JP2003004521A - Identification device for sound outside vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an identification device for sound outside a vehicle automatically determining whether an external sound is necessary for a driver or not without troubling him and, if determined necessary, surely informing the driver of its content by voice and/or image. SOLUTION: In the identification device, outputs of both a video equipment 10 and an audiovisual apparatus 14 installed inside a vehicle are optionally input to an amplifier 13, image signals are supplied to an image monitor 12 via an interrupt circuitry 11, and audio signals are input to a speaker 16 via an interrupt circuitry 15. Signals obtained from a microphone 17 are amplified by an amplifier 18, input to a speech recognition part 21 after the external sound component of which is extracted by a differential circuitry 20, converted into speech pattern data, and then input to a coincidence determination part 22 to be compared with the data of a specific identification parameter memory 23. When the data is determined necessary speech data, speech and image messages associated with the data are input to both the interrupt circuitry 11 and the interrupt circuitry 15 via an image conversion part 24 and a speech conversion part 25.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exterior of a vehicle useful for an in-vehicle environment in which a main sound emitted from a sound emitting unit and a sound outside the vehicle corresponding to a sound source output signal of a vehicle-mounted audio device are mixed. The present invention relates to a sound identification device.

[0002]

2. Description of the Related Art Radio reception, cassette tape reproduction, CD
It is widely practiced to install a vehicle-mounted audio device for playing and so on in a vehicle and driving the vehicle while listening to music and the like.
When the outside sound is heard while the vehicle is in the vehicle, the user determines whether the outside sound is the information that the user of the emergency vehicle or the like needs to take some action or unnecessary information such as unnecessary noise. However, when the outside sound is the necessary information, the driver can perform a sufficient manual operation to reduce the output of the on-vehicle audio equipment, or manually mute the speaker sound to sufficiently cope with the outside sound. The sound of the siren of an emergency vehicle and a warning sound from the vehicle, for example, a warning sound such as "Please change the course to the left because the emergency vehicle passes" are heard.

However, since such a corresponding operation is performed entirely by hand, it is difficult to recognize the presence of outside sound when the sound level emitted from the vehicle-mounted audio equipment is high, and the operation is not possible. In order to solve the problem that there is a situation in which an operation is interrupted and a manual operation cannot be performed, there is a vehicle outside sound identification device that can perform the corresponding operation by automatic determination without performing the manual operation.

This is because, for example, as in the "acoustic recognition device" disclosed in Japanese Patent Laid-Open No. 6-194218, the sound outside the vehicle is sampled by a microphone and compared with a plurality of prestored sound patterns, and the sound outside the vehicle is detected. Is, that is, whether it is road noise, a siren of an emergency vehicle, a warning sound of a railroad crossing, etc. When the vehicle outside sound is left as it is, for example, when the siren sounds of an emergency vehicle, the vehicle outside sound is easily heard by lowering or muting the level emitted by the vehicle-mounted audio device.

Further, there is a vehicle outside sound identification device constructed as shown in FIG. That is, it has a video device 1 and an audio device 4 installed in the vehicle, and the video device 1 is a DVD.
The audio device 4 is composed of a compatible car navigation device, TV tuner, etc., the audio device 4 is composed of a radio tuner, a cassette tape player, a CD player, etc., and supplies a video signal output from the video device 1 to the image monitor 2.
The audio signal output from the video device 1 and the audio signal output from the audio device 4 are input to the correction unit 3.

The output of the correction unit 3 is amplified by the amplifier 5 and then supplied to the speaker 6 to emit a predetermined sound. In this case, the correction unit 3 analyzes the output of the microphone 7 provided for collecting the sound outside the vehicle by the characteristic / level analysis unit 8 for the frequency distribution and level of the sound, and the outside sound needs to be siren ringing or the like. In the case of voice, the level of sound emitted from the speaker 6 is lowered or muted by "automatic correction" so that the outside sound of the vehicle can be easily heard.

Further, a user interface 9 is connected to the correction unit 3 so that the necessary sound can be heard reliably by adjusting the sound emission level and the frequency characteristic by "manual correction" to ensure the sound outside the vehicle. I am able to deal with.

