close

Вход

Забыли?

вход по аккаунту

?

JP2000324233

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2000324233
[0001]
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a
loudspeaker telephone equipped with a speaker for amplifying a received signal from a calling
party and a microphone for picking up our speech.
[0002]
2. Description of the Related Art In home telephones, in-vehicle telephones, mobile telephones
and the like, a speakerphone equipped with a hands-free function capable of making a call
without having a handset has become widespread. In such a speakerphone, voices amplified from
a speaker may be picked up by a microphone to cause echoes or howling, which may cause
deterioration of the speech quality. In order to suppress such echo and howling, a device such as
a voice switch or an acoustic echo canceler may be used. However, by adding these devices, there
is a problem that the circuit scale becomes large and the cost of the telephone increases.
Therefore, there is a demand for a low-cost apparatus for suppressing echo and howling at the
time of loud-speaking.
[0003]
A first conventional loudspeaker system will be described with reference to FIG. A speaker unit
101, a microphone unit 102, an echo suppressing means 103, and a handset / speech switching
09-05-2019
1
switch 104 are provided in this speakerphone. The speaker unit 101 is a speaker that amplifies
an incoming signal from the far-end speaker. The microphone unit 102 is a microphone that
picks up the voice of the near end speaker. The echo suppression means 103 suppresses the
voice of the far-end talker by the reception signal and the transmission signal, and outputs the
voice of the near-end talker as a transmission signal. The handset / speech switch switch 104 is a
switch for switching between a call using a handset (handset) 105 and a call using a
speakerphone.
[0004]
Normally, a call is made using the receiver 105. However, by switching to a loud call with the
handset / loudspeaker switch 104, a hands-free call can be made using the speaker unit 101 and
the microphone unit 102. In the hands-free state, the receiver 105 is moved from the upper part
of the speaker unit 101, so that the speaker unit 101 is not interrupted.
[0005]
Loudspeaking voice from the speaker unit 101 propagates through the space and the telephone
case and wraps around the microphone unit 102 to form a signal loop between transmission and
reception. In this case, echo and howling occur and the speech quality is greatly degraded. In
order to prevent such echo and howling, a voice switch or an acoustic echo canceller is provided
as the echo suppression means 103. In the case of a voice switch, the speech of the near end
speaker is detected by the output signal of the microphone unit 102 and the reception signal,
and the output signal of the microphone unit 103 is output as a transmission signal only when
the near end speaker speaks. Cuts off the loop between transmission and reception.
[0006]
Further, in the case of the acoustic echo canceller, an adaptive filter is provided, and a transfer
function between the speaker unit 101 and the microphone unit 102 is adaptively obtained by a
learning identification method or the like based on the reception signal and the transmission
signal. Then, the obtained transfer function is convoluted with the reception signal and
subtracted from the output signal of the microphone unit 102, whereby only the voice of the
near end speaker is output from the output signal of the microphone unit 102.
09-05-2019
2
[0007]
A second prior art loudspeaker system will be described with reference to FIG. This
speakerphone is described in JP-A-3-117148. A speaker unit 111, microphone units 112 and
113, and a differential operation circuit 114 are provided in this loudspeaker. The speaker unit
111 is a speaker that amplifies incoming voice. The microphone units 112 and 113 are
microphones for sound collection disposed at equal distances from the speaker unit 111. The
differential operation circuit 114 is a circuit that outputs the difference between the output
signal of the microphone unit 112 and the output signal of the microphone unit 113.
[0008]
Since the distance between the microphone unit 112 and the microphone unit 113 with respect
to the speaker unit 111 is equal, the received signal that has been amplified from the speaker
unit 111 is collected in the microphone units 112 and 113 in the same phase. When the
respective output signals are differentiated in the differential operation circuit 114, the
difference value is theoretically 0, and it is possible to suppress the voice wraparound of the farend speaker. On the other hand, since the distances between the microphone unit 112 and the
microphone unit 113 with respect to the utterance position of the near-end speaker are different
from each other, the voice of the near-end speaker picked up by the microphone units 112 and
113 is differentiated by the differential operation circuit 114. Even if this is the case, the amount
of attenuation due to the difference is small compared to the loud sound of the speaker unit 111.
