JP2002064389A - Noise removing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a noise removing device capable of removing noise without the sense of incompatibility in respect of every kind of external noise. SOLUTION: This device is provided with a noise detecting part for detecting a noise component, a noise cut part for cutting the high frequency components of the noise and a signal correcting part for correcting the signal of a noise cut part output and the signal correcting part as a signal preserving memory for temporarily preserving the noise cut part output, a signal predicting part for predicting a signal on the basis of either noise cut part output or temporary preserving memory output and a switch part for switching and outputting any one of noise cut part output and signal predicting part output. When a noise is detected in the noise detecting part, in the signal predicting part, a prediction signal is formed from an input signal, which contains no noise, stored in the temporary preserving memory and outputs this signal through the switch part.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise eliminator for eliminating intermittent external noise received by a receiving system for radio broadcasting or the like with high accuracy.
In particular, a normal signal is predicted from the signal state up to that point in the section affected by the noise, and the predicted signal is fitted into the section affected by the noise, so that the listener can remove the noise without discomfort. Related to the device.

[0002]

2. Description of the Related Art A noise removing apparatus employed in a conventional radio receiving system has a detecting section for detecting noise and a signal processing section for performing signal processing for noise removal based on the detection of noise. This signal processing unit performs pre-completion, performs diagonal complementation (connects a signal point before noise with a signal point after noise), and controls frequency characteristics of a noise section (for example, by applying a low-pass filter and applying a wide-range component). In FM broadcasts and the like, stereo reception is stopped, or the like, to remove external noise.

[0003]

However, with regard to the above-described processing for complementation, the noise itself can be removed, but if the width of the noise is wide, the section where the signal is missing is widened. This will cause discomfort. When the frequency characteristics are controlled, there is a disadvantage that a noise component remains. Even if these methods are combined, noise cannot be effectively removed from a wide range of noise.

Therefore, it is difficult to realize a noise elimination device that is effective against all kinds of noise and does not cause a sense of discomfort to a listener, and has not been achieved yet.

[0005]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems in the conventional noise elimination apparatus, and is effective for all kinds of noise and does not cause a listener to feel uncomfortable. It is intended to realize a removing device. In order to achieve the above object, according to the present invention, there is provided a noise detection unit that detects noise in an input signal, and a noise cut unit that cuts a high frequency component of noise from the input signal based on a noise detection result of the noise detection unit. Department and
A signal rectification unit for rectifying the output of the noise cut unit, the signal rectification unit comprising a signal storage memory for temporarily storing the output of the noise cut unit, and a prediction signal from the noise cut unit output signal in a normal state. Forming a prediction signal from the output signal of the signal storage memory when noise is detected by the noise detection unit; and outputting the noise cut unit when noise is detected by the noise detection unit. And a switch unit for outputting the output of the signal prediction unit.

In the noise elimination device having the above-described configuration, in a normal state, that is, when no noise is detected by the detection unit,
A normal signal that is not subjected to noise cut processing by the noise cut unit is output as it is via the changeover switch unit. On the other hand, when noise is detected by the detection unit, a high frequency component is removed from the waveform of the input signal by the noise cut unit based on the detection signal. The detection signal is simultaneously transmitted to the prediction unit and the changeover switch unit, and changes the operation mode in each unit. That is, the prediction unit changes its operation mode so as to perform signal prediction based on the output of the signal storage memory. The operation mode is changed so that the changeover switch unit also outputs the prediction unit output instead of the noise cut unit output.

As a result, the prediction unit predicts a signal based on a normal signal waveform before the occurrence of noise once stored in the signal storage memory, and outputs the signal to the changeover switch unit. The changeover switch unit outputs this signal instead of the output of the noise cut unit. As a result, since the signal in the noise-cut section of the input waveform is replaced by a signal predicted from a normal signal before the noise is input, the output signal does not include waveform distortion due to the noise cut. Therefore, unlike the conventional apparatus, the problem of the distortion of the output sound due to the noise removal or the uncomfortable feeling generated in the listener is solved. Further, it is possible to sufficiently cope with wide noise by adjusting the amount of memory in the signal storage memory.

