JP2008191742A - Auditory information notifying device and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable a vehicle occupant to the most quickly and surely recognize the presence of other vehicle around its own vehicle and its behavior. <P>SOLUTION: The sounds of other vehicle 1 around a vehicle collected by a plurality of microphones installed in the outer periphery of a body are corrected so as to be same as the sound of an object around the vehicle heard by a vehicle occupant with his or her both ears, and a warning sound 1 is generated. The position of the other vehicle 1 around the vehicle is detected, and a virtual sound source is arranged there, and a warning sound 2 to remind the vehicle occupant of the sound of the other vehicle 1 around the vehicle is output from the virtual sound source. Then, the warning sound 1 is synthesized with the warning sound 2 so that a synthetic sound can be output, and the synthetic sound is corrected to remove the influence of a space acoustic transmission system from a plurality of speakers arranged in a cabin to the vehicle occupant, and the corrected sound is amplified, and output from a plurality of speakers. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a hearing information notification device and a notification method.

2. Description of the Related Art An audio monitoring device for a vehicle is known in which a driver is notified of a situation around the host vehicle such as the approach of another vehicle by a sound image (see, for example, Patent Document 1).
In this device, the sound of an arbitrary surrounding vehicle is detected with two microphones (simply referred to as a microphone in this specification), and the relative direction of the surrounding vehicle is detected based on the time difference between the sounds of the surrounding vehicles input to the microphone. The driver is notified of the situation of the surrounding vehicle by outputting the sound of the vehicle to the position of the surrounding vehicle.

Prior art documents related to the invention of this application include the following.
Japanese Patent No. 2776092

  However, in the conventional apparatus described above, the sound around the vehicle collected by the microphone is output as it is from the speaker to the vehicle occupant, so the traveling sound of the own vehicle and the traveling sound of the surrounding vehicle are mixed, There is a problem that it is difficult to recognize the existence of a vehicle.

  A first notification sound is generated by correcting the sound of an object existing around the vehicle collected by a plurality of microphones installed on the outer periphery of the vehicle body to be equal to the sound of an object around the vehicle heard by both ears of a vehicle occupant. At the same time, the position of the vehicle peripheral object is detected, a virtual sound source is arranged at the position, and a second notification sound is output from the virtual sound source to remind the vehicle occupant of the sound of the vehicle peripheral object. Then, the first notification sound and the second notification sound are overlapped to output a synthesized sound, and a spatial acoustic transmission system between a plurality of speakers arranged in the passenger compartment and the vehicle occupant is generated for the synthesized sound. A correction for removing the influence is applied, and the corrected sound is amplified and output from a plurality of speakers.

  According to the present invention, it is possible to promptly and reliably recognize the presence and movement of other vehicles around the host vehicle with respect to the vehicle occupant.

  An embodiment in which the auditory information notification device of the present invention is mounted on a vehicle will be described.

<< Aural Information Notification Method of an Embodiment >>
With reference to FIG. 1, a method for notifying auditory information to a vehicle driver according to an embodiment will be described. In FIG. 1, an example will be described in which the presence of the other vehicle 1 approaching to the right rear of the host vehicle is notified to the vehicle occupant, particularly the driver, by two notification sounds. As the notification sound 1, a traveling sound that is actually emitted when the other vehicle 1 travels or a simulated sound that simulates the traveling sound of the other vehicle 1 is used. The traveling sound of the other vehicle 1 is subjected to approximate binaural sound collection by a plurality of microphones attached outside the vehicle, and corrected so that the actual traveling sound of the other vehicle 1 can be heard in both ears of the driver. Generate.

  On the other hand, the notification sound 2 arranges the virtual sound source at the position of the other vehicle 1 and sequentially changes the position of the virtual sound source according to the relative position between the other vehicle 1 and the host vehicle, while the virtual sound source sends the other vehicle to the driver. It is generated as a virtual sound reminiscent of one running sound. The sound source of the notification sound 1 and the virtual sound source of the notification sound 2 are arranged at the same position or close to each other, and are associated with the traveling sound of the other vehicle 1 or the simulated sound included in the notification sound 1 and the traveling sound of the other vehicle 1. The virtual sound, that is, the notification sound 2 is superimposed and synthesized and output to the driver.

<< Experiment for confirming the effect of the auditory information notification method according to one embodiment >>
In order to confirm that the recognition speed of the driver can be improved by the above-described method of notifying auditory information using the notification sound 1 and the notification sound 2, the following experiment was performed. As shown in FIG. 2, a simulated driving environment was constructed, and notification sounds 1 and 2 were output in a state in which the subject took driving action, and the reaction time at that time was investigated. First, a monitor is placed in front of the subject at a position offset to the left from the subject, and driving behavior is performed using this monitor and a driving simulator. Five LEDs 1 to 5 are arranged around the subject, and the flashing of these LEDs 1 to 5 reminds the subject of the movement of the other vehicle 1.

