JP5709973B2 - Noise removal device - Google Patents

Description

  The present invention relates to a hands-free control device.

  Conventionally, among audio devices, there is an audio reproduction device that determines a gain according to ambient noise. In such an audio reproducing apparatus, the frequency of noise is specified, a gain is obtained in consideration of the masking effect that is a characteristic of human hearing, and the frequency characteristics are corrected (see Patent Document 1, etc.).

Japanese Patent Laid-Open No. 8-70228

  In the noise removal method for correcting the frequency characteristics, the sound quality tends to be deteriorated by strongly removing noise. In the above technique, noise is excessively removed, and unnecessary sound quality may be deteriorated. The same applies to a hands-free control device that can make a call using a mobile phone or the like.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to suppress noise while suppressing deterioration of call sound quality in a hands-free control device.

In order to solve the above-described problems, a noise removal apparatus according to the present invention includes a microphone that inputs voice information, a noise removal unit that removes noise from the voice information input by the microphone, and a noise that has been removed by the noise removal unit. A noise removal apparatus comprising: a transmission unit configured to transmit the voice information to the mobile phone, wherein the first strength or the first strength is used to remove noise during a period in which the voice information is input by the microphone. Intensity selection means for selecting a second intensity lower than the intensity of the sound, and the intensity selection means removes the noise when the time elapsed since the start of voice information input by the microphone is the first time When the first intensity is selected as the intensity to be performed and the time elapsed since the start of the input of the voice information by the microphone is the second time exceeding the first time As the intensity of removing the noise and select the second intensity, the noise removal means is selected by the intensity selecting means, based on the first intensity or a second intensity input by the microphone It is characterized by removing noise from the voice information.

  According to the present invention, in the hands-free control device, it is possible to suppress noise while suppressing deterioration in call sound quality.

It is a figure which shows the outline | summary of a structure of a hands-free control system. It is a graph which shows a time-dependent change of the audibility level of noise. It is a figure which shows the characteristic of the noise removal of a hands-free control system. It is a flowchart of the noise removal process of a hands-free control system. It is a figure which shows the relationship between the call quality by a noise removal process, and an auditory level. It is a figure which shows the outline | summary of a structure of the hands-free control system by 2nd embodiment.

An in-vehicle hands-free control system 90 to which an embodiment of the present invention is applied will be described below with reference to FIGS. 1 to 5 do not show all the configurations of the in-vehicle hands-free control system 90, and for the sake of easy understanding, a part of the configuration is appropriately omitted.

  FIG. 1 is a diagram illustrating an overall configuration of an in-vehicle hands-free control system 90. FIG. 2 is a graph showing the change over time in the audibility level of noise. FIG. 3 is a diagram showing the noise removal characteristics of the in-vehicle hands-free control system 90. FIG. 4 is a flowchart of noise removal processing of the in-vehicle hands-free control system 90. FIG. 5 is a diagram showing the relationship between call quality and audibility level by noise removal processing.

  As shown in FIG. 1, the in-vehicle hands-free control system 90 includes an in-vehicle hands-free control device 1, a microphone 10 that passes input information to the in-vehicle hands-free control device 1, and an in-vehicle hands via an amplifier 21. A speaker 20 that passes output information from the free control device 1, a wireless communication device 30, and a mobile phone 40 that can communicate with the wireless communication device 30 wirelessly are configured. Note that the mobile phone 40 is a mobile phone that can communicate with the handsfree device 30 via the wireless communication device 30. For example, the wireless communication device 30 communicates with the mobile phone 40 by Bluetooth (registered trademark). The mobile phone 40 performs voice communication and data communication with the nearest base station by a predetermined communication method (for example, an IMT2000 communication method such as W-CDMA or Cdma2000).

  The in-vehicle hands-free control device 1 includes a hands-free control unit 2, an HFDSP (hands-free digital signal processor) 3, and an amplifier 4 that is a circuit that amplifies an input signal from the microphone 10.

