to the reading by inputting the kanji reading from a keyboard. For this reason, it is difficult to input kanji that cannot be read. Therefore, the kana-kanji conversion program uses
Characters that cannot be read can be searched by stroke number search or radical search.

In the stroke number search of the kana-kanji conversion program, the user inputs the stroke number of a character to be known. The kana-kanji conversion program searches for characters corresponding to the number of strokes input and displays all the characters. The user selects a required character from the displayed characters.

In the radical search of the kana-kanji conversion program, a user selects a radical of a character to be known from a radical list. The kana-kanji conversion program searches for a character having a radical input by the user and displays all of the characters.
The user selects a required character from the displayed characters.

[0007]

However, when performing a character search using a conventional kana-kanji conversion program, it is necessary to input the number of strokes in the stroke number search, and to select a radical in the radical search. Therefore, when inputting the number of strokes, the user needs to count and input the number of strokes of the search character.
In addition, in the radical search, the radical cannot be input by ordinary kanji conversion, so it is necessary to select and enter the radical of the search character from a preset radical list or the like. For this reason, it takes time to input for the search.

In the stroke number search, when there are many characters having the same stroke number, many character candidates are displayed. In such a case, necessary characters cannot be easily selected. The present invention has been made in view of the above points, and has as its object to provide an information search device, an information search method, and a storage medium that can easily search for character information that cannot be read.

[0009]

Means for Solving the Problems Claims 1 and 1 of the present invention
Nos. 2 and 13 allow the user to input a character, extract components constituting the character from the input character, and search for a character corresponding to the extracted component.

According to the first, twelfth and thirteenth aspects, a required character can be searched by inputting a character having parts included in the character to be searched. Therefore, the character can be searched by inputting the character, and the input operation for the search can be easily performed.

According to a second aspect of the present invention, a plurality of characters are input, and components constituting each of the input characters are extracted.

According to the second aspect, it is possible to input a plurality of characters including parts constituting the character to be searched, and it is possible to perform a search with higher accuracy.

According to a third aspect of the present invention, a search is made for a character including at least one component extracted from the input character.

According to the third aspect, by searching for a character including at least one component extracted from the input character, an input operation for the search can be easily performed.

According to a fourth aspect of the present invention, for each input character, a character which includes at least one part extracted from the character is searched.

According to the fourth aspect, for each input character, a character including at least one part extracted from the character is searched for, so that the input operation for the search can be easily performed.

According to a fifth aspect of the present invention, the searched character information is sorted according to the configuration of the parts extracted from the input characters and the configuration of the parts constituting the searched characters.

According to the fifth aspect, a character having a component configuration similar to the component configuration extracted from the character input from the searched characters is sorted to thereby obtain a character having a component configuration similar to the input character. Easy to search.

According to a sixth aspect of the present invention, the retrieved character information is sorted according to the difference between the number of parts constituting the input character and the number of parts constituting the retrieved character.

According to the sixth aspect, it is possible to narrow down the characters searched in the order in which the number of parts is approximated.

According to a seventh aspect of the present invention, the searched characters are sorted according to the difference between the arrangement position of the parts constituting the input character and the arrangement position of the parts of the searched character.

According to the seventh aspect, it is possible to narrow down the characters searched in the order in which the arrangement of the parts is approximated.

According to an eighth aspect of the present invention, a part constituting a character including a part to be excluded is extracted, and character information including the non-target part is excluded from the searched characters.

According to the eighth aspect, the characters to be excluded from the search information can be removed by inputting the characters including the parts to be excluded from the search information, so that the characters can be narrowed down easily.

According to a ninth aspect of the present invention, a character having the number of strokes that matches the number of strokes of the character to be searched is searched.

According to the ninth aspect, characters can be narrowed down by the number of strokes.

According to a tenth aspect, a difference in the number of strokes between the input character and the searched character is detected, and the searched characters are narrowed according to the difference in the number of strokes.

According to the tenth aspect, the characters can be narrowed down from the searched characters according to the difference in the number of strokes.

The eleventh aspect is characterized in that a character including all parts extracted from the input character is searched.

According to the eleventh aspect, a character to be searched can be narrowed down by searching for a character including all parts extracted from the input character.

[0031]

FIG. 1 is a block diagram showing a first embodiment of the present invention.

The character retrieval apparatus 1 of this embodiment is an apparatus for retrieving a kanji having a highly similar component configuration from the kanji component configuration input as a search kanji. Character search device 1
Are the input unit 2, the display unit 3, the RAM 4, the auxiliary storage device 5,
It comprises a CPU 6 and a bus 7.

The input unit 2 is composed of input devices such as a keyboard and a mouse, and is used for inputting by reading a search kanji and the like and various instructions. The display unit 3 has a CR
It is composed of a display such as T, LCD, etc., and displays a search kanji and similar kanji by displaying a search screen.

The RAM 4 is used as a working storage area for the CPU 6. The auxiliary storage device 5 is composed of a nonvolatile storage device such as a hard disk drive, and stores data such as a search program and character information. The programs and data stored in the auxiliary storage device 5 are stored in the RAM 4
And is processed by the CPU 6.

