JPS61250778A - Electronic translating device - Google Patents

Abstract

PURPOSE:To realize an electronic translating device in which it is not necessary to convert a unit manually by storing a conversion rate between mutual units beforehand and providing a unit translating part to convert a numerical value etc., inputted at the unit of the original language into a numerical value of a unit of translating language. CONSTITUTION:A central processing unit (CPU) 30 connected through an input output part 34 to a speech recognizing part 35 and a speech synthetic part 36 has a ROM 32 to store the converted value of the unit and a RAM to store temporarily the unit converted result besides a main memory 31 to execute the usual translating action. The CPU 30 translates the original language inputted from the speech recognizing part 35 by the usual translating action, and when the numerical value, etc., to need the conversion are present, the ROM 32 is referred to, converted into the unit of the translating language, and the converted result is once stored to the RAM 33 and the translation is proceeded.

Description

[Detailed description of the invention] The present invention relates to an electronic translation device.

In this specification, translation refers to (1) translating the language of one country into the language of another country, (2) translating an abbreviation into the original text even if it is in the same domestic language, or translating a standard Mutual conversion between words and dialects, characters and words, etc., (3) operations (general calculations and adding π to calculate “3”)
．． 14…”), the translated language refers to the language, numbers, formulas, etc. after translation, and the source language refers to the language, numbers, formulas, etc. before translation. For example, Japanese-English style In the electronic translation device, "Good morning" is the original language, and "Good morning" is the translated language. Language translation refers to translation that uses the units of the source language as is without unit conversion, and unit translation refers to translation that uses unit conversion to convert to the units of the target language.

Conventional electronic translation devices (hereinafter referred to as TLRDs) have only been capable of simply translating a source language into a target language.

For this reason, it had the disadvantage of being difficult to use due to its narrow range of applications.

The present invention eliminates such drawbacks and provides a TLRD that is versatile, highly practical, and easy to use without requiring any special ability or knowledge.
It provides:

An embodiment of the present invention will be described below with reference to FIGS. 1 and 2.

Fig. 1 is a block diagram of a TLRD which is an embodiment of the present invention, and Fig. 2 is a block diagram of its main parts. A microphone (hereinafter referred to as MIC) is connected to an electronic translation unit (hereinafter referred to as TLR) 4 which is set for Japanese-English translation via a speech recognition unit (hereinafter referred to as SA) 3, and 7 is a Percoll type speech synthesis unit (hereinafter referred to as MIC). (hereinafter referred to as SH) 5 and the amplifier (
This is a speaker (hereinafter referred to as SP) connected to the TLR 4 via the TLR 4 (hereinafter referred to as AMP) 6. 11 is an M1C connected to the electronic translation unit (hereinafter referred to as TLR) 14 which is set for English-Japanese translation via the VCA type AMP 12 and SA 13 as described above, and 17 is connected to the electronic translation unit (hereinafter referred to as TLR) 14 via the VCA type AMP 12 and SA 13.
This is an SP connected to the TLR 14 via.

8 and 18 are connected to TLRs 4 and 14, respectively, and are character display devices (hereinafter referred to as L) for displaying the original language and translated language in characters.
CD). 9.19 are MIC1 and 11 respectively
Detects the end of the original language audio input signal from TLR4.
This is a translation execution instruction section (hereinafter referred to as AS) for instructing the translation operation to the computer 14. Details of the technical content of AS9.19 have already been proposed by the present inventor in Japanese Patent Laid-Open Nos. 57-17081 and 57-100561. Figure 2 is a block diagram of the main parts of the TLRs 4 and 14, where 31 is the main memory (hereinafter referred to as MA), 32 is
Read-only memory (ROM) that stores unit conversion values such as INCH cm, 33 is a rewritable memory (RAM) that temporarily stores unit conversion results, and 30 is language translation, unit translation, and calculation. 34 is an input/output unit (hereinafter referred to as I/O) that also serves as an interface between the CPU 30 and external devices such as SA3.13 and SM5.15; , 36 are ports for connecting to SA3.13 and SM5.15, respectively.

