JPS61175851A - Character processor - Google Patents

Abstract

PURPOSE:To avoid the possibility of a wrong input caused by the peripheral noise, etc. of a microphone, by constituting a titled device so that a voice input mode can be turned off automatically, in case there is no voice input exceeding a prescribed time. CONSTITUTION:When a character processing operation is started, whether an input mode switch 32 is on or not is decided, and if it is on, a timer incorpo rated in a microprocessor MPU1 is started, and if it is off, the timer is returned to the start time point. When the timer is started, a voice input processing is executed in accordance with the processing procedure of a ROM13, and also whether there is a voice input or not is discriminated. In case when there is the voice input, the timer is reset, and the timer is started again. In case when there is no voice input, whether an elapsed time t1 from the timer start is larger than a prescribed time T1 (30sec-1min) or not is discriminated. If the former is larger, the input mode switch 32 is turned off, and if the former is not larger, the processing is returned to a step for discriminating whether there is a voice input or not once again.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to character processing devices such as electronic typewriters and word processors, and particularly to a character processing device that can perform voice input processing extremely efficiently.

[Prior Art] Character processing devices such as Japanese word processors having a kana-kanji conversion function are known to have a method for inputting the operator's voice (kana) in addition to a method for inputting kana characters from a keyboard device. ing.

The advantage of the voice input method is that documents can be input instead of keyboard input operations that require skill.

However, in character processing devices that use a voice input method, it is difficult to create an entire document using voice input alone. ) into a character corresponding to the case particle (e.g., "wa"), it is difficult to perform the task simply by voice input. Also,
There is also a problem unique to voice input work in that if no voice is uttered for a certain period of time during voice input work, there is a risk of incorrect input being made due to background noise or the like.

[Object] Therefore, it is an object of the present invention to solve the above problems and provide a character processing device that can perform voice input extremely efficiently.

[Example] The present invention will be described in detail below with reference to the drawings.

FIG. 1 shows an example of the configuration of the character processing device of the present invention, where 1 is a processing unit (MPU) in the form of a microprocessor, and 2 is an address bus for specifying the address of each input/output memory. , 3 is a bidirectional data bus used for transferring various data, and 4 is a control bus used for transmitting read, write, interwoven reception of each memory, data set timing, etc.

The MPU has a built-in timer.

5 is a keyboard provided with various keys such as alphabet keys, hiragana keys, numeric keys, recall keys, input 2 edit, save, and print keys.

Reference numeral 6 denotes a keyboard controller for controlling the keyboard 5, which encodes data input through the keyboard 5 and sends interwoven signals to the control bus 4. Note that the decoded key data is temporarily stored in the key buffer 6A, and is encoded by the controller 8 in response to access from the MPU 1.
Read into PU.

7 is a CRT controller which controls the CRT display device 8, display controller 9, character generator 10, refresh memory 11, and refresh memory controller 12.0 The display controller 9 refreshes the pattern signal from the character generator 10. Processing is performed according to display control signals such as brightness 2 inversion, underlining, blinking, etc. stored in memory 11, and the processed output is supplied to CRT display device M8 to output a display pattern.

The character generator 10 sequentially outputs, among the stored character patterns, character patterns for lines specified by inputting a code such as a character code or a line address of the character pattern.

The refresh memory 11 includes a CRT display device 8
The character code and display control code to be displayed on the screen are stored, and according to display instructions from the CRT controller 7,
The character code and display control code are repeatedly output. The refresh memory controller 12 controls the above-described refresh cycle for CRT display and the writing and reading of data from the MPU 1 in accordance with instructions from the CRT controller 7.

Reference numeral 13 denotes a read-only memory (ROM) which stores control procedures to be described later as well as various processing procedures such as character processing. 14 is a controller that controls the ROM 13.

Reference numeral 31 denotes a microphone for the operator to input voice, and the voice recognition device 30 performs processing such as frequency analysis, feature extraction, and pattern matching on the voice input through this microphone 31 to recognize monosyllables. do,
The voice recognition device 30 converts the voice signal into a character code and imports it into the device main body.In the frequency analysis, the voice recognition device 30 converts the frequency band of a certain monosyllable into the frequency band of a plurality of consecutive monosyllables that have been input. If the value is higher than that, an accent flag is set in the output character code and output.

15A is a random access memory (RAM),
It temporarily stores various data and also develops a dictionary for kana-kanji conversion in a predetermined area. Further, 15B is a RAM for storing created documents. 15G is a RAM having a key buffer 6A, a storage area for data sent from the speech recognition unit 30 (conversion buffer), a storage area for grammar processing 9 editing processing and tabulation processing routines, and a work area, which will be described later; is RAM1SA, R
This is an output buffer that temporarily stores the outputs of AM15B and RAM15G. 17 is RAM15A, RAM15B
and a controller for RAM 15G.

