Patent Grant
9384733
The present invention relates to a call registration device for an elevator.
In an elevator which performs car call registrations by button operations, car call registrations by button operations are sometimes difficult, for example, in the case where the user is visually handicapped, in the case where a user carries things, and in the case where the elevator is very crowded. Therefore, there has hitherto been known an elevator provided with a call registration device which permits call registrations by voice input without requiring manual button operations. When a voice uttered by a user is inputted to a voice input section, such as a microphone, this call registration device for an elevator recognizes the inputted voice and performs the registration of a call to a destination floor and the like uttered by the user. However, in such a call registration device for an elevator, due to the effect of environmental noises and the like in places where voice input is performed, there is a possibility that a wrong call is registered.
Therefore, in conventional call registration devices for an elevator, there are known call registration devices for an elevator which are such that a voice input section, such as a microphone, is provided in an elevator hall, human detection means which detects whether or not a user stands in front of this voice input section is provided, and only when this human detection means detects a user, the voice inputted to the voice input section is recognized and a call registration is performed (for example, refer to Patent Literature 1 and Patent Literature 2). In the call registration devices described in Patent Literature 1 and Patent Literature 2, an ultrasonic sensor, an infrared sensor, an optical sensor, and a mat sensor are used as human detection means.
The call registration device described in Patent Literature 2 is provided with a voice output section consisting of a speaker for notifying, by voice, a user of results of a call registration by voice recognition and the like. And so that a voice uttered from this voice output section does not interfere with voice recognition during a call registration, the recognition of the voice inputted to the voice input section is prohibited while the voice output section is outputting the voice.
In conventional call registration devices for an elevator using the voice recognition, in addition to the above-described call registration devices for an elevator, there are also known those provided with an indication device which indicates to a user the contents of recognition when the voice inputted to a voice input section is recognized (for example, refer to Patent Literature 3).
In fields other than call registration devices for an elevator, in a voice recognition control device which recognizes a voice inputted to a voice input section and executes an execution command, there have hitherto been known those which, in the case where a voice inputted to a voice input section has been recognized as a predetermined start command, starts the voice recognition of an execution command and the execution of the execution command for which voice recognition has been carried out (for example, refer to Patent Literature 4).
However, in the conventional call registration devices for an elevator described in Patent Literature 1 and Patent Literature 2, there arises a problem that when the human detection means has performed erroneous detection, also the voice recognition ceases to function rightly. In addition, in the conventional call registration device for an elevator described in Patent Literature 2, there arises a problem that because voice recognition is prohibited while the voice output section is outputting a voice, it is difficult for a user to get to know the timing at which a call registration by voice recognition is possible.
Furthermore, the conventional call registration device for an elevator described in Patent Literature 3 poses a problem that because erroneous registration is to be suppressed by indicating the contents of voice recognition on the indication device, it is impossible for a user whose sense of sight of is inadequate to check the contents of indication on the indication device and hence this function is ineffective as well as a problem that the manufacturing cost of products increases due to the provision of the indication device.
Furthermore, in the conventional voice recognition control device described in Patent Literature 4, because a start command for starting the voice recognition of an execution command is fixed, there arises a problem that depending on the condition of the device, periods of time when the device is used, and the like, the start command itself may become apt to be erroneously recognized or become unnatural.
The present invention was made to solve such problems as described above and provides a call registration device for an elevator which can suppress erroneous registration of calls by voices other than a voice uttered by a user for call registration and can start voice recognition for a call registration through the use of a start command which is not unnatural and is not easily erroneously recognized.
A call registration device for an elevator according to the present invention, which performs a call registration by voice recognition using a voice input section which receives the input of a voice uttered by a user of the elevator, comprises: a voice recognition section which stores beforehand voice data of a predetermined call registration command representing a destination floor or direction of the call registration and voice data of a predetermined start command which is used for starting voice recognition of a call registration command, and which is different from the call registration command and does not include the call registration command, wherein the voice recognition section also performs the voice recognition by judging whether or not a voice inputted to the voice input section matches either the voice data of the stored call registration command or the voice data of the start command; a call registration control section which, in the case where the voice inputted to the voice input section was judged to match the voice data of the call registration command, outputs to an elevator control device a call registration request to the destination floor or direction represented by the relevant matched call registration command; a start command changing part which changes, according to predetermined condition, the start command which becomes effective among the start commands stored by the voice recognition section, wherein in the case where the voice recognition section judges that the voice inputted to the voice input section matched the voice data of the start command made effective by the start command changing part, the voice recognition section starts the voice recognition as to whether or not the voice inputted to the voice input section matches the voice data of the call registration command.