  The subject is informed in advance that these blinks correspond to one of the four travel patterns of the other vehicle 1 shown in FIG. As an example of blinking LEDs 1 to 5 that remind the subject of the movement of the other vehicle 1, the LED 4 (subject can be visually recognized by the rearview mirror) is first blinked for 2 seconds at an arbitrary time. Next, in the case of pattern 1, the left LED 3 blinks, and in the case of pattern 2, the right LED 5 blinks. Note that the left and right LEDs 3 and 5 can be visually recognized by a door mirror. Further, when the movement of the other vehicle 1 is pattern 3, the LED 3 is switched from blinking to LED1 and when the movement of the other vehicle 1 is pattern 4, the LED5 is switched from blinking to LED2. When the blink pattern of these LEDs 1 to 5 is visually recognized and the other vehicle 1 moves to the left or right, or when the travel pattern of the other vehicle 1 can be recognized, the left or right button attached to the steering wheel is pressed. Asked.

  Further, using the virtual sound output speaker shown in FIG. 2, the notification sound 1 that simulates the traveling sound and the notification sound 2 (see FIG. 1) that is a virtual sound are superimposed on the position of the other vehicle 1 simultaneously with the movement of the blinking LED. Output (synthesize and output). The subject is also informed that the left or right button on the steering wheel may be pressed when the other vehicle 1 can recognize the left or right direction based on the sound information of the notification sounds 1 and 2. It was.

  In this experiment, the time until the vehicle 1 recognizes whether the other vehicle 1 has moved to the left and right from the start of the notification of auditory information and visual information and can react (hereinafter referred to as reaction time RT) is regarded as an evaluation characteristic. did. Twenty women who had healthy hearing and within 3 years of obtaining a license were selected as subjects, and notified 40 times per subject using 10 types of notification sounds.

《Notification sound 2 used in the experiment》
Table 1 shows the notification sound 2 used in the experiment.

  The characteristics of each group will be described. Group a is a non-changing sound, and is a group of sounds in which a constant rhythm or a constant melody is repeatedly reproduced for 12 seconds without being influenced by the distance or positional relationship between the host vehicle and another vehicle. However, spatial localization (localization and reproduction volume) is controlled during reproduction of any sound. Group b is a change sound, and is a group of sounds in which the rhythm and melody of the sound to be reproduced changes according to the distance and positional relationship between the host vehicle and the other vehicle. However, spatial localization is controlled during playback of any sound. Group c is a group that outputs notification sound 1 (road noise) and notification sound 2 simultaneously. The notification sound 2 is a sound included in the groups a and b, and is positioned as marking with a notification sound to other vehicle running noise. The sound of group c is notified by performing spatial localization processing together with noise.

表２において、各報知音のタイトルに付される○記号は変化音を表し、×記号は無変化音を表す。以下にそれぞれの報知音についてその特徴を説明する。 Since the notification sounds used in the group c are sounds belonging to either of the groups a and b, the notification sound materials used in the experiment are six types included in the groups a and b. Table 2 is a table in which six types of notification sounds are classified by sound type.

In Table 2, the symbol “O” attached to the title of each notification sound represents a changing sound, and the symbol “x” represents a non-changing sound. The characteristics of each notification sound will be described below.

  First, “symbol sound (rhythm)” is a series of notification sounds that have no change in pitch (the ones with a change in pitch are classified as melody) and are played with a rhythm in which one type of tone is repeated. It was. March_15 (no change) is a rhythm “bobobo, bobobo, bobobo” that repeats a unit sound of 196ms three times continuously (triplet) and then repeats the unit sound three times again with a rest of 609ms. , Bobobo ". The unit sound audibly has a pitch around 180 Hz. The notification sound does not change according to the distance or the like.

  Pon-Rhythm (change) is a rhythm “pawn, pawn, pawn, pawn, pawn” that repeats the sound of “pawn” obtained by applying an envelope processing of a decay system of 1.2 s to a 600 Hz sine wave. This repetition interval changes according to the distance between the host vehicle and the other vehicle. The repeat interval is designed to be longer as the distance is longer (far), and the repeat interval is shortened as the distance is closer. In the experiment, the repetition interval when the farthest was 1.7 s, the repetition interval when the closest was 0.5 s, and the repetition interval between them was adjusted at a pitch of 0.3 s. The sensibility feature of this rhythm aimed to be simple. Although it is a sine wave, it has been made a sound that is not harsh by taking a relatively long decay envelope.

  Next, a melody having a regularity with two or more pitch changes in a series of notification sounds was used. Pop-melody (no change) repeats a light and pop 1-bar (4 beats) melody, adding drum beats to make it more musical and nimble, with a cute-shaped light car running Produced the appearance of coming. The phrases are B · GC-A and B-GGC-A.

