JP4300596B2 - Car navigation system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an in- vehicle navigation system .
[0002]
[Prior art]
Conventionally, various navigation systems to be mounted on automobiles and the like have been developed. The navigation system is based on, for example, a large-capacity data storage means such as a CD-ROM in which road map data is stored, a current position detection means, and a road map in the vicinity of the detected current position read from the data storage means. Display means for displaying the information and displaying the route guidance to the destination, the current location, etc. on the display means.
[0003]
A user needs to input a destination name in advance to such a navigation system, but currently, a device has been developed in which a destination telephone number is automatically set by inputting a destination telephone number. ing.
[0004]
In addition, the navigation system needs to be able to operate without interfering with the driving of the car, so that not only the telephone number can be manually input but also the one that can recognize and input the spoken telephone number. Proposed.
[0005]
[Problems to be solved by the invention]
A navigation system that can input a spoken phone number is generally provided with a voice recognition device that has no restriction on the number of digits that can be recognized and that can recognize continuously spoken digits. Yes.
[0006]
When the speech recognition apparatus performs recognition of a spoken number, it may be erroneously recognized due to external noise before and after the speech section, and may recognize a number greater than the number of digits spoken. At this time, it is necessary to correct by some means, and for the user, the operation for performing this correction is very troublesome and may hinder driving of the automobile.
[0007]
The voice recognition device as described above is also provided in a so-called hands-free kit for enabling a phone while driving a car without having a mobile phone. A similar problem has arisen.
[0008]
The present invention has been proposed in view of such a situation, and an in- vehicle navigation system capable of improving operability by reducing erroneous recognition of consecutive numbers spoken and omitting correction operations. The purpose is to provide.
[0009]
[Means for Solving the Problems]
In order to solve the above-described problem, the present invention provides a vehicle navigation system that shows a route to a destination, map information, display means for displaying the current position and the destination on the map information, audio, When the signal input microphone and the voice signal input from the microphone are continuous numbers, continuous number recognition means for outputting a plurality of consecutive number candidates, and the number of digits specifying means for specifying the number of digits of the number And when there is no designation of the number of consecutive digits in the digit number designation means, the candidate with the most likely recognition result is selected from the candidates for the consecutive digit recognition means, and the digit number designation means When the number of consecutive digits is designated, among the plurality of consecutive number candidates output from the consecutive number recognition means, the number of digits designated by the number of digits designation means, And recognition probability is high Selection means for selecting an object, if the continuous number is a telephone number, and a control means for setting the destination based on the telephone number, the selecting means, the audio signal input from the microphone When the corresponding digit string is an area code constituting a telephone number, the number of digits designation means does not designate the number of digits of the consecutive digits, so that the recognition result that is the most recognized among the candidates for the consecutive digits recognition means In addition to selecting the correct possibility, the number-of-digits specifying means predicts and sets the number of consecutive digits input following the area code .
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention can be applied to, for example, the navigation system 1 configured as shown in FIG.
[0011]
The navigation system 1 is mounted on a car and displays the current position of the car to indicate a route to a destination. The navigation system 1 is configured by a voice recognition device 10 that can perform a predetermined operation by a driver's voice and the like, and that recognizes the voice, and a navigation device 20 that displays a current location based on the voice recognition result. .
[0012]
The voice recognition device 10 includes a hands-free microphone 11 provided in a driver's seat of an automobile, an analog / digital (A / D) converter 12 for digitizing a voice signal, an acoustic analysis circuit 13 for analyzing the voice signal, and a voice. And a voice recognition circuit 14 for recognizing the content of the signal.
[0013]
The hands-free microphone 11 has a single directivity, converts the driver's voice into an audio signal, and supplies it to the A / D converter 12. The hands-free microphone 11 is provided with a switch 11a. The user needs to turn on the switch 11a while speaking a number or the like. As a result, while the switch 11a is on, the voice recognition circuit 14 recognizes consecutive numbers uttered by the hands-free microphone 11. The A / D converter 12 digitizes the audio signal from the hands-free microphone 11 and supplies the audio signal to the acoustic analysis circuit 13.
[0014]
As shown in FIG. 2, the acoustic analysis circuit 13 includes a frequency analysis unit 31, an adaptation unit 32, a noise subtraction unit 33, and a feature extraction unit 34.
[0015]
The frequency analysis unit 31 performs frequency analysis of the audio signal using a general FFT as a frequency analysis method, and obtains a power spectrum of the audio signal.
[0016]
The adaptation unit 32 stores microphone adaptation data unique to the hands-free microphone 11 in a memory, and executes a microphone adaptation process for normalizing the characteristics unique to the microphone. The adaptation data is obtained so that the frequency characteristics of the hands-free microphone 11 are the characteristics of a standard microphone that is flat. That is, the microphone adaptation data differs depending on the type of microphone, and it is necessary to prepare each when using a plurality of microphones.
[0017]
The noise subtraction unit 33 obtains a spectrum of the noise section using a signal section that is not a voice section as a noise section, and performs a process of subtracting the noise spectrum in the voice section.