[0008]

In the conventional vehicle exterior sound identification device, in order to make the exterior sound easier to hear by "manual correction", the characteristic of the sound emission is determined under the driving condition that requires concentration. It has been found that the driver may not be able to concentrate on driving due to actions such as removing his / her hand from the steering wheel or taking a look away from various instruments in order to correct or mute the sound, and it may not be possible to drive comfortably. Even if
If you can not change the actual driving state and you can not release your hand, you can not make these corrections,
It becomes difficult to obtain the information accurately even if the outside sound is a required sound.

On the other hand, when the external sound is automatically detected and automatically corrected, the external sound is required for the user because the sound emission characteristic for canceling the external sound is given and the volume is corrected. It automatically corrects unconditionally without judging whether it is voice or unnecessary voice, so if the external sound is the necessary information that the driver wants to be taken, the above correction makes it difficult to hear in reverse. Therefore, there is a possibility that the user may miss the necessary information about the outside sound of the vehicle.

Therefore, an object of the present invention is to automatically determine whether the vehicle exterior sound is a vehicle exterior sound that is necessary or unnecessary and that the vehicle exterior sound is unnecessary and does not require the driver's hand. It is an object of the present invention to provide a vehicle exterior sound identification device capable of reliably notifying the driver of the content of the vehicle exterior sound with a sound and / or video when the vehicle exterior sound is generated.

[0011]

In order to solve the above problems, the vehicle exterior sound identification apparatus according to the present invention employs a characteristic configuration as described below.

(1) Sampling means for sampling a sound in an in-vehicle environment in which a main sound emitted from a sound emission unit and a sound outside the vehicle are mixed in correspondence with a sound source output signal of an on-vehicle audio equipment, and an emergency vehicle. From the audio device, a voice pattern storage unit that stores a plurality of voice pattern data such as a siren ringing sound and road noise, and a component corresponding to the main voice in the collected voice data collected by the collection unit. The vehicle exterior sound extraction means for obtaining the vehicle exterior sound component obtained by removing the vehicle exterior sound component based on the sound source output signal, and comparing the vehicle exterior sound component extracted by the vehicle exterior sound extraction means with the plurality of voice pattern data. , A judgment means for judging whether or not the sound is necessary sound, and when the judgment means judges that the sound is necessary sound, a message of sound and / or image corresponding to the data of the necessary sound is notified. Identifying device outside the vehicle sounds and a knowledge unit.

(2) Sampling means for sampling a sound in an in-vehicle environment in which a main sound emitted from a sound emitting unit and a sound outside the vehicle corresponding to a sound source output signal of an on-vehicle audio device are mixed, and an emergency vehicle. From the audio device, a voice pattern storage unit that stores a plurality of voice pattern data such as a siren ringing sound and road noise, and a component corresponding to the main voice in the collected voice data collected by the collection unit. The vehicle exterior sound extraction means for obtaining the vehicle exterior sound component obtained by removing the vehicle exterior sound component based on the sound source output signal, and comparing the vehicle exterior sound component extracted by the vehicle exterior sound extraction means with the plurality of voice pattern data. , A judgment means for judging whether or not the sound is necessary sound, and when the judgment means judges that the sound is necessary sound, a message of sound and / or image corresponding to the data of the necessary sound is notified. An exterior of the vehicle, which comprises a learning means and a characteristic correction means for converting a sound emission characteristic from the sound emission portion corresponding to the sound source output signal into an acoustic characteristic for making the exterior sound component extracted by the exterior sound extraction means easy to hear. Sound identification device.

(3) The vehicle outside sound identification device, wherein the voice pattern storage means of (1) or (2) is configured to have a learning function of the voice data collected by the collection means.

(4) The sampling means of the above (1) or (2) obtains a plurality of sampled data by voice detection members provided at a plurality of locations, and correlates data in which the correlation of the plurality of sampled data is added. A vehicle outside sound identification device configured to obtain.

(5) A vehicle outside sound identification device in which the sampling means of the above (4) is provided with a voice detection member inside and outside the vehicle.