18 was provided as necessary. 19 is a floppy disk drive; 20 is a controller for the floppy disk drive; 2
1 is a printer; 22 is a controller for the printer 21;

The character processing apparatus of the present invention having the above-described structure is operated by input from the keyboard 51 (not shown) or the microphone 31 (not shown). That is, when an input is made from the keyboard 5 or the microphone 31, an interwoven signal is supplied from the keyboard controller B or the voice recognition section 30 to the MPU I via the control bus 4, and character processing is performed according to the control procedure stored in the ROM 13. The present invention is characterized in that it performs character processing by voice input, as will be described later, but of course, depending on the voice input, the processing cannot be performed, and tasks that cannot be performed can be performed from the keyboard 5, for example. This is done by inputting data.

FIG. 2 shows an example of the key layout configuration of the keyboard 5, where 5A is a group of input keys for letters and numbers, 5B is a group of keys for selecting a processing mode for documents,
For example, print processing using the printer 21 is performed in response to pressing a "print" key. Furthermore, when the "voice input" key is pressed, the input mode switch 32 is closed so that the output of the voice recognition device 30 is connected to the data bus (
(turns on), the input mode switch 32 is controlled to be turned on or off by the NPU I. 5C is space key, 5
d is a group of cursor movement keys on the CR7 screen.

FIG. 3 shows an example of the voice input operation of the character processing device of the present invention. As shown in FIG. 3, in step Sl, it is determined whether or not there is voice input.
When "shi" is inputted, the speech recognition device 30 outputs a kana character code string (watashiwa) corresponding to the input, and the process proceeds to step S2.
In the case of , a character code string (Hanao) is output and the process advances to step S2.

In step S2, the character code from the speech recognition device 30 is input to the conversion buffer area in the RAM 15C. This buffer area has a capacity for, for example, 8 characters.Then, proceed to step S3, and select the tie in NP old? (re
trigable type). Next, the process proceeds to step S4, where it is determined whether character codes have been input to all addresses in the conversion buffer.

If character codes have been input into all the buffers, the process proceeds to step S5; otherwise, the process returns to step S1. In the above example, character codes have not been input to all addresses in the conversion buffer, so the process returns to step St. If there is no next input, the process proceeds to step S8, and it is determined whether the timer has timed out. The time from when the timer starts until it times out is, for example, 1 to
It can be about 2 seconds. If the timeout has not occurred in step S8, the process returns to step S1, and if the timeout has occurred, the process proceeds to step S7.

In step S7, a timeout code (-) is input next to the last character code in the conversion buffer, and the process proceeds to step S5.

In step S5, the character code string in the conversion buffer is extracted and converted into kana-kanji by referring to the dictionary in the RAM 15A. Similarly, a character code string (hanao) is converted to a sentence containing kanji and kana (hanao).

Next, the process proceeds to step S8, and it is determined whether the accent flag is attached to the sentence containing kanji and kana obtained in STELLA7'S5. If the accent flag is on, proceed to step S9; if not, proceed to step S1G
nisumu 0 For example, if you input a continuous monosyllable, such as "wa", "ta", "su", or "ka", there is no voice input until the timer times out (timeout code is added), and then the sound may end. When the pronunciation of ``ka'' is given a higher accent than before (in such cases, the pronunciation of ``ka'' is usually cut off).

In step S5, the sentence containing kanji and kana (Is it me?) is obtained, and then the process proceeds from step S8 to step S9, where the timeout code (=) after the sentence containing kanji and kana (Is it me?) is deleted and the mark "?" is obtained. ” after a sentence containing kanji and kana (i.e. 1, for example (Is it me?))
, proceed to step 910.

In step SIO, it is determined whether the last part of the sentence containing kanji and kana is "wa", "o", or "e", and if so (Yes), the process proceeds to step Sll, otherwise (
No) Proceed to step S12 0 For example, if the sentence containing kanji and kana is (wa wa) or (hanao), proceed to step Sll, and if it is (is it me?), proceed to step 5t12.

In step Sll, it is determined whether the timeout code (-) or the mark "?" follows the sentence containing kanji and kana, and if so, the process proceeds to step S13, and if not, the process proceeds to step S14.

For example, in the case of the kanji/kana mixed sentence (wa wa) 10 timeout code (=) and (hanao) + (u), the process goes to step S13 because they are applicable.

In step S13, the last characters "o", "wa" and "e" of the kanji-kana mixed sentence are replaced with the characters "wo", respectively.
0 For example, in the case of a sentence containing kanji and kana (Washi wa), it is converted to (Washi wa), and in the case of (Hanao), it is converted to (Hana wo). Ru.