  March_B20 (no change) repeats a simple phrase G-CGDGEG with a single tone color and a half measure (two beats). Dim-Guitar (change) is based on the melody design disclosed in Japanese Patent Laid-Open No. 11-208370. The Dim code (diminished code) arpeggio pattern is divided into four patterns according to the distance between the host vehicle and other vehicles. To change. The Dim code is a code that gives a sense of instability and urgency.

  Sus4-IceRain-updn (change) is a melody design in which an arrangement is added to the melody disclosed in Japanese Patent Laid-Open No. 11-208370. The basic idea of changing the arpeggio according to the distance to other vehicles is the same, but the code used is changed with less sense of pressure and being preferred by women, and the arpeggio rises and falls did. The code used is a Sus4 code (suspension code or suspended four code). Compared to normal major chords, there is some instability, but it is not as urgent as Dim chords. When the other vehicle approaches from behind, the arpeggio in the pitch increasing direction, and when the other vehicle moves away from the front, the arpeggio in the pitch descending direction is used. I tried to display the change of sound. The aim was to make the arpeggio sound beautiful by using a chime-type tone with a clear sound. In addition, we aimed to not give a sense of urgency as much as Dim-Guitar.

"Experimental result"
FIG. 4 shows the notification method of the notification sound and the average reaction time for each sound type. In the figure, the vertical axis represents the average reaction time [second], and the horizontal axis represents the sound type of the notification sound 2. The average reaction time when only the notification sound 1 is notified is indicated by a line, and the average reaction time when only the notification sound 2 is notified is notified by overlapping the notification sound 1 and the notification sound 2 on the right bar graph for each sound type. The average reaction time in each case is indicated by a left bar graph for each sound type. From the experimental results shown in FIG. 4, when the notification sound 1 and the notification sound 2 are overlapped and notified, the average reaction time is the shortest, and when the notification sound 2 is notified singly, the notification sound 1 is notified singly. The reaction time increases in this order. From this experimental result, it is apparent that the average recognition speed is the fastest according to the auditory information notification method of the embodiment in which the notification sound 1 and the notification sound 2 are overlapped for notification.

  FIG. 5 shows a notification sound notification method and a result of dispersion of reaction time for each sound type. In the figure, the vertical axis represents the dispersion value of the reaction time, and the horizontal axis represents the sound type of the notification sound 2. The notification value 1 and the notification sound 2 are shown in the right bar graph for each sound type with the dispersion value of the reaction time when only the notification sound 1 is notified as a line, and the dispersion value of the reaction time when only the notification sound 2 is notified. The dispersion value of the reaction time in the case where notification is repeated is shown by the left bar graph for each sound type. From the experimental results shown in FIG. 5, for the sound types Pop-melody and Dim-Guitar, when the notification sound 1 and the notification sound 2 are overlapped, the reaction time dispersion value is minimized. When the notification is made independently, the dispersion value of the reaction time increases in the order in which the notification sound 1 is notified alone. From this experimental result, for the sound types Pop-melody and Dim-Guitar, the auditory information notification method according to one embodiment in which the notification sound 1 and the notification sound 2 are overlapped and reported is most effective regardless of the subject. it is obvious.

  On the other hand, for the sound type Pop-Rhythm, the dispersion of the reaction time is smaller when the notification sound 2 is notified alone than when the notification sound 1 and the notification sound 2 are notified in an overlapping manner. For this reason, when using intermittent sounds (intermittent sounds with long durations of silence so that the vehicle occupant cannot recognize them as continuous sounds) like the sound type Pop-Rhythm, the dispersion of the reaction time increases. It was shown that.

  FIG. 6 shows the results of investigating the influence on the recognition time when the melody change is added or not according to the degree of approach of the other vehicle. In the figure, the vertical axis represents the response rate, and the horizontal axis represents the reaction time. In the figure, U4 was able to respond within 4 seconds from the start of information notification, U6 was able to respond within 6 seconds from the start of information notification, and U10 was able to respond within 10 seconds from the start of information notification. Each response rate is shown. This survey result shows that the response rate is higher when the melody change is not applied than when the melody change is applied. From this, it can be concluded that there is an effect of shortening the reaction time if the melody change is not applied according to the degree of approach of the other vehicle.

  In this embodiment, the frequency distribution of the sound of other vehicles around the own vehicle collected by a plurality of microphones is analyzed, and the type of the other vehicle is specified. Then, one or more elements of the rhythm, melody, harmony, and frequency characteristics of the notification sound 2 are determined according to the type of the other vehicle. Thereby, the recognition speed can be increased for other vehicles around the host vehicle.

  Further, in this embodiment, the second notification sound is a continuous sound that does not have a silent portion and continues continuously, or an intermittent sound that has a duration of a silent portion that is so short that the vehicle occupant cannot recognize it as an intermittent sound. did. In this specification, it is called continuous sound including both. For example, it is possible to improve the recognition performance of other vehicles in the vicinity of the host vehicle by using a continuous sound such as “Piroro Pirororo” instead of an intermittent sound such as “Pong, Pong, Pong”.

