JP3517306B2 - Telephone with automatic voice response function - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention, when an incoming call is received on a telephone, does not require manual operation on the telephone.
The present invention relates to a telephone with an automatic voice response function that makes a call by a voice response, and particularly to a telephone with an automatic voice response function that enters an incoming state only when a voice higher than a certain threshold level is input.

[0002]

2. Description of the Related Art In recent years, telephones that can make a call in a hands-free state have become widespread. Such a telephone is convenient when the user cannot release the hand because it is possible to talk with the other party without using the hand.

In this case, it is general to operate such a hands-free button to make such a hands-free call without holding a receiver. However, recently, a telephone has come to be provided which is in a call state by a voice response without operating a button of the telephone when an incoming call is received.

FIG. 8 shows an example of the configuration of the base unit side having such an incoming call reception function by voice response. When the ringer 1 detects a bell signal arriving on the line, the CPU 2 outputs a voice signal. The synthesizing circuit 3 is activated and the bell sound is output from the speaker 7. Normally, the bell signal is on for 1 second,
It is sent in a cycle of 2 seconds pause, but if voice is input from the microphone 11 within a certain time during the bell pause, the line is closed by the line closing circuit 9 composed of the transistors Tr1 and Tr2. Then, a call circuit is formed and a hands-free call state is established.

That is, an incoming call can be accepted without touching the operation buttons of the master unit, and the voice from the other party via the speech circuit 4 and the hands-free circuit 5 can be amplified by the amplifier 6 and heard from the speaker 7. Therefore, the response thereto can be amplified by the speech circuit 4 via the hands-free circuit 5 through the microphone 11 and transmitted to the other party.

When the caller goes on-hook during the hands-free call, a busy tone is sent from the line to the telephone set. When the busy tone detection circuit 10 detects the busy tone, the CPU 2 causes the line closing circuit 9 to close the line. To end the call.

FIG. 9 shows an example of the configuration applied to the slave side of the cordless telephone used with the master of FIG. 8, and the bell signal arriving on the line is the RF module 13 and antenna 14 on the master side. When the signal is taken into the antenna 20 and the RF module 21 on the handset side via the CPU 22, the CPU 22 determines the bell signal and drives the speaker 24 via the amplifier 23 to output the bell sound. When sound is input from the microphone 26 within a certain time during the bell rest period, the CPU 14 is operated from the antenna 14 and the RF module 13 on the master side via the antenna 20 and the RF module 21 on the slave side.
The signal is taken in by 2, and the line is closed by the line closing circuit 9 composed of the transistors Tr1 and Tr2 in the same manner as above on the side of the base unit, and a hands-free call state is made using the side of the handset unit. . By the way, FIG. 10 schematically shows a case where the response state can be established and a case where the response state cannot be established.

When the caller goes on-hook during the hands-free call, a busy tone is sent from the line, and the master unit detects the busy tone and requests the slave unit to end the call and automatically completes the call.

[0009]

However, in the above-mentioned telephone having the function of placing a call by the conventional voice, the threshold value for determining whether or not there is a voice response is fixed, so that ambient noise is generated. Since the threshold value must be set high in order to avoid a malfunction due to, the hands-free function cannot be operated unless a reply prompting the incoming call is made with a certain loud voice.

Further, if the surroundings are noisy and intermittent noise is generated, it may be erroneously recognized as a reply prompting an incoming call, which may cause a malfunction. That is, when the state of the room in which the telephone is placed is noisy, it may be recognized as a reply prompting an incoming call, and there is a risk that the incoming call state will occur.

The present invention has been made in consideration of such a situation, and an automatic voice response function capable of reliably operating even in a quiet place with a small voice and preventing malfunction even in a noisy environment. It is intended to provide an attached telephone.