(6) In the configuration of any one of the above (1) to (5), the detected output data of the image outside the vehicle is compared with a plurality of image data stored in advance, and the situation near the outside of the vehicle is voiced. And / or an exterior sound identification device provided with an exterior environment informing means for informing a message of an image.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to FIGS. First, a first embodiment of the present invention will be described with reference to FIG.
Video device 10 and audio device 14 installed in the car
The video device 10 includes a DVD-compatible car navigation device, a TV tuner, and the like, and the audio device 14 includes a radio tuner, a cassette tape player, and the like.
The video signal SV which is composed of a CD player or the like and is output from the video device 10 is supplied to the image monitor 12 via the interrupt circuit 11, and is output from the audio signal output from the video device 10 and the audio device 14. The generated audio signal is selectively input to the amplifier 13, and its output is input to the speaker 16 via the interrupt circuit 15 as the audio signal SA.

Audio signal output from the amplifier 13
SA is input to the circuit section a, and the interrupt circuit 11 and the interrupt circuit 15
The video interrupt signal V and the audio interrupt signal A from the circuit section b are input to each of the above. This circuit part a is
It has a microphone 17 provided in or on the vehicle, the output of which is amplified by an amplifier 18 and input to one input end of a difference circuit 20, and the delay circuit 19 to the output of the amplifier 13 at the other input end. The signal intervening is input. The delay circuit 19 is for matching the audio signal SA from the video device 10 or the audio device 14 with the signal from the microphone 17 on the time axis.

The output of the difference circuit 20 is input to the voice recognition unit 21, converted into voice pattern data of a form conforming to a predetermined method, and input to the coincidence determination unit 22 of the circuit unit b as the output of the circuit unit a. Specific identification parameter storage device 23
The degree of agreement with the voice pattern data called from

The notification signal for notifying the voice and / or image message output from the coincidence determination unit 22 is input to the image conversion unit 24 and the audio conversion unit 25, and the image conversion unit 24 outputs the notification signal. It is converted into a video interrupt signal V and supplied to the interrupt circuit 11, and converted into an audio interrupt signal A by the audio converter 25 and supplied to the interrupt circuit 15.

Further, the audio signal from the video device 10 or the audio device 14 is mixed with the main sound emitted from the speaker 16 via the amplifier 13 and the interrupt circuit 15 and the sound outside the vehicle to produce a sound in the vehicle interior environment. The sampling means for sampling is constituted by the microphone 17 or the like, and the speech pattern storage means for storing plural speech pattern data such as siren ringing noise or road noise of the emergency vehicle is constituted by the specific identification parameter storage device 23, and the microphone. A vehicle exterior sound extraction unit that obtains a vehicle exterior sound component obtained by removing the component corresponding to the main voice in the sampled voice data sampled at 17 based on the audio signal SA (sound source output signal) by the difference circuit 20. , Voice recognition unit 21.

The determination means for comparing the exterior sound component extracted by the exterior sound extraction means with a plurality of voice pattern data in the specific identification parameter storage device 23 to determine whether or not the voice is a necessary voice is a coincidence degree determination part. The image conversion unit 24 and the audio conversion unit 25 constitute an informing unit configured to include an audio message and an image message corresponding to the data when the coincidence determination unit 22 determines that the audio is a necessary audio. To be done.

Therefore, the sound of the in-vehicle environment in which the main sound from the speaker 16 and the sound outside the vehicle are mixed is detected by the microphone 17, and the component of the main sound, that is, the sound emission component (audio signal SA) from the speaker 16 is difference circuit. 20 is separated from the external sound and extracted, and the amplitude, wavelength, and
Based on the information such as the spectrum and the frame signal, the existing voice recognition means is used to analyze the voice component from the outside of the vehicle, and the coincidence degree determination unit 22 obtains the coincidence degree with the data of the specific identification parameter storage device 23, and the coincidence is obtained. If the sound is outside the vehicle, it is determined that the sound outside the vehicle is the required sound desired by the user.