  The second notification sound may be the above-described continuous sound and a sound having a constant frequency component fluctuation that is independent of the distance from the own vehicle to other vehicles around the own vehicle. In other words, the frequency of the continuous sound was not increased as the other vehicle approached. Thereby, the recognition performance of the other vehicle around the own vehicle can be improved.

  Further, the second notification sound may be the above-described continuous sound and a sound having a certain melody element irrelevant to the distance from the own vehicle to other vehicles around the own vehicle. Thereby, the recognition performance of the other vehicle around the own vehicle can be improved.

<< Configuration of One Embodiment >>
FIG. 7 is a block diagram showing a basic configuration of an embodiment, and FIG. 8 is a diagram showing a specific configuration of the embodiment. The sound input unit 100 is installed near the outer wall around the vehicle and collects sound around the vehicle. As the sound input unit 100, microphones 210_1 to 210_n can be used as shown in FIG. 8, and these microphones may be small enough to be attached to the outer wall of the vehicle, and a general dynamic microphone or the like is used. Can do. The sounds collected by these microphones 210_1 to 210_n are converted into electrical signals of appropriate magnitude by the filter-equipped microphone amplifier 220 shown in FIG.

  The first correction unit 110 processes the input electrical signals of a plurality of sounds into an appropriate state, converts the signal into a desired number of channels (the number of input channels of the second correction unit 150), and outputs it. The first correction unit 110 includes the AD converter 230, the arithmetic device 240, the storage device 245, and the DA converter 270 shown in FIG. The sound output from the first correction unit 110 needs to be theoretically equivalent to a desired sound at a “control point” set near the driver's both ears or near the passenger's head. Details of the configuration of the first correction unit 110 will be described later.

  The position detection unit 120 detects the position of an object such as another vehicle existing around the host vehicle. As the position detection unit 120, as shown in FIG. 8, a position sensor 250 that can detect a relative position of an object around the host vehicle with respect to the host vehicle, such as a millimeter wave radar, an ultrasonic sensor, a camera and an image processing device, can be used. .

  The virtual sound source generator 130 converts a monaural sound source or the like prepared in advance into a virtual sound source (notification sound 2) of two or more channels so that a sound image comes from the position detected by the position detector 120. As a generally known virtual sound source, there is a method using a binaural sound field reproduction system that superimposes transfer characteristics from a target position to both ears of a vehicle occupant and outputs headphones through two channels. The notification sound 2 at this time has the same number of channels as the notification sound 1.

  The synthesizing unit 140 adds the notification sound 1 and the notification sound 2 for each channel. At this time, the sound pressure levels of the notification sound 1 and the notification sound 2 may be adjusted to an appropriate ratio, for example, a ratio of 1: 1 in the energy region.

  When the second corrector 150 outputs the notification sound 1 and the notification sound 2 from the speakers 290_1 and 290_2 (see FIG. 8), the second correction unit 150 transmits spatial acoustics from the speakers 290_1 and 290_2 to, for example, control points near the driver's ears. The influence of the system is removed, and the output signal of the synthesis unit 140 is corrected so that the sound signal corrected by the first correction unit 110 is reproduced at the control point. The second correction unit 150 includes the AD converter 230, the arithmetic device 240, the storage device 245, and the DA converter 270 shown in FIG.

  The sound output unit 160 amplifies the sound signal output from the second correction unit 150 and outputs it from the speakers 290_1 and 290_2. The sound output unit 160 includes the amplification device 280 and speakers 290_1 and 290_2 shown in FIG. These amplifying device 280 and speakers 290_1 and 290_2 may be general ones that can be mounted on the vehicle.

  The distance detection unit 170 detects the distance between the host vehicle and the vehicle surrounding object based on the position of the vehicle surrounding object detected by the position detection unit 120. Although details will be described later, the generation of the notification sound 2 is stopped when the distance to the object around the host vehicle exceeds a predetermined threshold value. The distance detection unit 170 includes the arithmetic device 240 and the storage device 245 shown in FIG.

  The timer 180 measures the elapsed time from the start of notification of the notification sound 2, that is, the output continuation time. Although details will be described later, generation of the notification sound 2 is stopped or the output sound pressure level of the notification sound 2 is reduced to a predetermined value or less when the output duration time of the notification sound 2 exceeds a predetermined time. The timer unit 180 includes the arithmetic device 240, the storage device 245, and the timer 260 shown in FIG.

  The detection unit 190 detects the driver's head position, and changes the processing contents of the first correction unit 110 and the second correction unit 150 according to the head position. The detection unit 190 includes a head position detection device 265, a calculation device 240, and a storage device 245 shown in FIG.

<< Operation of One Embodiment >>
FIG. 9 is a flowchart showing the operation of the embodiment. The auditory information notification device of one embodiment performs this operation when the power is turned on. First, in step 110, initial setting is performed. At the time of initial setting, the filters used by the first correction unit 110 and the second correction unit 150 are read. At this time, a filter corresponding to the driver's head position detected by the detection unit 190 is selected. In step 120, the sound collected by the sound input unit 100 is input and amplified to an appropriate level.