[0012]

The invention according to claim 1 is
For phones with an automatic voice response function,
The ambient noise level is detected and detected every predetermined time.
The average value of the output noise level within a predetermined detection time is
Average value processing that is calculated while updating each time the noise level is detected
Means, and the average value updated by the average value processing means
Operation that compares with the latest detected ambient noise level
Means, and an output connection of the average value processing means and the computing means.
Threshold level that determines the threshold level based on the result
The decision means and the audio signal level of the audio frequency that prompts the incoming call
Judgment hand for judging whether or not the threshold level is exceeded
And a judgment is made after the bell signal arrives on the line.
A voice signal level of a voice frequency that prompts the incoming call by means of
When it is determined that the
It is characterized by being in a trust state.

According to a second aspect of the present invention, there is provided the above automatic voice response.
In the telephone with answer function, the arithmetic means is the average
The difference between the value and the latest detected ambient noise level
Based on the sound field determination means
The threshold level determining means is the sound field determining means.
When it is determined that the sound field fluctuation is large,
Determine the threshold level by adding a large reference value,
If the sound field judging means judges that the sound field fluctuation is small,
The threshold level is added by adding a relatively small reference value to the average value.
It is characterized by determining the rule. In addition, according to claim 3,
The invention, in the above-mentioned telephone with automatic voice response function,
The threshold level determining means is configured to detect the detected noise
When the absolute level is large,
Determine the threshold level by adding a relatively large reference value
However, the absolute level of the detected noise level is low.
In some cases, add a relatively small reference value to the average value before
It is characterized in that the threshold level is determined.

[0014]

[Action] In the automatic voice response function telephone of the present invention, electrostatic
The noise level in the surrounding area where the speaker is installed is fixed
(For example, every 1.5 seconds), the average value processing means
Therefore, the predetermined detection time at the detected noise level
The average value within (for example, 1.5 × 128 seconds) is the noise level.
It is requested while being updated every time a rule is detected. Also, the calculator
Depending on the tier, the average value updated from time to time and the latest surrounding noise
And the average value processing means and calculation are performed.
Based on the output result with the means,
Therefore, the threshold level is determined. And the judging means
The audio signal level of the audio frequency
Sound that prompts you to receive an incoming call
When the voice signal level of the voice frequency exceeds the threshold level
When the determination is made, the incoming call state is set.

Further, the calculation means is specifically required.
Difference between the average and the latest detected ambient noise level
Based on, the sound field determination means to determine the magnitude of the sound field fluctuation
This sound field determination means is provided as the threshold level determination means.
When it is judged that the sound field fluctuation is large, it is compared with the calculated average value.
Threshold level by adding a relatively large reference value (operating margin)
The sound field determination means determined that the sound field fluctuation was small.
In this case, add a relatively small reference value to the obtained average value.
To determine the threshold level. Furthermore, such a threshold
As the value level determining means, the detected noise level
When the absolute level is large, it is relatively
A large reference value is added to determine the threshold level and
If the absolute level of the generated noise level is small,
Threshold level by adding a relatively small reference value to the average value
You can also decide

Therefore, even if the surroundings are noisy and have intermittent noise, the determination is made based on the threshold value reflecting the situation of such noise in advance. It is possible to properly operate the automatic voice response function without being recognized as a voice prompting an incoming call.

[0017]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows an embodiment of a telephone with an automatic voice response function of the present invention. As shown in the figure,
The sound captured by the microphone 30 is amplified to a predetermined gain by the amplifier 31, and the bandpass filter (BPF) whose center frequency is set to 1 kHz, for example.
When the audio frequency is extracted by 32, the audio signal level of the audio frequency extracted by the A / D converter 33 is converted into a voltage.

Here, in a standby state in which no bell signal arrives (calls in) on the line, the ambient sound around the telephone is picked up by the microphone 30, and is picked up by the microphone 30. Similarly to the above, the sound has its voice band extracted by the band pass filter (BPF) 32, and is converted into the voice signal level voltage of the voice frequency extracted by the A / D converter 33.

