JPH09204497A - Method and device for recognizing handwritten character - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method and device for recognizing character with which operability/ processability is improved by extracting the same type of character strings from recognized candidates. SOLUTION: While successively fetching candidate character groups provided from character recognition, a same type character string segmenting part 3 segments cursive form of Japanese syllabary (KANA)/square form of KANA type character strings. The cursive form of KANA/ square form of KANA type character string is compared with a previously set word dictionary by a word matching part 6, and the coincident character string is extracted and exchanged with the recognized result of 1st candidate in the candidate character group. At such a time, a marker is added to the character matched with the word dictionary 7. AKANA/Chinese character (KANJI) converting part 4 performs KANA/KANJI conversion to the character string having no marker and adds the result to the candidate character group. Besides, the segmenting part 3 segments the KANJI/alphanumeral type character strings from the candidate character groups, these character strings are compared with the previously set dictionary 7 by the word matching part 6, the coincident character strings are extracted, and these character strings are exchanged with the recognized result of 1st candidate in the candidate character group. At such a time, a marker is added to a fixed word part.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a handwritten character recognition device for inputting characters by recognizing handwritten strokes, and more particularly to a handwritten character recognition method for improving the recognition rate.

[0002]

2. Description of the Related Art A conventional handwritten character recognition device is disclosed in Japanese Examined Patent Publication No. 63-4.
As described in Japanese Patent No. 9264, a user recognizes the above-mentioned character string "calculation key" in order to convert an incomplete compound word (for example, "calculation key") mixed with kanji and kana into a correct compound word ("calculation machine"). Had to tell. Further, as described in JP-A-60-7585, when performing word matching from the recognition result, only when the recognition results for a certain input are all the same character type,
That part was cut out as a word area.

[0003]

In the former case of the above-mentioned prior art, the user has to specify the character string to be converted into kana-kanji characters, and the consideration is insufficient in terms of operability.

Further, in the latter of the above-mentioned prior arts, there is a problem that a word region cannot be stably extracted when a recognition result of a plurality of character types is obtained for one input character.

An object of the present invention is to provide a character recognition method and apparatus which are excellent in operability / processability by extracting character strings of the same character type from recognition candidates.

Another object of the present invention is to provide a character recognition method and apparatus excellent in operability / processability by extracting a character string of the same character type from a recognition candidate and converting a kana character string into kana / kanji characters. To provide.

Still another object of the present invention is to extract character strings of the same character type from recognition candidates, convert hiragana / katakana character strings from them into kana-kanji characters, and perform word matching including the conversion result. The object of the present invention is to provide a character recognition method and device excellent in operability / processability.

[0008]

In order to achieve the above object, a handwritten character recognition apparatus according to the present invention inputs handwriting information of a handwriting and recognizes the input handwritten character.
So as to retain the selected first candidate character and the determined second candidate character by converting the selected recognition candidate character and converting the selected recognition candidate character to determine the second candidate character. It is the one.

According to another aspect, the handwritten character recognition method according to the present invention is a handwritten character recognition method for recognizing a character from a series of handwriting information of a handwriting, and for each input handwritten character. One or two or more recognition candidates are selected, the character strings of the same character type that are consecutive in the character arrangement order are cut out for each character recognition candidate, and the extracted character strings are collated with a word dictionary. A character string that matches a character string existing in the word dictionary is used as a recognition result.

As the start timing of the above-mentioned character string segmentation processing, after obtaining recognition candidates for a predetermined number of characters, after extracting delimiters such as punctuation marks based on the recognition candidates of each character, Alternatively, an arbitrary timing such as after obtaining the recognition candidate for each character can be selected.

Further, it is preferable to suppress clipping of a character string having a recognition candidate that is not the first character of a word as the first character.

Of the recognition candidates of each input character, only the recognition candidate having a small difference from the handwriting may be the target of the character string cutout process.

Another handwritten character recognition method according to the present invention is a handwritten character recognition method for recognizing a character from a series of handwriting information of handwriting, wherein one or more input handwritten characters are input. Recognition candidates of each character are selected, a character string of consecutive kana characters is cut out for the recognition candidates of each character, and each cut out kana character string is collated with a word dictionary to be present in the word dictionary. A kana character string that matches the kana character string is extracted, at least kana characters other than the extracted kana character string are converted into kana-kanji characters, and kanji obtained by the kana-kanji conversion and kanji characters in the recognition candidates are A character string of consecutive kanji in the order of arrangement is cut out, each cut-out kanji character string is collated with a word dictionary, and a kanji character string that matches the kanji character string existing in the word dictionary is extracted and extracted as described above. publication in Kana It is obtained so as to obtain a final recognition result based on the column and kanji string.