Next, the operation of this TLRD will be explained.

First, when you say to MIC1, "How much is this?", this audio signal is applied to SA3 via AMP2. SA3 converts this original language audio signal into an original language character signal called "korewaikradeska" and sends it to the MA of TLR4.
Temporarily write to part of 31. AS9 detects the end of the audio signal from MIC1 and transmits the translation execution instruction pulse to TLR.
4. By applying this pulse, TLR4 becomes MA
31. The CPU 30 translates the original language "Kore Waikura Deska" into a character signal in the translated language "HOW MUCH IS THIS?" and applies it to the SM5 via the I/O 34 and port 36. This signal is converted into a voice signal in the translation language "Hau matuchi is this" by SM5, and then A
It is sent to SP7 via MP6. Audio signal is AM
While the signal is being sent from P6, both AMP2 and 12 are in a state where the amplification operation is stopped (amplification degree is 0, CUT OFF).
Therefore, the input audio signal from MIC1.11 is SA3.
13 is not applied. Therefore, the translation language from SP7 will not be re-inputted and translated incorrectly. Details regarding this technology are provided by the inventor in Japanese Patent Application Laid-Open No. 56-1644.
This has already been proposed in Publication No. 74. Also, AMP6
When the output from the AMP 2.12 is reset and returned to its original state, the SM5 emits a transmission end sound (beep) via the SP7. The above is the language translation operation.

Next, when you say "Three Dollars" to the MIC 11, this original language input audio signal is sent to the SA via the AMP 12.
After converting to the character signal “3DOLLARS” at step 13, T
Applied to LR14.

The AS 19 operates the TLR 14 at the end of the audio input signal. The TLR 14 uses the CPU 30 and MA 31 to convert this into a character signal of the translation language "Sandru Des", and uses the RAM 33, ROM 32, and CPU 30 to
Furthermore, it is determined whether or not there is a unit that belongs to the source language (in this example, dollars) in the translated text, and if the unit is found, it is converted to a unit that belongs to the translation language (in this example, yen).

Therefore, the unit conversion value ($1=¥250,
1INCH=2.54CM, ■1=2.5 yen, 1mil
e=1.6km, 1 pound=approximately 453 g, etc.) are written in advance. As a result, the CPU 30
Read “$1 = 250 yen” from M32 and calculate with “3DOLLARS” from MA31 (multiply in this case)
Then, the character signal "750 yen" after the unit translation is written into the RAM 33.

In this case, the configuration is such that the numerical values before and after each unit and the converted value in the ROM 32 are calculated. And CPU30 is MA
A part of the sentence written in 31 is partially corrected to the content written in RAM 33, read out sequentially, and sent from I/O 34 to SM 15 via port 36. Therefore, SM1
5 is applied with a character signal in the translation language "750 Endes". Next, SM15 converts this into a voice signal saying "It's 750 yen" and sends it to SP17 via AMP16.
Send to. The LCD 18 displays two character signals of "3DOLLARS" and "750 ENDES".

Also, if you say to MIC1, "This needle is 7.5 cm," TLR4 will respond with "THIS NEEDLE IS7.
5cm,” and further translated as “THIS NEEDLE.”
IS ABOUT3INCHES,” translated into SP7
Say "This Needle Is About Three Inches". In this example, when converting, the second decimal place was rounded off, so 2.95...⇒3. In this way, when the result of division continues below the second decimal place, the second decimal place is rounded off to the quotient, and "ABOUT", "about" before the numerical value
and so on to compose the translation output.

In addition, although this example shows examples of conversion between dollar and yen and inch and cm, the conversion is not limited to these, and conversion between other units may also be possible. For example, FEET, m, kg
, g, POND, franc, German mark, penny, little, gallon, acre, carat, dozen, shaku, size, %
, kg/cm2, m/s, cm/s2, mile, yard, hectare, tsubo, anti, m2, cm3, go, momme, kan, ounce, pound, Western calendar - Showa, yen/km, ■/g, etc. …
…. In this way, in this embodiment, when language translation is performed, the units of the source language sentence are converted into the units of the translated language, which saves the effort of making separate calculations each time using a calculator or the like.