If the matching degree judging unit 22 judges that they do not match, the outside sound is recognized as unnecessary sound unnecessary for the user. In this case, the video interrupt signal V and the sound interrupt are recognized. Since the interruption signal A is not generated, no further notification is given, and when the coincidence determination unit 22 determines that the voice is necessary voice, the video interruption signal V and the voice interruption signal A are generated based on the information. As a result, the user is notified by both voice and image.

When the voice extracted from the outside sound is the siren of the emergency vehicle, the voice "the emergency vehicle is approaching" and the video signal from the video device 10
The car navigation display displayed on the image monitor 12 by the SV is interrupted by a character image "Emergency vehicle is approaching!", Or by interrupting with the original image erased to notify by both voice and image. Based on this notification, the user can take a reliable driving response to the outside sound of the siren or the like, for example, give up the route to the emergency vehicle, decelerate or temporarily stop.

In the first mode as described above, when the content of the vehicle exterior sound is necessary for the user, the image and the sound by the video device 10 and the audio device 14 are interrupted to notify that effect. In this example, the output of the original audio device 14 remains unchanged during the voice notification, but the output of the audio signal SA is output during the voice notification in response to the necessity / unnecessity of the external sound. An example of correcting the above will be described as a second embodiment with reference to FIG.

The second embodiment is a schematic circuit diagram of a vehicle outside sound identification apparatus according to the first embodiment described above with reference to FIG.
A part of the configuration shown in FIG. 2 is modified. The modification is that the circuit part b shown in FIG. 1 is removed and the circuit part c shown in FIG. 2 is provided instead. Further, a characteristic correction unit 26 is newly inserted between the output end of the amplifier 13 and the interrupt circuit 15.

The characteristic correction unit 26 substantially cancels the sound emission characteristic from the speaker 16 corresponding to the sound source output signal from the video device 10 or the audio device 14 so as to substantially cancel the outside vehicle sound component extracted by the circuit unit a. It constitutes a main part of a characteristic correction means for converting into acoustic characteristics.

The output end of the voice recognition unit 21 is the coincidence determination unit 2
7 is connected to the input end, and the coincidence determination unit 27 compares the stored data of the specific identification parameter storage device 28 and the output data of the voice recognition unit 21 to determine the coincidence.

A judging means for judging whether or not the sound is a necessary sound by comparing the outside sound component extracted by the outside sound extracting means (circuit part a) with a plurality of sound pattern data in the specific identification parameter storage device 28. Is configured by the coincidence determination unit 27,
The image conversion unit 29 and the audio conversion unit 30 constitute an informing unit that notifies an audio and image message corresponding to the data when the coincidence determination unit 27 determines that the audio is a required audio.

When the matching degree judging section 27 judges that the sound is a necessary sound, the output thereof is inputted to the characteristic / level analyzing section 31, and finally the characteristic of the sound emitted from the speaker 16 is outside the vehicle. When proper analysis is performed to obtain a characteristic that allows the sound to be heard well, and when it is determined that the sound is unnecessary, its output is input to the characteristic / level analysis unit 31, and finally the speaker 16 outputs the sound. An appropriate analysis is performed so that the characteristic of the voice that is generated is a characteristic that cancels the external sound, and the characteristic correction unit 26 is controlled by the output.

Therefore, the sound of the in-vehicle environment in which the main sound from the speaker 16 and the sound outside the vehicle are mixed is detected by the microphone 17, and the component of the main sound, that is, the sound emission component (audio signal SA) from the speaker 16 is difference circuit. 20 is separated from the external sound and extracted, and the amplitude, wavelength, and
Based on the information such as the spectrum and the frame signal, the voice recognition unit 21 which is an existing voice recognition means is used to analyze the voice component from the outside, and the coincidence degree determination unit 27 causes the coincidence degree with the data of the specific identification parameter storage device 28 to be determined. If they match, it is determined that the external sound is the necessary voice desired by the user.