The CPU 6 executes a search process according to the search program stored in the auxiliary storage device 5. Bus 7
Are the input unit 2, the display unit 3, the RAM 4, the auxiliary storage device 5,
It is connected to the CPU 6 and transfers addresses, data and control signals.

The search program and data of this embodiment are as follows:
It is also possible to provide the character search device 1 by recording it on a portable medium such as a CD-ROM or a floppy disk, or by holding it in a device accessible via a communication line.

When the search program and data are provided by being recorded on a portable medium, the character search apparatus 1 may be provided with a drive for accessing the medium and reading them. Further, when the search program and data are stored in another device, a communication control device such as a modem or a network card for accessing the device via a communication line may be provided.
If connected directly or indirectly to the bus 7, programs and data can be transferred between the components constituting the character search device 1.

Further, the retrieval program and data obtained in this way may be configured to be directly transferred to the RAM 4 and executed without being stored in the auxiliary storage device 5 once.

Here, the dictionary data stored in the auxiliary storage device 5 will be described.

The kanji dictionary data used in this embodiment has an SMF (Separation Management System) file structure. The SMF file is a file in which one character is divided into a plurality of parts and managed. Note that SM
Details of the F file are described in JP-A-7-64533.

FIG. 2 is a diagram for explaining the SMF file. FIG. 2A shows the entire character C1.
(B)-(E) are the figures which show the parts P1-P4 which comprise the character C1.

In the SFM file, for example, a character C1 of "love" is divided into parts P1 to P4. Part P1
To P4 are arrangement positions (X11, Y11), (X12, Y12), (X21, Y2) based on the reference position (0, 0) of the frame F, respectively.
1), (X22, Y22), (X31, Y31), (X32, Y32), (X41,
Y41) and (X42, Y42) are set. In the SFM file, a character pattern is created by designating a used part number and position information according to a character code. In the present embodiment, character search is performed from similar kanji and allomorphs using the component information of the SMF file.

Next, the dictionary data used in this embodiment will be described.

FIG. 3 is a diagram showing the data structure of dictionary data according to the first embodiment of the present invention.

The dictionary data 10 is stored in the auxiliary storage device 5 and is obtained from a font data section 11, a kanji code / part number dictionary 12, a part number / kanji code dictionary 13, a part coordinate dictionary 14, and a part number / kanji code dictionary 15. Be composed.

The font data section 11 stores character patterns corresponding to kanji codes. The kanji code is input as a logical code. The kanji code / part number dictionary 12 stores a part number group corresponding to the kanji code and component information of each part. With the kanji code / part number dictionary 12, component information constituting the kanji can be retrieved from the kanji code.

The part number / kanji code dictionary 13 stores kanji codes including part numbers for each logical number. With the part number / kanji code dictionary 13, a kanji code including a predetermined part number can be searched.

The component coordinates dictionary 14 stores component coordinates for each logical code. In the component coordinate dictionary 14, component information is stored for each component number. Number of parts / Kanji code dictionary 1
Reference numeral 5 stores the number of parts and the corresponding kanji code for each logical code. The kanji code corresponding to a predetermined number of parts can be acquired from the number of parts / kanji code dictionary 15.

Next, the kanji code / part number dictionary 12 will be described in detail.

FIG. 4 shows a kanji code / code according to the first embodiment of the present invention.
FIG. 4 shows a data configuration diagram of a part number dictionary.

The kanji code / part number dictionary 12 includes an unused area 21, a kanji code offset / part number area 22, and a part information area 23. In the unused area 21, the logical codes “0000 0000” to “0000”
“0327” is allocated. In the kanji code offset / part number area 22, the logical code "0000 0
328 ”to“ 0007 FFFF ”are allocated. In the component information area 23, the logical code "0008
0000 ”to“ xxxx xxxx ”are assigned.

No information is written in the unused area 21.

Each information area of the kanji code offset / number-of-parts area 22 includes an offset storage section 22a and a number-of-parts storage section 22b. The offset storage unit 22a stores an offset up to a logical code in which the component information of the input kanji code is stored. The number of parts constituting the kanji corresponding to the input kanji code is stored in the parts number storage unit 22b.

In each information area of the part information area 23, the part numbers # 1, # 2,.
.. and position information (x1, x2) indicating the size, (x
2, y2) are stored.

By referring to the kanji code / part number dictionary 12 using the kanji code, the number of parts can be detected from the part number storage unit 22b. Further, the logical code in which the component information is stored is designated from the offset stored in the offset storage unit 22a. The component information area 23 of the kanji code input by the logical code specified by the offset can be referred to. The part number and the part size of the corresponding kanji code can be detected from the part information area 23.

Next, the part number / kanji code dictionary 13 will be described in detail.

FIG. 5 shows a data structure diagram of the part number / kanji code dictionary according to the first embodiment of the present invention. The part number / kanji code dictionary 13 includes an unused area 31, an offset / kanji number area 32, and a kanji information area 33.