Further, the character recognition device according to the present invention is a handwritten character recognition device for recognizing characters from a series of handwritten handwriting information, and means for inputting handwritten handwriting information and each of the handwriting information from the handwriting information. Means for selecting one or more recognition candidates for the input character, means for cutting out character strings of the same character type that are continuous in the character arrangement order for the selected recognition candidates, and for each of the cut out character strings , A means for searching the word dictionary and a means for displaying a character string as a recognition result based on the search result.

This device may be provided with means for designating a character type when the character string of the same character type is cut out.
Preferably, the means for selecting a recognition candidate has a function of assigning a rank to a plurality of recognition candidates for each input character, and the rank is changed according to the search result.

In addition to the above-mentioned means, it is desirable to provide kana-kanji conversion means for converting the kana character string cut out by the means for cutting out the character string of the same character type into kanji.

[0017]

The above-mentioned character string cutting-out means cuts out a character string of the same character type by combining consecutive candidates of the same character type in the candidate character group of the recognition result. At this time, preferably, the input position information and the length or the last position information of the character string are added to the character string.

Specifically, if one candidate character A is used as a starting point and there is a character of the same character type in the candidate character group corresponding to the next input, they are combined to cut out a character string, and the character string is extracted. And output the above information to the character string buffer are repeated until the latest input character, and as a general rule, by performing the candidate character group for all input characters as the starting point, the necessary recognition from the above recognition result is performed. It becomes possible to extract character strings.

Further, the activation timing of the character string cutout means is as follows:
It can be performed when a phrase delimiter such as a particle or a punctuation mark appears in a candidate character. In any case, the above processing does not malfunction because the character string is cut out by the sequential processing.

The kana-kanji conversion means performs kana-kanji conversion on the hiragana / katakana character string cut out by the character string cutting means.

The means for searching the word dictionary, that is, the word matching means, matches the character strings of the same character type cut out by the character string cutting means with the word dictionary, and inputs the matched character strings. Is output as the first candidate of.

According to the present invention, with respect to an incomplete idiom mixed with kana and kanji, a character string of the same kind is cut out from a recognition candidate group and word matching is performed, thereby recognizing an input character with high accuracy from handwriting information. Will be possible.

After converting the recognized kana into kanji,
Cut out the kanji character string along with the kanji in the recognition candidates,
By performing word matching on this result, it is not necessary for the user to specify the kana-kanji conversion target, and a word mixed with kana-kanji can be automatically converted into an accurate kanji word.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 shows the construction of a handwritten character recognition apparatus including the present invention. The input unit 1 and the display unit 11 form an input integrated display mechanism that is superimposed on the same coordinate system. Input unit 1
Outputs a sequence of coordinate points in time series according to the movement of the pen for handwriting input. The display unit 11 displays the sequence of coordinate points obtained from the input unit 1 directly below it, and redisplays the clean copy information after processing such as recognition. The character recognition unit 2 cuts out the coordinate point sequence obtained from the input unit 1 for each character, obtains a similarity with a standard pattern dictionary (not shown) held in advance, and calculates the similarity. It outputs as a candidate character in descending order. The same character type character string cutout unit 3 outputs a character string of character type such as hiragana / katakana / kanji / alphanumeric characters together with the position information from the candidate character group obtained from the character recognition unit 2 in chronological order. The character string cutout unit 3 can specify the type of character string to be cut out by a character type designation register. The Kana-Kanji conversion unit 4
The hiragana / katakana character string obtained from the character string cutout unit 3 is converted into a kanji mixed with kanji by the conversion dictionary 5, and the resulting character string is added to the candidate character group obtained from the character recognition unit 2. The word matching unit 6 extracts a word from the same character type character string obtained from the character string cutout unit 3 including the converted character by using the word dictionary 7, and matches the matched character to the first character of the candidate character group. Output as one candidate. The document management unit 8 extracts the first candidates from the candidate character group obtained from the above processing, stores them in the document buffer 9, and stores the candidate character group in the candidate character buffer 10.

As the input unit 1, for example, a transparent electrostatic induction tablet, a pressure sensitive tablet or the like can be used. Display 1
As 1, a flat display such as a CRT or a liquid crystal / plasma can be used. These techniques are already known, and no detailed operation is described in this embodiment. Further, the character recognition unit 2 is not particularly limited in its recognition method, and for example, a method using an approximate feature of input handwriting as disclosed in JP-A-60-237580 can be used.