[0009]
SUMMARY OF THE INVENTION In the first conventional loud speaker, echo and howling due to
the wraparound of a received signal from the speaker unit 101 to the microphone unit 102 by
echo suppression means such as a voice switch and an acoustic echo canceler. I try to prevent.
When the voice switch is used, if the voice level of the near-end speaker is low, the voice switch
may malfunction, and part of the near-end speaker's voice may be interrupted. In this case, there
is a disadvantage that the call quality is degraded. When an acoustic echo canceler is used, nearend talker voices such as voice switches are not lost, but the circuit configuration becomes
complicated to obtain a sufficient echo suppression effect, making the loudspeaker telephone
expensive. It has the drawback of becoming
09-05-2019
3
[0010]
Further, although the complexity of the circuit configuration for howling suppression, which is
the problem of the first prior art, is improved in the second prior art loudspeaker, the end is
closer to the first prior art than the first prior art. There is a problem that the talk voice quality of
the speaker is degraded. FIG. 12 shows the frequency characteristics of the output signal of the
differential operation circuit 114 in the D1 direction of FIG. 11 as an example. Here, the distance
between the microphone units 112 and 113 is 20 cm. Since the phase difference between the
signals collected by the microphone unit 112 and the microphone unit 113 is small in the low
band, the difference value is greatly attenuated. Further, in the high frequency band, the two
microphone units cancel each other at a frequency that is an integral multiple of the half
wavelength of the sound wave, and the output value is greatly attenuated at that frequency.
Therefore, compared to the case where sound is picked up by the non-directional microphone of
the first conventional example, there is a shortage of bass and a partial lack of high frequencies,
so that there is a problem of causing deterioration of speech quality.
[0011]
The present invention has been made in view of such conventional problems, and solves the
drawbacks of echo suppression means such as voice switches and acoustic echo cancellers and
causes the quality of speech of near-end speakers to deteriorate. It is an object of the present
invention to realize a speakerphone capable of suppressing echo and howling.
[0012]
[Means for Solving the Problems] In order to solve such problems, the invention of claim 1 of the
present application receives the voice of the far-end speaker as a reception signal and sends the
voice of the near-end speaker A speaker phone for transmitting to the far-end speaker as a signal,
the speaker unit for expanding the received speech signal, and the speaker unit being provided
equidistantly from the speaker unit and at different distances from the near-end speaker First
and second microphone units for picking up the voice of the end speaker, difference calculating
means for calculating and outputting a difference value of output signals of the first and second
microphone units, and the difference calculating means And frequency correction means for
correcting the output signal with a predetermined frequency characteristic.
[0013]
09-05-2019
4
According to such a configuration, two nondirectional microphone units are arranged at equal
distances from the speaker unit, respectively, and loud voices are collected in the same phase.
Then, the voice of the far-end speaker is canceled by subtracting the output signals of these two
microphone units, and the frequency characteristic of the output signal is corrected with a
predetermined frequency characteristic, thereby degrading the speech quality. Can be reduced.
[0014]
The invention according to claim 2 of the present application is a speakerphone that receives the
voice of the far end speaker as a reception signal and transmits the voice of the near end speaker
as a transmission signal to the far end speaker, the reception signal A speaker unit for amplifying
the sound, and first and second microphone units which are provided at equal distances from the
speaker unit and at different distances from the near end talker and which picks up the voice of
the near end talker; Near end speaker voice detection means for detecting presence or absence of
the voice of the near end speaker by the reception signal and the output signal of the first or
second microphone unit, and near end by the near end speaker voice detection means When the
voice of the speaker is detected, the output signal of the first or second microphone unit is
output as it is, otherwise, the difference value between the output signals of the first and second
microphone units Calculate It is characterized in that it comprises a difference calculating means
for, the.
[0015]
According to such a configuration, two nondirectional microphone units are disposed
equidistantly from the speaker unit, and the loud voices are collected in the same phase.
When the far-end speaker only speaks, the voice of the far-end speaker is canceled by
subtracting the output signals of the two microphone units.