In the standby state, the bell signal presence / absence determining circuit 34 switches the switching circuit 35 to the terminal a side, and the voice band signal extracted by the band pass filter (BPF) 32 is subjected to the ambient noise level by the average value processing circuit 36. Is averaged. At the same time, the latest voice band signal is taken into the sound field comparison circuit 37 and compared with the noise level averaged by the average value processing circuit 36, and the determination result is the threshold value determination circuit 3.
8 determines the threshold level according to the magnitude of the fluctuation of the sound field in which the telephone is placed.

When a bell signal arrives (calls in) on the line, the bell signal presence / absence determining circuit 34 switches the switching circuit 35 to the terminal b.
The voice picked up by the microphone 30 is switched to the side (including, for example, a response “Hay”, which will be described later).
Similarly to the above, when the voice band is extracted by the band pass filter (BPF) 32 and the voice signal level of the voice frequency extracted by the A / D converter 33 is converted into a voltage, chattering is absorbed by the voice determination circuit 40. Waveform shaping is performed, and the presence or absence of a voice response such as "Hi" is determined based on the threshold level according to the fluctuation of the sound field determined by the threshold determination circuit 38. Outputs a trigger signal to activate the incoming call reception function.

The operation of the average value processing circuit 36 will now be described. When the bell signal is not received, ambient noise is captured by the microphone 30, output through the bandpass filter 32, and output by the A / D converter 33.
It is input to the average value processing circuit 36 via the terminal a of the switching circuit 35. When the output voltage from the A / D converter 33 represents the a _n, the average value processing circuit 36 outputs an average value by processing the input signal as indicated by the following equation.

[Equation 1]

In this embodiment, a new noise level signal is taken into the average value processing circuit 36 every 1.5 seconds (20 ms × 75), and 1.5 × 128 = 192 seconds = 3 minutes 1
It is designed to find the average noise level over a period of 2 seconds.

Next, the operation of the sound field comparison circuit 37 will be described. The purpose of this circuit is to determine the noise characteristics of the place where the telephone is installed and reflect it in the threshold setting.

For example, the noise in the room in which the air conditioner is operating is mainly the operating noise of the air conditioner, and although the noise is generated, its level fluctuation is small. On the other hand, when the TV is used in the same room, the sound from the TV becomes a noise source, and the noise level at the moment and the moment fluctuates considerably as the program progresses.

Considering such an example, when the noise level or the sound field fluctuates largely, the threshold value needs to have a margin more than that when the fluctuation fluctuates to prevent malfunction.

In the present embodiment, the average value processing circuit 36
Constantly takes in a new noise level voltage value, obtains a new average value, and then compares the noise level voltage value of the latest voice band signal with this average value, ± 0.04 of the average value.
When it is not within the width of V (± 3 dB in the case of microphone input), it is a sound field with large fluctuation, and within ± 0.04 V, it is determined to be a sound field with small fluctuation.

Considering that the situation of the sound field is constantly changing and the risk of erroneous determination, the threshold value is increased immediately when it is determined that the variation is large, and then the variation is changed. When the determination that the value is small is repeated 10 times, the threshold value that has been once increased may be returned to the original value.

Next, a method of determining the threshold value of the threshold value determining circuit 38 will be described. Previously, it was mentioned that the sound field comparison circuit 37 determines whether the fluctuation of the sound field is large or small and uses this to determine the threshold value.

As a result of experiments, we have found that another factor should be reflected in threshold determination. It is the magnitude of the absolute level of noise.

That is, the magnitude of the threshold value determined by the threshold value determination circuit 38 is determined by adding a certain operating margin to the average value obtained by the average value processing circuit 36 to determine the threshold value. The basic idea is that when the noise level becomes high, it is better to slightly increase the margin width to be added to the average value in order to reduce the risk of malfunction.

Also, due to human habits, when the surroundings become noisy, the voice uttered naturally becomes louder. Therefore, it seems that there is no discomfort even if the threshold value is raised practically.