The detailed operation of this apparatus will be described below.

FIG. 2 shows a flow of an example of processing when handwritten character input is performed using this apparatus.

In the initial setting A, the registers / buffers used in the following processes are initialized when the apparatus is started. Although not shown in FIG. 1, termination confirmation B is processing for terminating the character input function of this apparatus by a menu or the like. The input C is a process of the input unit 1, and when there is a handwriting input for character recognition, the coordinate information is passed to the next character recognition D. The character recognition D cuts out coordinate information for one character from the input handwriting, converts it into a preset data structure of a standard pattern dictionary, and outputs candidate character groups in descending order of similarity. The same-character-type character string cutout E is processing of the same-character-type character string cutout unit 3, and cuts out a character string of Hiragana / Katakana character type while sequentially fetching the candidate character groups obtained from the character recognition D. The word matching F is a process of the word matching unit 6, which compares a hiragana / katakana character string obtained from the character string segmentation E with a preset word dictionary, extracts a matched character string, and Is replaced with the recognition result of the first candidate of the candidate character group. At this time, a marker is added to a character matched with the word dictionary 7. The kana-kanji conversion G is a process of the kana-kanji conversion unit 4, in which the character string without the marker in the hiragana / katakana character string is converted to kana-kanji, and the converted character string mixed with kanji is added to the candidate character group. . The same character type character string cutout H cuts out a character string of Kanji / alphanumeric character type from the above candidate character group. Word alignment I
Compares the Kanji / alphanumeric character string obtained from the character string cut-out H with the word dictionary 7 set in advance as in the word matching F, extracts the matched character strings, and extracts them. The recognition result of the first candidate of the candidate character group is replaced. At this time, a marker is added to the confirmed word portion. The document management J registers the first candidate from the candidate character group obtained from the word matching I in the document buffer 9, displays the contents of the document buffer 9 on the display unit 11, and sets the candidate character group to the candidate character group. Store in buffer 10.

Hereinafter, the processing after the same character type character string cutout E, which is a main part of the present invention, will be described in detail.

<Cutting E of Characters of the Same Character Type> FIG. 3 to FIG. 6 show this character cutting process, FIG. 7 shows the status of each buffer at the same time when this book is written, and FIG. 8 shows the end of this process. The status of each buffer is shown below.

The flow of data is shown in FIGS. 8 and 9. First, the contents of the buffer used in this processing will be described.

In FIG. 7, the recognition buffer included in the character recognition unit 2 is an area for storing the character recognition result. For the sake of simplicity, the maximum number of candidate characters is 2, and for the subsequent kana-kanji conversion results, two character's worth of characters are stored. We have a spare area in the recognition buffer. The character string buffer is an area for storing the same character type character string to be cut out, and the registration number m of the cut out character string / starting position information X, Y / character string Bs / character string length L / each word flag small It is divided into areas. The word flag is a marker attached to a character string that matches a word dictionary in a word matching process described later. The word processing register indicates the input position of the character with the marker. The designated character type register is used for setting the character type to be cut out. The character string buffer, word processing register, and command character type register are included in the same type character string cutout unit 3.

Character type information may be added to each of the cut out character strings in the character string buffer.

Although the Y information is provided only for the first character of the cut-out character, the Y information may be provided for the second and subsequent characters. If you do this,
Although the storage capacity increases, the process of replacing the candidate ranks described later can be performed quickly.

The same character type character string cutout process is a process of finally outputting the contents of the character string buffer of FIG. 8 with respect to the input contents of the recognition buffer of FIG.

The process will be described below with reference to the flow of the same character type character string cutout process E in FIG. At this point, the designated character type register is set to Hiragana / Katakana.

Processes 3-1, 3-2: In FIG. 3, first, the input number Wps of the input serving as the starting point to be cut out (hereinafter referred to as starting point input) is set as the loop head is. The loop variable i executes the following processing from this is to the tail Wpe of the input character string. The processing for the first and second input characters (“ko”, “no”) in the recognition buffer of FIG.
Taking the case where the same character type character string is cut out from the input character as an example, Wps is 3 and the tail Wpe of the input character string is 12.

Process 3-3: Next, in order to set the number of starting points for clipping, the number of candidates for starting point input nmax is set to the number of candidate characters for starting point input nc. In the case of the third input of the recognition buffer of FIG. 7, nmax is 2 because it has two candidates of "shi" and "three".

Process 3-4: Processes up to the above nmax 3-
The same character type character string of 5 or less is cut out. When the processes up to nmax are completed, the number i of the combination source input is updated in the process 3-14, the previously registered number nold is initialized, and the process returns to the process 3-2.