In the unused area 31, the logical code "000"
The areas from “0000” to “0000 0327” are allocated. In the offset / kanji number area 32, logical codes "0000 0328" to "0007 FFFF"
Are allocated. The kanji information area 33 contains logical codes “0008 0000” to “xxxx xxx”.
An area “x” is allocated.

The unused area 31 is unused.

The offset / kanji number area 32 includes an offset storage section 32a and a kanji number storage section 32b.
The offset storage unit 32a stores an offset up to a logical code in which kanji information corresponding to a part number is stored. The kanji number storage unit 32b stores the kanji number of kanji including the input part number. In the kanji information area 33, the kanji code # 1 including the part of the designated part number,
$ 2 is stored.

The part number / kanji code dictionary 13 is referred to by the part number. The part number is input as a logical code. By referring to the part number / kanji code dictionary 13 with a logical code corresponding to the part number, the number of kanji including the part number specified in the kanji number storage section 32b can be detected. Further, the offset to the logical code storing the kanji information of the designated part number can be detected by referring to the offset storage unit 32a. The kanji information area is referred to by the detected offset, and the kanji code including the specified component can be referred to.

The component coordinate dictionary 14 stores component coordinates for each component number.

Next, the number of parts / kanji code dictionary 15 will be described in detail.

FIG. 6 shows a data structure diagram of the number of parts / kanji code dictionary according to the first embodiment of the present invention.

The parts number / kanji code dictionary 15 is composed of a maximum parts number area 41, an offset / kanji number area 42, and a kanji information area 43.

The area of the logical code “0000 0000” is assigned to the maximum number of parts area 41. The offset / kanji number area 42 has a logical code “0000”.
0008 ”to“ 0000 0080 ”. In the kanji information area 43, the logical code "0000"
0100 ”to“ xxxx xxxx ”.

The maximum number of parts area 41 stores the maximum number of parts. The offset / kanji number area 42 includes an offset storage section 42a and a kanji number storage section 42b.

The offset storage unit 42a stores an offset up to a logical code in which kanji information corresponding to a specified number of parts is stored. The kanji number storage unit 42b stores the number of kanji consisting of the specified number of parts.
The kanji information storage unit 43 stores kanji codes # 1 and # 2 of kanji composed of the designated number of parts.

The parts number / kanji code dictionary 15 is referred to by the parts number. The number of parts is input as a logical code. The number of kanji can be detected from the kanji number storage unit 42b by referring to the logical number corresponding to the number of parts in the parts number / kanji code dictionary 15. The maximum number of parts can be detected by referring to the number of parts / kanji code dictionary 15 with the logical code “0000 0000”.

The kanji information storage unit 43 can refer to the offset stored in the offset storage unit 42a. By referring to the kanji information storage unit 43, it is possible to refer to the kanji code of the kanji composed of the specified number of parts.

Next, functions executed by the CPU 6 will be described in detail.

FIG. 7 is a functional block diagram of the CPU according to the first embodiment of the present invention.

The function of the CPU 6 includes a kanji search section 51 and a cooperation information output section 52.

The kanji search section 51 executes a kanji search using the dictionary data 10 stored in the auxiliary storage device 5. The cooperation information output unit 52 outputs the kanji searched by the kanji search unit 51 as text or a character pattern.

Next, the kanji search section 51 will be described in detail.

The kanji search section 51 includes a search part information acquisition section 6
1. Result candidate extraction unit 62, non-target processing unit 63, component coincidence determination unit 64, total stroke count determination unit 65, component score determination unit 6
6, a component number search unit 67, an individual component determination unit 68, and a character area difference determination unit 69.

The retrieval part information acquiring section 61 performs the kanji code / part number dictionary 12 according to the kanji input from the input section 2.
, The part number and the part position size information constituting the input kanji are searched. The result candidate extraction unit 62 refers to the part number / kanji code dictionary 12 according to the part number acquired by the search part information acquisition unit 61 and extracts the kanji code of the kanji including the part number as a search result candidate. The extraction result of the search candidate extraction unit 62 is a result candidate string X
It is stored in the RAM 4.

The search candidate string extracting unit 62 extracts a kanji code to be determined as a search result based on the part number and the position information acquired by the search part information acquiring unit 61. In addition,
The extraction of the result candidates includes an extraction method including at least one or more part numbers constituting each search kanji in all kanji strings,
There is an extraction method including a part number constituting each search kanji.

The non-target processing unit 63 includes a result candidate extracting unit 62
The kanji code including the part of the search kanji specified as the target is excluded from the kanji code group acquired in the step. The part matching degree determination unit 64 ranks the result candidate kanji code groups in descending order of the number of matching parts and the number of matching part areas from the kanji code groups acquired by the result candidate extraction unit 62.

The total stroke number judging section 65 extracts a kanji code in which the stroke number difference matches the part number acquired by the retrieval part information acquiring section 61. The part number search unit 67 acquires a kanji code group having the number of parts that matches the number of parts of the searched kanji.