The input signal includes an AM broadcast reception signal, an FM broadcast reception signal, a television audio signal, a digital television audio signal, and the like. The prediction unit is configured using an adaptive digital filter (ADF).

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIGS. 1 and 4 to 7 are block diagrams showing various examples of a noise elimination device according to a first embodiment of the present invention. Relates to the basic configuration of the present invention. 2 and 3 are waveform diagrams for explaining the operation of the noise elimination device shown in FIG. Hereinafter, each embodiment will be described.

(Embodiment 1-1) The embodiment shown in FIG. 1 is a noise eliminator according to a basic configuration of the present invention. In FIG. 1, reference numeral 1 denotes a signal input unit, for example, AM, F
An M radio signal or a TV broadcast audio signal is received via an antenna or the like. Reference numeral 2 denotes an ADC (analog-to-digital converter) for converting an input signal from analog to digital, and reference numeral 3 denotes a DSP process for performing various digital processes including a process for removing noise on the input signal converted to a digital signal. Department.

In this embodiment, a DSP processing unit 3 cuts noise by digital processing from an input signal based on a noise detection unit 4 composed of a high-pass filter or the like and a noise detection signal detected by the detection unit 4. The apparatus includes a noise cut section 5 and a signal straightening section 6 for straightening the signal after the noise cut. The signal correction unit 6 stores a noise-cut input signal for a predetermined time, that is, a signal storage memory 7 for delaying the input signal, a prediction unit 8 for predicting a signal, and an output of the noise cut unit 5 or an output of the prediction unit 8. And a changeover switch section 9 for selecting and outputting any one of the above. The prediction unit 8 is configured by, for example, an adaptive (adaptive) digital filter (ADF), and this filter has a function of realizing a frequency characteristic approximate to the characteristic of a previous signal.

Next, referring to the waveform diagrams of FIGS. 2 and 3,
The mechanism of noise removal in the noise removal device of FIG. 1 will be described. A waveform a in FIG. 2 shows a signal waveform at the output stage of the input unit 1. When this noise elimination device is used for an automobile radio, a received signal of a radio broadcast includes intermittent noise such as a pulse noise generated from an automobile engine as shown in the figure. This input signal is input to the noise detection unit 4 of the DSP processing unit 3, and a noise component is detected (waveform c).

The noise cut section 5 removes the high frequency component of the signal from the input waveform by digital processing based on the noise component detected by the detection section 4. As an example of the processing, a cut flag is generated from the noise component detected by the detection unit 4, and the noise is cut by removing the high-frequency component from the waveform of the input signal using the flag as a timing signal in the noise cut unit 5. I do. A waveform b indicates an output waveform of the noise cut unit 5. This waveform b
Is input to the changeover switch unit 9, the signal storage memory 7, and the prediction unit 8 of the signal correction unit 6.

In the noise detection section 4, an ADF flag is further formed based on the detected noise waveform. Waveform diagram d
Is such an ADF output from the noise detection unit 4.
3 shows a waveform of a flag signal. The flag signal d is input to the prediction unit 8 and the changeover switch unit 9 as a timing signal. The waveform e in FIG. 2 shows the output waveform of the signal storage memory 7. As is clear from the comparison between the waveforms b and e, the signal storage memory 7 delays the waveform b for a predetermined time and inputs the delayed waveform to the prediction unit 8. The adaptive digital filter (ADF) in the prediction unit 8 forms a prediction signal from the waveform e when the input waveform has noise, that is, when the flag signal is input from the noise detection unit 4. As described above, since the waveform e is a signal obtained by delaying the waveform b, no noise component is included at the timing when the flag is set. Therefore, the adaptive digital filter can form a prediction signal from a portion of the waveform e that does not include a noise component.

A waveform f in FIG. 3 shows a signal waveform inside the adaptive digital filter, and the filter updates the filter coefficient so that the signal waveform approaches zero. Waveform g
Shows the predicted signal waveform obtained as described above. This signal g is input to the changeover switch unit 9 and is switched to the waveform b when the input waveform has noise, that is, at the timing when the flag is set, and is output.