In the call registration device for an elevator of the present invention, it is possible to suppress erroneous registration of calls by voices other than a voice uttered by a user for call registration and to produce the effect that it is possible to start voice recognition for a call registration through the use of a start command which is not unnatural and is not apt to be erroneously recognized, either.
Embodiment 1
In
At least either in the car 1 or each of the halls 2 is provided with a call registration device main body 3 by use of which an elevator user registers a desired call.
The voice input section 3a is intended for receiving the input of a voice uttered by a user. This voice input section 3a consists of a device which converts the sound waves of, for example, a microphone to an electrical signal. And the voice inputted to the voice input section 3a is converted by this voice input section 3a to a voice signal and is outputted.
The voice signal outputted from the voice input section 3a is inputted to the voice recognition section 3b. The voice data of a predetermined call registration command is stored beforehand in this voice recognition section 3b. And the voice recognition section 3b performs voice recognition processing by performing comparative checking of the voice inputted to the voice input section 3a against the voice data stored beforehand in the voice recognition section 3b.
The call registration command voice data is voice data of call registration command (waveform data on voice signals) which consists of a vocabulary representing destination floors of call registration (for example, “first” and “first floor” representing the first floor and “second” and “second floor” representing the second floor. Also, the vocabulary may be a floor name, such as “main floor” representing the main floor, and the names of representative facilities present on the floor instead of a floor number).
In the call registration device main body 3 installed in the hall 2, a vocabulary representing the direction of a call (for example, “up” representing the upward direction and “down” representing the downward direction) may be included as a call registration command.
The voice recognition section 3b first cuts out an utterance division which is a voice part uttered by a user from the voice signal outputted from the voice input section 3a. And the voice recognition section 3b performs the comparative checking of the voice signal of this cut-out utterance division against the voice signal of the voice data of the call registration command, thereby performing voice recognition by judging whether or not the voice inputted to the voice input section 3a matches any of the call registration command voice data.
Results of the voice recognition by this voice recognition section 3b are outputted to the call registration control section 3c. The call registration control section 3c is communicatively connected by a communication path 5 to an elevator control device 4 which governs the control of the whole operation and action of the elevator. The call registration control section 3c outputs a call registration request to the elevator control device 4 on the basis of the results of the voice recognition outputted from the voice recognition section 3b.
In the case where results of the voice recognition in the voice recognition section 3b reveal that the voice inputted to the voice input section 3a matched the call registration command, the call registration control section 3c outputs a call registration request to the destination floor or direction corresponding to this matched call registration command to the elevator control device 4. The elevator control device 4 which received the call registration request from the call registration control section 3c performs the registration of a call for the destination floor or direction designated by the relevant call registration request. At the same time, the call registration control section 3c outputs information regarding the destination floor or direction of the outputted call registration request to the voice output section 3d.
This voice output section 3d is intended for informing a user of various kinds of information regarding the condition of the elevator including results of voice recognition and results of call registration. The voice output section 3d has a speaker and the like which sounds voices. And the voice output section 3d can inform a user of various kinds of information by broadcasting a voice announcement which is set beforehand from this speaker and the like.
As described above, in the case where a call registration command was recognized by the voice recognition section 3b, a call registration request for the destination floor or direction corresponding to the relevant call registration command is outputted from the call registration control section 3c to the elevator control device 4. At the same time, information regarding the destination floor or direction of the relevant call registration request is outputted from the voice call registration section 4 to the voice output section 3d. The voice output section 3d which received the information regarding the destination floor or direction of this call registration request informs a user by voice announcement of results of voice recognition, i.e., results of call registration on the basis of this received information. As a concrete voice announcement on this occasion, for example, “We've registered X-floor” (X is a numeral representing a destination floor) is conceivable.