[0018]
Since the feature extraction unit 34 has the power spectrum as a result of the frequency analysis as it is, there are many parameter dimensions, and the amount of calculation of the speech recognition circuit 14 increases. The feature extraction unit 34 uses, for example, a cepstrum, and particularly performs a compression process using a mel cepstrum weighted with auditory characteristics on the frequency axis in speech recognition, and outputs coded cepstrum coefficients and mel cepstrum coefficients. Then, it is supplied to the voice recognition circuit 14.
[0019]
The voice recognition circuit 14 recognizes the voice input to the handsfree microphone 11 from the cepstrum coefficient and the mel cepstrum coefficient based on the HMM (Hidden Markov Model) method. That is, the voice recognition circuit 14 recognizes the voice input to the handsfree microphone 11 using a model (HMM) stored in advance in the memory 14 a and supplies the recognition result to the system controller 28.
[0020]
Specifically, as shown in FIG. 3, the speech recognition circuit 14 includes a continuous number recognition processing circuit 41 that recognizes continuous numbers based on a cepstrum coefficient or the like from the acoustic analysis circuit 13, and a plurality of recognized plural numbers. A recognition candidate selection circuit 42 for selecting a predetermined one from candidates and a digit number designating circuit 43 for designating the number of digits of numbers are provided.
[0021]
For example, as shown in FIG. 4, the continuous numeral recognition processing circuit 41 outputs a plurality of continuous numeral candidates in order from the most probable recognition result. In FIG. 4, the recognition result of three consecutive digits that is candidate number 1 is most likely to be correct, and the possibility of being correct becomes lower as candidate numbers 2, 3,. The continuous number recognition processing circuit 41 supplies such a plurality of continuous number candidates to the recognition candidate selection circuit 42.
[0022]
The recognition candidate selection circuit 42 can most correctly recognize the recognition result among a plurality of continuous number candidates supplied from the continuous number recognition processing circuit 41 when the number of consecutive digits is not specified by the digit number specifying circuit 43. The serial number is selected and this serial number is supplied to the navigation device 20. Further, when the number of consecutive digits is designated by the digit number designating circuit 43, the continuous number recognition processing circuit 41 uses the number of digits designated by the digit number designating circuit 43 from among a plurality of candidates for consecutive digits. Then, the most probable recognition result is selected and this continuous number is supplied to the navigation device 20.
[0023]
The number-of-digits specifying circuit 43 specifies the number of consecutive digits to be supplied next to the recognition candidate selecting circuit 42 based on the continuous numbers selected and output by the recognition candidate selecting circuit 42.
[0024]
Specifically, the case where the telephone number of Tokyo 23 wards is recognized will be described as an example. The telephone number of Tokyo 23 wards consists of two digits with the area code “03” followed by four digits plus four digits.
[0025]
When the user utters “03” to the hands-free microphone 11, the continuous numeral recognition processing circuit 41 supplies a plurality of continuous numeral candidates to the recognition candidate selection circuit 42 based on the supplied cepstrum coefficient or the like. At this time, since the number-of-digits specifying circuit 43 does not initially specify the number of consecutive digits, the recognition candidate selecting circuit 42 has the most likely recognition result among a plurality of consecutive number candidates, for example, “ Select “03” and output. Based on “03” output from the recognition candidate selection circuit 42, the number-of-digits designating circuit 43 determines that the next consecutive number input is four digits, and designates the number of consecutive digits as four digits. For example, when the first input is “06”, the digit number designating circuit 43 designates the next three digits, and when the first input is “0492”, the second digit designation is performed. That is, the next number of digits is designated according to the first area code.
[0026]
Next, when the user utters “1234” to the hands-free microphone 11, the recognition candidate selection circuit 42 similarly supplies a plurality of consecutive number candidates to the recognition candidate selection circuit 42. The recognition candidate selection circuit 42 selects and outputs “3456” having the number of digits designated by the digit number designation circuit 43 and the most likely recognition result among a plurality of candidates for consecutive numbers. Here, even if a certain consecutive number is candidate number 1, the recognition candidate selection circuit 42 determines that a number other than four digits is incorrect. For example, as shown in FIG. 4 described above, even if the consecutive number of the candidate number 1 is “123”, it is determined as incorrect and “3456” of the candidate number 4 that is four digits and most likely correct. Select to output.
[0027]
Then, the number-of-digits specifying circuit 43 determines that the next consecutive number to be input is four digits based on “3456” output from the recognition candidate selection circuit 42, and designates the number of consecutive digits as four digits. To do. In the same manner, the last four-digit consecutive numbers spoken are also recognized.
[0028]
As described above, the speech recognition circuit 14 can improve the speech recognition rate by adding a condition for selecting a speech recognition result by performing speech recognition processing in consideration of information related to the number of consecutive digits. it can. That is, it is possible to find the optimum continuous number from the conditions of both the recognized continuous number and the number of digits of the number.
[0029]
The navigation device 20 includes a GPS antenna 21, a GPS receiving circuit 22, a CD-ROM drive 23 that reads map data from a CD-ROM (Compact Disc-Read Only Memory), an autonomous sensor 24, and voice. A response unit 25 for outputting, a display unit 26 for displaying a current position and the like together with a map, a mobile phone connection unit 27 for connecting to a mobile phone (not shown), and a system controller 28 for controlling the whole are provided.