  In step 130, a first correction process is performed. That is, the sound signal is discretely converted, correction filter processing and channel integration processing are performed, and the corrected sound signal, that is, the notification sound 1 is output. On the other hand, in parallel with the first correction process, the position of the other vehicle around the own vehicle is detected in step 111, the distance from the own vehicle to the other vehicle is detected in the subsequent step 112, and the distance to the other vehicle is set in advance. Determine whether it is below the threshold. If the distance to the other vehicle is greater than the threshold value, the process returns to step 111, and if the distance to the other vehicle is equal to or less than the threshold value, the process proceeds to step 113. In step 113, notification sound 2 that sounds like a sound coming from the position of another vehicle is generated, and the process proceeds to step 135.

  In step 135, notification sound 1 and notification sound 2 are added for each channel. In step 140, a second correction process is performed. That is, correction filter processing for canceling crosstalk between the speakers 290_1 and 290_2 and the control point (near the driver's both ears) is performed on the synthesized sound signal of the notification sound 1 and the notification sound 2. In step 150, the sound signal after the second correction processing is amplified, and in step 170, the sound signal is output from the speakers 290_1 and 290_2.

  In step 175, it is confirmed whether or not the output time of the notification sound 2 being measured by the timer 180 exceeds a predetermined time set in advance, or whether the distance from the other vehicle exceeds a predetermined value set in advance. When the output duration time of sound 2 exceeds the predetermined time, or when the distance to the other vehicle exceeds the predetermined distance, the process proceeds to step 176, the output of the notification sound 2 is stopped, and the process returns to step 111. Note that the notification sound 2 is set for each other vehicle around the host vehicle, and the output duration of the notification sound 2 is also measured for each other vehicle around the host vehicle.

  When the output time of the notification sound 2 does not exceed the predetermined time and the distance from the other vehicle does not exceed the predetermined distance, the process proceeds to step 180 to detect the driver's head position. If a change in the head position is detected, the process proceeds to step 190, the filters of the first correction unit 110 and the second correction unit 150 corresponding to the head position are selected, and the process returns to step 120. If there is no change in the driver's head position in step 180, the process directly returns to step 120.

  If the signal in the series of processing is a continuous signal, AD conversion processing and DA conversion processing are performed before and after the first correction processing and second correction processing. If the signal is a discrete signal, these AD conversion processing and DA conversion processing are performed. The first correction process and the second correction process are performed without performing them. In the l-th correction process and the second correction process, it is desirable to apply a discrete filter from the viewpoint of cost or the like, but a filter capable of continuous processing may be applied.

<< Detailed Configuration and Processing of First Correction Unit >>
The detailed configuration of the first correction unit 110 and the first correction process will be described. At least two microphones 210_1 to 210_n are installed, and these microphones are sufficiently spaced so that the sound coming from one sound source has a large difference for each microphone in characteristics such as time, frequency, and sound pressure level. Make the arrangement. Ideally, it is desirable that two sounds input from two microphones are established as binaural sound sources. Actually, since the sound recorded by the microphone attached to the outer wall of the vehicle is different from the sound recorded by the microphone arranged at the pinna position such as the dummy head, pseudo binaural recording is performed.

  For example, as shown in FIG. 10, a dome-shaped casing (similar in size to a human head) having a shape close to a dummy head on the roof of a vehicle, and pseudo auricles attached to the left and right sides of the casing Attach a microphone in the ear canal and collect sound. For example, as shown in FIG. 11, pseudo auricles are attached to the left and right outer walls of the vehicle, and a microphone is attached to the ear canal of these auricles to collect sound. In such a configuration, although the input signal is different from the actual binaural recording, a sound close to the driver can be obtained in that a sound image is formed in a target direction. At this time, it is easier for the microphone to be as close to the wall as possible, so that the difference between the sound signals input to the left and right microphones is likely to occur.

The input sound collected by two microphones is pseudo binauralized using the method described with reference to FIGS. That is, the sound is close to the sound collected by the microphone installed in the dummy head or the head and torso simulator pinna. When the first correction unit 110 performs filtering on the sound source collected in this state, it is desirable to perform processing so that the content of the sound source is not changed. For example, in FIG. 10, the input from the left microphone is X1, the input from the right microphone is X2, and the filters of the first correction unit 110 (first correction processing) applied to each input are H1, H2, If the outputs from the first correction unit 110 (first correction process) are Y1 and Y2,
Y1 = X1 · H1 (1),
Y2 = X2 · H2 (2)
Can be expressed as At this time, since Y1 and Y2 may be equivalent to X1 and X2, the filters H1 and H2 have the tap number 1 and the coefficient 1, or the tap number N is any time (and the same time in H1 and H2). It is desirable to use a time delay filter in which a coefficient of 1 stands. For example,
H1 = H2 = 1 (3),
H1 = H2 = [0, 1, 0, 0, 0, 0, 0] (N = 7 taps, time delay 1 tap)
... (4)
It can be realized by such a configuration. The first correction unit 110 only needs to be adjusted so that the traveling noise of the other vehicle is localized at a relative position with respect to the host vehicle, and a correction method other than the method described above may be used.