Therefore, in this embodiment, the margin width is changed assuming that the noise level is large or small with the noise level of 70 dB as the boundary.
Also combining the determination results of the sound field comparison circuit 37 described above,
The threshold is determined as follows.

When the noise level is less than 70 dB (equivalent to standard microphone output-62.5 dB) in the sound level meter and the fluctuation of the sound field is small, threshold value = average value + 0.14 V (6 dB in terms of microphone input level). When the noise level is less than 70 dB in the sound level meter and the fluctuation is large, threshold = average value + 0.40 V (13 dB in terms of microphone input level) When the noise level is 70 dB or higher in the sound level meter and the fluctuation is small , Threshold value = average value + 0.20 V (8 dB in terms of microphone input level) When the noise level is 70 dB or more in the sound level meter and fluctuation is large, threshold value = average value + 0.60 V (at microphone input level) Speaking of 20dB)

Furthermore, the voice determination operation in the voice determination circuit 40 is performed as follows.

That is, as shown in FIG. 2 (a), when a signal waveform "Hai" is taken in during a bell rest period between the bell sound and the bell sound, a half wave is generated as shown in FIG. 2 (b). After being rectified, the voltage is converted as shown in FIG.
However, it is obtained as the output of the A / D converter 33, and the threshold value shown in FIG. 7D is determined by the threshold value determination circuit 38.

Then, in FIG.
It is assumed that the signal level is read at an interval of 0 ms (sampling period) and, for example, the next data string shown on the left side of is obtained. Here, “H” is a high level when the threshold is exceeded, and “L” is a low level below the threshold. H H L L H H → (chattering absorption) H
H H H H H H H H L L L H → (chattering absorption)
H H H H H L H In both of the above cases, chattering is absorbed by the voice determination circuit 40 and converted into the right data string.

That is, when absorbing chattering,
In both of the above, L and L in the data string on the left side are ignored up to twice, and if they continue three times or more, they are recognized as "L".

As a result, if all the data strings are obtained as "H" as in the above right data string, it is determined to be a voice response. The chattering absorption operation will be described in more detail later.

Further, the timing of automatic response is as follows.
This is as shown in FIGS. 3 and 4. That is, an input signal that exceeds a threshold during a bell pause between the bell and the bell,
For example, in the case of "Hi", the time from the moment when the threshold value is exceeded to the moment when it falls below the threshold value is measured, and it is checked whether the time t2 is between 123 ms and 760 ms. If the duration is within this time t2, it is an incoming call.

In other words, even if there is an input signal exceeding the threshold value between the times t1 and t3, if the duration of the input signal does not satisfy the condition of 123 ms to 760 ms,
The "Hi" during that time will be ignored.

That is, as shown in FIG. 4, even when the magnitude of the input signal falls below the threshold level for a short time on the way, as described above, H H L L H H → (chattering absorption) H H
The duration of all “H” s on the right side of H H H H is 123 ms to 760
If it is within the ms time, the incoming state is established. H H H L L L H → (chattering absorption) H
In the case of H H H H L H, if the duration from the right data string “H” to “L” is within the time of 123 ms to 760 ms, it becomes the incoming state and the duration until it becomes “L”. Is 12
If the time is 3 ms or less or 760 ms or more, no call is received.

That is, in the extended portion (long sound) between the words "Hi", the sound pressure decreases, and the sound pressure tends to fall below the threshold value. Therefore, the voice determination circuit 40 is set so as not to determine that it is a voice response unless a signal exceeding a continuous threshold value is input.
The chattering absorption described above facilitates the determination of the presence or absence of a voice response.

Next, the operation of the telephone having the automatic voice response function having such a configuration will be described with reference to FIGS.

First, the CPU is initialized and the time T1 for voice processing of the CPU is set to 20 ms (steps 501 and 502). If A1 is set in (step 503) and A1 is not 75 in (step 504) in order to average the ambient noise and compare the sound fields, that is, 20 ms × 75 = 1.5 seconds If has not elapsed, move to (Step 509),
If the call signal is not input, (step 50
A1 is increased by 1 in 5).