The individual part search unit 67 searches for a kanji code group having the same number of parts. The individual component determination unit 68 compares the arrangement and stroke count difference of each component of the search kanji component with respect to the kanji code group extracted by the component score search unit 67, and determines the kanji whose arrangement and stroke count difference satisfy a predetermined condition. Get codes as result candidate kanji code group. The character area difference determination unit 69
From the kanji code groups acquired by the part matching degree determination section 64, the overlap with the search kanji is checked, and the result candidate kanji code groups are ordered in ascending order of the area difference.

The cooperation information output unit 52 is a text output unit 7
1 and a character pattern generation unit 72.

The text output unit 71 outputs the kanji code selected and instructed from the search result as text information such as shift JIS or Unicode which can be handled by word processing software or the like. The character pattern generation unit 72 outputs the kanji code selected and instructed from the search result as image data such as bitmap and contour (polygon) data that can be handled by word processing software or the like.

FIG. 8 is a flowchart of a kanji search process according to the first embodiment of the present invention, and FIG. 9 is a diagram showing an operation screen according to the first embodiment of the present invention.

The kanji search processing is performed in the following step S1-1.
To S1-11.

Step S1-1 is a process for inputting a search kanji. Step S1-1 is performed by inputting Chinese characters into the search Chinese character input unit 101 of the operation screen 100 shown in FIG. The search kanji input unit 101 receives one or more kanji including at least one component constituting the kanji to be searched. For example, when searching for the kanji “ambiguity”, the user only has to input a relatively simple kanji such as the kanji “ai” having a part similar to the kanji to be searched.

In step S1-2, a process for instructing execution of a search is performed. Step S1-2 is executed by operating the search button 102 on the operation screen 100 shown in FIG. In step S1-2, search conditions to be executed in the subsequent search processing are determined. Step S1-
2 stores the search kanji input in step S1-1 as a kanji string K in the RAM 4.

Here, the kanji string K held in the RAM 4 will be described. FIG. 10 shows a data structure diagram of a kanji string according to the first embodiment of the present invention.

The kanji character string K has a structure in which kanji codes are listed as shown in FIG.

Step S1-3 is a process for acquiring search component information, which is executed by the search component information acquisition unit 61 shown in FIG. In step S1-3, step S1-1
The kanji code / part number dictionary 12 shown in FIG. Based on the reference result of the kanji code / part number dictionary 12, the part number and the part position information of each kanji code of the kanji string K are obtained. The acquired part number is stored in the RAM 4 as a part number string X.

Here, the part number string X held in the RAM 4 will be described.

FIG. 11 shows a data structure diagram of a part number sequence according to the first embodiment of the present invention.

As shown in FIG. 11, the part number sequence X has a structure in which the part numbers 1 to m of the parts constituting the kanji of the kanji code are arranged for each of the kanji codes 1 to I.

Also, component position information (x1, y1), (x
2, y2) is held in the RAM 4 as a position information sequence Y.

Here, the position information sequence Y held in the RAM 4 will be described.

FIG. 12 shows a data structure diagram of the position information sequence according to the first embodiment of the present invention.

As shown in FIG. 12, the position information sequence Y includes position information (x1, y1) for each of the part numbers 1 to m,
(X2, y2) is stored. The position information (x1, y1) indicates the coordinates of the upper left corner of the component frame. The position information (x2, y2) indicates the coordinates of the lower right corner of the component frame. Note that the component frame indicates a frame in which components are stored.

In step S1-4, the end of the number sequence X is determined. In step S1-4, the result candidate extracting unit 6 shown in FIG.
2 is executed. Step S1-4 is a loop processing for sequentially taking out one part number m from the part number string X from the top and continuing the processing of step S1-5 for the number of number strings X.

Step S1-5 is a process for acquiring a group of kanji codes of part number m. Step S1-5 is executed by the result candidate extraction unit 62 of FIG. Step S1-
5 retrieves a kanji code group including the part number m by referring to the part number / kanji code dictionary 13 based on the part number extracted in step S1-3. The extracted kanji code group is temporarily stored in the RAM 4 as a result candidate string Z. Note that a kanji code group included in all kanji strings of at least one of the part numbers constituting each kanji of the kanji string K may be stored in the RAM 4 as a result candidate string Z.

In step S1-6, the end of the result candidate sequence Z is determined. Step S1-6 is executed by the part coincidence determining section 64 shown in FIG. Step S1-6 is:
One kanji code I is sequentially extracted from the result candidate string Z from the top, and the number of result candidate strings Z is equal to the number of steps S1-7 to S1-
9 is a loop process for performing the process shown in FIG.

Step S1-7 is a process for acquiring the component information of the kanji code I. Step S1-7 corresponds to FIG.
Is performed by the part matching degree determination unit 64 shown in FIG. In step S1-7, the kanji code I refers to the kanji code / part number dictionary 12 shown in FIG.