  The hands-free control unit 2 is a processing unit that controls the in-vehicle hands-free control device 1, and is configured by a processing device such as an ASIC (Application Specific Integrated Circuit). However, the present invention is not limited to this, and the hands-free control unit 2 executes a program read from a ROM (Read Only Memory) to a RAM (Random Access Memory) by a general-purpose processing device such as a CPU (Central Processing Unit). It may be a thing. When the voice communication (call) is started by the mobile phone 40, the hands-free control unit 2 processes voice information between the microphone 10, the speaker 20, and the mobile phone 40. That is, the hands-free control unit 2 receives information input from the microphone 10 and passes it to the mobile phone 40 as input information. In addition, the hands-free control unit 2 receives information input from the mobile phone 40 and passes it to the speaker 20 as output information.

  The HFDSP 3 removes noise from the input analog electric signal with a strength according to an instruction from the hands-free control unit 2 to obtain a digital signal. Then, the HFDSP 3 outputs the digital signal after noise removal to the wireless communication device 30.

  More specifically, the HFDSP 3 includes an A / D converter 5 that converts an analog electric signal input via the microphone 10 or the like into a digital signal, and noise during utterance from the digital signal (the time during which the speaker is speaking). The speech canceling noise canceller 6 and the speechless noise canceller 7 for removing noise during speechless time (time when the speaker is not speaking) from the digital signal. Further, the HFDSP 3 includes a noise cancellation setting unit 8 that receives the instruction from the hands-free control unit 2 and sets the noise cancellation strength for the speech noise canceller 6 and the non-speech noise canceller 7.

  The noise cancellation setting unit 8 is configured by an arithmetic processing device such as an ASIC. However, the present invention is not limited to this, and the noise cancellation setting unit 8 may execute a program read into a RAM (Random Access Memory) by a general-purpose processing device such as a CPU (Central Processing Unit). When the noise cancellation setting unit 8 receives designation of the intensity of noise cancellation from an external processing unit such as the hands-free control unit 2 or the like, the noise cancellation setting unit 8 is designated for the utterance noise canceller 6 and the no-utterance noise canceller 7. Set the strength.

  The utterance noise canceller 6 removes a predetermined noise component from the voice information according to the set noise cancellation strength when in the utterance state (that is, when voice information having a predetermined frequency characteristic is detected). Output. On the other hand, the no-speech noise canceller 7 has a predetermined noise component from the input signal information according to the set noise cancellation strength when it is not in an utterance state (that is, when voice information having a predetermined frequency characteristic is not detected). Is output. The utterance noise canceller 6 and the non-utterance noise canceller 7 are configured by an arithmetic processing unit such as an ASIC, for example.

  Here, the temporal change of the human auditory level with respect to noise will be described with reference to FIG. In general, it is known that the human auditory level gradually decreases with the passage of time when the same kind of sound is heard for a predetermined period of time. The graph 100 shows this visually. In the graph 100, time (unit: second) is plotted on the horizontal axis 110, and noise audibility level (unit: phone) is plotted on the vertical axis 120. In the graph 100, a curve 130 is shown with respect to how the audibility level changes when time elapses after starting to listen to a predetermined sound as noise. That is, the audibility level rises remarkably until the elapsed time is about 1 second, but the audibility level starts gradually decreasing after about 1 second, and gradually decreases until about 100 seconds elapses. In other words, a listener's subjective opinion is that a certain sound (for example, noise in a vehicle) is anxious when he / she starts to listen, but gradually becomes uncertain as time passes. In the first embodiment according to the present invention, paying attention to the characteristics, the noise cancellation strength is suppressed as much as possible to ensure the sound quality of the call.

  FIG. 3 is a diagram showing a graph 200 showing a change in the intensity of noise cancellation in the first embodiment according to the present invention. In the graph 200, time (unit: second) is plotted on the horizontal axis 210, and noise cancellation level (unit: decibel) is plotted on the vertical axis 220. The graph 200 shows a noise cancellation level condition 230 when a predetermined time has elapsed since a call was started via the in-vehicle hands-free control device 1. In other words, noise cancellation is strongly applied until the elapsed time is about 1 second, but the noise cancellation level starts to gradually decrease from the elapsed time 1 second to 100 seconds, and gradually cancels the noise until about 100 seconds elapses. The level goes down. Then, when 100 seconds have elapsed, the noise cancellation level stops decreasing and a constant noise cancellation level is maintained.