<< Detailed Configuration and Processing of Second Correction Unit >>
Next, the detailed configuration of the second correction unit 150 and the second correction process will be described. FIG. 12 is a diagram for explaining the concept of correction processing by the second correction unit 150. At the control points C1 and C2 set near the driver's both ears, the second correction unit 150 is configured so that the synthesis unit 140 reproduces the input signals Y1 and Y2 obtained by adding the notification sound 1 and the notification sound 2 for each channel. Configure. Between the input signals Y1 and Y2 from the synthesizer 140 and the reproduction signals (observation signals) Z1 and Z2 at the control points C1 and C2,
Y1 = Z1 (5),
Y2 = Z2 (6)
It is desirable that this relationship is established. Therefore, in this embodiment, the influence of the spatial acoustic transmission system is removed using a correction filter. Hereinafter, this filter is referred to as an “inverse filter”.

The configuration and calculation method of the inverse filter used in the second correction unit 150 will be described with reference to FIG. For each arbitrary frequency, when the input signal Yn, the inverse filter Hmn (m is the sound source number, n is the control point number), the reproduction signal (observation signal at the control point) Zn, and the spatial transfer characteristic Fnm, the relationship between the elements is as follows: ,
F ・ H ・ Y = Z (7)
It is expressed. In the equation (7), F, H, Y, and Z are vectors, and are represented by determinants of F = [Fnm], H = [Hmn], Y = [Yn], and Z = [Zn]. At this time, in order to satisfy the relationship of the above expressions (5) and (6),
F ・ H = I (8)
(I is a unit vector). Therefore, in order to derive H from F, H = F ⁻ (9) for each arbitrary frequency.
Should be calculated. In equation (9), [•] ⁻ represents a general inverse matrix of the matrix [•]. For example, the calculation method described in “Application of inverse filter design using minimum norm solution to transoral system: Proceedings of the Acoustical Society of Japan, pp495-496 (1998)” can be employed.

  By the way, it has been pointed out that it is generally difficult to hear the notification sound due to a change in the hearing state of the driver (for example, a sound of a specific frequency cannot be heard). Even when the sound is localized in a specific direction as in the embodiment, the influence of the frequency component cannot be ignored in one of the localization cues, and in order to design the system robustly, the bandwidth should be as wide as possible. It is desirable to use a notification sound. Therefore, by designing the notification sound 2 so as to positively use frequency components that are not included in the notification sound 1, a robust system that does not depend on the driver can be realized, and recognition of other vehicles around the host vehicle can be realized. Performance can be improved.

  For example, when the energy distribution in the frequency band of the notification sound 1 includes many sounds of less than 1 kHz, the notification sound 2 may be designed to include many frequency components of 1 to 2.5 kHz. At this time, it is effective that the sound of the notification sound 2 is selected by the sound of the notification sound 1. As an example, the arithmetic device 240 and the storage device 245 shown in FIG. 8 are used, and the frequency analysis is performed at the time when the notification sound 1 is input, and the frequency of the notification sound 2 stored in the storage device 245 in advance is notified. What is necessary is just to make it the method of shifting and transmitting so that many bands which are not included in 1 may be included.

<< Output timing of notification sound 2 >>
If the notification sound 2 is continuously output, there is a risk of a mental burden on the driver. Therefore, in this embodiment, the output timing of the notification sound 2 is controlled. Now, as shown in FIG. 14, a region 1 having a radius r [m] centered on the own vehicle is set, and other vehicles in this region 1 are detected. When another vehicle 1 enters the area 1, a virtual sound source is set at the position of the other vehicle 1 and a notification sound 2 is output. The region 1 may be determined according to performance requirements such as the detection distance of the position detection sensor, or determined according to the relationship between the sound pressure level of the notification sound 1 and the driver's hearing, that is, the distance at which the notification sound 1 can be heard. Also good.

  In the example shown in FIG. 14, since the other vehicle 1 is in the region 1, the notification sound 2 is assigned to the other vehicle 1, and the output of the sound is started with the notification sound 1. On the other hand, since the other vehicle 2 is not in the area 1, the notification sound 2 is not allocated. As shown in FIG. 15, when the other vehicle 1 leaves the area 1, the notification sound 2 for the other vehicle 1 is stopped. At this time, since the other vehicle 2 is in the area 1, the notification sound 2 is allocated to the other vehicle 2, and the output of the sound is started overlapping the notification sound 1.

