KR100564742B1 - Apparatus for converting characters into voice and method of the same - Google Patents

Abstract

Disclosed are an apparatus and method for converting text to speech. The voice data storage unit stores voice data in which sound pieces corresponding to texts are recorded. The keyboard receives text from the user. The character processing unit extracts and outputs Korean text corresponding to a character input through the keyboard unit from a predetermined character conversion table. The voice converter reads the synthesized unit matching the Korean text inputted from the text processor from the voice data storage unit, and converts the read synthesized unit into voice. The voice output unit outputs the converted voice. According to the present invention, by reading the vowel consonants and vowels of the Korean syllables or by converting the input characters into Korean syllables, the visually confirming the input characters such as the visually impaired by the people with handicap Identify spelling.

Speech synthesis, syllables, letters, keyboard, spelling

Description

Apparatus for converting characters into voice and method of the same}

1 is a block diagram showing the configuration of a text-to-speech device according to the present invention;

2 is a view showing an example of voice data stored in the voice data storage unit 150;

3 is a block diagram showing a detailed configuration of the character processing unit 120;

4 is a block diagram showing the detailed configuration of the Hangul syllable conversion unit 310,

5 is a block diagram showing the detailed configuration of the numeric Hangul conversion unit 320,

6 is a block diagram showing the detailed configuration of the English Hangul conversion unit 330,

7 is a block diagram showing the detailed configuration of the symbol Hangul conversion unit 340,

8 is a block diagram showing a detailed configuration of the voice conversion unit 130,

9 to 44 are views illustrating one example of a syllable conversion table;

45 is a diagram showing an example of a consonant conversion table;

46 is a view showing an example of a vowel conversion table;

47 is a view showing an example of a number conversion table;

48 is a diagram showing an example in which an entry of an uppercase conversion table and an entry of a lowercase conversion table are configured into one table; and

Fig. 49 is a diagram showing an example of the symbol conversion table.

The present invention relates to a voice conversion device, and more particularly, to a voice conversion device for converting spelling of Hangul syllables input through a keyboard into voice.

Conventional devices for converting text to speech perform text-to-speech conversion by converting into representative consonants and representative vowels as described in Table 1 and Table 2 for Korean compound vowels and vowel consonants. Table 1 lists representative consonants for Korean final consonants, and Table 2 describes representative vowels for Korean neutral vowels.

Korean Consonant Representative consonants A, ㄳ, ㄺ, rr, ㄳ / ㄱ / ,, ㄵ, ㄶ /N/ ㄹ, ㄼ, ㄽ, ㄾ, ㅀ / ㄹ / ㄷ, ㅅ, ㅆ, ㅈ, ㅌ, ㅌ, ㅎ / C / ㅂ, ㅄ, ㅍ / ㅂ / ㅁ, ㄻ / ㅁ /

Neutral vowels in Korean syllables Representative collection of neutral vowels in Korean syllables ㅐ, ㅔ / ㅐ / ㅒ, ㅖ / ㅒ / ㅙ, ㅚ / ㅚ /

It is difficult to discriminate between "bulletin board" and "initiation board" from the sound values of the synthesized sound generated by the apparatuses for converting conventional characters into voices based on the pronunciation principles shown in Tables 1 and 2. Of course, the conversion of the phonetic consonants and the representative vowels of the device for converting the existing text into voice is not wrong, which is reasonable based on Korean standard phonology and pronunciation rules. However, there is a problem that cannot be used in an application portion in which spelling information must be accurately detected from a sound value.

In other words, if a device for converting a conventional text into a voice is used for a visual lover's document editor, the user should input "e". If the user incorrectly inputs "e", the conventional voice converter is said to be "Ae". Is given. However, the "E" and "Ae" in the Korean vowels are difficult to identify by pronunciation, so you do not know whether you have entered "Ae" or "E". As a result, the visually impaired person has no choice but to generate a large number of misspellings and cannot detect and correct errors.

