JPH0738661B2 - Automatic voice return telephone - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention relates to a telephone equipped with an automatic return function means for the purpose of fighting back unwanted calls.

[Conventional technology]

As one of the methods of repelling unwanted calls, there is a method of temporarily recording the voice of the other party who has called the unwanted call and sending it back to the other party as it is.

In order to realize this function, there is a telephone provided with a voice recording medium of an IC recording system and a recording / playback key which have a short access time. With this telephone, the telephone user who recognizes that it is a nuisance call records the voice of the other party by pressing the record key, and then presses the play key when the other party has finished talking, and then presses the recorded party. The method of playing back and returning the sound is adopted.

On the other hand, Japanese Patent Application No. 62-50292 (Japanese Unexamined Patent Publication No.
3-217860), which has a line signal level detection circuit that detects the presence or absence of voice, so that it automatically detects when the other party has finished talking and records / records.
It is a telephone that can switch the playback timing. With this telephone, when the owner recognizes that the answered call is a nuisance call, the owner need only press the nuisance call repulsion key.

FIG. 2 shows a nuisance call repulsion operation flow in this telephone. When the voice detection result of the line signal level detection circuit becomes sound, the recording operation is started. Then, when the voice detection result becomes silent and the silent state continues for the utterance end detection time Te or longer, the recording state is switched to the reproduction state, and the content recorded until then is reproduced.

When the reproduction is finished, when the voice detection result becomes voiced again, the recording is started, and the above operation is repeated thereafter.

[Problems to be Solved by the Invention]

The utterance end detection time Te corresponds to the time required from the end of utterance by the other party to the start of reproduction by the telephone. Therefore, if the utterance end detection time Te is set too long, the utterance end detection time Te is extended and the nuisance call repulsion effect is reduced. Therefore, the utterance end detection time Te is often set relatively short.

On the other hand, it is possible that the other party ends the call during the nuisance call repulsion operation. In this case, it is necessary for the telephone to open the line as soon as possible. When the other party of the call ends the call, a lockout tone is sent from the exchange. To detect the lockout tone, the length of the basic signal transmission section of the lockout tone (hereinafter referred to as ON section),
It is based on information such as the frequency of the basic signal and the length of the signal rest period (hereinafter referred to as the OFF period).

Here, if the utterance end detection time Te is set to be shorter than the pause time of the lockout tone, the automatic voice return function works for the lockout tone. That is, the lockout tone basic signal recorded in the lockout tone ON section is reproduced in the lockout tone OFF section. For this reason, not only can the OFF section length not be measured accurately, but when the lockout tone ON section starts next time, the playback signal from the telephone and the signal from the exchange will be superposed, so the ON section length will also increase. It cannot be measured accurately. That is, it is impossible to detect the lockout tone.

Therefore, the following method is used to detect the lockout tone during the automatic voice return operation.

Speaking end detection time Te is locked out Tone pause time Toff
Longer than. That is, the utterance end detection time Te is set to satisfy Te> Toff (1).

The standard value of the lockout tone pause time Toff is 0.5 seconds. Therefore, if the utterance end detection time Te is set so as to satisfy the expression (1), there is a delay between the end of utterance by the other party and the return of the voice, which reduces the nuisance call repelling effect. .

[Means for Solving the Problems]

A zero-crossing counter is provided to detect the fundamental frequency of the input signal.

Normally, the utterance end detection time Te is set to a value short enough to obtain the effect of repelling nuisance calls, and when recording is being performed when it is determined to be voiced, the signal fundamental frequency is detected by the zero-crossing counter. I do.

When the voiced section ends and transitions to the silent state, the duration of the voiced section and the level / fundamental frequency of the signal that constitutes the voiced section are checked, and these values are locked. If there is a strong possibility that the out tone is on,
Set the utterance end detection time Te longer than the lockout tone OFF time, and set it to a value that allows lockout tone detection. In order to do this, the automatic voice return telephone has a voice recording medium, the function of recording the line signal on the voice recording medium by the line signal recording means, and the voice recorded on the voice recording medium by the voice reproducing means. The line signal level measuring means for detecting the presence / absence of voice depending on the level of the line signal level is provided, and the line signal is recorded on the voice recording medium when the line signal is detected as sound. , After the detection result transitions from voiced to silence, when the silent state continues for a predetermined utterance end detection time or longer,
An automatic voice-replying telephone that replays and sends out the voice recorded on the voice recording medium to the line, and repeats the above operation from recording after the end of the replay, has a zero-crossing counter to detect the signal fundamental frequency in the voiced section. , Normally, a utterance end detection time that is shorter than the signal pause time of the lockout tone is set, and it is determined that the recorded voiced section corresponds to the ON period signal of the lockout tone based on its duration and signal fundamental frequency. At times, an automatic voice-replying telephone having a control unit for setting the detection end time of utterance to be longer than the pause time of the lockout tone was constructed.