Process 3-5: The character type K of the candidate character for inputting the starting point (hereinafter, starting point character) is set. Since the first starting character of the third input of the recognition buffer in FIG. 7 is "shi", the character type K is katakana.

Process 3-6: Next, it is determined whether or not the character type K matches the character type set in the designated character type register. In the example, the character type K is katakana by the above processing 3-5, and the designated character type register is set to hiragana / katakana. Therefore, the determination result is affirmative, and the processing proceeds to the following processing. When the result is negative, the candidate number n of the starting point input i is updated in the process 3-15, the process returns to the process 3-4, and the processes 3-5 and below are performed for the next starting character.

The character code is JIS / shift JIS.
Since they are assigned to character types according to the above, a character type classification table as shown in FIG. 12 is created from those character code tables. In this character type classification table, the character type is divided into eight tables, and the code (shown in hexadecimal) of each table is the head character code of each character type.

Process 3-7: Next, the starting point character and position /
Store the length in the string buffer. This processing is shown in FIG.
7 shows the detailed flow.

Process 3-7-1: First, the starting character is stored in Bs [m] of the character string buffer.

Process 3-7-2: At the same time, the character position (X,
Y) / length L is stored. Here, X [m] corresponds to the starting point input number i, and Y [m] corresponds to the starting point candidate number n. In the example, the character position of the starting character "shi" is X [m] =
3, Y [m] = 1 and length L = 1.

Process 3-7-3: Finally, the character string registration number m is updated.

Process 3-8: Next, returning to FIG. 3, to check how many characters of the same character type continue, the input number js of the candidate character group to be combined (hereinafter, the combined input number 9 Initialization is performed to the next input number after the current input number i. In the case of this processing example, since i is 3, js is 4.

Process 3-9: Next, the same-character-type character string search and concatenation process is started. The following processing is repeated until the combined input number j comes after the input. However, if the value of the word processing register corresponding to the position of j is ON or if the characters of the same character type are interrupted, the candidate number of the starting character is updated (process 3-1-4). Then, the process returns to the process 3-4. In the example of FIG. 8, since the word processing register is in the initial state, the following processing is executed until the above j changes from 4 to 12.

Process 3-10: To obtain a character that can be combined next, from the group of candidate characters (hereinafter, combined input character) corresponding to the combined input number j, the same character type as the character type K set in the above process 3-5. Search for candidates.

FIG. 5 shows the detailed flow of processing 3-10.

Processing 3-10-1 and -2: First, for the initialization of the parameters, the number of discoveries nf indicating that there are one or more candidates for the same character type is 0, and the combination flag Cflag [] is turned off.

Process 3-10-3: Next, in order to set the number of combining processes, the total number njmax of candidate characters input to the combining destination is set.

From FIG. 7, the candidate character of the fourth input is "nu / s" in the case of the input number j = 4 of the connection destination, and njmax
Is 2.

Process 3-10-4: After that, the search for the same type character is performed n times.
Repeat jmax times.

Process 3-10-5: Candidate character code for combination [j]
The character type setting of [nj] (= “nu”) is performed by the same procedure as the processing 3-7. At this time, K [j] [nj] becomes katakana.

Process 3-10-6: Next, the starting character type K [j]
[Nj] is compared with the combination destination character type K [j] [nj], and if they match, the process proceeds to flag setting processing. In this example, the starting character type K [i] [n] and the joining destination character type K [j] [nj] are both katakana and match.

Process 3-10-7: Combining destination candidate character code [j]
To save that [nj] can be combined with the starting character, the combination flag Cflag [nj] is turned on.

Process 3-10-8: Next, the found number nf is updated in order to store the combined number. If the character types do not match in the process 3-10-6, the direct connection destination candidate number is updated (3-10-9), and the process returns to the process 3-10-4.

As a result of this processing 3-10, the connection destination characters "nu" and "su" are both katakana character types, and since they match the above origin character types, the combination flag Cflag [1-2] is turned on and the number of found characters. nf becomes 2.

Returning to FIG. 3, process 3-11: If the found number nf is 0 or more, the process proceeds to a character string combination / storage process 3-12. If nf is 0, the candidate number of the starting character is updated in process 3-15 and the process 3 is performed.
Return to -4. In this example, nf is 2 and processing 3-1
Proceed to 2.

Process 3-12: Here, the combined character is connected to the starting character and registered in the character string buffer.
Therefore, the following processing is performed according to the set combination flag and the number of found pieces. FIG. 6 shows the detailed flow of processing 3-12.