In the case where although a voice was inputted to the voice input section 3a, the inputted voice did not match any of the call registration commands which had been set beforehand and no call registration command was recognized by the voice recognition section 3b, the voice output section 3d outputs a voice announcement to the effect that the voice recognition section 3b failed in the recognition of the call registration command. As a concrete voice announcement on this occasion, for example, “We couldn't recognize the command” is conceivable. On the basis of this voice announcement from the voice output section 3d, a user can get to know whether or not the destination floor or direction desired by himself or herself was recognized.
In addition to the call registration command voice data, predetermined start command voice data is also stored beforehand in the voice recognition section 3b. This start command voice data is voice data of a start command consisting of a specific vocabulary for starting the voice recognition of a call registration command. The vocabulary of this start command is set to be different from a vocabulary consisting of a call registration command and does not include a vocabulary constituting a call registration command. Here, words of greetings usually used in daily conversations (for example, “Good morning”, “Good afternoon”, “Good evening”) are set as start commands. In this manner, words of greetings usually used in daily conversations are used as a start command, whereby voice recognition becomes possible in natural dialogs.
In normal times, the voice recognition section 3b does not perform the checking of a voice inputted to the voice input section 3a against a call registration command, that is, the voice recognition section 3b does not perform the recognition of a call registration command. Instead of this, in normal times, the voice recognition section 3b performs the checking of a voice inputted to the voice input section 3a against a start command, that is, the voice recognition section 3b performs the recognition of a start command. And in the case where the checking of a voice inputted to the voice input section 3a against a start command (recognition of a start command) was successful, the checking of a voice inputted to the voice input section 3a against a call registration command (recognition of a call registration command) is started.
Therefore, in normal times, only after a user utters a start command and this start command is recognized by the voice recognition section 3b, a call registration by the utterance of a call registration command becomes possible. When after the recognition of a start command by the voice recognition section 3b, a call registration command is recognized and a call is registered, the voice recognition section 3b suspends the recognition of a call registration command until a start command is recognized again. Also in the case where after the recognition of a start command by the voice recognition section 3b, a predetermined given time elapses without the recognition of a call registration command, the voice recognition section 3b suspends the recognition of a call registration until a start command is recognized again. On this occasion, the voice output section 3d makes a voice announcement to the effect that the receiving of a call registration command is finished.
Or it may be ensured that until the voice recognition section 3b recognizes a start command, even when a call registration command is recognized, a call registration request is not outputted from the call registration control section 3c so that a call is not registered.
A plurality of vocabularies different from each other are set in a start command. That is, a plurality of start commands different from each other are stored beforehand in the voice recognition section 3b. And the voice recognition section 3b is provided with a start command changing part 3e which changes and selects a start command which becomes effective among these multiple start commands under a predetermined condition. And the predetermined condition under which the start command changing part 3e changes an effective start command is, here, to which time division of the day fixed beforehand the present time belongs. As a result of this, it is possible to change a start command in such a manner as to suit a time division of the day and it becomes possible to perform voice recognition by a natural dialog suited to a time division of the day which a user uses.
Specifically, in the morning time division of the day, “Morning” and “Good morning” are set as start commands. In the daytime time division of the day, “Good afternoon” is set as a start command. In the night time division of the day, “Good evening” is set as a start command. Here, the time divisions of the day may include hours overlapping together. In the hours where the time divisions of the day overlap each other, as a matter of course, two start commands effective in a plurality of time divisions of the day to which the overlapping hours belong, are effective.
In the example of
In this embodiment, the call registration device for an elevator acts according to a series of actions of flow shown in
First, in normal times, the call registration device main body 3 is in a state of waiting for a start command (Step S1). And when a voice uttered by a user is inputted to the voice input section 3a (Step S2), in Step S3 the voice recognition section 3b ascertains whether or not the voice inputted to the voice input section 3a in Step S2 matches a predetermined start command which is changed and selected by the start command changing part 3e and is effective at present. In the case where the inputted voice did not match the start command, the flow of actions returns to Step S1.