[Work]

As described above, the present invention can be placed on a telephone equipped with an automatic return function for the purpose of repulsing a nuisance call, and can recognize that the other party has ended the call without impairing the nuisance call repulsion effect. It is possible to shorten the time for which the line is invalidly captured after the call ends.

〔Example〕

FIG. 1 shows a block diagram of an embodiment of the present invention. 1 is a ringer that rings when a ringing signal arrives; 2 is a loop opening / closing part for opening / closing the line; 3 is a hook switch connected in parallel with the loop opening / closing part for opening / closing the line depending on an off-hook / on-hook state of a handset; 4 is a communication circuit ,
5 is a handset, 6 is a handset that is installed in the handset and is connected to the receiving terminal (R) of the communication circuit, and 7 is a transmission terminal (T) of the communication circuit that is installed in the handset
To the receiver, 8 is a line signal recording unit connected to the reception terminal (R) of the communication circuit, 9 is a voice recording medium on which voice is recorded by the line signal recording unit, and 10 is voice in the voice recording medium. A voice reproduction unit for reproducing and transmitting to the line from the transmission terminal (T) of the communication circuit, 11 is a line signal level detecting unit connected to the line and detecting the level of the line signal, 12 is connected to the line and zero crossing of the line signal Zero-crossing counter to measure the number, 13 is a nuisance call repulsion key, 14 is a control unit, 15 is a timer A for measuring the elapsed time from the start of the nuisance call repulsion process, 16 is a voice recording medium in the nuisance call repulsion process A timer B for measuring the elapsed time from the start of the recording to 17 and a timer C for measuring the duration of the voiced / non-voiced sound of the other party's voice in the nuisance call repulsion processing. The timers A to C start counting according to a reset instruction from the control unit 14, and the count value can be read out to the control unit 14.

Further, FIG. 3 shows a specific configuration of the line signal level measuring circuit and an interface example with the control unit, and FIG. 4 shows a specific configuration of the zero-crossing counter and an interface example with the control unit. In FIG. 3 and FIG. 4, the control unit 14
Data is read from the A / D converter 18 and the UP counter 19. Regarding the UP counter 19, the count value is reset immediately after reading the data.

The operation of the configuration shown in FIG. 1 will be described below.

When a ringing signal arrives on the line, the ringer 1 sounds. When the telephone user off-hooks the handset 5, the hook switch 3 is closed and current is supplied to the call circuit 4, so that the transmitter 7 becomes valid and the line signal from the user receiver 6 can be heard. it can. Here, when the user recognizes that the incoming call is a nuisance call, the nuisance call repulsion key 13 is pressed. At this time, the loop opening / closing unit 2 that has been open until then is closed, and line supplementation continues even if the user on-hooks the handset.

5 (a), (b), and (c) are flowcharts for explaining the nuisance call repulsion processing procedure after the loop opening / closing unit is closed. FIG. 5 (a) shows the flow of the entire processing,
(C) illustrates the details of the silence determination process and the recording process included in the overall process. Hereinafter, FIG. 5 (a),
The operation will be described in accordance with (b) and (c).

First, the timer A15 is reset and started. The timer A is a timer that counts the time after the nuisance call repulsion process is started, and is necessary for automatically ending the process when the timeout time Tprc has elapsed since the process was started.

Next, a silent decision process is performed. This process is for determining whether or not the other party starts utterance, and forcibly disconnects the line when the other party remains silent for Tmugon time and does not start utterance. Note that the voiced / non-voiced determination for determining whether the other party speaks and is in a voiced state, or whether or not the other party is in a voiced state without vocalizing is performed as follows. The read value Vin from the line signal level detection unit 11 and the voice detection level threshold Vth stored in the control unit, and the read value Zin from the zero crossing counter 12 and the voice detection zero crossing number threshold stored in the control unit. The Zths are compared with each other, and when Vin exceeds Vth, or when Zin exceeds Zth, the input signal is considered as sound, and the other states are considered as silence.