When A1 becomes 75 and time is up,
That is, when 20 ms × 75 = 1.5 seconds has elapsed, the A / D input is averaged and the sound field is compared to determine the threshold value (steps 503 to 509).

On the other hand, in step 509, when the call signal is input, the bell ringing is started, and ringing and rest are repeated. During the bell ringing pause period, B1 for setting t1 in FIG. 3 is set to "0", and the A / D input threshold is reached until B1 reaches 5 times (t1 = 20 ms × 5). If the value is not exceeded and the bell signal is not sounded again, the next step can be proceeded to (steps 510 to 516). When the A / D input exceeds the threshold value or a bell signal comes in midway, the process returns to (step 509) and a series of operations is repeated.

If it is determined in step 517 that the A / D input is greater than the threshold value, the "Hi" take-in time t2 is started (step 51).
In step 9), the value of c1 is set to "0" which is the initial value of the number of L (steps 518 and 519).

Here, c1 is a coefficient provided for absorbing chattering. Chattering means that a signal level exceeding the threshold value must be input from the A / D during t2 in FIG. However, when the word “Hi” is A / D-converted and taken in, the sound pressure of the extended portion (long sound) of the word “Hi” tends to decrease, and it tends to fall below the threshold value. By inputting this lower data, the response is actually ignored. However, it is chattering absorption that makes it easier to respond by compensating for this lower limit.

In FIG. 2d, it is assumed that the result of the data string shown on the left such as the following is obtained by reading at intervals of 20 ms (sampling period). Here, "H" is a high level when the threshold is exceeded, and "L" is a low level when the threshold is exceeded. H H L L H H → (chattering absorption)
H H H H H H H H H L L L H → (chattering absorption)
H H H H H L H In the above, chattering is absorbed by the voice determination circuit 40 of FIG. 1 and converted into the data string on the right side. That is, when chattering is absorbed, L in the data string shown on the left side is ignored up to twice and is recognized as L if it continues three times or more.

As a result, L becomes 2 as shown on the right side.
By obtaining all as "H" even after repeated times, it is possible to dramatically determine whether or not there is a voice response. If the above description is expressed in steps, t2 starts and (c5 = 0) is set in (step 519).

Next, it is determined whether or not the A / D input is larger than the threshold value, and when it is determined that it is smaller, c1 = c1 +1 is set by (step 512) (that is, L of chattering c1 is L). Will be added by one). In fact, c1 = 1. If c1 is not 3, the process returns to the threshold value judgment step (step 520).

When it is judged that the value of the A / D input re-acquired is small, c1 = is further determined by (step 521).
c1 + 1. Therefore, c1 = 2. Since c1 = 3 is not satisfied in (step 522), the threshold value is determined again in (step 520). next,
If "H" of A / D input is taken in (Step 5
The determination in 20) is YES, and the process returns to (step 519), so that c1 = 0 is cleared. Thus
"L" is ignored up to twice (steps 519 to 52).
2).

Next, when "L" continues three times, that is, when c1 = 3, the time t2 is 123 ms≤t2≤76.
Only when it is within the range of 0 ms, it is possible to shift to the next step (steps 520 to 523). If the time t2 is not within the range of 123 ms ≤ t2 ≤ 760 ms, or if a bell signal is input midway, (step 509).
Then, the series of operations is repeated.

At step 524, D1 for setting t3 in FIG. 3 is set to "0". Only when there is no A / D input exceeding the threshold value and there is no bell ringing in the middle, D1 is incremented by 1 in (step 528).
Only when D1 = 5, that is, t3 = 20 ms × 5 = 100 ms, there is no bell ringing and there is no A / D input exceeding the threshold value, the next step can be proceeded (steps 524 to 528).

