JP2658476B2 - Document Braille device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a document braille device, and more particularly to a document braille device for creating a braille document intended for a visually impaired person.

[Conventional technology]

The conventional document Braille converting apparatus has been totally dependent on a person who does not have a visual impairment to perform a conversion operation into a Braille document for a visually impaired person. That is, various conversion operations from selection of a document to be converted to management of the converted document have been supported by the activities of volunteers who have no visual impairment.

[Problems to be solved by the invention]

In the conventional document braille apparatus described above, the user of the converted document and the operator performing the conversion are completely separated from each other, and the user and the operator are different from each other. In spite of misalignment in quantity and quality, Braille documents are not managed sufficiently even if Braille is turned on, and Braille documents are dispersed or obtained by users who need them. It has the disadvantage that it is difficult.

SUMMARY OF THE INVENTION It is an object of the present invention to select and confirm various readings of a document by speech for each phrase, and to automatically convert the reading into Braille, so that a visually impaired person can freely perform the work of converting the document into Braille. It is to provide a Braille device.

[Means for solving the problem]

The document braille converting apparatus according to the present invention comprises: an input device (1-1) for reading an input sentence; a voice synthesizing unit (1-4) for synthesizing and outputting a voice; and a plurality of one-to-two reading candidates. A dictionary (1-5) for storing words, an input / output buffer (1-8) for holding data, and an input sentence from the input device (1-1) are decomposed into words and converted into words (1-5). A segmentation unit (1-2) for assigning all the reading candidates for the word based on the word, grouping the word into a phrase, and holding the phrase in the input / output buffer (1-8); An interface unit (1-7) to which a braille creation instruction is input, and all reading candidates held in the input / output buffer (1-8) are sequentially output as speech from the speech synthesis unit (1-4) for each phrase. , Decides one of the readings of the phrase according to the reading candidate decision instruction The input / output buffer (1-8) is stored in the input buffer (1-8), and the input / output buffer (1-8) is stored in the dictionary (1-5). A braille creating unit (1-9) for converting the held phrase reading to braille.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the document braille apparatus of the present invention.

As shown in FIG. 1, the document braille device of the present embodiment
An input device 1-1 for reading and inputting a target document; a dictionary 1-5 for machine processing; a segmentation unit 1-2 for separating input data into segments using the dictionary 1-5; Speech synthesis section 1-4 that outputs reading candidates as speech, determination section 1-3 that branches processing, interface section 1-7 that exchanges information with the user, and dictionary registration section 1 that performs dictionary registration -6, and a Braille creating unit 1-9 that converts a phonetic symbol string corresponding to the final reading that has been correctly answered into a Braille string and launches it.
And an input / output buffer 1-8 for storing data exchanged between the above devices and units.

Therefore, the input of the document is performed using the input device 1-1. If the document is a printed matter, the document is read and input by an OCR or a document proofreading device. At this time, an error may be present in the input sentence.

Then, the segmentation unit 1-2 uses the dictionary 1-5 to divide the input sentence into segments and to pass them to the judgment unit 1-3.
The determining unit 1-3 includes a voice synthesizing unit 1-4 and a dictionary registering unit 1- in response to an instruction from the user through the interface unit 1-7.
6. Activate any of the Braille creating units 1-9.

The speech synthesizing unit 1-4 outputs, as a speech, a candidate for reading a phrase being processed based on a phonetic symbol string for the phrase, which is information from the dictionary 1-5.

Further, the dictionary registration unit 1-6 performs dictionary registration of dictionary information including pairs of phonetic symbols and parts of speech, based on input data and instructions from the user, for a character string to be registered in the dictionary.

On the other hand, the Braille generator 1-9 converts the phonetic symbol string into a Braille string.

FIG. 2 is a flowchart showing an example of a data flow in this embodiment. Next, the operation of this embodiment will be described in accordance with the data flow shown in FIG.

First, in step 2-1, the sentence "The empty set is a subset of all the sets" input to the input / output buffer 1-8 by the input device 1-1 is sent to the segmentation unit 1-2. , [Empty set], [all], [of the set], [subset. ] Is divided into four clauses. A phrase is a word string that starts with a part of speech word other than the delimiter and ends with the delimiter, using punctuation marks, particles, and auxiliary verbs as delimiters.
Or, it is one of an adverb, conjunction, or verb when no delimiter comes after. In addition, the word here refers to the dictionary 1
In the dictionary 1-5, one kanji is registered in advance, and one or a plurality of reading candidates are stored in each word. However, since the empty set and the subset are a plurality of words, they are not initially registered.