  FIG. 4 shows a flow of processing for realizing a change in the intensity of noise cancellation in the embodiment of the present invention. The noise removal process is started by the hands-free control unit 2 when a call of the mobile phone 40 is started.

  First, the hands-free control unit 2 starts measuring time (step S001). Specifically, the hands-free control unit 2 instructs a timing unit of the hands-free control device 1 (not shown) to start timing.

  Next, the hands-free control unit 2 determines whether or not the call has ended (step S002). When the call is ended (“Yes” in step S002), the hands-free control unit 2 instructs the noise cancellation setting unit 8 of the HFDSP 3 to end the noise cancellation process (step S003). Then, the noise removal process ends.

  If the call has not ended (“No” in step S002), the hands-free control unit 2 selects the next processing content according to the elapsed time (step S004). Specifically, the hands-free control unit 2 selects and implements processing according to any of the elapsed time being less than 1 second, 1 second to 100 seconds, and exceeding 100 seconds.

  If the elapsed time is less than 1 second (“elapsed time <1 second” in step S004), the hands-free control unit 2 samples the noise level of the audio signal input via the microphone 10 (step S005). . Then, the hands-free control unit 2 returns the process to step S002.

  If the elapsed time is 1 second to 100 seconds (“1 second ≦ elapsed time ≦ 100 seconds” in step S004), the hands-free control unit 2 sets the noise cancellation level (initial noise cancellation) in the noise cancellation setting unit 8 in the call. It is determined whether or not (level) has been set (step S006).

  When the initial noise cancellation level has not been set (“No” in step S006), the hands-free control unit 2 determines the initial noise cancellation level according to the sample audio information sampled in step S005 (step S007). Specifically, when the noise level of the sample audio information is stable (for example, the amount of change in the noise level is within minus 10% to plus 10%), the hands-free control unit 2 The initial noise cancellation level is determined according to a predetermined correspondence according to the average value of the noise level sampling period.

  Then, the hands-free control unit 2 instructs the noise cancellation setting unit 8 to set the initial noise cancellation level determined in step S007 (step S008).

  Next, the hands-free control unit 2 instructs the HFDSP 3 to start noise cancellation processing (step S009). When receiving the instruction, the utterance noise canceller 6 and the no-utterance noise canceller 7 of the HFDSP 3 start operation to perform noise cancellation on the digital audio signal output from the A / D converter 5. Then, the hands-free control unit 2 returns the process to step S002.

  When the initial noise cancellation level has been set (“Yes” in step S006), the hands-free control unit 2 determines to reduce the noise cancellation level by a predetermined amount according to the elapsed time (step S010). Specifically, the hands-free control unit 2 determines the amount of noise cancellation level reduction according to the elapsed time acquired from the time measuring unit. In the determination, the hands-free control unit 2, for example, reduces the noise cancellation level from the upper limit to the lower limit of the noise cancellation level evenly in 100 seconds (that is, proportional to the passage of time). Determine the amount of reduction. In this case, since the reduction amount uses a decibel that is a relative value, the actual noise cancellation amount is not always proportional to the passage of time.

  Next, the hands-free control unit 2 instructs the noise cancellation setting unit 8 to set a noise cancellation level to which the reduction amount of the noise cancellation amount determined in step S010 is applied (step S011). Then, the hands-free control unit 2 returns the process to step S002.

  When the elapsed time exceeds 100 seconds (“100 seconds <elapsed time” in step S004), the hands-free control unit 2 returns the process to step S002, that is, when the HFDSP3 speaks without setting the noise cancellation amount. The noise canceller 6 and the no-speech noise canceller 7 continue noise cancellation with the already applied noise cancellation level.