The A / D input exceeds the threshold value,
When the bell signal comes in on the way, the process returns to (step 509) to repeat a series of operations.

As described above, when the voice signal "Hi" which satisfies the times t1, t2 and t3 is taken in, the call is received and the call state is set (steps 529 and 530).

FIG. 7 shows another embodiment in which the configuration of the telephone with automatic voice response function shown in FIG. 1 is changed.
The difference from FIG. 1 is that the voice level can be detected even during the period when the bell sound is present.

That is, when an incoming call is generated on the line and a ring signal is generated by the ring circuit 42 in response thereto,
Predetermined level (normal, large, medium, small 3 by amplifier 43
It is amplified to the stage 44) and then output from the speaker 44. At this time, the phase correction circuit 41 adds an opposite phase signal to the bell signal output from the ring circuit 42, whereby the amplifier 31
Since the bell signal output to the above is canceled, the bell signal from the amplifier 31 is not output.

Therefore, in this case, the bell signal output is not included in the output of the amplifier 31 even when an incoming call occurs on the line and the bell is ringing, so that the above-mentioned "hi" is generated. Since the response timing is not limited to the bell rest period, the responsiveness is further improved and the operability can be improved.

As described above, in each of the embodiments, the ambient noise in which the telephone is installed is constantly detected by the A / D converter 33 which is the noise level detecting means, and the average which is the threshold level determining means. The value processing circuit 36, the sound field comparison circuit 37, and the threshold value determination circuit 38 determine the threshold level according to the noise level, and after the bell signal arrives on the line, the voice determination circuit 40 prompts an incoming call. When it is determined whether the voice signal level of the frequency exceeds the threshold level and the voice signal level of the voice frequency that prompts an incoming call exceeds the threshold level, the incoming state is set.

Therefore, only when the audio signal level of the audio frequency for prompting the incoming call exceeds the threshold level, the incoming state is set, and the case where the surroundings are noisy and intermittent noise is included. However, since the threshold value reflects the situation in advance, noise will not be recognized as a voice prompting an incoming call, and the automatic voice response function can be operated properly, so the hands The malfunction of the free function is prevented.

Further, the threshold level is divided into two cases, that is, a case in which the ambient noise is large and a case in which the ambient noise is small, and is set in four ways, that is, the noise variation is large or small in each state. Can be recognized more accurately.

In this embodiment, the case where the automatic response is performed when the voice signal level of the voice frequency exceeds the threshold value has been described, but the present invention is not limited to this example, and the voice signal level of the voice frequency is detected. In this case, since the frequency spectrum is also detected, it is possible to accurately identify whether or not the voice is a real voice (for example, a human voice from a television), and thus it is possible to more reliably prevent malfunction. Become. Furthermore, in order to avoid erroneous recognition, voice detection after ringing a bell is performed by detecting the frequency spectrum, so that only the voice of a registered person or only a specific word is accepted. Well, even in this case, it is possible to more reliably prevent malfunction by preventing incoming calls unless the threshold level is exceeded.

[0064]

As described above, according to the telephone with automatic voice response function of the invention, the telephone is installed.
The noise level of the surroundings is every predetermined time (for example, 1.5 seconds
Every) and detected by the average value processing means
A predetermined detection time at the noise level (for example, 1.5 ×
The average value within 128 seconds is updated every time the noise level is detected.
It is required while being asked. In addition, the calculation means can be used at any time.
Compare the new average value with the latest ambient noise level
The output result of this average value processing means and calculation means is
Based on the threshold level determination means, the threshold level
Is decided. Then, the judging means prompts for an incoming call.
The audio signal level of the audio frequency exceeds the threshold level
It is determined whether or not the audio signal of the audio frequency that prompts the incoming call
When it is determined that the level exceeds the threshold level,
Believed.