The acquired part information is a part number and position information. The acquired part number is held in the RAM 4 as a part number string x, and the position information is held as a position information string y. Step S1-8 includes a part number string X and a part number string x
This is the process of comparing the part numbers with the above. Step S1-8
Is executed by the part coincidence determining section 64 shown in FIG.

In step S1-8, the part numbers of the part number string X and the part number string x are compared to determine the number of matching parts. The determined number of matching parts is stored in the RAM 4 as a matching parts number sequence α in the order corresponding to each kanji code.

In step S1-9, the position information sequence Y is compared with the position information sequence y. Step S1-9 is executed by the part coincidence determining section 64 shown in FIG. Step S
Step 1-9 performs a process of comparing the position information sequence Y and the position information sequence y to obtain a matching comparison area. The calculated area is represented by R as a comparison area sequence β in the order corresponding to each kanji code.
AM4.

Here, the comparison area sequence β will be described.

FIG. 13 is a diagram showing the data structure of a comparative area column according to the first embodiment of the present invention.

As shown in FIG. 13, the comparison area column β stores a comparison area which is an area that matches the position information of the retrieved kanji code for each of the kanji codes 1 to I in the candidate column. It can be determined that the larger the comparison area, the larger the area that matches the search kanji. Note that an area difference may be obtained instead of the comparison area.

FIG. 14 is a diagram for explaining the operation for obtaining the area difference according to the first embodiment of the present invention.

As shown in FIG. 14, the difference between the retrieved kanji and the result candidate is the area difference, and it can be determined that the smaller the area difference, the greater the degree of matching.

In step S1-10, the search result candidate sequence Z is sorted in the order of the index. Step S1-10
Is executed by the part coincidence determining unit 64 shown in FIG. In step S1-10, the kanji codes of the result candidate sequence Z are rearranged in the order of higher similarity from the matching component count sequence α and the comparison area sequence β. Alternatively, the result candidate sequence Z may be rearranged with reference to only one of the matching component number sequence α and the comparison area sequence β.

Step S1-11 is a process for displaying the result candidate string Z on the result display unit 104 of the operation screen as a search result. In step S1-11, kanji codes are sequentially obtained from the result candidate string Z, and corresponding character data is extracted from the font data stored in the font data section 11 shown in FIG. 3 and displayed in kanji. The search result is displayed on the search result display section 104 of the operation screen 100 shown in FIG.

Note that the number of Chinese characters to be displayed in the search result may be set in the search number input section 103 of the operation screen 100 shown in FIG. In step S1-11, a process is performed to display the number of search results set by the search number input unit 103. In order to introduce the search result into another application, an operation screen 100 shown in FIG.
Of the kanji displayed on the search result display unit 104, the user may place the pointer on the desired kanji and double-click the operation button of the mouse to output the cooperation information.

Next, an actual search operation according to the first embodiment of the present invention will be described.

FIG. 15 is a diagram for explaining the operation of the first embodiment of the present invention.

FIG. 15 shows the search kanji “go”, “grass”,
This shows a search operation when "interest" is input. FIG. 15 shows an example of the input search kanji character string K and search candidates obtained as search results. The input search kanji character string K and the component numbers of the components of each search kanji character are shown in the top row. 8 shows an example of the search candidates obtained in step S1-5 in the leftmost column. FIG. 15 shows search candidates classified into those that include all the search kanji strings K and those that include the part numbers of the search kanji strings K.

As in the present example, "go",
If "Grass" and "Ri" are entered, at least one of the components of the search kanji "Go", at least one of the search kanji "Grass", and at least one of the search kanji "Ri" As an example of a candidate result including all of one part, characters such as “Kurashi”, “Kurashi”, and Chinese character C11 are extracted.

Further, as a candidate including a component used in any of the search kanji “go”, “grass”, and “ri”,
“Tower”, “To”, “Shi”, “Oh”, Kanji C12,
“Toki”, “sub” and the like are extracted as candidates.

For example, a search candidate "shu" containing a part that matches each of the search kanji characters matches the component numbers "101" and "103" of the search kanji "go", and the structure of the search kanji "grass" This indicates that the part number matches the part number “345” and matches the constituent part “215” of the search kanji character “Ri”.

The search candidate “to” including a component used in any of the search kanji matches the part number “101” of the component of the search kanji, as well as the structure of the search kanji “go”. It can be seen that the part numbers “205” and “301” of the part and the part number “345” of the component part of the search kanji “Kusa” also match.

Next, a second embodiment of the present invention will be described.

FIG. 16 is a flowchart of a search process according to the second embodiment of the present invention, and FIG. 17 is a diagram showing an operation screen according to the second embodiment of the present invention. 8, the same components as those in FIGS. 8 and 9 are denoted by the same reference numerals, and description thereof will be omitted.