A waveform h in FIG. 3 shows an output waveform of the changeover switch unit 9. As shown, the noise component included in the waveform b has been replaced by the output of the adaptive digital filter, and as a result, the noise component included in the input signal is effectively removed. As described above, the present embodiment is characterized in that when noise occurs, the signal before the occurrence of noise stored in the signal storage memory is input to the ADF, thereby performing interpolation based on a normal signal waveform in the interpolation section. I have.

(Embodiment 1-2) In the embodiment shown in FIG.
Analog processing is applied to the noise cut unit 5 and digital processing is applied to the noise detection unit 4. Reference numeral 20 denotes a digital-to-analog converter (DAC). (Embodiment 1-3) The embodiment shown in FIG.
Applies analog processing, and the noise cut unit 5 applies digital processing.

(Embodiment 1-4) In the embodiment shown in FIG.
Analog processing is applied to both the noise detection unit 4 and the noise cut unit 5. Next, the effects of the signal processing circuits of the above embodiments are compared. In the embodiment 1-1, relatively large noise is removed in advance by the noise cut unit 5, and thereafter, noise remaining in the prediction unit 8 is removed. Therefore, the accuracy and stability of noise removal in the prediction unit 8 are improved. Further, also in the changeover switch section 9, the performance at the time of switching between the original signal including noise and the prediction signal is improved by removing the unnecessary signal in the noise cut section 5.

In the embodiment 1-1, the noise detection unit 4
Since the digital signal processing is performed on all the components from the switch unit 9 to the changeover switch unit 9, the DSP processing amount is the largest compared with the other embodiments. However, it is possible to obtain fine precision for various control signals, and the noise removal performance is the most. improves. Example 1
In 2, the control signal generated by the detection unit 4 can be obtained with high precision. In the first to third embodiments, a digital filter for noise detection is not required.
The amount of P processing can be reduced. Further, depending on the application, since the detection unit 4 is outside the DSP processing unit 3,
It becomes possible to set the sampling frequency of the DSP low.

In the case of the embodiment 1-4, the DSP processing amount can be reduced most. Although depending on the application, the DSP sampling frequency can be set low because the detection unit is outside the DSP processing unit 3. (Embodiment 1-5) FIG. 7 shows a fifth embodiment of the first embodiment of the present invention, and particularly shows a specific configuration of the noise detector 4 of FIG. As shown in the figure, the noise detection unit 4 includes a signal storage memory 4a and a prediction unit 4 including an ADF filter.
b, the prediction unit 4b includes a filter unit 4c and a coefficient update unit 4
d and a subtractor 4e.

When a signal a '(see FIG. 1) is input from the input terminal of the noise detector 4, a stationary signal Y appears at the output terminal of the filter unit 4c of the prediction unit 4b, while the subtracter An unsteady signal E appears as the output of 4e. Therefore, when pulse noise is mixed in the input signal a ', more accurate noise detection becomes possible by using the subtracter output E appearing as an unsteady signal as a noise detection signal.

Second Embodiment Hereinafter, examples in which the noise eliminator according to the first embodiment of the present invention is applied to an FM broadcast receiving system or an application having an MPX circuit in general will be described with reference to various examples. I do. (Example 2-1) FIG. 8 is a block diagram showing an FM broadcast receiving system of a first example according to the second embodiment of the present invention. In the figure, 100 is an FM broadcast signal receiving circuit,
Reference numeral 200 denotes a signal processing circuit, which includes a detection unit 300, a noise removal device 400 and a noise removal device 400 according to the first embodiment of the present invention.
PX circuit 500 is included. Note that L and R are M
5 shows a signal after stereo separation by the PX circuit 500.

When the noise elimination device according to the first embodiment of the present invention is applied to FM broadcasting, not only the configuration of the illustrated embodiment but also various configuration methods exist. Regarding the noise removal of FM broadcasting, ignition pulse noise and multipath noise are typical as noise types.
Since it may be as short as 0 μsec, the conventional noise canceller system (prefix interpolation only) has a certain effect. Of course, the use of the noise elimination device of the present invention is superior in distortion characteristics, but it is also true that there is not much difference in audibility.