If the matching degree judging section 27 judges that they do not match, the outside sound is recognized as unnecessary sound unnecessary for the user. In this case, the video interrupt signal V and the sound interrupt are recognized. Since the built-in signal A is not generated, no special notification is given, and the sound emission characteristic is set to cancel the external sound. If the coincidence determination unit 27 determines that the voice is a required voice, the video interrupt signal V and the voice interrupt signal A are generated based on the information and the interrupt circuit 11 and the interrupt circuit 15 are generated. By the interrupt input, the user is notified by both voice and image.

Simultaneously with such notification, the characteristic correction unit 26 controlled by the characteristic / level analysis unit 31 makes the sound finally emitted from the speaker 16 a characteristic that allows the outside sound to be heard well.

Therefore, when the sound extracted from the sound outside the vehicle is the siren of the emergency vehicle, the sound "the emergency vehicle is approaching" and the video signal SV from the video device 10 cause the image monitor 12 to display. It is possible to notify by both voice and image by interrupting the text image "Emergency vehicle is approaching!" To the car navigation display displayed in, or by interrupting with the original image erased, Based on this notification, the user can take a reliable driving response to the outside sound of the siren or the like, for example, give the course to an emergency vehicle, decelerate or temporarily stop, and at the same time, emit sound from the speaker 16. Since the sound has a characteristic that the outside sound can be heard well, the outside sound can be clearly heard.

The specific identification parameter storage device 23 or the specific identification parameter storage device 28 used in the first or second embodiment as described above is provided with a plurality of voices such as siren ringing noise and road noise of an emergency vehicle. Although the voice pattern storage means for storing the pattern data is configured, the voice pattern storage means may be configured to have a learning function of the voice data sampled by the sampling means. The sampling means constituted by the circuit section a shown in FIG. 1 or FIG. 2 is composed of a microphone 17 or the like provided at one place in the vehicle, but is a voice detection member (microphone) provided at a plurality of places. It may be configured to obtain a plurality of collected data and obtain correlation data in which the correlation of the plurality of collected data is added.

This embodiment will be described below as a third embodiment with reference to FIG. In FIG. 3, each of the first circuit portion a1 and the second circuit portion a2 has the same microphone 17 and amplifier 1 as in the circuit portion a (see FIG. 1 or 2).
8, a delay circuit 19 and a difference circuit 20. The microphones 17 of the first circuit portion a1 and the second circuit portion a2 are installed at the front portion and the rear portion of the vehicle, respectively. Each of the outputs is input to the voice recognition unit 32, and the voice component from the outside of the vehicle is “incorporated into the direction of the outside voice” based on the information such as the amplitude, wavelength, spectrum, and frame signal of the voice component from the outside of the vehicle. The degree of coincidence with the data of the specific identification parameter learning storage device 34 is analyzed by the degree of coincidence / direction determination unit 33 in the next stage, and if they coincide, it is determined that the outside sound is a necessary voice desired by the user. To do.

The specific identification parameter learning storage device 34 is
A plurality of sound data to be judged are stored with a learning function, and when an attempt is made to cross a railroad track, for example, the outside sound of the vehicle, that is, the warning sound of the railroad crossing, will be listened carefully. Here, the level crossing sound is stored in advance and learned by the component of the vehicle exterior sound extracted by the first circuit unit a1 or the second circuit unit a2, so that the degree of matching can be determined more reliably and in accordance with the actual sound. Can be made. It is possible to learn not only the siren sound and the leveling sound but also other sounds, for example, the ringing tone of the telephone, and it is possible to recognize various telephone ringing tones and melodies.

Therefore, the sound of the in-vehicle environment in which the main sound from the speaker 16 and the sound outside the vehicle are mixed is the two microphones 17.
Detected by the main voice component, that is, the speaker 16
The sound emission component (audio signal SA) from the vehicle is separated and extracted from the vehicle exterior sound by the difference circuit 20, and the existing voice recognition means is based on the information of the amplitude, wavelength, spectrum, frame signal, etc. of the signal components of these two systems. The voice component from the outside is analyzed by using the voice recognition unit 32, and the matching degree with the data having the learning function extracted from the specific identification parameter learning storage device 34 by the matching degree / direction determination unit 33 is determined in the direction of the sound. The sound outside the vehicle is determined to be the necessary voice desired by the user when both are in agreement.