When the near-end speaker speaks, the subtraction is stopped and only the signal of one
microphone unit is output. Thus, echo generation of the voice of the far-end speaker is prevented
and howling is suppressed. Also, since the signal of one of the microphone units is always output
as a transmission signal, it is possible to make a smooth call without losing the voice of the nearend speaker.
09-05-2019
5
[0016]
The invention according to claim 3 of the present application is a speakerphone that receives the
voice of the far-end speaker as a reception signal and transmits the voice of the near-end speaker
as a transmission signal to the far-end speaker. The first and second microphone units provided
at different distances from the user and picking up the voice of the near-end speaker, and the
reception signal and the output signal of the first or second microphone unit A near-end speaker
voice detection unit that detects the presence or absence of the voice of the near-end speaker,
and the first or second voice of the near-end speaker is detected by the near-end speaker voice
detection unit; Difference output means for outputting the output signal of the microphone unit
as it is, otherwise calculating and outputting the difference value between the output signals of
the first and second microphone units, and the near-end speaker voice detection When the voice
of the near-end speaker is detected by the means The line segment connecting the phase
inversion means for inverting the phase of the reception signal and outputting it otherwise and
outputting the reception signal as it is, and the first and second microphone units as the
symmetry axis; The line segment connecting the first speaker unit, which is provided on one of
the perpendicular lines of the line segment across the axis of symmetry and which amplifies the
received speech signal, and the first and second microphone units is the axis of symmetry, And a
second speaker unit disposed on the other of the line perpendicular to the axis of symmetry and
for amplifying the output signal of the phase inverting means.
[0017]
According to such a configuration, by arranging the two speaker units and the two microphone
units at equal intervals from the respective speaker units, the sound is amplified from the two
speakers to the respective microphone units. Sound is picked up in the same phase.
When the far-end speaker only speaks, the far-end speaker's voice is amplified by amplifying the
far-end speaker's voice in the same phase from both speaker units and subtracting the output
signals of the two microphone units. Play out. When the near-end speaker speaks, the far-end
speaker's voice whose phase is inverted is amplified from one of the speaker units, and only the
signal of one of the microphone units is output. In this case, the far-end speaker's voice amplified
from the two speaker units is canceled at the sound collection point of the microphone unit for
picking up the near-end speaker's voice. Therefore, at the time of double talk in which the far-end
speaker and the near-end speaker speak simultaneously, only the near-end speaker's voice can be
output as a transmission voice, and a smooth call can be performed by the further echo
suppression effect. it can.
09-05-2019
6
[0018]
According to the invention of claim 4 of the present application, in the loudspeaker telephone set
according to claim 2 or 3, the near end speaker voice detection means calculates a short time
average level of an output signal of the first microphone unit. Calculating means; second level
calculating means for calculating the short-time average level of the reception signal; dividing
means for dividing the output value of the first level calculating means by the output value of the
second level calculating means The division result of the division means is compared with a
predetermined threshold value to determine the magnitude, and when the division result is large,
it is determined that the voice of the near-end speaker is present, and when the comparison
result is small, the near-end speaker And threshold determination means for determining that
there is no voice.
[0019]
According to such a configuration, the short time average level of the signal collected by one of
the microphone units is divided by the short time average level of the reception signal, and the
near end talk is performed when the division result is larger than a predetermined threshold
value. The control signal is output as an utterance state of the person, otherwise, as an utteranceless state.
This makes it possible to detect the speech of the near-end speaker with a simple configuration.
[0020]
BEST MODE FOR CARRYING OUT THE INVENTION A loudspeaker according to each embodiment
of the present invention will be described below with reference to FIGS. In the following
embodiment, since the speech communication state is described, the handset / phoned
changeover switch provided in the telephone is omitted, and only the speech communication
state will be described.
[0021]
09-05-2019
7
(First Embodiment) FIG. 1 is a schematic view of a main part of a loudspeaker according to a first
embodiment of the present invention. This speakerphone comprises a speaker unit 11,
microphone units 12 and 13, difference calculation means 14, and frequency correction means
15. The speaker unit 11 is a speaker that amplifies the incoming call signal 16. The microphone
units 12 and 13 are disposed at equal distances d from the speaker unit 11 and are microphones
that pick up the output sound of the speaker unit 11 and the voice of the near-end speaker. The
difference calculating means 14 is a circuit that outputs a difference between the output signal of
the microphone unit 12 and the output signal of the microphone unit 13. The frequency
correction means 15 is a circuit for correcting the output signal of the difference calculation
means 14 with a predetermined frequency characteristic. Thus, the transmission signal 17 is sent
from the frequency correction means 15 to the far-end speaker.