In one case, the noise width is considerably long and may be several milliseconds. This will surely be discomfort remains for between conventional pre and system to be removed-interpolation will fly the sound too long, can not be completely remove the noise. As described in the description of the operation of the embodiment of FIG. 1, the noise elimination device of the present invention can remove noise without discomfort to a listener even in the case of a wide noise.
Optimal for this problem. Therefore, in the system shown in FIG. 8, sufficient noise removal is performed.

(Embodiment 2-2) The embodiment shown in FIG. 9 is different from the embodiment shown in FIG. 8 in that the signal detected by the detection section 300 is introduced into the DSP processing circuit 200 via the AD converter 2. Thus, the noise removal device 400 and the MPX circuit 500 can both be DSP-processed. MP
Since the X circuit is a circuit that is easy to perform DSP processing due to its configuration, it can be integrated with the noise elimination device 400 of the first embodiment and stored in one chip. This increases the efficiency of the entire receiving system and leads to a reduction in the cost of the system.

Embodiment 2-3 The embodiment shown in FIG.
A system in which the signal correction unit 6 according to the first embodiment of the present invention is applied to noise removal of a pilot signal used in the MPX circuit 500 will be described. A 19 kHz pilot signal is extracted from the FM signal 1f (1g) and is generated as a 38 kHz signal (1h).

When pulse noise or the like occurs, the pilot signal is also affected by this, and this also causes signal noise after stereo separation. Therefore, if the pilot signal is MP
By inputting the signal to the signal correcting section 6 before inputting to the X circuit 500, noise in the pilot signal can be removed, and more accurate demodulation of the stereo signal can be performed. (Embodiment 2-4) The embodiment of FIG. 11 specifically shows the DSP processing unit 200 of the embodiment shown in FIG.
It is characterized in that the signal correction unit 6 is arranged at the preceding stage of the X circuit 500. Thus, unlike the embodiment 2-5 described later, only one ADF is required, and the DSP processing amount is reduced. Further, since the ADF is used before the MPX circuit, the number of taps of the ADF can be set smaller than when the ADF is used in the audio band.

(Embodiment 2-5) The embodiment shown in FIG.
This shows another specific example of the DSP processing unit of the embodiment shown in FIG. 1, which is a system in which a total of two signal correction units 6 are provided at the subsequent stage of the MPX circuit 500 independently for L and R of stereo.
In this system, two ADFs are required by arranging the signal correction unit 6 at the subsequent stage of the MPX circuit 500.
Since noise processing is performed independently for L and R, it is effective in maintaining stereo performance. In addition, when compared with the embodiment 2-4 in FIG. 11, the number of taps needs to be increased in order to use the ADF in the audio band (low frequency), but the prediction characteristics are improved.

Embodiment 2-6 The embodiment shown in FIG.
FIG. 10 shows still another specific example of the DSP processing unit shown in FIG. 9, in which a signal correction unit (ADF,
(Including memory and changeover switch) for (L + R),
(L-R) shows a system having two in total. 14 and 15 show a specific configuration of the MPX circuit 500 according to the present embodiment.

14 and 15, 500a is a low-pass filter, 500b is a high-pass filter,
500c indicates a mixer, and 500d indicates a matrix portion. In both the configurations of FIGS. 14 and 15, the signal adjuster 6 applies A to the main (MAIN) signal.
Since the DF is applied, the sound performance is improved when the system is operated in monaural.

On the other hand, in the processing of the sub (SUB) signal, the example shown in FIG. 14 has an advantage that the number of ADF taps is reduced. In the example shown in FIG.
Since the processing in (1) corresponds to the voice band, a large number of ADF taps are required, but on the contrary, the predictability is excellent. (Embodiment 2-7) The embodiment shown in FIG.
The SP processing unit is characterized in that a signal correction unit 6 is provided before and after the MPX unit 500. As a result, the present device has all the effects seen in Examples 2-4, 2-5, and 2-6, and more effective performance is obtained as noise removal.

Embodiment 3 In this embodiment, in the noise elimination device shown in Embodiment 1 or the FM reception system shown in Embodiment 2, the DSP
Provided is a system for downsampling processing in a high frequency band and a low frequency band in a processing unit. (Embodiment 3-1) In the embodiment shown in FIG. 17, for example, a noise detection unit, a noise cut unit,
A downsampling unit 3c is provided between a high-frequency band processing unit 3a including an MPX unit and a low-frequency band processing unit 3b including an ADF unit, for example.