In this example, once the translation is made (without unit conversion),
Although an example of performing unit conversion is shown, the present invention is not limited to this, and it is also possible to perform unit conversion all at once, or output the first translation (before unit conversion) from LCD 8.18 and SP 7.17, respectively. Thereafter, the second translated text (after conversion) may be output from the LCD 8.18 and the SP 7.17, respectively. In addition, in this example, MIC1.1
1, Voice input according to SA3.13 and SM5.15,
Although an example of audio output using SP7.17 is shown, the example is not limited to this, and it may be either one of them, key input, OCR input (character input - character recognition), or character output using a printer such as a word processor. Good too.

In this example, the amplification degree of AMP6.12 is changed to AMP2.1.
The device was designed to have a larger width than the one shown in Fig. 6.
The distance between C1 and SP17 is the same as the maximum arm length (usually), so it is short, and conversely, the distance between this Japanese person and an American (who is generally tall) who is having an English conversation via this device, and MIC11,
The distance from SP7 is longer than this, so even if you have a conversation at the same volume, the Japanese voice will be louder in MIC1, and the voice from SP17 will be audible louder, but on the other hand, the American voice will be louder. does not fall significantly into MIC11, and SP7
Since it is not possible to hear the voice from another person, a good translated conversation is not possible unless the amplification degree of AMP12 is relatively greater than that of AMP2, and the amplification degree of AMP16 is greater than that of AMP6. It is also possible to provide a relative gain difference by attenuating not only the amplification degree of AMP but also the volume.

This also applies when this device is interposed in the telephone line.
The same thing may be required depending on the location of the RD.

Another embodiment of the present invention will be described using FIG.

FIG. 3 is a block diagram of a TLRD which is another embodiment of the present invention.

In the figure, reference numerals 51 and 61 are MICs connected to a composite speech recognition unit (hereinafter referred to as MSA) 53, which includes an SA for specific speakers and an SA for non-specific speakers, via AMPs 52 and 62, respectively. 54 is a bidirectional type (for both English-Japanese and Japanese-English) that translates the source language character signal from the MSA 53 into the language selected by the translation execution instruction unit (hereinafter referred to as TS) 63 and applies the translated language character signal to the SM 55. It is TLR.

57 is an SP connected to the SM 55 via the VCA type AMP 56. 58 is connected to TLR54 and the source language/
LCD for character display of both translated languages, 60 is T
This is a key input unit (hereinafter referred to as KEY) for directly applying character signals, operation commands, etc. to the LR 54. TS6
3 selects a translation language based on the speech recognition rate of the MSA 53 and gives a translation instruction to the TLR 54, which is a technique already proposed by the present inventor in Japanese Patent Laid-Open No. 142479/1983.

In this embodiment, the TS 63 also has the role of selecting the translation language and controlling the amplification degree of the AMP 56 by switching it from large to small depending on the type of the selected language. Also, AMP
The degree of amplification of AMP 62 is greater than that of AMP 52.

Next, the operation of this device will be explained.

First, let's consider the case of renting a car in the United States. When a Japanese person whose voice is registered in advance says, "How much does it cost?" to the MIC 51, this voice signal is applied to the SA for specific speakers of the MSA 53 via the AMP 52. At the same time, it is applied to the speaker-independent SA of the MSA 53 via the MIC 61 and AMP 62. In this case, since Japanese voices are registered in advance in the SA for specific speakers, the recognition rate is higher than in the SA for non-specific speakers. The TS 63 discriminates (detects) this information, and sends only the character signal from the SA for a specific speaker from the MSA 53 to the TLR 54.
4 performs Japanese-English translation operation, and TS63 performs AMP5
6 is controlled to increase the degree of amplification.