On the other hand, in the case where the inputted voice matched the start command, the flow of actions proceeds to Step S4. In this Step S4, the voice output section 3d sends a notice to the effect that voice recognition of a call registration command has been started. As this concrete voice announcement, for example, “Please speak your destination floor” is conceivable. And in succeeding Step S5, the call registration device main body 3 starts waiting for a destination floor (a call registration command). That is, the voice recognition section 3b starts the recognition of a call registration command.
And when a voice is inputted to the voice input section 3a in this state of waiting for a destination floor (Step S6), the flow of actions proceeds to Step S7, where the voice recognition section 3b ascertains whether or not the voice inputted in Step S6 matches a predetermined call registration command. In the case where the inputted voice did not match the call registration command, the flow of actions proceeds to Step S8, where the voice output section 3d sends a notice about a failure in the recognition of the destination floor (call registration command) by the announcement “We couldn't recognize the voice.” And the flow of actions returns to Step S1.
On the other hand, in the case where in Step S7 the inputted voice matched the call registration command, the voice recognition section 3b succeeded in the recognition of the call registration command and the flow of action proceeds to Step S9. In this Step S9, a call registration request to the destination floor of the relevant recognized call registration command is sent from the call registration control section 3c to the elevator control device 4, whereby the elevator control device 4 is informed of the destination floor of the relevant recognized call registration command.
And when a call registration is performed in the elevator control device 4, the flow of actions proceeds to Step S10, where the voice recognition section 3b of the call registration device main body 3 finishes the voice recognition. And in succeeding Step S11, the voice output section 3d sends a notice about the destination floor of the registered call (for example, the announcement “We've registered X-floor” (X is a numeral representing a destination floor) is made), and the call registration device main body 3 returns to the state of waiting for a start command in Step S1.
The voice recognition section 3b may be formed from a microcomputer, a memory and the like. On this occasion, a memory may be built in the microcomputer. Furthermore, the configuration may be such that part of the functions of the call registration control section 3c, the voice input section 3a, and the voice output section 3d is provided by a microcomputer and the like. By using a microcomputer and the like in this manner, it is possible to constitute the call registration device by general parts and, therefore, it is possible to reduce the manufacturing cost.
As a predetermined condition for changing an effective start command by the start command changing part 3e, the use of the operating condition of the elevator is conceivable in addition to time divisions of the day.
Furthermore, in this embodiment, the start command changing part 3e is provided in the voice recognition section 3b of the call registration device main body 3. However, the place where this start command changing part 3e is provided is not limited to the voice recognition section 3b alone; for example, this start command changing part 3e may be provided within the elevator control device 4.
In the call registration device for an elevator configured as described above, it is ensured that in the case where a predetermined start command has been recognized, the voice recognition of a call registration command is started. Therefore, a user can easily grasp the timing at which a call registration command is to be uttered and it is possible to suppress erroneous registration of calls by voices other than the voice uttered by the user for call registration. And furthermore, because it is possible to automatically change a start command according to the operating condition and periods of use time for an elevator, it is possible to start voice recognition for call registration by a start command which is not unnatural and is not apt to be erroneously recognized, either.
The present invention can be used in a call registration device for an elevator which performs call registration by voice recognition using a voice input section which receives the input of a voice uttered by an elevator user.