[0022]
The frequency correction means 15 is constituted by, for example, a low pass filter, and
multiplies an input signal by a predetermined frequency characteristic and outputs the result.
The far end talker's received speech signal 16 is amplified by the speaker unit 11 to our sound
field, but since the distance d between the speaker unit 11 and the microphone units 12 and 13
is equal, the microphone units 12 and 13 have the same phase. The sound is picked up, and
when the difference calculating means 14 calculates the difference value of those signals, it
theoretically becomes 0. For this reason, it is possible to suppress that the voice of the far-end
speaker output from the microphone units 12 and 13 is mixed into the transmission signal 17.
[0023]
On the other hand, the speech of the near-end speaker is collected by the microphone units 12
and 13. However, since the distance between the speech position and the microphone units 12
and 13 is usually different, the sound is collected by the microphone units 12 and 13. Even when
the speech signal of the near-end speaker is differentiated by the difference calculating means
14, the attenuation amount is small compared to the speech of the far-end speaker. Therefore,
the output of the difference calculating means 14 can be used as the transmission signal 17.
[0024]
FIG. 2 shows a change in the output level of the difference calculation means 14 when the angle
09-05-2019
8
θ from the straight line connecting the sound sources to the microphone units 12 and 13 is
changed. Here, the distance L of the microphone units 12 and 13 is 20 cm, and the level ratio is
shown with θ = 0 ° as a reference. As shown in the figure, bi-directionality is formed by the two
microphone units 12 and 13 and the difference calculating means 14, and the sensitivity is
greatly attenuated in the 90 ° direction, ie, in the left-right direction where the speaker unit 11
is located. There is.
[0025]
On the other hand, due to phase interference, the frequency characteristics for the collected
voice of the near-end speaker in the difference calculation means 14 are as shown in FIG. 12, and
the sound quality is degraded compared to the case where the nondirectional microphone alone
receives Will occur. Particularly in the case of transmitting voice, important voice information is
included in a band of 1 kHz or less, and the attenuation of the low frequency as shown in FIG. 12
may cause a reduction in speech intelligibility. Therefore, the frequency correction characteristic
as shown in FIG. 3 is set in the frequency correction means 15, and in the output signal of the
difference calculation means 14, the frequency characteristic is corrected so as to increase the
gain on the low band side. FIG. 4 shows the frequency characteristics after correction, and it was
possible to improve the frequency characteristics of the low-frequency component of the nearend speaker's voice suppressed by phase interference. Since the frequency characteristic of FIG.
12 is determined by the distance L between the microphone units 12 and 13, the correction
frequency characteristic according to the distance is set in the frequency correction unit 15.
[0026]
Second Embodiment Next, a loudspeaker according to a second embodiment of the present
invention will be described with reference to FIGS. FIG. 5 is an explanatory view showing the
configuration of the main part of the loudspeaker according to the present embodiment. This
speakerphone comprises a speaker unit 51, microphone units 52 and 53, near end speaker voice
detection means 54, and difference calculation means 55. The speaker unit 51 is a speaker that
amplifies the incoming call signal 56. The microphone units 52 and 53 are non-directional
microphones disposed at equal distances d from the speaker unit 51 and collecting the output
sound from the speaker unit 51 and the voice of the near-end speaker.
[0027]
09-05-2019
9
The near end speaker voice detection means 54 receives the reception signal 56 and the output
signal of the microphone unit 52, detects whether or not the near end speaker is speaking, and
gives the detection result to the difference calculation means 55. It is. The difference calculating
means 55 inputs the voice signal collected by the microphone units 52, 53, and when the near
end talker speech detecting means 54 detects the speech of the near end talker, the microphone
unit 52 or the microphone unit The signal of 53 is output as the transmission signal 57 as it is. In
other cases, the difference value of the output signals of the microphone units 52 and 53 is
calculated and output.
