JPS6022227A - European text processor - Google Patents

Abstract

PURPOSE:To improve the transmission efficiency by detecting words and the start/end of a sentence out of a text containing only input small letters and converting the head alphabet of a word and the head alphabets of the words set immediately the start and the end of a sentence into capital letters. CONSTITUTION:A text containing only small letters given from a character recognizer, etc. is fed to an input terminal 1, and the extracted 2 words are stored to a word memory 4 at a character output part 3 as well as to a proper noun detecting part 5. The part 5 retrieves a proper noun table 6 with the extracted words to detect the proper nouns in an input text. Then the part 5 delivers a detection signal to a character conversion part 9 when the proper noun is detected. At the same time, the sentence start/end detecting parts 7 and 8 detect the start and end of a stentence respectively and sends a detection signal to the part 9. The part 9 converts the head alphabet of the detected proper noun and the word extracted at the end or immediately after the end of a sentence into capital alphabets and outputted 10.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to a Roman text processing device that automatically converts English text consisting only of uppercase or lowercase letters into text containing a mixture of uppercase and lowercase letters.

[Prior art]

In a character recognition device, when character recognition is performed character by character, it is extremely difficult to distinguish uppercase and lowercase characters such as 0, P, and S, and there is a high probability of erroneous recognition. Therefore, conventionally,
For post-processing of character-by-character recognition.

In this process, misrecognized characters are corrected using context and other factors. However, although the character recognition device is expensive, there is a problem in that it cannot sufficiently reduce errors in recognizing uppercase and lowercase letters.

If a means could be obtained to automatically convert English text consisting only of uppercase or lowercase letters into mixed case text, it would be possible to configure a character recognition device to recognize characters as only uppercase or lowercase letters. I can do it,
It has the potential to lower device costs and reduce case recognition errors. Furthermore, even in systems that transmit English text, it is possible to transmit text consisting only of uppercase or lowercase letters and convert it to mixed case text on the receiving side, potentially improving transmission efficiency.

〔the purpose〕

The present invention has been made in view of these points, and its purpose is to provide a Roman text processing device that automatically converts English text consisting only of uppercase or lowercase letters into mixed case text. be.

〔Example〕

An embodiment of the present invention will be described with reference to FIG.

Text consisting only of lowercase (or uppercase) letters from a character recognition device, telex, computer, etc. is input to input terminal 1.
is input. The word extractor 2 extracts words from the input text and outputs the extracted words as a lowercase code string. The extracted words are stored in the word memory 4 of the character output section 3 and are also input to the proper noun detection section 5.

6 is a proper noun table storing proper nouns. The proper noun detection unit 5 detects proper nouns in the input text by searching the proper noun table 6 for the extracted words, and outputs a detection signal when a proper noun is detected. 7
is a sentence start detection unit that detects the start of a sentence (STX) from the input text, and outputs a detection signal when the start of a sentence is detected. Reference numeral 8 denotes a confounding detection unit that detects the end of input text (period, question mark, exclamation mark, etc.), and outputs a detection signal when detected.

Each detection signal output from the proper noun detection section 5, sentence start detection section 7, and entanglement detection section 8 is input to the uppercase conversion section 9 in the text output section 3. The uppercase conversion unit 9 detects the proper noun detected by the proper noun detection unit 5 and the beginning of a sentence by the sentence start detection unit → from among the words extracted by the word extraction unit 2 and stored in the word memory 4. Only the first letter of the word extracted immediately after the word is extracted, or the word extracted immediately after the entanglement is detected by the entanglement detection unit 8, is converted into an uppercase letter. The converted words are sent from the output terminal 10 to a printer or the like.

An example of conversion is shown in FIG. In this figure, (a) is an input text, which is converted into an output text as shown in (b). That is, the first letter of a sentence or the first letter of a proper noun (the circled letter in FIG. 2) is converted to an uppercase letter.

Another embodiment of the present invention will be described with reference to FIG.

In this figure, since 1 to 11 are the same as the corresponding parts in the previous embodiment, their explanation will be omitted, and only the other parts will be explained below.

11 is an italic word table in which words to be printed in italic font are registered. The italic detection unit 12 is
An italic word table 11 is searched for the word sent from the output terminal 10, and if the word should be printed in an italic font, a detection signal is given to the CG selector 13. 1
4 is a roman character generator that generates a roman character pattern, and 15 is an italic character generator that generates an italic character pattern. When the CG selector 13 is not given a detection signal from the italic detector 12, the CG selector 13 causes the Roman character generator 14 to generate a Roman character pattern corresponding to each character code of the word sent from the output terminal 10. When a detection signal is given from the italic detector 12, the italic character generator 15 generates an italic character pattern corresponding to each character code of the word.
It outputs it to the print buffer 16. Print buffer 16
The character pattern data obtained is outputted from the terminal 17 to a plotter or the like.

As described above, according to this embodiment, a specific word can be converted into an allograph.

〔effect〕

As shown in the embodiments described above, according to the present invention, a Roman text consisting only of uppercase or lowercase letters can be automatically converted into a text containing mixed uppercase and lowercase letters. Therefore, if the Roman text processing device according to the present invention is used, for example, as a post-processing device for a character recognition device, the character recognition device can be configured to recognize only uppercase or lowercase letters on a character-by-character basis. It is possible to realize a character recognition device that does not generate recognition errors. Furthermore, if the Roman text processing device according to the present invention is installed at the receiving terminal of a data transfer system, the sending terminal only needs to transfer text consisting of uppercase or lowercase letters, which simplifies the sending terminal and improves transmission efficiency. can be raised.

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram showing one embodiment of the present invention, and FIG.
The figure shows an example of input text and output text, and FIG. 3 is a schematic block diagram showing another embodiment of the invention.

Claims (1)

[Claims] (1) A Roman text processing device that converts a Roman input text consisting only of uppercase or lowercase letters into a mixed case text, which includes a word extraction unit that extracts words from the input text, and a word extraction unit that extracts words from the input text, and a sentence start part that extracts words from the input text. a sentence start detection unit that detects the end of a sentence from an input text, a proper noun table that stores proper nouns, and a proper noun table that stores the words extracted by the word extraction unit. A proper noun detection unit that detects proper nouns by searching, and a word extracted immediately after the beginning of a sentence is detected by the sentence start detection unit among the words extracted by the word extraction unit or the word above. For words extracted immediately after the end of a sentence is detected by the end-of-sentence detector, only the first letter of the word is output as uppercase and the remaining letters are output as lowercase.For other words, all letters are output as lowercase.Text output A Roman text processing device having a section.