JPS62138989A - On-line character recognizing device - Google Patents

Abstract

PURPOSE:To easily discriminate a character with the different type of a character and the similar shape by means of a simple method by providing a character type symbol string dictionary to accommodated the character type symbol string concerning the special character with the different type of the character and the similar shape beforehand. CONSTITUTION:A stroke data extracting part 2 extracts the stroke data of a character inputted from a tablet 1, and a character characteristic recognizing part 5 compares and collates the stroke recognizing result by a stroke recognizing part 3 with the data accommodated at a character characteristic dictionary 6, determines the character, from the character recognizing result, judges the type of the character and accommodates it to a character type accommodating memory 8. the special character with the different type of the character and the similar shape, reads the type of the character of the front and rear from a memory 8, and determines the character by collating the special character with a character type symbol string dictionary 7 to describe the sequence of the type of the character. For example, when the front and the rear are not KATAKANA (square form of JApanese syllabary), they are not 'ka' of KATAKANA, and are the 'KA' of KANJI (Chinese character), when the front or the rear is KATAKANA, it is the 'ka' of KATAKANA.

Description

[Detailed Description of the Invention] (Field of Industrial Application) The present invention relates to an online character recognition device that recognizes handwritten Japanese text containing kanji and kana in real time, and particularly relates to an online character recognition device that recognizes characters of different character types with similar shapes in real time. The present invention relates to an online character recognition device that can easily and accurately identify characters.

(Prior Art) Examples of characters of different character types with similar shapes include the following.

Kanji katakana kanji numerals katakana O ragana katakana power and power, two and two, hedo to.

Kanji katakana minus long sound kanji
Chikaratakachiou and Kou, - and -1 mouth
and mouth.

Hiragana Katakana Numbers Alphabets Symbols Kanji ri and ri, 0 and
O, # and I.

3 Personal name English letters Katakana × and X and me.

Conventionally, online character recognition devices that recognize handwritten character strings containing a mixture of kanji, hiragana, katakana, alphabets, numbers, symbols, etc. in real time have been used to distinguish between characters with similar shapes among different character types as described above. The following method was used.

(1) The scribe checks the display immediately after entering characters,
If a character of a different type than the one inputted is displayed, press the next candidate key to input the desired character.

(2) In addition to inputting the desired character or sentence, the scribe is also required to input which character type the character or writing style belongs to.

(3) Analyze the input character string based on the sentence structure and determine whether the character is a kanji, hiragana, katakana, symbol, etc.

(Problems to be solved by the invention) However, among the methods described below, (1) and (2)
The method (2) has the disadvantage that the scribe's operations are troublesome, and the method (1) has the disadvantage that the software required for analysis is complicated and the cost of the equipment is high.

The present invention was made in order to eliminate the drawbacks of the prior art, and it does not require any troublesome operation.
in a simple way without using complicated software.
An object of the present invention is to provide an online character recognition device that can easily distinguish between "characters of different character types and similar shapes."

(Means for Solving the Problems) The present invention provides a tablet for inputting handwritten characters;
a recognition unit that recognizes input characters by comparing character data from the tablet with predetermined reference character data; a recognition character memory that stores recognition results by the recognition unit; and information in the recognition character memory. The target is an online character recognition device that recognizes in real time sentences containing a mixture of characters such as kanji, hiragana, katakana, alphanumeric characters, and symbols. To solve the problem, write the input character string as a character type symbol string a! A character type storage memory and a character type symbol string dictionary are provided in which character type symbol strings for specific characters of different character types and similar shapes are stored in advance, and when the specific character is input, By comparing and collating the input character type symbol string in the character type storage memory with the character type symbol string in the character type symbol string dictionary, the character type is determined and character recognition is performed.

(Operation) When a scribe writes handwritten characters on the tablet, the tablet sends the character data to the recognition unit. The recognition unit performs character recognition by comparing and collating the received character data with predetermined standard character data, that is, data stored in a character feature dictionary, and also determines the character type symbol of the recognized character. and stores it in the character type storage memory.

On the other hand, the character type symbol string dictionary stores character type symbol strings for "specific characters of different character types and similar shapes" set under possible conditions based on the sentence structure. When a specific character is selected as a candidate for recognition, the recognition unit compares the data in the character type storage memory with the data in the character type symbol string dictionary, determines the character type, and recognizes the specific character. The recognized characters are then stored in the recognized character memory, and the interface section outputs the stored information to the outside. Therefore, even when the scribe inputs a specific character as described above, no tedious operations are required, and the character can be easily and accurately recognized.