Process 3-12-1: In order to save the current registration number m, the current registration number is saved in m '.

Process 3-12-2, -3: To combine character strings, multiply the number of character strings nold generated in the previous combining process by the number of Cflag [nj] of the combination destination candidate character group that is turned on. The number of connections ncon is set in order to perform the number of times. In the example of FIG. 8, since the katakana character type is the first combination, nold is 1, and the combination flag Cflag [] is 2 on (the number of discoveries = 2) from the process 3-9. Is 2.

Process 3-11-4: First, the combination flag Cflag [n
j] is determined.

Process 3-11-5: If the result of determination is that it is on, the previously registered character string Bs [m'-nold] is combined with the concatenation destination character [j] [nj] and stored in the character string buffer. .
At this time, the position and length of the character string are also stored in the same manner as in the process 3-6. After that, the registration number m is updated.
At present, the character string Bm′-nold] is “si” and the candidate character code for connection is “nu / su”. Therefore, two character strings “sinu” and “cis” are stored in the character string buffer Bs [m]. And Bs [m + 1] are additionally registered. Then process 3-
Return to 12-3.

Process 3-12-6: When the combined registration of character strings is completed, the previously registered number nold is updated to the combined number ncon, and this process is completed. In this example, nold is 2.

Returning to FIG. 3, after the character strings are combined / stored in the process 3-13, the combination destination input number j is updated, and the process 3 is executed.
Return to -9.

As shown in the character string buffer of FIG. 8, the same character type character string cut-out process E as described above is performed, for example.
It can be seen that the katakana character string of "shi / shinu / cis / sinute /.../ system /.../ Seikei" is cut out starting from the third input first candidate "shi".

In this processing E, since the hiragana / katakana character string is cut out, the character strings such as "ko / kono / no" sa / san / sanki / n / nki / ki "are also cut out at the same time. . Here, there is also a meaningless character string "n / nki" or the like because the same character type character string is simply cut out from the candidate character group stored in the recognition buffer. In these methods, if a character that does not become the first character of a word, such as a syllabary or a beating sound, is not adopted as a starting character, and a method for suppressing clipping is used,
The meaningless character string can be prevented from being output, and the processing time can be shortened.

When the starting character is set (process 3-3)
In addition, by comparing with the index of the word dictionary or the kana-kanji conversion dictionary, and when there is no corresponding character index, the candidate number n is updated in advance, and the value is used as the initial value of the loop in the processing 3-4. Can be shortened.

Further, by limiting the maximum word length of each character type, it is possible to suppress the generation of unnecessary character strings and reduce the processing time. In this case, after the character strings are combined / stored in process 3-12, when the character string length L reaches the maximum word length, the combining process is terminated, and the starting character is updated in process 3-15. .

At the time of cutout, by executing the cutout only on the candidate characters having a high similarity equal to or less than a preset value according to the degree of difference (not shown) in the recognition result, it is possible to suppress the generation of unnecessary character strings. it can.

<Word Matching F> The hiragana / katakana character string cut out in the process E is compared with the word dictionary 7 to find a matching character string. In this example, "system" is a matching character string, and the "su" in the recognition buffer is replaced with the first candidate as shown in FIG. 9, and a character string of up to 4 characters starting from "si" in the character buffer is used. On the other hand, the word flag is set, and the on-flag which means word determination is set in Rt [3-6] of the word processing register. Here, "sun" can be interpreted as a foreign word of "sun", for example,
It can also be used to read kanji such as math ... ". Therefore, when a word is found, if those characters are not in the kana-kanji variable reading dictionary, the word is determined, and if there is, it is suspended and misalignment is suppressed. In this example, the "sun" is suspended.

Further, since the number of words used by one user is generally limited, the determined word is temporarily stored,
The processing time can be shortened by preferentially comparing the next word matching. Also, the processing time can be shortened by holding the word lengths in the word dictionary and performing matching processing only for words having the same word length.

When a word is found, by selecting the first candidate according to the dissimilarity of the corresponding candidate characters, for example, by always selecting the character string with the minimum sum of the dissimilarities as the first candidate, it is possible to stabilize the operation. Good results can be obtained.

<Kana Kanji Variable G> A character string mixed with Kanji by using a conversion dictionary for the input determined in the process F, that is, the Hiragana / Katakana character string excluding the input character string corresponding to the word processing register Rt [3-6]. Convert to. A known technique is used for the kana-kanji variable technique, and the result is stored in the spare area of the recognition buffer as shown in FIG. At this time, since the character string length may be shortened by the conversion, a null symbol is set in the spare area of that part. In the example of FIG. 10, a null symbol is set in the column of “n” from the conversion effect “arithmetic” of “sun”.