In step S2-1, a kanji having parts included in the kanji to be excluded is input. Step S2-1
Is executed based on the kanji input to the non-target input unit 201. In the non-target input unit 201, one or more kanji characters partially similar to the kanji characters to be excluded from the search results when the number of search result candidates increases is input. For example,
In the case where many kanji including “Kusakanmuri” such as kanji C21 (Kusakanmuri Ai) as shown in FIG.
A relatively simple kanji such as "grass" may be input to 1.

Step S2-2 is a process for acquiring non-target part information. Step S2-2 is executed by the non-target processing unit 63 shown in FIG.

In step S2-2, the kanji input by the non-target input unit 201 is set to R as a non-target kanji string K '.
AM4 and the same processing as in step S1-3 is performed.

That is, the part number and part position information of each kanji code included in the non-target kanji string K 'are acquired, and these are stored in the RAM 4 as the part number string X'.

Step S2-3 is a process for comparing the part number string X with the part number string X '. Step S2-3
Is executed by the non-target processing unit 63 shown in FIG. In step S2-3, if there is a matching part number by comparing the part number string X and the part number string X ', the corresponding part number and position information are deleted from the part number string X and the position information string Y. I do.

According to this embodiment, kanji including non-target parts can be deleted from search candidates, so that search candidate kanji can be reduced and search can be performed easily.

Next, a third embodiment of the present invention will be described.

FIG. 18 is a flowchart of a search process according to the third embodiment of the present invention, and FIG. 19 is a diagram showing an operation screen according to the third embodiment of the present invention. 8, the same components as those in FIGS. 8 and 9 are denoted by the same reference numerals, and description thereof will be omitted.

Step S3-1 is a process for inputting the number of strokes. In step S3-1, the number of strokes input to the stroke number input unit 201 of the operation screen 100 shown in FIG. 17 is input. The number of strokes input to the stroke number input unit 201 is determined in step S
The difference between the search kanji input in 1-1 and the total number of strokes of the kanji to be searched is input. For example, when the kanji character "ki" is input as a search kanji for the purpose of searching for the kanji character "", the difference between the total stroke number "8" of the kanji character "" and the total stroke number "4" of the kanji character "ki" is obtained. The user may input "4" as the number of strokes.

In step S3-2, a process of acquiring a kanji code group including all parts of the searched kanji is performed. At this time, in step S3-2, first, the part number / kanji code dictionary 12 is referred to from the part number of the part number string X constituting the search kanji input in step S1-1, and the part number m
Kanji code group including. Next, a part number group is extracted from the extracted kanji code group by referring to the kanji code / part number dictionary 11. Then, when the extracted part number group includes all the part numbers of the part number string X, it is stored in the result candidate string Z. The extracted kanji code group is temporarily stored in the RAM 4 as a result candidate string Z.

According to the above operation, for example, even when the search kanji character “animation” is input in order to search by “animation” + “number of strokes”, the component parts “heavy” + “number of strokes + 3” or “power” + “number of strokes” It is no longer necessary to extract up to the kanji of the condition of “difference +9”. Therefore, necessary kanji can be narrowed down.

In step S3-3, it is determined whether or not the difference in the number of strokes between the part number sequence X and the part number sequence x is equal to the number of strokes S. Step S3-3 is executed by the total stroke number determination unit 65 shown in FIG. In step S3-3, the part number sequence X and the part number sequence x are compared to determine the number of strokes. The method of calculating the number of strokes from the part number is obtained from the upper three digits of the part number. For example, the number of strokes of the part number “131” is “1”, and the number of strokes of the part number “1001” is “10”. Note that the part numbering rules are based on Japanese Patent Application Laid-Open No.
No. 4533.

In step S3-3, if the determined difference in the number of strokes matches the number of strokes S, the process proceeds to the next kanji code in the result candidate string Z. If the determined number of strokes does not match the number of strokes S in step S3-3, step S3-4 is executed.

Step S3-4 is a process of excluding the kanji code I from the result candidate string Z. Step S3-4
Is executed by the total stroke number determination unit 65 shown in FIG. In step S3-4, the kanji code I for which the difference in the total number of strokes is determined to be different in step S3-3 is deleted from the candidate row Z obtained in step S3-2.

According to this embodiment, unnecessary kanji can be deleted in accordance with the total number of strokes, so that the number of candidate kanji can be reduced and kanji search can be easily performed.

Next, a fourth embodiment of the present invention will be described.

FIG. 20 is a flowchart of a search process according to the fourth embodiment of the present invention, and FIG. 21 is a diagram showing an operation screen according to the fourth embodiment of the present invention. 8, the same components as those in FIGS. 8 and 9 are denoted by the same reference numerals, and description thereof will be omitted.

Step S4-1 is a process for inputting the image number difference S and the position error D. Step S4-1 corresponds to FIG.
1 and a position error input unit 402 shown in FIG.
This is executed based on the difference in the number of strokes input to.

Step S1--
The difference S in the total number of strokes between the search kanji input in step 1 and the kanji to be searched is input. For example, if the kanji character “my” is input as a search kanji character to search for the kanji character “already”, the stroke count difference “0” is input.