  The above is the processing flow of the noise removal processing. In the noise removal process, noise cancellation is not started when the elapsed time is less than 1 second, but this starts the noise cancellation process with a predetermined initial intensity once at the start of a call as shown in FIG. It may be a thing. Processing With the noise removal processing, the HFDSP 3 can perform noise cancellation processing with different strengths by changing the strength of noise cancellation according to the elapsed time from the start of the call. In addition, after the elapse of a predetermined time (which is 100 seconds in the above flow, but is not limited thereto, for example, a predetermined time between 60 seconds and 180 seconds or a variable time may be used), HFDSP3 Keeps the noise cancellation intensity constant. Therefore, the user of the hands-free control device 1 can obtain the noise removal result and the speech quality as shown by the graph 300 in FIG.

  FIG. 5 is a diagram illustrating a relationship between call quality and noise audibility level by noise removal processing. By applying the noise removal process according to the embodiment of the present invention, as shown in the graph 300, the noise audibility level is maintained at a predetermined level and the call sound quality is improved as time passes. be able to. That is, it can be said that the call sound quality can be improved by weakening the strength of the noise removal processing with the passage of time as the human hearing level becomes slow.

The first embodiment according to the present invention has been described above. According to the first embodiment, it is possible to suppress noise while suppressing deterioration in call sound quality. For example, the call sound quality can be improved by reducing the strength of the noise removal processing with the passage of time as the human hearing level is slowed down.
The first embodiment can be variously modified. For example, the present invention can be applied not only to noise cancellation processing on the input side but also to noise cancellation processing on the output side. An in-vehicle hands-free control system 90 to which another embodiment of the present invention is applied will be described below with reference to FIG.

  FIG. 6 is a diagram showing an outline of the modified second embodiment. In the second embodiment, the basic configuration is the same as that of the first embodiment, but for the voice information transmitted from the other party on the output side, that is, the mobile phone 40, the speech noise canceller 6 and The noise canceling process by the no-speech noise canceller 7 is performed. Specifically, the HFDSP 3 includes a D / A converter 9 that outputs an analog signal to an amplifier 21 to which the speaker 20 is connected. The D / A converter 9 converts the digital audio signal from which noise has been removed by the utterance noise canceller 6 and the no-utterance noise canceller 7 into analog audio information. When the wireless communication device 30 receives the voice information received by the mobile phone 40, the wireless communication device 30 delivers the voice information to the speech noise canceller 6 and the speechless noise canceller 7.

  The second embodiment has been described above. According to the second embodiment, it is possible to suppress noise while suppressing deterioration of call sound quality for voice information received from a call partner.

  Not limited to the second embodiment, in the noise removal processing of the first embodiment, the process of step S005 for sampling is performed until the noise information of the sampled audio information is stabilized, and the start of the noise cancellation process is delayed. You may do it. More specifically, sampling is performed until a change in the noise level of the sampled audio information within a predetermined time converges within a predetermined range (for example, within minus 20% to plus 20%), and noise cancellation processing is performed after the convergence. Start. Alternatively, noise cancellation may be performed at a predetermined noise cancellation level from the start of the call, and the start of noise cancellation level reduction may be delayed until the noise level converges. A graph 230a in FIG. 3 shows an example in which the start of reduction of the noise cancellation level is delayed until the noise level converges in this way. This makes it possible to set a more appropriate initial noise cancellation strength.

  Further, in the noise removal processing of the first embodiment, the noise cancellation strength, the noise cancellation reduction range, the start point of the noise cancellation processing, and the like are changed according to the mobile phone 40 model, communication carrier, communication method, and the like. You may do it. For example, the hands-free control unit 2 acquires model information and the like from the mobile phone 40, and when the mobile phone 40 adopts a communication method having a noise canceling function, the hands-free control unit 2 starts the noise removal process 1 You may make it control so that it may delay intentionally rather than the time of second progress, or a noise removal process may not be started. In addition, for example, the hands-free control unit 2 may acquire the type of carrier from the mobile phone 40 and correct (change and set) the noise cancellation intensity according to the carrier. For example, if it is a predetermined carrier, as shown in the graph 230b of FIG. 3, the initial value of noise cancellation may be set to half the initial value of other carriers. Of course, the hands-free control unit 2 may accept the carrier type input from the caller via the microphone 10 or the like and acquire the information to set the carrier type. By doing so, the mobile phone 40 can perform appropriate noise cancellation processing.