1 car
2 hall
3 call registration device main body
3
a voice input section
3
b voice recognition section
3
c call registration control section
3
d voice output section
3
e start command changing part
4 elevator control device
5 communication path
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2011/057352 | 3/25/2011 | WO | 00 | 9/11/2013 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2012/131839 | 10/4/2012 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 4534056 | Feilchenfeld et al. | Aug 1985 | A |
| 4558298 | Kawai et al. | Dec 1985 | A |
| 4590604 | Feilchenfeld | May 1986 | A |
| 4958707 | Yoneda et al. | Sep 1990 | A |
| 5255341 | Nakajima | Oct 1993 | A |
| 5952626 | Zaharia | Sep 1999 | A |
| 6003637 | Kim et al. | Dec 1999 | A |
| 6223160 | Kostka et al. | Apr 2001 | B1 |
| 6397976 | Hale et al. | Jun 2002 | B1 |
| 6510924 | Bauer et al. | Jan 2003 | B2 |
| 6828918 | Bowman et al. | Dec 2004 | B2 |
| 7254559 | Florance et al. | Aug 2007 | B2 |
| 7299187 | Tahara et al. | Nov 2007 | B2 |
| 7373300 | Bangalore et al. | May 2008 | B1 |
| 7451085 | Rose | Nov 2008 | B2 |
| 7516190 | Kurganov | Apr 2009 | B2 |
| 7711565 | Gazdzinski | May 2010 | B1 |
| 8352246 | Lloyd | Jan 2013 | B1 |
| 8996381 | Mozer et al. | Mar 2015 | B2 |
| 20020036122 | Fayette | Mar 2002 | A1 |
| 20040060777 | Takeuchi | Apr 2004 | A1 |
| 20040230436 | Sugawara et al. | Nov 2004 | A1 |
| 20050080630 | Mills et al. | Apr 2005 | A1 |
| 20050114141 | Grody | May 2005 | A1 |
| 20050252723 | Takeuchi | Nov 2005 | A1 |
| 20060074658 | Chadha | Apr 2006 | A1 |
| 20060144645 | Yamakawa | Jul 2006 | A1 |
| 20060224383 | Lee et al. | Oct 2006 | A1 |
| 20060235684 | Chang | Oct 2006 | A1 |
| 20070185702 | Harney et al. | Aug 2007 | A1 |
| 20070225983 | Maringo et al. | Sep 2007 | A1 |
| 20080059186 | Mowatt et al. | Mar 2008 | A1 |
| 20080140415 | Shostak | Jun 2008 | A1 |
| 20080165937 | Moore | Jul 2008 | A1 |
| 20080167871 | Kim et al. | Jul 2008 | A1 |
| 20080177542 | Yamamoto | Jul 2008 | A1 |
| 20090204410 | Mozer et al. | Aug 2009 | A1 |
| 20090326937 | Chitsaz | Dec 2009 | A1 |
| 20100131275 | Paek | May 2010 | A1 |
| 20100174546 | Kim et al. | Jul 2010 | A1 |
| 20100332236 | Tan | Dec 2010 | A1 |
| 20110054900 | Phillips | Mar 2011 | A1 |
| 20110125503 | Dong et al. | May 2011 | A1 |
| 20110166855 | Vermeulen et al. | Jul 2011 | A1 |
| 20110168499 | Takeuchi | Jul 2011 | A1 |
| 20120075184 | Madhvanath | Mar 2012 | A1 |
| 20140006034 | Takeuchi | Jan 2014 | A1 |
| 20140278435 | Ganong, III | Sep 2014 | A1 |
| Number | Date | Country |
|---|---|---|
| 1 247378 | Oct 1989 | JP |
| 3 98967 | Apr 1991 | JP |
| 4 129976 | Apr 1992 | JP |
| 6 255911 | Sep 1994 | JP |
| 2000 80828 | Mar 2000 | JP |
| 2000 194393 | Jul 2000 | JP |
| 2007 186280 | Jul 2007 | JP |
| 2010 254437 | Nov 2010 | JP |
| Entry |
|---|
| Combined Office Action and Search Report issued Jan. 6, 2015 in Chinese Patent Application No. 201180069079.8 (with partial English language translation and English Translation of Category of Cited Documents). |
| Office Action issued Mar. 5, 2015 in Korean Patent Application No. 10-2013-7028088 (with partial English language translation). |
| International Preliminary Report on Patentability and Written Opinion issued Oct. 10, 2013, in International application No. PCT/JP2011/057352. |
| International Search Report Issued Dec. 13, 2011 in PCT/JP11/057352 Filed Mar. 25, 2011. |
| Office Action issued Dec. 16, 2014 in Korean Patent Application No. 10-2013-7028088 (with English language translation). |
| Number | Date | Country | |
|---|---|---|---|
| 20140006034 A1 | Jan 2014 | US |