The technical problem to be achieved by the present invention is to solve the problem that can not identify the correct spelling information of the syllables only by the sound value of the syllables of the Korean syllable unit for reading and the method of combining them and the keyboard using the same An apparatus and method for converting Korean characters input through a voice into voice are provided.

Another technical problem to be solved by the present invention is to solve the problem of not being able to identify the correct spelling information of syllables only by sound values of Korean syllables. The present invention provides a computer-readable recording medium that records a program for executing a method of converting Hangul characters inputted through a keyboard into a voice.

In order to achieve the above technical problem, a text-to-speech device according to the present invention, the keyboard unit for receiving a character from the user; A character processing unit which extracts and outputs Korean text corresponding to a character input through the keyboard unit from a predetermined character conversion table; A voice data storage unit for storing voice data in which sound pieces corresponding to texts are recorded; A voice conversion unit for reading a synthesis unit matching the Korean text input from the text processing unit from the voice data storage unit and converting the read synthesis unit into a voice; And a voice output unit for outputting the voice.

In order to achieve the above another technical problem, a text-to-speech method according to the present invention comprises the steps of: (a) receiving a text from a user; (b) extracting Korean text corresponding to the input characters from a predetermined character conversion table and outputting the extracted text; (c) extracting a synthesis unit corresponding to the Korean text from speech data in which a sound piece corresponding to a character is recorded, and converting the extracted synthesis unit into speech; And (d) outputting the voice.

According to the present invention, by reading the vowel consonants and vowels of the Korean syllables or by converting the input characters into Korean syllables, the visually confirming the input characters such as the visually impaired by the people with handicap Identify spelling.

Hereinafter, with reference to the accompanying drawings will be described in detail a voice conversion apparatus and method according to the present invention.

1 is a block diagram showing the configuration of a text-to-speech device according to the present invention.

Referring to FIG. 1, the text-to-speech device 100 according to the present invention includes a keyboard unit 110, a text processing unit 120, a voice conversion unit 130, a voice output unit 140, and a voice data storage unit. It consists of 150.

The keyboard unit 110 is a component that receives a Korean character or a special character on the keyboard from the user. The character processor 120 converts a character input through the keyboard 110 into a proper Korean text string. The voice conversion unit 130 reads the synthesis unit matching the text string input from the text processing unit 120 from the voice data storage unit 150 and converts the read synthesis unit into voice. The voice output unit 140 outputs the converted voice. The voice data storage unit 150 stores voice data in which sound pieces corresponding to texts are recorded.

2 is a diagram illustrating an example of voice data stored in the voice data storage unit 150. Referring to FIG. 2, the voice data 200 includes Korean syllable syllables 210, Korean consonants 210, Korean vowels 230, Korean Bok consonants 240, Korean vowels 250, and Korean numerals. The phoneme 260, an English uppercase phoneme 270, an English lowercase phoneme 280, and a symbol Korean phoneme 290 are included.

Korean syllable syllables 210 are composed of voice signals of representative syllables for 11177 Korean syllables based on Unicode 3.0. The Korean consonants 220 and the Korean vowels 230 consist of Korean consonants and Korean vowels for the vowels. In addition, the Korean Bokjayum consonant 240 and the Korean Bokchoeum consonants 250 are composed of Hangul vowels pronounced by unwinding the Bokja and vowels, respectively. The Korean numeral phoneme 260, the English uppercase phoneme 270, and the lowercase English phoneme 280 are composed of Korean pronunciations by solving numbers, English uppercase letters, and English lowercase letters. Sign Hangeul 290 is composed of Hangul pronunciation for the symbols that can be input on the keyboard.

3 is a block diagram showing the detailed configuration of the character processing unit 120.

Referring to FIG. 3, the character processor 120 includes a Hangul syllable converter 310, a numeric Hangul converter 320, an English Hangul converter 330, and a symbol Hangul converter 340.

