JPS581821B2 - Japanese data input device - Google Patents

Description

[Detailed Description of the Invention] This invention aims to easily input Japanese data written in kanji, hiragana, katakana, etc. by using symbols, alphanumeric characters, and kana character recognition. It is related to the device.

OCR (Optical Channel) is used for conventional symbol/character recognition.
offline recognition of handwritten or printed characters (recognition of characters that have already been written or printed) using a digitizer, etc., and online recognition of handwritten characters using a digitizer (recognition that takes into account stroke order while handwriting). ).

For both recognition methods, the recognition technology for symbols, alphanumeric characters, and katakana has almost been completed, and the technology for hiragana character recognition has also been completed to some extent.

However, the recognition of the most important kanji in Japanese notation is: ■ Kanji are more complex than other characters, so high reading resolution is required. ■ Kanji has many types, so high discrimination ability is required. This makes it more difficult in terms of both hardware and software compared to the knowledge of symbols, alphanumeric characters, katakana, and hiragana, and not only is it difficult to achieve a high recognition rate, but the reading recognition equipment is also very expensive. There were drawbacks.

This invention was made to eliminate these drawbacks, and is designed to easily input Japanese data including Kanji characters by recognizing only symbols, alphanumeric characters, and katakana characters.

This invention will be explained in detail below.

In this specification, kana means hiragana and katakana, and in the case of hiragana, it is written as "kana" or "hiragana".
In the case of katakana, it will be expressed as ``kana'' or ``katakana.''

FIG. 1 is an information flow diagram showing the configuration of an embodiment of the present invention in a block diagram. 1 is Japanese data including kanji, and 2 is Japanese data written in katakana with identification symbols representing the characteristics of kanji. 3 is a recognition section that recognizes the pattern information from the scanner section 2; 4 is a conversion section that converts the katakana written Japanese recognized by the recognition section 3 into a sentence containing kanji and kana; .

Figure 2a shows an example of a handwritten or printed Japanese text used in this invention, and Figure 2b shows an example of the Japanese text used in this invention.
This figure shows an example of a Japanese sentence expressed as a katakana sentence with symbols using the input keyboard section, which will be described later.

FIG. 3 shows details of an example of the input keyboard section.

In Fig. 3, the input keyboard section 1A is the keyboard section 1B.
and a kanji information adding keyboard 1C, the kana writing board 1B has a katakana key 1B1, and the kanji information adding keyboard 1C has a pattern key 1C for adding additional information representing the characteristics of the kanji. place name) key 1C
2, a separator key 1C3, and a feed key 1C4.

Again in FIG. 2b, 5 is a pattern representing the characteristics of a kanji character (hereinafter referred to as a feature identification symbol) added by the input extractor 1A of FIG. 3, and 6 is a katakana character, which is a Japanese The data is separated by feature identification symbol 5.

Japanese handwritten or printed in the form shown in Figure 2b (hereinafter referred to as katakana text with symbols) is converted into pattern information by the scanner part 2, and then converted into katakana text with symbols by the recognition unit 3. is converted into a katakana code and a feature identification symbol 5 corresponding to the character, and further converted word by word into a character (kanji) code corresponding to a sentence containing kanji and kana.

That is, the conversion file in the conversion unit 4 is registered in advance so that conversion can be performed uniquely.

In the case of "election" above, the conversion file is scanned with the signal code "■senkiyo■", and if it matches the pre-registered code for "election", the kanji corresponding to "election" is detected. This is how the code is read.

Next, various application examples of the system configuration according to the present invention will be shown.

In FIG. 4, a terminal device 10 has various functions, namely a scanner section 2, a recognition section 3, a conversion section 4, and a transmission control section 7, which connect a front end processor 8 and a central processing section via a communication line 11. The character code is sent to a center 12 having a device 9, where it is recognized and converted into a Japanese sentence mixed with Kanji and Kana.

In FIGS. 5 to 7, pattern information is sent from the terminal device 10 using facsimile codes, character codes are sent to the center 12 via the communication line 11, and kanji and kana are mixed in the communication line 11 or at the center 12. This allows the recognition and conversion of the text into Japanese sentences.

In FIGS. 8 and 9, characteristic identification symbols and kana character codes are sent to the center 12 via the communication network 11, and are recognized and converted into Japanese sentences containing kanji and kana in the communication network 11 or at the center 12. It is something.

In this way, the present invention enables not only local processing but also systems with various configurations using lines to be realized.

Note that the present invention is applicable to both online recognition and offline recognition, and the characteristic identification symbol is not limited to those described above, but may be any symbol that can be distinguished from other symbols or characters.

Also, it goes without saying that if it's only kana, it's fine to write only kana.

As explained above, the present invention has many advantages as described below.

■ Since the data is written in parts using alphanumeric and kana characters and a limited number of about 10 types of simple symbols, the burden on humans inputting data is extremely small, regardless of whether it is handwritten or typed.

■ Because it recognizes alphanumeric kana and about 10 simple symbols, conventional recognition technology can be used as is, and advanced kanji recognition is not required, so the scanner and recognition unit can be made inexpensive and the recognition rate can be improved. will improve.

■ Since the word unit, that is, the odd-numbered and even-numbered feature identification symbols and the kana character string sandwiched between these feature identification symbols, is converted into a sentence containing kanji and kana, the recognition unit Even if % recognition is not achieved, there is a possibility of correct conversion on a word-by-word basis, and there is no need to improve the recognition rate more than necessary.

According to this invention, alphanumeric and kana characters are recognized and converted into kana sentences with symbols using a few feature identification symbols.
It is possible to input Japanese data that is inexpensive and has a high recognition conversion rate using conventional alphanumeric and katakana character recognition technology, and because functions can be easily distributed, it is possible to configure an efficient system. has advantages.

[Brief explanation of the drawing]

FIG. 1 is an information flow diagram showing the configuration of an embodiment of this invention in a block diagram, FIG. 2 a shows an example of Japanese text used in this invention, and FIG. Figure 3 showing an example of the Japanese sentence a expressed in katakana sentences with symbols.
The figure shows details of the input keyboard section, and FIGS. 4 to 9 are block diagrams showing configuration examples of the system according to the present invention. In the figure, 1 is Japanese data, 2 is a part of the scanner, 3 is a recognition unit, 4 is a conversion unit, 5 is a feature identification symbol, and 6 is a katakana character.

Claims (1)

[Claims] 1. A part of the scanner that optically reads handwritten or printed Japanese kana data character strings, which are formed by adding characteristic identification symbols representing the graphical characteristics of the kanji before and after the pronunciation kana written in kanji, and this scanner. A recognition unit that identifies the odd-numbered and even-numbered feature identification symbols read in part and the kana character strings interposed between the feature identification symbols, and a summary of the kana character strings, feature identification symbols, and kanji codes. It consists of a conversion unit having a conversion file, searches for the conversion file using the information of the recognition unit, and if it matches, outputs the kanji code of the conversion unit to convert the handwritten or printed Japanese data. A Japanese data input device characterized by converting the text into the corresponding kanji code.