If the coincidence / direction determining unit 33 determines that they do not match, the outside sound is recognized as unnecessary sound unnecessary for the user. In this case, the image converting unit 35 and the sound are recognized. Since the video interrupt signal V and the audio interrupt signal A are not generated via the conversion unit 36, no special notification is given. If the coincidence / direction determination unit 33 determines that the voice is required voice, the video interrupt signal V and the voice interrupt signal A are generated based on the information, and the interrupt circuit 11 and the interrupt circuit are generated. The interrupt input to 15 notifies the user by both voice and image.

Therefore, when the sound extracted from the vehicle exterior sound is the siren of the emergency vehicle, the sound "the emergency vehicle is approaching from behind" and the video device 10
A video signal SV from the car navigation display displayed on the image monitor 12 is interrupted by the text image "Emergency vehicle approaching from behind!" It is possible to perform notification by both of the images, and based on this notification, the user can perform reliable driving response to external noise such as siren, for example, transfer the route to an emergency vehicle or respond to deceleration or temporary stop, In addition, it is possible to notify by the direction of the sound outside the vehicle, in this case, the rear, but from the front or the side by both the voice and the image.

The vehicle exterior sound identification apparatus according to the first to third embodiments as described above is an example of detecting the vehicle exterior sound and notifying the vehicle with appropriate voice and image when the required voice is detected. To such a device, an outside-vehicle environment notifying means for comparing the detected output data of the outside-vehicle image with a plurality of pre-stored image data and notifying the vicinity situation outside the vehicle of a voice and / or image message is added. It may be provided, which will be described as a fourth embodiment with reference to FIG.

The microphone 44 and the voice recognition processing section 45 shown in FIG. 4 have the same configuration and operation as the "vehicle outside sound identification device" in any of the first to third embodiments. Therefore, the description thereof will be omitted and only different parts will be described. The portion added in this embodiment is the circuit portion e, which is an infrared camera 37 for detecting an image outside the vehicle.
Is amplified by the amplifier 38 and input to the coincidence determination unit 40 via the image recognition unit 39. This coincidence determination unit 4
0 is for comparing the detected image output data with a plurality of image data stored in advance in the specific identification parameter storage device 41 to determine the degree of coincidence of the proximity situation outside the vehicle.

For example, it is possible to judge the situation where another vehicle is approaching in the vicinity of the vehicle, there is an obstacle such as a fence in the front, and when the coincidence judging section 40 judges that the notification is necessary, The notification data is output to the image conversion unit 42 and the audio conversion unit 43, and the video interrupt signal V1 and the audio interrupt signal A1 generated by the image conversion unit 42 and the audio conversion unit 43 are generated to generate the interrupt circuit 11. Is input to the interrupt circuit 15, and the situation near the vehicle is notified by voice and image.

Therefore, in the same manner as in the above-mentioned embodiment, the notification by the voice and the image is given, and when the image obtained by the infrared camera 37 is the approaching vehicle, it is said that the vehicle is approaching. The sound is emitted from the speaker 16 via the interrupt circuit 15, and at the same time, a character image "Vehicle approaching!" Is displayed on the image monitor 12 via the interrupt circuit 11.

In the four embodiments described above, the "outside vehicle sound" is a sound emitted from the outside of the vehicle, such as a siren or a railroad crossing warning sound. The sound is a generic term for sounds other than the “speaker 16 that is emitted by the output from the video device 10 or the audio device 14,” and means, for example, the sound of an in-car telephone, the crying sound of an infant on board the vehicle, and the like. To do. Therefore, it is possible to voice-recognize that the infant sleeping in the back seat cries and to inform the driver of the content.

Further, it goes without saying that the notification made based on the result of the image recognition may be both voice and image, only voice, or only image. Totally optional.

[0049]

As is clear from the above description, the vehicle exterior sound identification apparatus according to the present invention determines whether the vehicle exterior sound is a vehicle exterior sound that is necessary or unnecessary for the driver. The automatic judgment is performed without bothering the user, and when the required external sound such as the siren of an emergency vehicle or a warning sound, the content is surely notified to the driver by voice and / or video. It is possible to provide a device for identifying a vehicle exterior sound that can be heard, prevent an important vehicle exterior sound from being missed, and safely handle the information.