As a result, a character signal of the translation language "WHAT'S THE RATE?" is applied to the SM55.

The SM 55 converts this into an audio signal called "Father the Late" and sends it to the SP 57 via the AMP 56.
In response, an American said to MIC61, “TWELV
E DOLLARSPER DAY. ", this audio signal is sent to the MS via AMP62, MSA53's SA for unspecified speakers, MIC51, and AMP52.
It is applied to both SAs of the specific speaker SA of A53. At this time, American voices are not registered, so S for unspecified speakers is used.
A has a better recognition rate. The TS63 detects this and sends the output of the SA for unspecified speakers from the MSA53 to the TLR54.
At the same time, the TS63 causes the TLR54 to perform English-Japanese translation and obtain the main translated sentence "Ichinichi 12 Dollar Des", but after determining the presence or absence of the source language unit "Dollar" in the translated sentence, the TLR54 converts the word "Dollar" into Japanese. Previous number “12” and RO
Multiply by the conversion rate "250" from M, translate "12 dollars" into "3000 yen" in yen, and correct a part of the main translation. After this, write "Ichinichi 3000" in SM55. Apply the character signal "Endes", and SM5
From No. 5, an audio signal saying "It's 3,000 yen a day" is obtained. This audio signal is then transmitted to the AMP56 (at this time T
S63 sends it to SP57 via the AMP56 (the degree of amplification is set smaller than the previous time). Although an example of $→¥ is shown here, the conversion is not limited to this, and conversion of ¥→$ is also possible. By doing this, Japanese people can
51 and say, "How much does it cost?" (In this case, there is no source language unit in the source language sentence, so TLR54 does not perform any calculations for conversion, but merely performs language translation.)
P57 loudly chanted “WHAT’S THE RA.
TE? ” In response, the Americans responded with MIC6.
1, “TWELVE DOLLARS PER
DAY. ” (In this case, “DOLLAR” is included in the source language sentence.
Since there is a unit in the source language called "S" (S indicates plural), not only language translation but also calculations for unit conversion and unit translation are executed), SP57 softly says "1". It's 3,000 yen a day.'' Therefore, instead of literally translating it as ``3 dollars,'' as in the past,
By storing the exchange rate in advance in memory such as ROM, RAM, floppy disk, tape, card bubble, etc., the calculation 3 dollars x 250 yen/dollar = 3000 yen can be calculated by T.
It is executed inside the LR54 and performs not only so-called language translation, which can output "3000 yen", but also unit translation.

The case where voice input is performed using the MIC51.61 has been described, and next, the case where the voice input is performed using the KEY60 will be described.

A character signal of the original language is applied to the TLR 54 via the KEY 60 and output to the SP 57 via the SM 55 and AMP 56. At this time, the type of translation language is selected by the KEY 60, and the translation operation by the TLR 54 is executed.
This selection output is sent to the AMP via TLR54 and TS63.
The degree of amplification of 56 is changed depending on the size,
This operation is the same as the voice input from the MIC51.61 above. However, when inputting with this KEY60, the first mode is only to translate the source language into the translated language. For example, "HOW MUCH"
IS THE CHARGE FOR THAT RO
OM? As an answer to the question “TWENTY-F
When you key in the source language "IVE DOLLARSPER NIGHT," TLR54 displays "Hitoban 2.
The translated language "5 Dollar Des" is output and displayed on the LCD 58, and the audio with the same content is output on the SM 55 and AMP 5.
It is uttered from SP57 via 6.

When the operator further manipulates KEY60 to request the second mode, TLR54 will display "25 dollars" as the second mode.
is calculated using a yen exchange rate of 250 yen/dollar and translated as "6,250 yen." As a result, LCD58 displays “Hitoban 6”.
250 Endes” is displayed, and SM55, AMP
56, you can hear a voice from SP 57 saying, ``It's 6,250 yen per night.'' Therefore, MIC51.
When inputting voice from 61, direct SM55, AMP56
, the operation is different from the voice output from the SP57 that says, "It's 6,250 yen per night." This is KEY
60 for language learning with this device, or when conversion rates are constantly changing and accurate conversion values are required for business purposes, you may need units or conversion rates that are not registered in memory. This is because in some cases, it may be better to operate the KEY 60 to output in the first mode.