Note that the voice detection level threshold Vth needs to be set to a sufficiently large value so that ambient noise, line noise, and the like of the communication partner are not erroneously detected as voice. When Vth is set to such a value, a consonant having a low level such as a fricative sound is not detected as a sound, so it is detected based on the zero-crossing number information. The fricative has a high fundamental frequency, so Zth
Needs to be set to a sufficiently large value so that ambient noise and line noise are not erroneously detected as voice.

Note that this silence determination process is performed when the other party terminates the call and does not immediately send a lockout tone, but when the present embodiment is connected to an exchange that is in a silent state. It also has the effect of shortening the time.

When the utterance of the other party is confirmed, the timer B is reset and then the recording process is started. When recording is started, the line signal recording unit 8 stores the line signal in the voice recording medium 9. The timer B monitors whether or not the voice recording medium 9 is full during the recording process. When the voice recording medium is full, the recording is finished and the reproduction process is started without waiting for the other party to finish the vocalization. This is the timer needed to start.

When the recording process is started, the timer C is reset. The timer C is for measuring the length of the voiced period or the silent period of the other party's voice. However, since the other party's voice is always in the voiced state immediately after the recording process is started, the voiced period length is measured.

During the recording process, the count value of the timer B is compared with the voice time length Tfull that can be continuously recorded on the voice recording medium 9, and when the count value of the timer B becomes Tfull or more, the voice recording medium 9 is Since it is full, the recording operation is stopped and the reproduction is started.

Note that Tfull must be longer than the lockout tone ON time. This is because if the recording time is shorter than the ON time of the lockout tone, the audio recording medium will become full when the lockout tone arrives, and playback will start, and the lockout tone cannot be detected. Is.

Even after the recording is started, the control unit continuously determines whether or not the other party is speaking by the sound / silence determination. While the other party is speaking, that is, while the voiced / non-voiced determination result is determined to be "voiced", the read value Vin from the line signal level detection unit 11 is compared with the level thresholds V1 and V2. . V1, V
2 shows that the level of the lockout tone ON time measured by the line signal level detector is V in consideration of the switchout lockout tone transmission level and variations in subscriber line loss.
Set it so that it can be inserted between 1 and V2.

Since the lockout tone level is generally higher than that of the other party's voice transmitted via a relay system, V1 is higher between V1 and Vth.

When the read value Vin from the line signal level measuring unit satisfies V1 ≦ Vin ≦ V2 (2), the control unit sets the zero-crossing counter 12
Compare the read value Zin from the basic signal number of the lockout tone. Since the signal fundamental frequency of the lockout tone varies depending on the exchange, the fundamental signal frequency of the lockout tone measured by the zero-crossing counter is between the zero-crossing number thresholds Z1 and Z2 in consideration of this variation. Set Z1 and Z2 in advance. When the expression (2) is satisfied and Z1 ≦ Zin ≦ Z2 (3) is satisfied, it can be determined that the lockout tone ON period signal has arrived. If either (2) or (3) does not hold, the timer C is reset. As a result, the count value of the timer C becomes equal to the length of time during which the signal judged to be the ON signal of the lockout tone has continuously arrived.

When the result of the voiced / non-voiced determination changes from "voiced" to "voiceless", the count value of the timer C is substituted into a variable called Ton. At this time, Ton corresponds to the duration of the lockout tone ON signal immediately before changing to the silent state.

After the count value of the timer C is assigned to Ton, the timer C is reset and the measurement of the silent period length is started.

After that, Ton is compared with the time thresholds Ton1 and Ton2. Ton
1, Ton2 is ON for the lockout tone sent by the exchange
The minimum and maximum values of the standard duration of the period. When it is confirmed that Ton1 ≦ Ton ≦ Ton2 (4) holds for the measured Ton, it is determined that the immediately preceding sound period is the ON period of the lockout tone. on the other hand,
When (4) does not hold, it can be determined that the lockout tone has not arrived.

For Ton, the control unit sets the utterance end detection time Te to a constant Te2 when (4) is satisfied, and sets Te to a constant Te1 when Expression (4) is not satisfied.

Here, TE1 is a sufficiently small value so that the time to start playing the voice just recorded is sufficiently short and the nuisance call repelling effect is large. Generally, the value is shorter than the pause time Toff of the lockout tone.

On the other hand, Te2 is a value sufficiently longer than the pause time Toff of the lockout tone. That is, Te2 is a value that satisfies Te2> Toff (5).

In the silent state, the count value of the timer C, that is, the duration of the silent state of the opponent is compared with the utterance end detection time Te. Then, when the duration of the other party's silent state is Te or more, it is determined that the other party's utterance has ended, the recording process is terminated, and the recorded content of the audio recording medium 9 is reproduced by the audio reproduction unit 10, and the communication circuit 4 to the line.