To search for the character C13 "Michi (two points)", enter "Michi (one point)" as a search kanji, and enter the characters C13 "Michi (two points)" and "Michi ( (1 point) is input.

In the position error input section 402, a position error D allowed between the position of the retrieved kanji part input in step S1-1 and the position of the retrieved kanji part is set.

In step S4-2, a process for acquiring a group of kanji codes corresponding to the number of part number strings X is executed. In step S4-2, the part number search unit 67 shown in FIG.
Executed in In step S4-2, the kanji code group having the same number of parts is searched by referring to the parts number / kanji code dictionary 15 shown in FIG. The searched kanji code is
The result candidate string Z is stored in the RAM 4.

In step S4-3, the difference in the number of strokes between
It is determined whether it is within. Step S4-3 is executed by the individual component determination unit 68 shown in FIG. Step 4-
In step 3, the part numbers (number of strokes) are sequentially taken out from the part number string X and the part number string x and compared with each other to determine whether the difference in the number of strokes is within the difference S in the number of strokes.

If it is determined in step S4-3 that the difference between the number of strokes in all of the part number string X and the part number string x is within S, step S4-4 is executed. In step S4-3, if there is any image having a difference in the number of strokes larger than S, step S3-4 is executed, and the image is excluded from the candidate sequence Z. Step S
In step 4-3, only the difference between the numbers of strokes may be determined, or the total number of strokes of each component may be determined.

[0146] In step S4-4, the arrangement difference between the components is D.
It is determined whether it is within. Step S4-4 is executed by the individual component determination unit 68. In step S4-4, position information is sequentially extracted from the position information sequence Y and the position information sequence y, and each is compared to determine whether or not the position error is within an allowable range.

FIG. 22 is a diagram for explaining a position error determination method according to the fourth embodiment of the present invention.

As shown in FIG. 22, the coordinates (x1, y1), (x1 ', and x2) of the rectangle F1 of the component P1 forming the "road" and the rectangle F1' of the component P1 'forming the "road" y
1 ′) and coordinates (x2, y2), (x2 ′, y) of the lower right corner
2 ′) is obtained. If the ratio of the difference between the end points and the coordinates of the entire character is within D, step S1-11 is executed, and the kanji in the candidate string Z is displayed on the operation screen 10 shown in FIG.
0 is displayed in the candidate row display section 104.

If the ratio between the difference between the end points and the coordinates of the whole character is larger than D, step S3-3 is executed, and the corresponding kanji code I is excluded from the candidate string Z.

According to this embodiment, kanji candidates can be narrowed down according to the difference in the number of strokes and the position error of each part.

As described above, according to the present invention, it is possible to search for a kanji by using the similar kanji that can be easily input and the SMF file information even if the kanji is difficult to input because the reading is not understood. .

In each of the embodiments described above, the search for Japanese kanji has been described. However, the searchable characters are not limited to Japanese kanji. Any language and characters can be applied as long as they are managed as.

Also, the component information is not limited to the format of the SMF file information, and it is sufficient that the component information is managed in a format that can realize the present invention.

[0154]

As described above, claims 1 and 1 of the present invention.
According to 2 and 13, a required character can be searched by inputting a character having a part included in the character to be searched. Therefore, there is a feature that the character can be searched by inputting the character, and the input operation for the search can be easily performed.

According to the second aspect, it is possible to input a plurality of characters including parts constituting the character to be searched, and it has a feature that the search can be performed with higher accuracy.

According to the third aspect, there is an advantage that the input operation for the search can be easily performed by searching for a character including at least one part extracted from the input character.

According to the fourth aspect, for each of the input characters, a character including at least one part extracted from the character is searched, so that the input operation for the search can be easily performed. Has features.

According to the fifth aspect, a character having a component configuration similar to the component configuration extracted from the character input from the searched characters is sorted, so that a character having a component configuration similar to the input character is sorted. It has features such as easy search.

According to the sixth aspect, it is possible to narrow down the characters searched in the order in which the number of parts is approximated.

According to the seventh aspect, it is possible to narrow down the characters searched in the order in which the arrangement of the parts is similar.

According to the eighth aspect, since characters to be excluded from the search information can be removed by inputting characters including components to be excluded from the search information, it is possible to easily narrow down the characters. .

According to the ninth aspect, there is an advantage that characters can be narrowed down according to the number of strokes.

According to the tenth aspect, there is an advantage that characters can be narrowed down from the searched characters according to the difference in the number of strokes.

According to the eleventh aspect, it is possible to narrow down the characters to be searched by searching for characters including all parts extracted from the input characters.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first embodiment of the present invention.

FIG. 2 is a diagram for explaining an SMF file.

FIG. 3 is a data configuration diagram of dictionary data according to the first embodiment of the present invention.

FIG. 4 is a data configuration diagram of a kanji code / part number dictionary according to the first embodiment of the present invention.

FIG. 5 is a data configuration diagram of a kanji code / part number dictionary according to the first embodiment of the present invention.

FIG. 6 is a data configuration diagram of the number of parts / kanji code dictionary according to the first embodiment of the present invention.