Characters input through the keyboard unit 110 is transmitted to the character processing unit 120 via the voice conversion unit 130. The Hangul syllable converting unit 310 converts the input Hangul syllable into a Hangul string without ambiguity between the Bokja consonant and the vowel vowel. The numeric Hangul conversion unit 320 converts the input number into Hangul word. The English Hangul conversion unit 330 converts the input English letters into Korean words by distinguishing between uppercase and lowercase letters. The symbol Hangul conversion unit 340 converts the input special characters into Hangul word. At this time, the input characters pass through the Hangul syllable conversion unit 310, the numeric Hangul conversion unit 320, the English Hangul conversion unit 330, and the symbol Hangul conversion unit 340, each conversion unit input The type of character is determined by the code value of the character to determine whether the received character is converted.

4 is a block diagram illustrating a detailed configuration of the Hangul syllable converter 310.

Referring to FIG. 4, the Hangul syllable conversion unit 310 includes a Hangul identification unit 410, a consonant confirmation unit 420, a syllable conversion unit 430, a syllable conversion table storage unit 440, and a consonant conversion unit 450. ), A consonant conversion table storage unit 460, a vowel conversion unit 470, a vowel conversion table storage unit 480, and Hangul output unit 490.

The Hangul checking unit 410 determines whether or not the Hangul is based on the code value of the character input from the voice converter 130. The Hangul checking unit 410 outputs the input character to the syllable converting unit 430 when the input character is a Korean syllable, and outputs the input character to the consonant checker 420 when the input character is a Korean alphabet. If the input character is not a Korean syllable, the Hangul identification unit 410 outputs the input character to the numeric Hangul conversion unit 320. The consonant checker 420 checks whether a consonant of a character input from the Hangul checker 410 is checked. The consonant checker 420 outputs the input characters to the consonant converter 450 if the input characters are consonants, and outputs the input characters to the vowel converter 470 when the input characters are vowels.

The syllable converter 430 extracts the representative syllables and the read-out text of the syllable corresponding to the input character (that is, the Hangul syllable code) from the syllable conversion table stored in the syllable conversion table storage unit 440 to obtain the Hangul text. Create 9 to 44 show examples of syllable conversion tables.

The consonant converter 450 extracts the consonant reading text corresponding to a character (that is, a consonant code) input from the consonant conversion table stored in the consonant conversion table storage unit 460. 45 shows an example of a consonant conversion table.

The vowel converting unit 470 extracts the read-out text of the vowel corresponding to the input character (ie, vowel code) from the vowel conversion table stored in the vowel conversion table storage unit 480. 46 shows an example of a vowel conversion table.

The Hangul output unit 490 outputs Hangul text received from each of the converters 430, 450, and 470.

Meanwhile, each of the converters 430, 450, and 470 provided in the Hangul syllable converter 310 may be configured as one Hangul converter. In this case, each storage unit 440, 460, 480 is connected to the Hangul conversion unit, the Hangul conversion unit to the character input by the confirmation signal received from the Hangul confirmation unit 410 and the consonant confirmation unit 420 Perform the conversion by selecting the corresponding conversion table. In addition, when converting Hangul syllables by one converter, it is not necessary to include the Hangul output unit 490.

5 is a block diagram showing the detailed configuration of the numeric Hangul conversion unit 320.

Referring to FIG. 5, the numeric Hangul conversion unit 320 includes a numeric confirmation unit 510, a numeric conversion unit 520, and a numeric conversion table storage unit 530.

The number confirmation unit 510 confirms whether the input character is a number. If the input character is not a number, the number checking unit 510 outputs the input character to the English Hangul conversion unit 330. The number converter 520 extracts and outputs Hangul text corresponding to the input character (ie, a numeric code) from the number conversion table stored in the number conversion table storage unit 530. 47 shows an example of a number conversion table.

6 is a block diagram showing the detailed configuration of the English Hangul conversion unit 330.

Referring to Figure 6, the English Hangul conversion unit 330 is a case conversion unit 610, uppercase conversion unit 620, uppercase conversion table storage unit 630, lowercase conversion unit 640, lowercase conversion table storage unit 650, and the Hangul output unit 660.