However, if you need mutual translation between Japanese, English, and English, such as when traveling to the United States, and you only need to know the approximate value (which is the case for many general tourists), you can use the MIC51.61 to enter voice input.

Also, when using MIC51.61, the first mode and the second
The microcomputer's software program can also be modified to cause the audio output to be delivered sequentially in the mode.

The LCD 58 displays the original language input character signal to the TLR 54 and the translated language output character signal from the TLR 54, respectively.

The respective amplification degrees of AMP52 and AMP62 are small for AMP52 and large for AMP62. The reason is the same as in the previous embodiment. The amplification degree of the AMP 56 is reset after the end of audio output transmission.

In the above two embodiments, an example of translation between Japanese and English was shown, but the translation is not limited to this, and bidirectional or unidirectional translation between other languages is also possible. You may also do this with In addition, it is possible to perform not only currency conversion but also all weights and measures reciprocal calculations.
No. 23625, No. 56-165166, No. 55-16
No. 6325, No. 55-187062 No. 55-1762
No. 96, No. 57-175135, No. 58-18470
No. 9, it is also possible to add technologies already proposed elsewhere to achieve high added value.

The TLRD of the present invention may be used by connecting to a television, radio, telephone, communication device, tape recorder, etc., or may be used by having these built-in.
A), electronic translation department (TLR), speech synthesis department (SM), A
The S, TS, VCA, etc. may be configured with a one-chip LSI, or the AS, TS, VCA core roll, etc. may be executed using an SA, TLR, or SM microcomputer. TLR shown in Figures 1 and 2
Although D shows an example in which SA, TLR, SM, etc. are housed in one housing, these may be divided.As described above, according to the present invention, it is possible to perform a more complete translation. , the hassle of manual unit conversion (calculation using a calculator, solovan, etc.) can be greatly reduced.

[Brief explanation of drawings]

Fig. 1 is a block diagram of a TLRD which is one embodiment of the present invention, Fig. 2 is a block diagram of the same main part, and Fig. 3 is a block diagram of a TLRD which is another embodiment of the present invention.1.11 ...Microphone (MIC), 2.12...VCA amplifier (AMP), 3.13...
Speech Recognition Department (SA), 4.14...Electronic Translation Department
(TLR), 5.15...Speech synthesis unit (SM), 6
．． 16...Output amplifier (AMP), 7.17...Speaker (SP), 8.18...Character display device (
LCD), 9.19... Translation operation execution instruction section (AS),
31. ...Main memory (MA), 32. ...Memory (ROM), 33. ...Memory (RAM), 30 ...Central processing unit (CPU). 34...Input/output unit (I/O), 35.36...Port, 51.61...Microphone (MIC), 52.62...Input amplifier (AMP), 53...Multiple speech recognition unit (MSA) ), 54...Two-way electronic translation department (
TLR), 55...Speech synthesis unit (SM), 56...VCA amplifier (AMP), 57...Speaker (SP), 58...Character display unit (LCD), 60...Key input unit (KEY) , 63... Translation language selection section (TS).

Claims (3)

[Claims] (1) It has a source language input section for inputting the source language and a memory section that stores exchange rates between units in advance, and numbers, characters, symbols, etc. input in the units of the source language are stored in the memory. The numerical value of the unit of translation language, based on the conversion rate of
An electronic translation device comprising an electronic translation unit that performs unit translation for converting into characters, symbols, etc., and a translated language output unit that outputs the translated language. (2) The electronic translation device according to claim 1, wherein the electronic translation unit performs language translation from a source language to a translation language and also performs unit translation. (3) According to claim 2, the electronic translation section is capable of selecting and transmitting either a translated language output as a unit of the original language or a translated language output after unit translation. The electronic translation device described.