[Brief description of drawings]

FIG. 1 is a block circuit diagram showing a schematic circuit configuration of a vehicle outside sound identification apparatus according to a first embodiment of the present invention.

FIG. 2 is a block circuit diagram showing a schematic circuit configuration of a vehicle exterior sound identification apparatus according to a second embodiment of the present invention.

FIG. 3 is a block circuit diagram showing a schematic circuit configuration of a vehicle exterior sound identification apparatus according to a third embodiment of the present invention.

FIG. 4 is a block circuit diagram showing a schematic circuit configuration of a vehicle exterior sound identification apparatus according to a fourth embodiment of the present invention.

FIG. 5 is a block circuit diagram showing a schematic circuit configuration of an example of a conventional vehicle exterior sound identification device.

[Explanation of symbols]

10 Video equipment 11,15 Interrupt circuit 12 image monitor 13,18 amp 14 audio equipment 16 speakers 17 microphone 19 Delay circuit 20 Difference circuit 21 Speech recognition unit 22 Consistency judgment unit 23 Specific identification parameter storage device 24 Image converter 25 voice converter

Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) G10L 21/02

Claims (6)

[Claims] 1. A sampling means for sampling a sound in an in-vehicle environment in which a main sound emitted from a sound emitting portion and a sound outside the vehicle corresponding to a sound source output signal of an on-vehicle audio device are mixed, and an emergency vehicle. A voice pattern storage unit that stores a plurality of voice pattern data such as a siren ringing sound or road noise, and a component corresponding to the main voice in the sampled voice data sampled by the sampling unit from the acoustic device. Exterior sound extraction means for obtaining a vehicle exterior sound component obtained by removing based on the sound source output signal, comparing the vehicle exterior sound component extracted by the vehicle exterior sound extraction means and the plurality of voice pattern data, Determining means for determining whether or not the voice is necessary voice; and a notification for notifying a voice and / or image message corresponding to the data of the voice when the voice is required voice by the determining means. A vehicle exterior sound identification device comprising: 2. A sampling means for sampling a sound in an in-vehicle environment in which a main sound emitted from a sound emitting section and a sound outside the vehicle are mixed in response to a sound source output signal of an on-vehicle audio device, and an emergency vehicle. A voice pattern storing means for storing a plurality of voice pattern data such as a siren ringing sound or road noise, and a component corresponding to the main voice in the collected voice data collected by the collecting means from the acoustic device. The vehicle exterior sound extraction means for obtaining the vehicle exterior sound component obtained by removing the vehicle exterior sound component based on the sound source output signal, and comparing the vehicle exterior sound component extracted by the vehicle exterior sound extraction means with the plurality of voice pattern data. A determination unit for determining whether or not the voice is a required voice, and when the determination unit determines that the voice is the required voice, a voice and / or image message corresponding to the data of the required voice is notified. An informing unit, and a sound emission characteristic from the sound emission unit corresponding to the sound source output signal,
An exterior sound identification device, comprising: a characteristic correction means for converting the exterior sound component extracted by the exterior sound extraction means into an acoustic characteristic that makes it easier to hear. 3. The vehicle exterior sound identification device according to claim 1, wherein the voice pattern storage means is configured to have a learning function of the voice data collected by the collection means. . 4. The collecting means is configured to obtain a plurality of collected data by voice detection members provided at a plurality of locations and to obtain correlation data in which a correlation of the plurality of collected data is taken into consideration. The vehicle exterior sound identification device according to claim 1 or 2. 5. The vehicle exterior sound identification device according to claim 4, wherein the sampling means is provided with voice detection members inside and outside the vehicle. 6. A vehicle outside environment informing means is additionally provided for comparing detected output data of an image outside the vehicle with a plurality of image data stored in advance to inform the vicinity of the vehicle outside by a voice and / or image message. The vehicle outside sound identification device according to any one of claims 1 to 5.