The case confirmation unit 610 determines whether the input letter is in English capital or lowercase. The upper and lower case character checking unit 610 outputs the input character to the uppercase conversion unit 620 if the input character is English uppercase, and outputs the input character to the lowercase conversion unit 640 if the input character is lowercase English. If the input character is not an alphabetic character, the case confirmation unit 610 outputs the input character to the symbol Hangul conversion unit 340.

The uppercase conversion unit 620 extracts and outputs Hangul text corresponding to the English uppercase letter (ie, English uppercase code) input from the uppercase conversion table stored in the uppercase conversion table storage unit 630. The lowercase conversion unit 640 extracts and outputs Hangul text corresponding to an English lowercase letter (ie, an English lowercase code) input from the lowercase conversion table stored in the lowercase conversion table storage unit 650. The Hangul output unit 660 outputs the Hangul text received from the uppercase converter 620 and the lowercase converter 640. FIG. 48 shows an example in which the entry of the uppercase conversion table and the entry of the lowercase conversion table are configured into one table.

Meanwhile, each of the conversion units 620 and 640 included in the English Hangul conversion unit 330 may be configured as one English conversion unit. In this case, each of the storage unit 630, 650 is connected to the English conversion unit, the English conversion unit selects a conversion table corresponding to the character input by the confirmation signal received from the case confirmation unit 610 conversion operation Do this. In addition, it is not necessary to include the Hangul output unit 660 in the case of converting English characters by one conversion unit.

7 is a block diagram showing the detailed configuration of the symbol Hangul conversion unit 340.

Referring to FIG. 7, the symbol Hangul conversion unit 340 includes a symbol confirmation unit 710, a symbol conversion unit 720, and a symbol conversion table storage unit 730.

The symbol confirmation unit 710 confirms whether the input character is a symbol. The symbol conversion unit 720 extracts and outputs Hangul text corresponding to the input character (ie, symbol code) from the symbol conversion table stored in the symbol conversion table storage unit 730. If there is no character input in the symbol conversion table, the character processing unit 120 ends the character processing process. 49 shows an example of the symbol conversion table.

As mentioned above, the character processing unit 120 described with reference to FIGS. 3 to 7 determines a type of a character by a code value of a character in which a separate character discriminating unit is input, and converts the character into a conversion unit corresponding to the type of the determined character. It may be configured to deliver the input character. In this case, the input characters are input only to the converters 310 to 340 corresponding to the type of characters. In addition, it is not necessary to include the identification unit 410, 420, 510, 610, 710 in each of the conversion unit (310 to 340).

8 is a block diagram illustrating a detailed configuration of the voice converter 130.

Referring to FIG. 8, the voice converter 130 includes a voice search unit 810, a speech rate converter 820, and a voice combination unit 830.

When the Hangul text is input from the character processor 120, the music search unit 810 searches for a sound piece corresponding to the input Hangul text among various sounds 210 to 290 stored in the voice data storage unit 150. The speech rate changing unit 820 changes the speech rate in order to reduce the speech time of the retrieved sound and to increase the response speed to the user. In particular, the speech rate of the Korean syllables read out is faster than the representative syllables. In addition, the utterance conversion unit 820 converts the upper and lower case letters, the pitch of the upper case and the volume of the lower case, the lower pitch of the lower and lower volume of the English phonetic pronunciation. Sound pieces whose speech rate is converted are combined into a single voice signal sequence by the sound piece combining unit 830.

The invention can also be embodied as computer readable code on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, and may also be implemented in the form of a carrier wave (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

Although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the specific preferred embodiments described above, and the present invention belongs to the present invention without departing from the gist of the present invention as claimed in the claims. Various modifications can be made by those skilled in the art, and such changes are within the scope of the claims.