[0028]
Since the difference value of the microphone units 52 and 53 is output by the difference
calculating unit 55 when the near end speaker does not speak, the microphone units 52 and 53
and the difference calculating unit 55 shown in FIG. Such a directivity is formed, and the voice of
the far-end talker who is loudened from the speaker unit 51 is largely attenuated, and it is
possible to suppress the mixing into the transmission signal 57.
[0029]
On the other hand, when the near-end speaker is speaking, the near-end speaker speech
detection means 54 detects the speech of the near-end speaker, and the difference operation of
the difference calculation means 55 is stopped. The output signal is output as the transmission
signal 57 as it is.
In this case, it is possible to suppress the deterioration of the frequency characteristic of the low
band part of the near-end speaker's voice due to the phase interference as shown in FIG.
[0030]
FIG. 6 is an explanatory view showing an arrangement example of a speaker unit and a
microphone unit in the case where the loudspeaker according to the first and second
embodiments is applied to an in-vehicle loudspeaker. FIG. 6 shows an example in which the
speaker unit 62 is incorporated in the central portion of the handle 61 or in the central portion
of the instrument panel, and the microphone units 63 and 64 are incorporated in the arms of the
handle 61 or left and right portions of the instrument panel. The distances between the speaker
09-05-2019
10
unit 62 and the microphone units 63 and 64 are arranged to be equal. By arranging the speaker
unit and the microphone in the in-vehicle speakerphone as described above, the same effects as
those of the first and second embodiments can be realized, and the driver's hand on the
telephone is released from the telephone. Contribute to safe driving.
[0031]
(Third Embodiment) Next, a loudspeaker according to a third embodiment of the present
invention will be described with reference to FIG. FIG. 7 is an explanatory view showing the
configuration of the main part of the speakerphone according to the present embodiment. This
speakerphone comprises speaker units 71 and 72, microphone units 73 and 74, near end
speaker voice detection means 75, difference calculation means 76, and phase inversion means
77. The speaker unit 71 is a speaker that amplifies the incoming call signal. The speaker unit 72
is a speaker that amplifies the output sound of the phase inverting means 77. The microphone
units 73 and 74 are non-directional microphones disposed at equal distances d from the speaker
units 71 and 72 and collecting the output sound from the speaker units 71 and 72 and the voice
of the near-end speaker.
[0032]
The first microphone unit 73 and the second microphone unit 74 are provided at different
distances from the near end talker. The first speaker unit 71 has a line segment connecting the
first and second microphone units 73 and 74 as an axis of symmetry, and is provided on one of
perpendicular lines of the line segment with the axis of symmetry interposed therebetween. The
second speaker unit 72 has a line segment connecting the first and second microphone units 73
and 74 as an axis of symmetry, and is provided on the other of the perpendiculars of the line
segments with the axis of symmetry interposed therebetween.
[0033]
The near end speaker voice detection means 75 is a voice detection means for detecting the
speech of the near end speaker from the received signal and the output signal of the microphone
unit 73. The difference calculation means 76 outputs the output signal of the microphone unit
73 or 74 as it is when the near-end speaker voice detection means 75 detects the speech of the
near-end speaker, and in the other cases, the difference calculation means 76 It is a circuit that
09-05-2019
11
calculates and outputs a difference value. The phase inversion means 77 is a circuit which inverts
the phase when the near end speaker voice detection means 75 detects the speech of the near
end speaker, and otherwise outputs the input signal as it is.
[0034]
When the near-end speaker does not speak, the phase inverting means 77 outputs the received
signal as it is to the speaker unit 72, so the received voice is amplified from the speaker units 71
and 72 in the same phase. Further, since the difference calculating means 76 calculates the
difference between the output signals of the microphone units 73 and 74, the directivity as
shown in FIG. 2 is formed as in the first and second embodiments, and the reception signal is
mixed in the transmission signal. Can be suppressed.