  Further, if the other vehicle continues to stay in the region 1 shown in FIGS. 14 and 15, the driver will hear the notification sound 2 endlessly. The notification sound 2 is different from the notification sound 1 in that a mental burden is large for the driver to continue to listen. Therefore, the timer 260 shown in FIG. 8 measures the output continuation time from the output start time t0 of the notification sound 2 for each other vehicle, and stops the notification sound 2 of other vehicles whose output continuation time exceeds the preset time te. Or the sound pressure level is lowered to a level that the driver does not care about. For example, as shown in FIG. 16, when the other vehicle 1 still remains in the region 1 even when the output duration time of the notification sound 2 exceeds the set time te, the notification sound 2 for the other vehicle 1 is temporarily displayed. Stop or reduce the sound pressure level. When the other vehicle 1 accelerates and moves from position P1 in the region 1 to P2 at time t1 (> te), that is, the change in the distance in the region 1 from the own vehicle to the other vehicle 1 is greater than or equal to a predetermined value. In this case, the output of the notification sound 2 is started again or the sound pressure level is restored, and the output duration time of the vehicle 1 other than the timer 260 is reset.

  Next, an example in which the filter is switched in accordance with the change in the driver's head position will be described. The first correction unit 110 and the second correction unit 150 change the first correction unit 110 and the second correction unit 150 according to the change in the head position because the reproduction accuracy and the like are reduced due to the change in the driver's head position. It is desirable to do. After the head position of the driving vehicle is detected by the head position detection unit 190 shown in FIG. 7, the filter of the first correction unit 110 and the filter of the second correction unit 150 stored in advance in the storage device 245 are displayed on the driver's head. Select according to the head position.

  Specifically, a filter obtained from a transfer function from the microphone to the driver's binaural position at the head position P1 when the driver's seat is moved to the front and at the head position P2 when the driver's seat is moved to the rearmost part. G1 and G2 are stored in advance, and a filter of the first correction unit 110 corresponding to the position of the seat is selected. Similarly, as a selection example of the second correction unit filter, the driver's binaural from the microphone at the head position P1 when the driver's seat is moved to the front and the head position P2 when the driver's seat is moved to the rearmost part. The inverse filters H1 and H2 obtained from the transfer function up to the position are stored in advance, and the inverse filter of the second correction unit 150 is selected according to the position of the seat.

  Thus, according to one embodiment, the sound of other vehicles around the vehicle that hears the sounds of other vehicles existing around the vehicle collected by a plurality of microphones installed on the outer periphery of the vehicle with both ears of the vehicle occupant. To generate a notification sound 1 and detect the position of the other vehicle around the vehicle, place a virtual sound source at the position, and send the sound of the other vehicle around the vehicle to the vehicle occupant from the virtual sound source. The notification sound 2 to be associated is output. Then, the notification sound 1 and the notification sound 2 are overlapped to output a synthesized sound, and with respect to this synthesized sound, a plurality of speakers arranged at positions different from the plurality of microphones in the vehicle interior to the vehicle occupant. Since the correction was made to remove the influence of the spatial acoustic transmission system and the corrected sound was amplified and output from multiple speakers, the presence and movement of other vehicles in the vicinity of the host vehicle were quickly identified for the vehicle occupant. , You can be sure to recognize.

  According to one embodiment, the output of the notification sound 2 is started when the position of the other vehicle around the vehicle falls within a predetermined distance from the own vehicle, and the position of the other vehicle around the vehicle is changed from the own vehicle. Since the output of the notification sound 2 is stopped when a predetermined distance or more is left, it is possible to prevent an increase in the load on the vehicle occupant due to the output of the notification sound 2 for a long time.

  Furthermore, according to one embodiment, the output duration of the notification sound 2 is measured, and when the output duration exceeds a predetermined time, the output of the notification sound 2 is stopped or the sound pressure level of the notification sound 2 is set. Since it is reduced to a predetermined value or less, it is possible to prevent an increase in the load on the vehicle occupant due to the continuous output of the notification sound 2 for a long time.

  Furthermore, according to one embodiment, the head position of the vehicle occupant is detected, and the correction contents of the first correction process and the second correction process are changed according to the head position of the vehicle occupant. Sounds of other vehicles around the host vehicle can be faithfully and accurately reproduced at the positions of both ears of the vehicle occupant.

The figure explaining the alerting | reporting method of the auditory information with respect to the vehicle driver of one embodiment The figure which shows the experimental installation for confirming the effect by the alerting | reporting method of the auditory information of one embodiment The figure which shows the movement pattern of the other vehicle around the own vehicle The figure which shows the notification method of the notification sound of the experimental result and the average reaction time for each sound type The figure which shows the result of the dispersion | distribution of the notification time of the notification sound of an experimental result, and the reaction time for every sound classification The figure which shows the result of investigating the influence on the recognition time with and without the melody change depending on the degree of approach of other vehicles The block diagram which shows the basic composition of one embodiment The figure which shows the specific structure of one embodiment. Flowchart showing operation of one embodiment Diagram showing an example of microphone installation Diagram showing another example of microphone installation The figure for demonstrating the concept of the correction process by a 2nd correction | amendment part. The figure explaining the structure and calculation method of an inverse filter used in the 2nd correction part The figure explaining the notification start timing of the notification sound 2 The figure explaining the notification stop timing of the notification sound 2 The figure explaining the output continuation time restriction | limiting of the notification sound 2