<Same Character Type Character String Cutout H> The same processing as the above-mentioned same character type character string cutout E is performed except that the designated character type is Kanji / alphanumeric. As a result, as shown in the character string buffer in FIG. 11, "total / total difference / calculation /.../
The character string "Calculator / Calculator / ..." is cut out.

<Word Matching I> The same processing as the processing word matching F is performed except that the designated character type is kanji / alphanumeric. As a result, the matching result of the two words "calculation" and "calculator" is obtained. Here, the word with the longest match is prioritized, and “computer” is set as the confirmed word. Therefore, as shown in FIG. 11, "arithmetic" and "machine" are replaced with the first candidate in the recognition buffer, and an on-flag indicating word determination is set in Rt [8-11] of the word processing register.

<Document Management J> A character to be finally displayed is selected from the recognition buffer Br / word processing register Rt updated by the above processes A to I and stored in the document buffer 10 / candidate character buffer 9. The word processing register Rt is sequentially read from the beginning, and undetermined input is TE / NI / OR / HA /
Particles such as No. and punctuation marks such as punctuation are preferentially stored in the document buffer 10. The fixed input is stored in the document buffer while deleting the input that is a null symbol. The candidate character group corresponding to the input stored in the document buffer 10 is stored in the candidate character buffer 9. As a result, the character string “This system is a computer” is displayed on the display unit 11.

In the above description of the embodiments, the kana and kanji character strings are cut out separately, but they can also be executed simultaneously.

[0082]

According to the present invention, since a character string of the same character type can be automatically extracted from a character string input by handwriting, there is an effect of improving operability.

By converting the hiragana / katakana part of the extracted character string into kana-kanji, kanji whose memory is ambiguous can be automatically output, so that the operability is improved.

Further, since the correct character can be automatically selected from the similar characters by word matching the extracted character strings, the character recognition rate and the operability are improved.

Then, the Hiragana / Katakana part of the character string extracted from the incomplete compound word mixed with Kana-Kanji character is converted into Kana-Kana character, and the correct compound word is extracted from the extracted character string of the same character by word matching. Since the incomplete idiom mixed with the Chinese characters can be correctly converted by automatically generating it, there is an effect of improving operability.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of a character recognition device according to an embodiment of the present invention.

FIG. 2 is a flow chart of processing of the character recognition device of FIG.

FIG. 3 is a flowchart of the same character type character string cutout processing.

FIG. 4 is a flow chart of a character string segmentation process of the same character type.

FIG. 5 is a flow chart of the same character type character string cutout processing.

FIG. 6 is a flowchart of the same character type character string cutout process.

FIG. 7 is an explanatory diagram of a buffer / register in a processing process.

FIG. 8 is an explanatory diagram of a buffer / register in a processing process.

FIG. 9 is an explanatory diagram of a buffer / register in a processing process.

FIG. 10 is an explanatory diagram of a buffer / register in a processing process.

FIG. 11 is an explanatory diagram of a buffer / register in a processing process.

FIG. 12 is an explanatory diagram of a character type table.

[Explanation of symbols]

1 ... Input unit, 2 ... Character recognition unit, 3 ... Same character type cutout unit,
4 ... Kana-Kanji conversion unit, 5 ... Conversion dictionary, 6 ... Word matching unit,
7 ... Word dictionary, 8 ... Document management unit, 9 ... Document buffer, 1
0 ... Candidate character buffer

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] January 8, 1997

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Name of invention

[Correction method] Change

[Correction contents]

Title: Handwritten character recognition method and apparatus

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

1. A handwriting character recognition method for recognizing a character from handwriting information of a handwriting, wherein a recognition candidate character is selected for an input handwriting character, and a character string including the recognition candidate character is a kanji character. When it is composed of Kana characters, it is converted into Kanji using a conversion dictionary, and the Kanji string consisting of the Kanji obtained as the conversion result and the Kanji that constitutes the character string is used as the word dictionary. A handwritten character recognition method characterized by collating and using a kanji string that matches a kanji string existing in a word dictionary as a recognition result.

2. A handwriting character recognition method for recognizing a character from handwriting information of a handwriting, wherein recognition candidate characters are selected for input handwriting characters, and some of the selected recognition candidate characters are selected. For the characters of, the character types are converted using the conversion dictionary, and the character string consisting of the characters obtained as the conversion result and the recognition candidate characters that have not been converted is compared with the word dictionary and exists in the word dictionary. A handwritten character recognition method, characterized in that a character string that matches the character string to be used is the recognition result.