Therefore, in step 2-3, all reading candidates are
The data is read from the dictionary 1-5 and stored in the input / output buffer 1-8. Then, the determination unit 1-3 starts the process for each clause.

First, the judging section 1-3 determines that the empty set of the first clause is
In response to this, the phonetic symbol {so'ra} 'SuOugoOu {Wa} of the first candidate reading is extracted from the input / output buffer 1-8 and passed to the speech synthesis unit 1-4. No. 4 performs audio output for the audio symbol {so'ra} 'SuOugoOu {Wa}. However, since the pronunciation is not correct, the user can hear the sound output one after another by instructing the determination unit 1-3 from the interface unit 1-7 to the next candidate.

And the phonetic symbols of the third candidate reading ｛ku'Ou｝ ｛'SuOu
By instructing dictionary registration for voice output by goOu {Wa}, the dictionary registration unit 1-6 allows the delimiter {W
空 ku'Ou｝ ｛'SuOugoOu｝, which is the accent of ｛ku'Ou｝ ｛' SuOugoOu｝ except for a｝
In addition to registering the noun of u｝ as a noun,
3 is the input / output buffer 1-8, and the determined phonetic symbol {kuOu
Su'OugoOu｝ is held.

Next, the user instructs the next section from the interface section 1-7 to the determination section 1-3, thereby shifting to [all] processing of the next section.

When the processing of all the clauses is completed in this way, in step 2-4, a definite phonetic symbol for the sentence "The empty set is a subset of all the sets."｛KuOuSu'OugoOuWa'｝ ｛su, beteno｝ ｛bubuOnS
u, OugoOudeOaru.｝ is obtained.

Then, the user instructs the determination unit 1-3 from the interface unit 1-7 to create the braille, so that step 2 is performed.
-4, the braille creation unit 1-9 outputs the phonetic symbol {kuOuSu'OugoO
A braille string corresponding to uWa,｝ ｛su, beteno｝ ｛bubuOnSu, OugoOudeOaru.｝ can be generated.

〔The invention's effect〕

As described above, the document braille apparatus of the present invention selects and confirms various readings of a document by speech for each phrase, and automatically converts the readings into Braille, so that the visually impaired can convert the reading of the document. This has the effect that Braille work can be performed freely.

In addition, the document braille device of the present invention stores pronunciations in the form of phonetic symbol strings, and divides the document into phrases to process the text, so that it is easy to determine the pronunciation, so that the braille accuracy is improved. This has the effect that it can be greatly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a document braille converting apparatus according to the present invention, and FIG. 2 is a flowchart showing an example of a data flow in the present embodiment. 1-1: input device, 1-2: segmentation unit, 1-3 ...
Judgment unit, 1-4 ... Speech synthesis unit, 1-5 ... Dictionary, 1-
6 ... Dictionary registration unit, 1-7 ... Interface unit, 1-
8 ... I / O buffer, 1-9 ... Braille creation unit.

Claims (1)

(57) [Claims] An input device for reading an input sentence, a voice synthesizing unit for synthesizing and outputting a voice, and a plurality of words to which one or a plurality of reading candidates are stored. A dictionary (1-5), an input / output buffer (1-8) for holding data, and an input sentence from the input device (1-1) are decomposed into words. A phrase segmentation unit (1-2) that assigns all reading candidates, combines words into phrases, and holds the words in the input / output buffer (1-8), a reading candidate determination instruction, and a Braille creation instruction The input unit (1-7) and all the reading candidates held in the input / output buffer (1-8) are sequentially output as speech from the speech synthesizing unit (1-4) for each phrase. Determines one of the phrase readings according to the determination instruction of the input / output buffer (1 8) and a determination unit (1-3) for registering the determined word reading in the dictionary (1-5), and a determined phrase stored in the input / output buffer (1-8) in response to the Braille creation instruction. And a Braille creating unit (1-9) for converting the reading of the text into Braille.