FIG. 7 is a functional block diagram of a CPU according to the first embodiment of the present invention.

FIG. 8 is a flowchart of a kanji search process according to the first embodiment of the present invention.

FIG. 9 is a diagram showing an operation screen according to the first embodiment of the present invention.

FIG. 10 is a data configuration diagram of a kanji string according to the first embodiment of the present invention.

FIG. 11 is a data configuration diagram of a part number sequence according to the first embodiment of this invention.

FIG. 12 is a data configuration diagram of a position information sequence according to the first embodiment of this invention.

FIG. 13 is a data configuration diagram of a comparison area column according to the first embodiment of the present invention.

FIG. 14 is a diagram for explaining an operation for obtaining an area difference according to the first embodiment of the present invention.

FIG. 15 is a diagram for explaining the operation of the first embodiment of the present invention.

FIG. 16 is a flowchart of a search process according to a second embodiment of the present invention.

FIG. 17 is a diagram showing an operation screen according to the second embodiment of the present invention.

FIG. 18 is a flowchart of a search process according to a third embodiment of the present invention.

FIG. 19 is a diagram showing an operation screen according to the third embodiment of the present invention.

FIG. 20 is a flowchart of a search process according to a fourth embodiment of the present invention.

FIG. 21 is a diagram illustrating an operation screen according to a fourth embodiment of the present invention.

FIG. 22 is a diagram for explaining a position error determination method according to a fourth embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 character search device 2 input unit 3 display unit 4 RAM 5 auxiliary storage device 6 CPU 7 bus 10 dictionary data 11 font data unit 12 kanji code / part number dictionary 13 part number / kanji code dictionary 14 part coordinate dictionary 15 part number / kanji Code dictionary 51 Kanji search unit 52 Linked information output unit 61 Search component information acquisition unit 62 Result candidate extraction unit 63 Non-target processing unit 64 Part match degree determination unit 65 Total stroke count determination unit 66 Component count determination unit 67 Component count search unit 68 Individual Component determination unit 69 Character area difference determination unit 71 Text output unit 72 Character pattern generation unit

Claims (14)

[Claims] An input unit for inputting a character; a component extracting unit for extracting a component constituting the character from the character input by the input unit; and a character corresponding to the component extracted by the component extracting unit. A character search device comprising: a search means for searching. 2. A plurality of characters are input by the input unit, and the component extracting unit extracts components constituting each of the input characters.
Description character search device. 3. The character search device according to claim 1, wherein the search unit searches for a character including at least one component extracted from the input character. 4. The character search device according to claim 2, wherein the search means searches for each of the input characters for a character including at least one component extracted from the character. 5. A number-of-parts sorting means for sorting characters searched by said search means according to a structure of parts extracted by said part extraction means and a structure of parts constituting characters searched by said search means. 2. The method according to claim 1, wherein
The character search device according to any one of claims 1 to 4. 6. The number of parts for sorting the characters searched by the search means according to the difference between the number of parts extracted by the part extraction means and the number of parts constituting a character searched by the search means. 2. The apparatus according to claim 1, further comprising a sorting unit.
The character search device according to any one of claims 1 to 5. 7. A component position for sorting the characters searched by the search means according to the difference between the arrangement position of the parts extracted by the part extraction means and the arrangement position of the parts of the characters searched by the search means. 7. The character search device according to claim 1, further comprising a sorting unit. 8. A non-target component extracting unit for inputting a character including a component to be excluded from the input unit, extracting a component constituting the character, and extracting the target from the character information retrieved by the retrieval unit. 2. A non-target processing unit that excludes a character including a part extracted by the external product extraction unit from a search target.
The character search device according to any one of claims 1 to 7. 9. A stroke number input unit for inputting the stroke number of a character to be searched, and a stroke number search for searching for a character having a stroke number matching the stroke number input by the stroke number input unit among the characters searched by the search unit. 9. The character search device according to claim 1, further comprising: means. 10. A stroke count difference detecting means for detecting a stroke count difference between a character input by the input means and a character searched by the search means, and the stroke count difference detected according to the stroke count difference detected by the stroke count difference detecting means. 10. The character search device according to claim 1, further comprising: a number-of-images difference narrowing-down unit that narrows down character information searched by the search unit. 11. The character search device according to claim 1, wherein the search means searches for characters including all parts extracted from the input characters. 12. An input procedure for inputting a character, a component extraction procedure for extracting a component constituting the character from the character input in the input procedure, and a character corresponding to the component extracted in the component extraction procedure. And a search procedure for searching. 13. An input procedure for causing a computer to input a character, a component extraction procedure for extracting a component constituting the character from the character input in the input procedure, and a component extracted in the component extraction procedure. And a computer-readable recording medium storing a program for executing a search procedure for searching for characters. 14. The program according to claim 1, wherein a plurality of characters are input in the input procedure, and the component extracting step extracts components constituting each of the input characters. 14. The computer-readable recording medium according to claim 13.