Further, the calculation means is specifically obtained.
Difference between the average and the latest detected ambient noise level
Based on, the sound field determination means to determine the magnitude of the sound field fluctuation
This sound field determination means is provided as the threshold level determination means.
When it is judged that the sound field fluctuation is large, it is compared with the calculated average value.
Threshold level by adding a relatively large reference value (operating margin)
The sound field determination means determined that the sound field fluctuation was small.
In this case, add a relatively small reference value to the obtained average value.
To determine the threshold level. Furthermore, such a threshold
As the value level determining means, the detected noise level
When the absolute level is large, it is relatively
A large reference value is added to determine the threshold level and
If the absolute level of the generated noise level is small,
Threshold level by adding a relatively small reference value to the average value
You can also decide

Therefore, only when the frequency level of the voice prompting the incoming call exceeds the threshold level, the incoming state is set, and even when the surroundings are noisy and intermittent noise is included. The automatic voice response function can be properly operated without being recognized as a voice prompting an incoming call, and a malfunction of the hands-free function can be prevented.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a telephone with an automatic voice response function of the present invention.

FIG. 2 is a waveform diagram for explaining the operation of the telephone with an automatic voice response function of FIG.

FIG. 3 is a waveform diagram for explaining the operation of the telephone with automatic voice response function of FIG.

4 is a waveform diagram for explaining the operation of the telephone with the automatic voice response function of FIG.

5 is a flowchart for explaining the operation of the telephone with the automatic voice response function of FIG.

6 is a flow chart for explaining the operation of the telephone with automatic voice response function of FIG.

FIG. 7 is a block diagram showing another embodiment when the configuration of the telephone with automatic voice response function of FIG. 1 is changed.

FIG. 8 is a block diagram showing a base unit side of a conventional telephone with an automatic voice response function.

FIG. 9 is a block diagram showing a handset side of a conventional telephone with an automatic voice response function.

FIG. 10 is a diagram for explaining the operation of the telephone with an automatic voice response function of FIG.

[Explanation of symbols]

30 microphone 31 amplifier 32 band pass filter (BPF) 33 A / D converter 34 Bell signal presence / absence determination circuit 35 Switching circuit 36 Average value processing circuit 37 Sound field comparison circuit 38 Threshold value determination circuit 39 Wave shaping circuit 40 Voice judgment circuit 41 Phase correction circuit

Continuation of front page (56) References JP-A-4-220849 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04M 1/00 H04M 1/24-1/253 H04M 1 / 58-1/62 H04M 1/66-1/82

Claims (3)

(57) [Claims] 1. A noise level around a telephone is installed.
The bell is detected every predetermined time, and the detected noise level
The average value within the predetermined detection time of
The average value processing means obtained while updating, and the average value and the latest detection value updated by the average value processing means.
A calculation means for comparing the emitted noise level of the surroundings, a threshold level determination means for determining a threshold level based on the output results of the average value processing means and the calculation means, and a voice for prompting an incoming call. A determination means for determining whether or not the voice signal level of the frequency exceeds the threshold level, the voice signal level of the voice frequency prompting the incoming call by the determination means after the arrival of a bell signal on the line A telephone with an automatic voice response function, which is set to an incoming call state when it is determined that the threshold level is exceeded. 2. The calculation means is configured to calculate the average value and the latest value.
The sound based on the difference from the detected ambient noise level of
The sound field determining means for determining the magnitude of the field fluctuation is provided, and the threshold level determining means is
When it is judged that the field fluctuation is large, the average value is relatively large.
A threshold value to determine the threshold level.
If the field determination means determines that the sound field fluctuation is small, the average
Add a relatively small reference value to the threshold level
The automatic sound according to claim 1, wherein the automatic sound is determined.
Telephone with voice response function. 3. The threshold level determining means is a detection means.
If the absolute level of the noise level is large,
The threshold value is obtained by adding a relatively large reference value to the average value.
The level is determined and the absolute level of the detected noise level is determined.
If the bell is small, the average value is relatively small.
A value is added to determine the threshold level.
With the automatic voice response function according to claim 1 or 2.
Telephone.