3. A handwriting character recognition method for recognizing a character from handwriting information of a handwriting, wherein a recognition candidate character is selected for an input handwriting character, and kana-kanji conversion is performed on the recognition candidate character. , The Kanji obtained as a result of conversion, the Kanji originally existing as the recognition candidate character, and the kana character combination existing originally as the recognition candidate character are collated with the word dictionary and matched with the character string existing in the word dictionary. A handwritten character recognition method, characterized in that a recognized character string is used as a recognition result.

4. A means for inputting handwriting information of a handwriting, a means for selecting a recognition candidate character for a handwriting character from the input handwriting information, and a character string consisting of the recognition candidate character is a kanji character or a kana character. In the case of kana characters, if kana characters are used, a conversion dictionary is used to convert the kanji into a kanji string consisting of the kanji that is obtained as a result of the conversion and the kanji that constitutes the character string. A handwritten character recognition device comprising: a matching unit; and a unit that outputs a Kanji string that matches a Kanji string existing in a word dictionary as a recognition result.

5. A means for inputting handwriting information of a handwriting, a means for selecting a recognition candidate character for a handwriting character from the input handwriting information, and a part of the selected recognition candidate characters. For, the means for converting character types using a conversion dictionary, the means for matching a character string consisting of the characters obtained as a result of conversion and the unrecognized recognition candidate characters with a word dictionary, and a word dictionary And a means for outputting a character string that matches the character string existing in 1. as a recognition result, a handwritten character recognition device.

6. A means for inputting handwriting information of a handwriting, a means for selecting a recognition candidate character for a handwriting character from the input handwriting information, and a means for performing kana-kanji conversion on the recognition candidate character. , Kanji obtained as a result of conversion, Kanji existing as a recognition candidate character, Kana existing as a recognition candidate character, means for collating a character string that originally exists as a recognition candidate character with a word dictionary, and a character string existing in the word dictionary And a means for outputting a character string that matches with the recognition result as a recognition result.

[Procedure 3]

[Document name to be amended] Statement

[Correction target item name] 0008

[Correction method] Change

[Correction contents]

[0008]

[MEANS FOR SOLVING THE PROBLEMS] To achieve the above object
According to the first aspect of the present invention,
A handwriting character recognition method that recognizes characters
For the calligraphy characters, select the candidate characters for recognition and
When the character string consisting of consists of Kanji and Kana characters
For kana characters, use the conversion dictionary to convert them to kanji.
And the Kanji obtained as the conversion result and the character string are constructed.
The Kanji string consisting of the Kanji to be formed is collated with the word dictionary
A kanji string that matches a kanji string existing in the dictionary is set as the recognition result.
A handwritten character recognition method is provided.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0009

[Correction method] Change

[Correction contents]

A second aspect of the present invention for achieving the above object
According to the aspect, a handwriting for recognizing a character from handwriting information of the handwriting
It is a character recognition method, for the handwritten characters that have been input,
Select the recognition candidate character and select the selected recognition candidate character.
Then, for some characters, use the conversion dictionary
Is converted, and the characters obtained as a result of conversion are converted into
A character string consisting of recognition candidate characters
The character string that matches with the character string existing in the word dictionary
Is a handwritten character recognition method that is characterized by
Provided.

[Procedure Amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0010

[Correction method] Change

[Correction contents]

A third aspect of the present invention for achieving the above object.
According to the aspect, a handwriting for recognizing a character from handwriting information of the handwriting
It is a character recognition method, for the handwritten characters that have been input,
Select recognition candidate characters, and use kana kanji for the recognition candidate characters.
After conversion, the kanji and recognition candidate sentences obtained as the conversion result
Kanji that originally existed as characters, or whether they were originally as recognition candidate characters
Character string that is a combination of existing kana characters from
Characters that match the calligraphy and match a character string that exists in the word dictionary
A handwritten character recognition method characterized by using columns as recognition results
Is provided.

[Procedure correction 6]

[Document name to be amended] Statement

[Correction target item name] 0011

[Correction method] Change

[Correction contents]

A fourth aspect of the present invention for achieving the above object.
Aspects are as follows.

[Procedure amendment 7]

[Document name to be amended] Statement

[Correction target item name] 0012

[Correction method] Change

[Correction contents]

That is, means for inputting handwriting information of handwriting
And recognition of handwritten characters from the input handwriting information
Characters consisting of recognition candidate characters and means for selecting candidate characters
If the column consists of Kanji and Kana
For that, using the conversion dictionary, a means to convert to Kanji,
The Kanji obtained as a conversion result and the character string are constructed.
A means for matching a kanji string consisting of kanji with a word dictionary, and
Kanji strings that match the Kanji strings existing in the word dictionary are recognized as recognition results.
And a means for outputting the handwritten text
A character recognition device is provided.