Explanation of symbols

100 sound input unit 110 first correction unit 120 position detection unit 130 virtual sound source generation unit 140 synthesis unit 150 second correction unit 160 sound output unit 170 distance detection unit 180 timer 190 detection units 210_1 to 210_n microphone 240 arithmetic device 245 storage device 250 Position sensor 260 Timer 265 Head position detection device 290_1-290_n Speaker

Claims (11)

A plurality of microphones provided around the vehicle body for collecting sounds around the vehicle;
A plurality of speakers arranged in the passenger compartment;
First correction means for generating first notification sound by correcting the sound of an object existing around the vehicle collected by the plurality of microphones to be equal to the sound of the vehicle peripheral object heard by both ears of a vehicle occupant When,
Object position detecting means for detecting the position of the object around the vehicle;
Virtual sound output means for arranging a virtual sound source at a position of the vehicle peripheral object, and outputting a second notification sound reminiscent of a sound of the vehicle peripheral object from the virtual sound source to a vehicle occupant;
Combining means for superposing the first notification sound and the second notification sound to output a synthesized sound;
Second correction means for performing correction to remove the influence of a spatial acoustic transmission system between the plurality of speakers and a vehicle occupant on the synthesized sound;
An auditory information notification device comprising: output means for amplifying the output of the second correction means and outputting the amplified output from the plurality of speakers. The auditory information notification device according to claim 1,
The virtual sound output means starts the output of the second notification sound when the position of the object around the vehicle falls within a predetermined distance from the own vehicle. In the auditory information notification device according to claim 1 or 2,
The virtual sound output means stops the output of the second notification sound when the position of the object around the vehicle is more than a predetermined distance from the own vehicle. In the auditory information notification device according to any one of claims 1 to 3,
Comprising time measuring means for measuring the output duration of the second notification sound,
The virtual sound output means stops the output of the second notification sound when the output duration time of the second notification sound exceeds a predetermined time, or sets the sound pressure level of the second notification sound to a predetermined value or less. An auditory information notification device characterized by reducing. In the auditory information notification device according to any one of claims 1 to 4,
A head position detecting means for detecting a head position of a vehicle occupant;
The first correction means and the second correction means change the correction processing according to the head position of the vehicle occupant, and the auditory information notification apparatus is characterized. In the auditory information notification device according to any one of claims 1 to 5,
An object type determining means for analyzing the sound of the vehicle peripheral object collected by the plurality of microphones and determining the type of the vehicle peripheral object;
The virtual sound output means determines one or more elements of a rhythm, a melody, a harmony, a timbre, and a frequency characteristic of the second notification sound according to the type of the object around the vehicle. Information notification device. In the auditory information notification device according to any one of claims 1 to 6,
The second notification sound is a continuous sound that does not have a silent part and continues continuously, or is an intermittent sound that has a duration of a silent part that is so short that the vehicle occupant cannot recognize it as an intermittent sound. Information notification device. In the auditory information notification device according to any one of claims 1 to 6,
The second notification sound is a continuous sound that does not have a silent portion and continues continuously, or an intermittent sound that has a duration of a silent portion that is so short that the vehicle occupant cannot recognize it as an intermittent sound. An auditory information notification device characterized by having a constant frequency component variation that is independent of a distance to an object around the vehicle. In the auditory information notification device according to any one of claims 1 to 6,
The second notification sound is a continuous sound that does not have a silent portion and continues continuously, or an intermittent sound that has a duration of a silent portion that is so short that the vehicle occupant cannot recognize it as an intermittent sound. An auditory information notification device having a certain melody element irrelevant to a distance to a surrounding object. In the auditory information notification device according to any one of claims 1 to 6,
The auditory information notification device, wherein the frequency component of the second notification sound is different from the frequency component of the first notification sound. A first notification sound is generated by correcting the sound of an object existing around the vehicle collected by a plurality of microphones installed on the outer periphery of the vehicle body to be equal to the sound of the vehicle peripheral object heard by both ears of a vehicle occupant A first correction process to
A virtual sound output process of detecting a position of the vehicle peripheral object, arranging a virtual sound source at the position, and outputting a second notification sound reminiscent of a sound of the vehicle peripheral object from the virtual sound source to a vehicle occupant;
A synthesis process for superposing the first notification sound and the second notification sound to output a synthesized sound;
Performing a second correction process for correcting the synthesized sound to remove the influence of a spatial acoustic transmission system between a plurality of speakers arranged in the passenger compartment and the vehicle occupant;
A method for notifying an auditory information, wherein the sound after the second correction processing is amplified and output from the plurality of speakers.

Images