(Example) An example of the present invention will be described below. FIG. 1 is a block diagram showing the configuration of the online character recognition device of this embodiment. This online character recognition device is compatible with tablet 1
, stroke data extraction unit 2, stroke recognition unit 3, stroke number dictionary 4, character feature recognition unit 5, character feature dictionary 6, character type symbol string dictionary 7, character type storage memory 8, determined character memory 9, display unit 10, and It is composed of an interface section II. The tablet 1 is used by a scribe to input characters by hand, and the method -C includes an electrostatic coupling method, an electromagnetic coupling method, a resistive film method, etc., and any method may be used in the present invention. Stroke data extraction unit 2 is tablet 1
Extract the stroke data of the characters input. The stroke recognition section 3 performs stroke recognition by comparing and collating the stroke data extracted by the stroke data extraction section 2 with the stroke data stored in the stroke number dictionary 4 and determining a stroke number. The character feature recognition unit 5 determines the character by comparing the stroke recognition result by the stroke recognition unit 3 with the data stored in the character feature dictionary 6, and also determines the character type of the character from the character recognition result. The data are sequentially stored in the seed storage memory 8. The character feature recognition unit 5 further reads out the character types before and after the character from the character type storage memory 8 when the character feature dictionary 6 alone cannot determine "a specific character with a different character type and similar shape." , the character is determined by comparing the specific character with the character type symbol string dictionary 7 that describes the order of character types. The determined character memory 9 stores the characters determined by the character feature recognition section 5. Determined character memory 9
The character information stored in is used by the display unit IO and the interface unit 11 for output to the outside. Among the above elements, a part of the character feature recognition unit 5, a character type symbol string,
! The elements and their functions other than the door and the character type storage memory 8 are generally the same as those of the conventional online character recognition device for kanji, kana, and Japanese characters based on the stroke feature extraction method.

In contrast, the feature of the present invention is that the character type is determined by adding a character type symbol string dictionary 7 for "specific characters of different character types and similar shapes".

Next, an example of the contents of the character type symbol string dictionary 7 is shown in FIG. 2, and a procedure for determining a character type using the dictionary will be explained.

As described above, the character feature recognition unit 5 recognizes characters, and also determines character type symbols for the recognized kanji, hiragana, katakana, alphabetic characters, numbers, and symbols, and stores them in the character type storage memory 8. On the other hand, regarding "characters similar in shape but different in character type", possible symbol string conditions based on the sentence structure are registered in advance in the character type symbol string dictionary 7 as shown in the second example.

When that character is recognized as a candidate, the character string registered in the character type symbol string dictionary 7 and the data in the character type storage memory 8, which stores the character types written up to the character before that character, are By comparing, it is determined whether the character is a kanji, hiragana, katakana, alphabet, number, or symbol. If the character cannot be determined based on the symbol string conditions up to the preceding character, a mark is placed on that character, and the character type is determined by further comparing it with the character type written after the character.

Hereinafter, an example of the above-mentioned symbol string condition will be shown, and a method for determining the character type will be specifically explained.

(1) Distinguishing between the katakana character "power" and the kanji character "power" The probability that the character before or after the katakana character "power" in the sentence string is not katakana is small. If the first or second part is not katakana, it is written as the kanji ``power'' instead of the katakana ``°force''. If the first or second part is katakana, it is written as ``power'' in katakana.

Therefore, when the character feature recognition unit 5 selects the katakana word “power” or the kanji word “power” as a candidate, the character feature recognition unit 5 automatically accesses the contents of the character type symbol string dictionary 7, and the character type storage memory 8 Check the contents.

If the character type of the previous character in the contents of the character type storage memory 8 is katakana, the dictionary No. 1-1-1 in which the character type symbol string condition for determining the character type is registered as shown in Fig. 2 Since it corresponds to , it is determined to be the katakana word for “power.” Furthermore, if the character type of the previous character is something other than katakana, it cannot be determined yet, so for example, it is determined as a kanji, a mark is attached, and the next input character is recognized. After recognizing a certain number of characters, the process returns to the above-mentioned recognition of the marked characters, and then the type of the written character is read out from the contents of the character type storage memory 8 and checked by comparing it with the character type symbol string dictionary 7. At this time, if the character type written next to the character in question is katakana, katakana "
``Power'' and corrects the result that was previously recognized as a kanji. Also, if it is other than katakana, the mark is deleted, leaving the result unchanged as a kanji.