[0035]
On the other hand, when the near-end speaker speaks, the phase inverting means 77 inverts the
phase of the received signal by 180 ° and outputs it. Since the microphone units 73 and 74
have the same distance from the speaker units 71 and 72, respectively, the sound reception
signal of the positive phase amplified from the speaker unit 71 and the sound amplification from
the speaker unit 72 at the sound receiving points of the microphone units 73 and 74. The
received signals in the reverse transmission cancel each other out and theoretically become 0. On
the other hand, the distance between the receiving point of the present speaker and the speaker
units 71 and 72 is usually different, and attenuates only a little with respect to the amount of
attenuation in the microphone units 73 and 74. At this time, since the difference calculating
means 76 outputs the output of the microphone unit 73 or 74 as a transmission signal as it is, it
is possible to suppress the deterioration of the frequency characteristic of the near end speaker's
voice due to the phase interference as shown in FIG.
[0036]
(Fourth Embodiment) A specific embodiment of the near-end speaker detection means used in
the second and third embodiments will be described as a loudspeaker in the fourth embodiment
of the present invention. FIG. 8 is a block diagram of the near end speaker voice detection means.
This near-end speaker voice detection means is comprised of a first level calculation means 81, a
second level calculation means 82, a division means 83, and a threshold value judgment means
09-05-2019
12
84. The first level calculating means 81 is a level calculating means for calculating the short time
level of the output signal of the microphone unit. The second level calculating means 82 is a level
calculating means for calculating the short time level of the incoming call signal from the far end
speaker. The dividing means 83 is a circuit that divides the output of the first level calculating
means 81 by the output of the second level calculating means 82. The threshold discriminating
means 84 is a circuit which compares the output of the dividing means 83 with a predetermined
threshold and discriminates the magnitude thereof.
[0037]
An example of an output signal of the dividing means 83 at the time of a call is shown in FIG.
Since the transfer characteristic between the speaker unit and the microphone unit is steady in a
state where the near-end speaker does not speak, that is, from time t1 to t2, the microphone
output approximates the reception signal, and the short time level of the microphone output The
output of the dividing means 84 obtained by dividing by the short time level of the reception
signal becomes almost constant. On the other hand, when the near-end speaker utters, the nearend speaker's voice is largely mixed into the microphone output, and the output of the dividing
means 84 increases as shown in the near-end speaker's utterance section from time t2 to t3.
Therefore, by measuring in advance the output of the dividing means 84 in the state where the
near end talker is not generated, and setting the threshold appropriately, the near end talker can
be determined according to the magnitude of the output of the division means 83 Can detect the
presence or absence of the utterance of
[0038]
As described above, according to the loudspeaker telephone set of claim 1, two microphone units
are arranged at equal intervals from the loudspeaker unit for loudspeaker, and the output signals
of these microphone units are determined in advance, and then determined in advance. By
correcting with the frequency correction characteristic, mixing of the reception signal to the
transmission signal is prevented, and echo and howling are suppressed with a simple
configuration without using a complex configuration such as voice switch and acoustic echo
canceller. can do. Further, by the differential output of the two microphone units, it is possible to
suppress the deterioration of the near-end speaker's voice due to the phase interference, and it is
possible to make a smooth call.
[0039]
09-05-2019
13
According to the second aspect of the present invention, the two microphone units are arranged
at equal intervals from the speaker unit for the loudspeaker, and the near-end speaker is
differentiated only when the near-end speaker is not speaking The echo and howling can be
suppressed with a simple configuration when the near-end speaker does not speak without
causing deterioration of the near-end speaker's voice when the speaker speaks.
[0040]
According to the third aspect of the present invention, the microphone units are arranged at
equal intervals from the two speaker units and each speaker unit, and when the near-end speaker
does not speak, the two speakers are arranged. The reception signal can be prevented from being
mixed into the transmission signal by amplifying the reception signal in the same phase from the
unit and subtracting the outputs of the two microphone units.
When the near-end speaker speaks, near-end talk is performed by expanding the opposite-phase
received signal from one of the speakers so that the received signals are canceled at the
collection points of the two microphone units. Even when a person speaks, echo and howling can
be suppressed with a simple configuration without causing deterioration of the near-end
speaker's voice.
09-05-2019
14
Документ
Категория
Без категории
Просмотров
0
Размер файла
26 Кб
Теги
jp2000324233
1/--страниц
Пожаловаться на содержимое документа