[0030]
The GPS receiving circuit 22 receives a measurement signal from a GPS satellite via the GPS antenna 21 and analyzes this signal to detect a current position. The data indicating the current position is latitude and longitude data at the location. The GPS receiving circuit 22 supplies the latitude and longitude data to the system controller 28.
[0031]
When the position data is supplied from the GPS receiving circuit 22, the system controller 28 controls the CD-ROM drive 23 to read road map data in the vicinity of the position indicated by the position data. The system controller 28 displays the road on the display unit 26 based on the road map data read from the CD-ROM drive 23, and outputs the voice of the current location such as “currently xx” to the response unit 25. Let
[0032]
Further, the system controller 28 sets the destination based on the spoken telephone number based on the recognition result of the voice recognition circuit 14, inputs the spoken place name, and switches the display mode of the display unit 26. It can be performed. For example, when the driver turns on the switch 11a to generate a place name in the hands-free microphone 11, the system controller 28 recognizes the place name via the hands-free microphone 11, the acoustic analysis circuit 13, the voice recognition circuit 14, and the like. Then, the place name data is read from the CD-ROM drive 23 and the position of the place name is marked on the display unit 26.
[0033]
On the other hand, the autonomous sensor 24 is supplied with a pulse signal corresponding to a vehicle speed supplied from an engine control computer (not shown) of the automobile, calculates an accurate traveling speed of the automobile based on the pulse signal, The traveling direction is detected by the output of a gyro sensor (not shown). Furthermore, the current position is measured by autonomous navigation from a position determined by the speed and direction of travel.
[0034]
The mobile phone connection unit 27 connects the mobile phone and the navigation system 1. Thereby, the user can make a phone call while talking with the hands-free microphone 11, for example.
[0035]
As described above, when the navigation system 1 recognizes a telephone number spoken for setting a destination, for example, the voice recognition process is performed by taking into account information regarding the number of consecutive digits. The recognition result selection condition is added to improve the speech recognition rate, so that the troublesome operation for correction at the time of erroneous recognition is eliminated and the operability can be further improved.
[0036]
Needless to say, the present invention is not limited to the above-described embodiments, and can be applied to various devices that need to recognize spoken numbers.
[0037]
【The invention's effect】
As described above in detail, according to the in-vehicle navigation system according to the present invention, the selection means is configured such that the numeric string corresponding to the audio signal input from the microphone is the area code constituting the telephone number. Since there is no designation of the number of consecutive digits in the digit number designating means, the candidate having the most likely recognition result is selected from the candidates for the consecutive digit recognition means, and the digit number designating means Since the number of consecutive digits input following the external station number is predicted and set, the optimum one can be selected from a plurality of consecutive digits, resulting in erroneous recognition of consecutive digits spoken. As a result, the operability can be improved by omitting the correction operation.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a specific configuration of a navigation system to which the present invention is applied.
FIG. 2 is a block diagram showing a specific configuration of an acoustic analysis circuit provided in a voice recognition device that constitutes the navigation system.
FIG. 3 is a block diagram showing a configuration of a speech recognition circuit provided in the speech recognition apparatus.
FIG. 4 is a diagram illustrating an example of a processing result of a continuous numeral recognition processing circuit included in the speech recognition apparatus.
[Explanation of symbols]
1 navigation system, 41 continuous digit recognition processing circuit, 42 recognition candidate selection circuit, 43 digit number designating circuit

Claims (4)

In the car navigation system that shows the route to the destination,
Map information and display means for displaying the current position and the destination on the map information;
A microphone for inputting audio signals;
When the voice signal input from the microphone is a continuous number, continuous number recognition means for outputting a plurality of continuous number candidates,
A digit number specifying means for specifying the number of digits of the number;
If the number of consecutive digits is not specified by the number of digits specifying means, the candidate of the recognition result that is most likely to be correct is selected from the candidates for the consecutive digits recognition means, and the number of digits specifying means continues When there is designation of the number of digits , among a plurality of candidates for consecutive digits output from the consecutive digit recognition unit, the number of digits designated by the digit number designation unit, and A selection means for selecting one having a high recognition probability;
Control means for setting the destination based on the telephone number when the continuous number is a telephone number;
With
In the case where the number string corresponding to the audio signal input from the microphone is an area code constituting a telephone number, the selection means does not specify the number of consecutive digits in the digit number designation means. While selecting the most likely recognition result from among the candidates for the continuous number recognition means, the number-of-digits specifying means predicts the number of consecutive digits that are input following the area code. In- vehicle navigation system characterized by setting . The number-of-digits designating unit designates the number of consecutive digits output next from the sequence number recognition unit based on the number of digits of the sequence number selected by the selection unit. The in-vehicle navigation system described. Switch means indicating that an audio signal is input;
The in-vehicle navigation system according to claim 1, wherein the microphone inputs an audio signal while the switch means is being pressed. The in-vehicle navigation system according to claim 1, wherein when a voice signal input from the microphone is place name information, a position corresponding to the place name information on the map information is marked.

Images