In the following explanation, the number of characters after which a specific marked character should be recognized again can be written in the character type symbol string and 7, or it can be done up to three characters later as shown in Figure 2. If you need to check it, you can fix it to 3 characters. In this check, if even one character with the same group number (leftmost symbol) of No. in Figure 2 is selected as a candidate, the character with the same group number (1-1-1, 1-1- 2,1-
2) or (2-1-1°2-1-2.2-2) all perform the same check.

Furthermore, the format of the character type symbol string dictionary 7 is not limited to the one shown in Figure 2; in the case of discrimination when there are two types of characters, either one
A character type symbol string may be defined for each type, and if the character type is not applicable, it may be determined that the character type is another type. In the same way, the following characters can also be determined, and an example of the conditions is shown below.

輯）Process in the same way as in (1) for distinguishing between the katakana ``工'' and the kanji ``工''.

(3) Distinguish between katakana "guchi" and kanji "guchi" using the same method as in (1).

(4) 3 characters to distinguish between Katakana “2” and Kanji numeral “2” +
If there is no katakana character before or after ``+rr'', use the Kanji numeral °"2°". 3 characters 1) 1
If katakana is included before or after 3 characters, use katakana no 2.

(5) Distinguishing between hiragana “su°°” and katakana “su” If there is a katakana after it, use katakana “υpa”, otherwise use hiragana “su”.

(6) Distinguishing between hiragana “he” and katakana “hepa” 3
If there is a katakana character before or after the first three characters, it will be ``8°'' in katakana, and if it is not, it will be ``°'' in hiragana.

The above-mentioned conditions are just examples, and various other conditions can be considered.

For example, when ``characters with different character types and similar shapes'' are consecutive, such as ``Inakahekaerimas'', they are written in katakana.

In short, ``characters with different character types and similar shapes are stored in advance in the character type symbol string dictionary 7 for each character type string, and the input character type array is compared with the conditions. Decide on one candidate.

(Effects of the Invention) As described in detail above, according to the present invention, there is provided a character type storage memory that stores a written character string as a character type symbol string, and a character type storage memory that stores a written character string as a character type symbol string, and a character type storage memory that stores a written character string as a character type symbol string. A character type symbol string dictionary was set up, and for a specific character, the character type was determined by comparing and collating data from both methods. Therefore,
When recognizing a specific character, it is no longer necessary to perform the cumbersome operations of pressing the next candidate key or inputting separately which character type the character belongs to, unlike in conventional technology. No additional software is required. Therefore, it is possible to easily and inexpensively distinguish between "characters of different character types with similar shapes", and there is an advantage that the first character candidate character can be automatically obtained without pressing the next candidate key.

[Brief explanation of drawings]

1. FIG. 1 is a block diagram showing the configuration of an embodiment of an online character recognition device according to the present invention, and FIG. 2 is a diagram showing an example of the contents of a character type symbol string dictionary in the device of the above embodiment. DESCRIPTION OF SYMBOLS 1... Tablet, 2... Stroke data extraction part 3... Stroke recognition part, 4... Stroke number dictionary, 5... Character feature recognition part, 6... Character feature dictionary, 7.
...Character type symbol string dictionary, 8...Character type storage memory, 9...
- Determined character memory, IO...display section, 11...interface section.

Claims (1)

[Scope of Claims] A tablet for inputting handwritten characters; a recognition unit that recognizes input characters by comparing and comparing character data from the tablet with predetermined reference character data; and the recognition unit. It has a recognized character memory that stores recognition results, and an interface section that outputs the information in the recognized character memory to the outside, and can read sentences containing a mixture of characters such as kanji, hiragana, katakana, alphanumeric characters, and symbols in real time. An online character recognition device uses a character type storage memory that stores the input character string as a character type symbol string, and a character type symbol string that stores in advance the character type symbol strings for specific characters of different character types with similar shapes. A character type symbol string dictionary is provided, and when the specific character is input, by comparing and collating the input character type symbol string in the character type storage memory and the character type symbol string in the character type symbol string dictionary. An online character recognition device characterized by determining character type and performing character recognition.