  Further, in the noise removal processing of the first embodiment, when a call is started, the hands-free control unit 2 acquires call history information from the mobile phone 40 and uses the same call destination as the previous call destination. If there is a predetermined time (for example, 3 minutes or less) from the end of the previous call, the initial noise cancellation level determined in step S007 is determined to be lower than normal (for example, noise at the end of the previous call). (The graph 230c in FIG. 3 sets the initial value of noise cancellation to a half of the normal initial value, reduces the reduction rate, and the like. In other words, an example is shown in which the amount of reduction in noise cancellation intensity per elapsed time is the same as usual. In this way, when there is a possibility that the caller maintains the audibility level corresponding to the previous call, the caller may not feel the deterioration of the call sound quality due to the change of the noise cancellation level. become.

In the above, this invention was demonstrated centering on embodiment. The features described in the above embodiment do not need to have all of them, and any feature may be provided. Moreover, you may combine freely the characteristic to provide. In the above-described embodiment, the example in which the present invention is applied to the in-vehicle hands-free control device has been described. Can do. To summarize the above-described techniques, the technique according to the present invention is described in A. A hands-free control device that delivers input voice information received by a voice input device as input information to a mobile phone, and delivers output voice information received from the mobile phone as output information to a voice output device, Noise removal means for performing noise removal processing at a predetermined intensity, and noise removal intensity setting means for setting the intensity of noise removal processing in the noise removal means, wherein the noise removal intensity setting means includes the intensity, It can be said that the present invention relates to a hands-free control device including changing according to the elapsed time from the start of a call of the mobile phone. B. A. above. It can be said that the noise-removing strength setting means relates to a hands-free control device characterized in that the strength decreases as the elapsed time becomes longer. In addition, C.I. A. above. Or B. If the elapsed time is within a predetermined time, the noise-removing intensity setting means decreases the strength as the elapsed time becomes longer, and sets the elapsed time to a predetermined time. Beyond this, it can be said that the present invention relates to a hands-free control device characterized in that the strength is kept constant. D. C. above. It can be said that the predetermined time is a time of 60 seconds or more and 180 seconds or less, which is related to the hands-free control device shown in FIG. In addition, E.I. A. above. ~ D. In the hands-free control device shown in any one of the above, the noise removal strength setting means samples the input voice information, and when the noise level converges to a constant level, the noise removal strength setting means sets the noise removal strength according to the sampling result. It can be said that the present invention relates to a hands-free control device characterized by determining and setting an initial value. F.F. A. above. ~ E. The noise-removing intensity setting means obtains the previous call destination and call end time from the call history of the mobile phone, and the call destination is the previous call destination. When the time from the call end time to the call start time is within a predetermined time, the initial value of the strength of the noise removal process is determined and set lower than the normal initial value. It can be said that it relates to a hands-free control device characterized by the above. G. A. above. ~ E. The noise-removing intensity setting means obtains the previous call destination and call end time from the call history of the mobile phone, and the call destination is the same as the previous call destination. When the time from the call end time to the call start time is within a predetermined time, the initial value of the strength of the noise removal process is determined to be set lower than a normal initial value. It can be said that it relates to the featured hands-free control device. H. F. above. Or G. In the hands-free control device shown in any one of the above, the noise removal strength setting means sets the initial value of the strength of the noise removal processing to the same strength as the noise removal strength at the end of the previous call. It can be said that the present invention relates to a hands-free control device. In addition, I.I. A. above. ~ H. A hands-free control device according to any one of the above, wherein the noise removal strength setting means corrects the noise removal strength according to a carrier of the mobile phone. It can be said that it is a thing.