According to the speech conversion apparatus and method of the input characters according to the present invention, by performing the speech conversion by the Hangul text that is converted to the reading method of the input characters, the visually impaired by the voice corresponding to the characters input through the keyboard Visually confirm the input characters, such as those with a handicap can identify the correct spelling of the Hangul syllable input. In addition, in the case of unread Korean syllables, it is possible to convert the speech syllables of the representative syllables and the consonant of the original syllables converted into representative syllables and the unreadable Hangul text of the vowels into different speech rates. Can be easily distinguished.

Claims (16)

A keyboard unit for receiving characters from a user; A character processing unit which extracts and outputs Korean text corresponding to a character input through the keyboard unit from a predetermined character conversion table; A voice data storage unit for storing voice data in which sound pieces corresponding to texts are recorded; A voice conversion unit for reading a synthesis unit matching the Korean text input from the text processing unit from the voice data storage unit and converting the read synthesis unit into a voice; And It includes; voice output unit for outputting the voice, The character processing unit, Hangul syllable conversion unit for converting the bokja consonants and vowel ambiguities of the input Hangul characters into Hangul string; Numeral Hangul conversion unit for converting the input number into Hangul word; English Hangul conversion unit for converting the input alphabet to Hangul word; And Symbol Hangul conversion unit for converting the input of the special characters on the keyboard Hangul phrases. delete The method of claim 1, The character processing unit checks the type of the input character based on the code value of the input character and converts the input character into the Hangul syllable conversion unit, the numerical Hangul conversion unit, the English Hangul conversion unit, and the symbol Hangul. Text-to-speech device characterized in that it further comprises a character verification unit for selectively outputting to the conversion unit. The method of claim 1, The Hangul syllable converter, A Hangul identification unit for checking whether the Hangul syllables of the input character are based on the code value of the input character; A consonant confirmation unit confirming whether consonants and vowels of the input characters are based on code values of the input characters; A syllable conversion table storage unit for storing a syllable conversion table in which a read text for Hangul syllables is recorded; A consonant conversion table storage unit configured to store a consonant conversion table in which read-out text for Hangul consonants is recorded; A vowel conversion table storage unit for storing a vowel conversion table in which the read text of the Korean vowel is recorded; A syllable converting unit for extracting and outputting a read-out text corresponding to a code value of an input character identified as the Hangul syllable from the syllable conversion table; A consonant converter for extracting and outputting a read-out text corresponding to a code value of the input character identified as the Korean consonant from the consonant conversion table; And And a vowel conversion unit for extracting and outputting the read-out text corresponding to the code value of the input character identified by the Hangul vowel from the vowel conversion table unit. The method of claim 1, The numerical Hangul conversion unit, A number checking unit for checking whether or not the number of input characters is based on a code value of the input character; A number conversion table storage unit for storing a number conversion table in which the read text for the number is recorded; And And a numeric conversion unit for extracting and outputting the read-out text corresponding to the code value of the input character identified as the number from the number conversion table. The method of claim 1, The English Hangul conversion unit, An alphabetic character checking unit that checks whether uppercase or lowercase letters of the input character are based on the code value of the input character; An uppercase conversion table storage unit configured to store an uppercase conversion table in which read text for English uppercase letters is recorded; A lowercase conversion table storage unit configured to store a lowercase conversion table in which read text for English lowercase letters is recorded; An uppercase conversion unit for extracting and outputting a read-out text corresponding to a code value of the input character identified by the uppercase letters from the uppercase conversion table; And And a lowercase conversion unit for extracting and outputting the read-out text corresponding to the code value of the input character identified by the lowercase letters in the lowercase conversion table. The method of claim 1, The symbol Hangul conversion unit, A symbol confirmation unit which checks whether a symbol of an input character is based on a code value of the input character; A symbol conversion table storage unit which stores a symbol conversion table in which the read text of the symbol is recorded; And And a symbol conversion unit for extracting and outputting the read-out text corresponding to the code value of the input character identified as the symbol from the symbol conversion