Before the count value of timer C reaches the set value of Te,
When the result of the voice / silence determination changes from “silence” to “voice”, the count value of the timer C is assigned to a variable called Toff, and then the timer C is reset to measure the duration of the voice state. To start.

When the lockout tone arrives, the utterance end detection time Te is longer than the lockout tone pause time Toff.
Since it is set to Te2, the recording process is not ended and the playback is not started during the pause period of the lockout tone.

The variable Toff is compared with the time thresholds Toff1 and Toff2. Toff
1 and Toff2 are the minimum and maximum values in the standard of the pause period duration of the lockout tone sent by the exchange. now,
When it is confirmed that Toff1 ≤ Toff ≤ Toff2 (6) holds for the measured Ton, the immediately preceding silent period is determined to be the lockout tone pause period. When the end-of-speech detection time Te is set to Te2 and the expression (6) is satisfied, the lockout tone has three types of the ON period duration, the signal fundamental frequency during the ON period, and the pause period length. Now that all the information has been confirmed, it can be determined with high probability that the lockout tone has arrived. Therefore, at this time, the loop opening / closing part is opened, the recording is ended, and the nuisance call repulsion operation is ended.

Further, when the expression (6) is not satisfied and the pause time length does not match the condition of the lockout tone, or the utterance end time Te is not set to Te2, that is, the ON period signal indicates the lockout tone. If the condition is not satisfied, the process for the voiced state is performed again.

After the silent state length exceeds the utterance end time Te and transitions to the playback operation, or after the audio recording medium 9 is full and the playback operation is performed, the total processing time satisfies the timeout condition or the lockout tone. The above process is repeated until is detected.

〔The invention's effect〕

As described above, according to the present invention of claim 1, in a telephone having an automatic return function for the purpose of repulsing a nuisance call, it is possible to detect that the other party has ended the call without impairing the effect of repulsing the nuisance call. It will be possible. Therefore, there is an effect that it is possible to shorten the line supplement invalid time after the other party ends the call without lowering the repelling effect.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an embodiment of the present invention, FIG. 2 is a flowchart of a conventional nuisance call repulsion operation of a telephone having an automatic voice function, and FIG. 3 is a configuration and control unit of a line signal level measuring circuit. Figure showing the interface example with
FIG. 5 is a diagram showing a configuration of a zero-crossing counter and an example of an interface with a control unit. FIG. 5 (a) is a flowchart a of an embodiment of the present invention, and FIG. 5 (b) is a flowchart of an embodiment of the present invention. FIG. 5B and FIG. 5C are flowcharts c of one embodiment of the present invention. 1 ... Ringer, 2 ... Loop opening / closing part, 3 ... Hook switch, 4 ... Communication circuit, 5 ... Handset, 6 ... Handset, 7 ... Transmitter, 8 ... Line signal recording part, 9 …… Voice recording medium, 10 …… Voice reproduction unit, 11 …… Line signal level detection unit, 12 …… Zero-crossing counter, 13 …… Troubleshooting key, 14 …… Control unit, 15 …… Troubleshooting process Timer A, 16 for measuring the elapsed time from the start of ……, Timer B, 17 for measuring the elapsed time since recording on the voice recording medium was started in the nuisance call repulsion processing, …… Timer C, 18 …… A / D converter, 19 …… UP for measuring the presence / absence of the other party's voice in processing
counter.

Claims (1)

[Claims] 1. A voice recording medium, a function of recording a line signal on a voice recording medium by a line signal recording means, and a function of transmitting a voice recorded on the voice recording medium to a line by a voice reproducing means. At the same time, it has a line signal level detection means for detecting the presence or absence of voice depending on the level of the line signal level. When the line signal is detected as sound, the line signal is recorded on a voice recording medium, and the detection result is from sound to silence. When the silent state continues for more than the predetermined occurrence end detection time after transition to, the sound recorded on the sound recording medium is played back and sent to the line, and after the end of playback, the above operation is repeated from recording The telephone has a zero-crossing counter to detect the signal fundamental frequency in the voiced section, and usually sets the occurrence end detection time shorter than the signal pause time of the lockout tone, and records. When it is determined that the voiced section that has been generated corresponds to the ON period signal of the lockout tone based on its duration and signal fundamental frequency, it shall have a control unit that makes the vocalization end detection time longer than the pause time of the lockout tone. An automatic voice-replying telephone that features.