DESCRIPTION OF SYMBOLS 1 ... Vehicle-mounted hands-free control apparatus, 2 ... Hands-free control part, 3 ... HFDSP, 4 ... Amplifier, 5 ... A / D converter, 6 ... Noise canceller at the time of speech, 7 ... Noiseless noise canceller, 8 ... Noise cancellation setting section, 10 ... Microphone, 20 ... Speaker, 21 ... Amplifier, 30 ... Wireless communication device, 40 ... Mobile phone

Claims (6)

A microphone for inputting audio information,
Noise removing means for removing noise of voice information input by the microphone;
A noise removal apparatus comprising: transmission means for sending voice information that has been noise-removed by the noise removal means to a mobile phone;
Intensity selecting means for selecting a first intensity or a second intensity lower than the first intensity as an intensity for removing noise while audio information is being input by the microphone,
The intensity selecting means selects the first intensity as an intensity for removing the noise when the time elapsed from the start of input of the audio information by the microphone is the first time, and the audio information is received by the microphone. When the time elapsed since the start of input is the second time exceeding the first time, the second intensity is selected as the intensity for removing the noise,
The noise removing unit removes noise of voice information input by the microphone based on the first intensity or the second intensity selected by the intensity selecting unit. . Wherein the noise removal means according to claim 1, characterized in that said depending on the level of the noise sampled until the predetermined time elapses from the start of the voice information input by the microphone, setting the first intensity The noise removal apparatus described in 1. Voice receiving means for receiving voice information from a mobile phone;
Noise removing means for removing noise of voice information received by the voice receiving means;
A speaker that outputs audio information from which noise has been removed by the noise removing unit, and a noise removing device comprising:
Intensity selection means for selecting a first intensity or a second intensity lower than the first intensity as an intensity for removing noise while audio information is output by the speaker,
The intensity selecting means selects the first intensity as an intensity for removing the noise when the time elapsed since the start of output of the audio information by the speaker is the first time, and the audio information is output by the speaker. When the time elapsed since the start of output is the second time exceeding the first time, the second intensity is selected as the intensity for removing the noise,
The noise removing means removes noise of voice information received by the voice receiving means based on the first intensity or the second intensity selected by the intensity selecting means. apparatus. Wherein the noise removal means, wherein depending on the level of the noise sampled until the predetermined time elapses from the audio information is initiated outputted by the speaker, claim 3, characterized in that setting the first intensity The noise removal apparatus described in 1.   A noise removal method for a noise removal apparatus,
  The noise removing device is
  A microphone for inputting audio information,
  A transmission means for sending audio information from which noise has been removed to a mobile phone; and a control means,
  The control means includes
  An intensity selection step of selecting a first intensity or a second intensity lower than the first intensity as an intensity for removing noise while audio information is being input by the microphone;
  A noise removing step of removing noise of the voice information input by the microphone based on the first intensity or the second intensity selected in the intensity selecting step;
  Run
  In the intensity selection step, when the time elapsed since the start of input of audio information by the microphone is the first time, the first intensity is selected as the intensity for removing the noise, and the audio information is output by the microphone. When the time elapsed since the start of input is the second time exceeding the first time, the second intensity is selected as the intensity for removing the noise.
The noise removal method characterized by the above-mentioned.   A noise removal method for a noise removal apparatus,
  The noise removing device is
  Voice receiving means for receiving voice information from a mobile phone;
  A speaker that outputs audio information from which noise has been removed;
  Control means,
  The control means includes
  An intensity selection step of selecting a first intensity or a second intensity lower than the first intensity as an intensity for removing noise while audio information is being output by the speaker;
  A noise removing step for removing noise of the voice information received by the voice receiving means based on the first intensity or the second intensity selected in the intensity selecting step;
  Run
  In the intensity selection step, when the time elapsed since the start of output of audio information by the speaker is the first time, the first intensity is selected as the intensity for removing the noise, and the audio information is output by the speaker. When the time elapsed since the start of output is the second time exceeding the first time, the second intensity is selected as the intensity for removing the noise.
The noise removal method characterized by the above-mentioned.

Images