table. A keyboard unit for receiving characters from a user; A character processing unit which extracts and outputs Korean text corresponding to a character input through the keyboard unit from a predetermined character conversion table; A voice data storage unit for storing voice data in which sound pieces corresponding to texts are recorded; A voice conversion unit for reading a synthesis unit matching the Korean text input from the text processing unit from the voice data storage unit and converting the read synthesis unit into a voice; And It includes; voice output unit for outputting the voice, The voice conversion unit, A piece search unit for searching for a piece corresponding to the Hangul text input from the text processing unit from the voice data stored in the voice data storage unit; And And a sound combiner for combining the sound pieces into one voice signal string and outputting the sound pieces. The method of claim 8, The voice conversion unit further comprises a speech rate changing unit for selectively changing the speech rate, pitch, and volume of the retrieved piece according to the type of the character. A keyboard unit for receiving characters from a user; A character processing unit which extracts and outputs Korean text corresponding to a character input through the keyboard unit from a predetermined character conversion table; A voice data storage unit for storing voice data in which sound pieces corresponding to texts are recorded; A voice conversion unit for reading a synthesis unit matching the Korean text input from the text processing unit from the voice data storage unit and converting the read synthesis unit into a voice; And It includes; voice output unit for outputting the voice, The voice data is composed of Korean syllable syllables composed of voice signals for representative syllables of Korean syllables, Korean consonant syllables composed of voice signals for Korean consonants, Korean vowel syllables composed of voice signals for Korean vowels, and Korean Bokja Hangul Bokjaum consonants composed of one voice signal, Hangeul Vowel consonants composed of voice signals pronounced by Korean vowel vowels, Numeric consonants composed of Korean pronunciation voice signals for numbers, English uppercase consonants composed of voice signals pronounced by English capital letters, and English A voice conversion device comprising at least one of an English lowercase phoneme consisting of a phonetic signal decoded by lowercase letters, and a phoneme phoneme consisting of a Korean phonetic voice signal for a symbol that can be input on a keyboard. (a) receiving a text from a user; (b) extracting Korean text corresponding to the input characters from a predetermined character conversion table and outputting the extracted text; (c) extracting a synthesis unit corresponding to the Korean text from speech data in which a sound piece corresponding to a character is recorded, and converting the extracted synthesis unit into speech; And (d) outputting the voice; In step (b), (b1) checking a type of the input character based on a code value of the input character; And (b2) when the input character is a Hangul character, converts the Hangul string of the Hangul character and the ambiguity of the vowel vowels, and if the input character is one of a number, an English letter, and a special character, corresponds to the input character. And converting the input text into a Hangul word based on a predetermined text conversion table. delete The method of claim 11, The character conversion table may include a syllable conversion table in which read-out text is written for Hangul syllables, a consonant conversion table in which read-out text is written in Hangul consonants, and a vowel conversion table in which read-out text is recorded in Hangul collection. , Numeric conversion table with read text for numbers, uppercase conversion table with read text for uppercase letters, lowercase conversion table with read text for lowercase letters, and solvers for symbols And at least one or more of the symbol conversion tables in which the read text is recorded. The method of claim 11, Step (c) is, searching for a piece corresponding to the Hangul text from the voice data; And (c2) combining the sound fragments into a single voice signal string and outputting the sound signal; The method of claim 14, Before the step (c2), further comprising the step of selectively changing the speech rate, pitch, and volume of the searched sound piece according to the type of the input character. (a) receiving a text from a user; (b) extracting Korean text corresponding to the input characters from a predetermined character conversion table and outputting the extracted text; (c) extracting a synthesis unit corresponding to the Korean text from speech data in which a sound piece corresponding to a character is recorded, and converting the extracted synthesis unit into speech; And (d) outputting the voice; In step (b), (b1) checking a type of the input character based on a code value of the input character; And (b2) when the input character is a Hangul character, converts the Hangul string of the Hangul character and the ambiguity of the vowel vowels, and if the input character is one of a number, an English letter, and a special character, corresponds to the input character. And converting the input character into a Hangul word based on a predetermined character conversion table. 2. The computer-readable recording medium having recorded thereon a program for executing a voice conversion method of a character in a computer.

Images