In step S12, it is determined whether the sentence after the kanji/kana mixed sentence is a timeout code (-), and if so, (Y
es) Proceed to step S15, otherwise (No
) Proceed to step S14 0 For example, in the case of (Is it me?) mentioned above, the answer is NO, so proceed to step S14.
For example, if the high-pitched voice input (pronunciation) is "Yu" or "ni" and the input is interrupted until a timeout occurs, a timeout code (-) is added to the character code (step S7), Kana-Kanji conversion is performed (step S5), and the kanji-kana mixed sentence (on the floor) + (u) is obtained, and then the process goes through step 51G and proceeds to step S12, where the kanji-kana mixed sentence is followed by (u). It is determined that this is the case, and the process proceeds to step S15.

In step S15, the timeout code (-)
Determine whether the sentence before is a noun, case particle, conjunctive particle, or adverbial particle. If so (Yes), proceed to step S1B; otherwise (No) proceed to step S17.0 For example, a sentence containing kanji and kana ( In the case of )+(u) on the floor, since (ni) is a case particle, the answer is Yes and the process proceeds to step S18.

In step S1B, the timeout code (=) after the kanji/kana mixed sentence is converted to a comma ",', and the process proceeds to step S14. 0For example, the kanji/kana mixed sentence (on the floor)
+ For (LJ), the sentence is a mixture of kanji and kana (on the floor)
is obtained.

In step S17, the timeout code (u) after the sentence containing kanji and kana is converted to a period "," and the process proceeds to step S14.0For example, the voice input (pronunciation) is "o".
If the timer is not input after that and the timer times out, a timeout code (u) is added to the one-character code (step S7), and kana-kanji conversion is performed (step S5): Kanji-kana mixed. Text commercial
)+(u) is obtained, and then the process proceeds to step S1H via step SIQ, where it is determined that the sentence containing kanji and kana before (-) is not a noun, case particle, conjunctive particle, or adverbial particle, and the process proceeds to step S17. Proceeding, in step S17, it is determined that the part before the (-) is the final form, and (u)
Convert to a period “, ′ and write a sentence with kanji and kana (put,)
is obtained.

In step S14, the kanji-kana mixed sentence obtained as described above is outputted to the output buffer 1B, output processing such as display on a CRT is performed, and the process returns to step S1. ) as “wa” da” shi” wa” / time when the timer times out / “ha” na” o” / time when the timer times out / “yu” or “ni” / time when the timer times out / “o” Just by saying “ku”, you can mix kanji, kana, jiri (I put flowers on the floor,
) is obtained (see Figure 5), and the pronunciation “wa” is obtained.
If you input ``su''ka'' with ``ta'' (when the ``ka'' is accented), you will get a sentence with kanji and kana (Is it me?) (see Figure 6).

FIG. 4 shows another example of the operation of the character processing device of the present invention.
As shown in FIG. 4, in step S41, it is determined whether the input mode switch is on, and if it is on, step S4
Proceed to 2, and if it is off, return to Chi. In step S2, a built-in timer is started in the MP old, and the process proceeds to step S43.
The voice input process explained in the figure is performed (step S1
to step S14), and then proceeds to step S44. In step S44, it is determined whether or not there is voice input. If so, proceed to step S45, reset the timer, and return to step S42. If there is no voice input, the process proceeds to step 94B, and the elapsed time 11 from the timer start is determined to be the predetermined time T! (e.g. 30 seconds to 1 minute)
Determine whether it is larger. If it is larger, the process advances to step S47 and the input mode switch is turned off. If it is not larger, the process returns to step S44.

As described above, the voice input mode can be automatically turned off when there is no voice input for a certain period of time or more, and therefore, it is possible to effectively avoid the possibility of erroneous input due to microphone surrounding noise or the like.

Note that the automatic off operation of one or more input modes requires, for example, determining whether the voice input information is a final type, starting a timer if so, and inputting the final type voice input information for a certain period of time. For example, the input mode switch can be turned off.

[Effects] As explained above, according to the present invention, voice input processing can be performed extremely efficiently.

[Brief explanation of drawings]

FIG. 1 is a diagram showing an example of the configuration of a character processing device of the present invention, and FIG. 2 is a diagram showing the arrangement of a keyboard. FIG. 3 is a diagram showing an example of the voice input operation of the character processing device of the present invention, FIG. 4 is a diagram showing another example of the voice input operation of the character processing device of the present invention, and FIGS. 5 and 6 are diagrams of the present invention. FIG. 1...MPU, 13...ROM, 15A, 15B, 15G...RAM, 3G...speech recognition device. y) lh L18""s l5r=H,'"5bi1→f
i−+=ヱ” 7”y) F, <'/4
47'>) U cRT, %i Multiple input of Figure 6 (A2 Senri 11'l "Two L?"

Claims (1)

[Scope of Claims] Speech recognition means for recognizing voice input information and converting it into character information; identification means for identifying breaks in the character information from the voice recognition means; and character recognition means based on the information from the identification means. 1. A character processing device comprising: means for processing information.