[Procedure amendment 8]

[Document name to be amended] Statement

[Correction target item name] 0013

[Correction method] Change

[Correction contents]

A fifth aspect of the present invention for achieving the above object.
Aspects are as follows.

[Procedure amendment 9]

[Document name to be amended] Statement

[Correction target item name] 0014

[Correction method] Change

[Correction contents]

That is, means for inputting handwriting information of handwriting
And recognition of handwritten characters from the input handwriting information
A method of selecting candidate characters and a method of selecting the selected recognition candidate characters.
For some of the characters, use the conversion dictionary
A means to perform seed conversion and the resulting character
And a recognition candidate character that has not been converted
With the word dictionary and the sentences existing in the word dictionary
A means to output a character string that matches the character string as a recognition result
Provided with a handwritten character recognition device characterized by
Is done.

[Procedure amendment 10]

[Document name to be amended] Statement

[Correction target item name] 0015

[Correction method] Change

[Correction contents]

A sixth aspect of the present invention for achieving the above object.
Aspects are as follows.

[Procedure amendment 11]

[Document name to be amended] Statement

[Correction target item name] 0016

[Correction method] Change

[Correction contents]

That is, means for inputting handwriting information of handwriting
And recognition of handwritten characters from the input handwriting information
Means to select candidate characters and kana for recognition candidate characters
Kanji conversion means and kanji obtained as a result of conversion,
Kanji and recognition candidate characters that originally exist as recognition candidate characters
As a combination of kana characters that originally existed as
A means of matching columns to a word dictionary and the sentences present in the word dictionary
A means to output a character string that matches the character string as a recognition result
Provided with a handwritten character recognition device characterized by
Is done.

[Procedure amendment 12]

[Document name to be amended] Statement

[Correction target item name] 0017

[Correction method] Change

[Correction contents]

[0017]

According to the present invention, the input handwritten characters are not
First, the recognition candidate character is selected.

[Procedure amendment 13]

[Document name to be amended] Statement

[Correction target item name] 0018

[Correction method] Change

[Correction contents]

A part of the selected recognition candidate characters
Regarding the characters, the character type is converted using the conversion dictionary.
It is.

[Procedure Amendment 14]

[Document name to be amended] Statement

[Correction target item name] 0019

[Correction method] Change

[Correction contents]

The conversion of the character type is performed even if it is the kana-kanji conversion.
I do not care.

[Procedure Amendment 15]

[Document name to be amended] Statement

[Correction target item name] 0020

[Correction method] Change

[Correction contents]

Then, the character obtained as the conversion result,
For the character string consisting of the unconverted recognition candidate characters
String that matches the word dictionary and exists in the word dictionary
The character string that matches is taken as the recognition result.

[Procedure Amendment 16]

[Document name to be amended] Statement

[Correction target item name] 0021

[Correction method] Change

[Correction contents]

In this way, for example, the recognition candidate character
If "calculation key" is selected as
You will be able to obtain a "calculator".

[Procedure amendment 17]

[Document name to be amended] Statement

[Correction target item name] 0022

[Correction method] Change

[Correction contents]

Further , kana-kanji conversion for recognition candidate characters
And then the kanji and recognition obtained as a result of conversion.
Kanji that originally existed as candidate characters, and recognition candidate characters
A string that is a combination of kana characters that originally existed
You may make it collate with a word dictionary.

[Procedure amendment 18]

[Document name to be amended] Statement

[Correction target item name] 0023

[Correction method] Change

[Correction contents]

In this way, the sentence to be checked against the word dictionary
The number of strings is increased, and as a result, the recognition accuracy of the strings is improved.
You.

Front Page Continuation (72) Inventor Masaki Miura 4026 Kuji Town, Hitachi City, Ibaraki Prefecture, Hitachi Research Laboratories, Ltd. (72) Inventor Toshihiko Matsuda 1 Ikegami, Haruoka-cho, Owariasahi-shi, Aichi Hitachi Ltd. Asahi factory

Claims (1)

[Claims] 1. A means for inputting handwriting information of a handwriting, a means for recognizing an input handwritten character to select a first candidate character, and a means for converting the selected recognition candidate character to a second A handwritten character recognition device comprising: means for determining a candidate character; and means for holding the selected first candidate character and the determined second candidate character.