As a general idea of noise elimination, it is desirable to detect noise at a higher sampling frequency. By doing so, even high-frequency noise can be detected. However, since the band to be used is determined in the audio band and MPX processing,
By converting to a sampling frequency corresponding to the block, the DSP processing amount can be reduced. Therefore, the present embodiment in which downsampling is performed between the high frequency band processing unit and the low frequency band processing unit is effective.

FIG. 18 shows an example of an approximate DSP processing amount when downsampling is not performed (Example 1-1) and when downsampling is performed (Example 3-1). The conditions of the illustrated rough example are: before downsampling: 192 kHz sampling after downsampling: 48 kHz sampling Downsampling rate: 1/4. As is clear from the figure, the processing after down-sampling has a sampling frequency of 1/4, so even if the same processing calculation is performed, the processing amount MOPS is 1
/ 4.

(Embodiment 3-2) The embodiment shown in FIG.
This shows a system in which a downsampling unit 3b is provided at a stage subsequent to the noise cut unit of the apparatus according to the embodiment 1-1 shown in FIG. This system is effective when the ADF processing is performed in the audio band, and the DSP processing amount in the signal correction unit 6 is greatly reduced.

(Embodiment 3-3) In the embodiment shown in FIG. 20, in the DSP signal processing section of the apparatus shown in Embodiment 2-4,
The down-sampling unit 3b is provided after the signal correction unit 6 and before the MPX unit 500. In the apparatus of the present embodiment, the DSP processing amount in the MPX unit 500 is reduced. The present apparatus is also effective as a system for removing higher-frequency noise, that is, for performing noise detection, cutting, and ADF processing in a high-frequency region.

(Embodiment 3-4) In the embodiment shown in FIG. 21, in the DSP signal processing section of the apparatus shown in Embodiment 2-5,
The down-sampling unit 3 b is provided after the MPX unit 500 and before the signal correction unit 6. In the apparatus according to the present embodiment, the DSP processing amount in the signal adjusting units 6 is greatly reduced. The system in this case is, for example, FM
In the case of broadcasting, the sampling frequency before downsampling can be set to about 200 kHz, and the sampling frequency after downsampling to about 50 kHz,
An ideal configuration is possible in terms of feasibility.

(Examples 3-5 and 3-6) FIGS. 22 and 2
In each of the embodiments shown in FIG. 3, the downsampling units 3ba to 3ba in the apparatus shown in FIGS.
3 bd is provided. Note that FIGS. 22 and 23 show a configuration in which the downsampling unit 3b is provided at all the locations where downsampling is possible, but it is not necessary to perform downsampling at all locations.

In the apparatus shown in FIGS. 22 and 23, the downsampling units 3bb and 3bc are different from the apparatus shown in FIG.
Thus, the processing amount of the matrix unit 500d can be further reduced. Further, in the downsampling units 3ba and 3bd, the processing amount of each signal adjusting unit 6 in the ADF is reduced. Embodiment 4 Hereinafter, an embodiment in which the noise removal apparatus of the present invention is applied to an AM broadcast receiving apparatus will be described.

(Embodiment 4-1) The embodiment shown in FIG.
This embodiment shows an embodiment in which the noise removal device shown in FIG. 1 is applied to the AM reception signal input from the AM detection input means 1A after the AD conversion, and the noise is removed. AM
Regarding the noise removal of broadcasting, typical types of noise include ignition pulse noise and vehicle electrical system switch noise, and the system of the present invention is very effective for both of them, as in FM broadcasting reception.

By the way, the noise width that appears in the AM output is
, Which is about 500 μsec to just over 1 msec, and the pre-interpolation process alone cannot perform high-performance noise removal to the extent that the presence of noise is not noticed. On the other hand, if the noise elimination device of the present invention is used, accurate noise elimination without discomfort can be achieved. (Embodiment 4-2) The embodiment shown in FIG. 25 shows a case where the noise eliminators of the first, second and third embodiments of the present invention are applied to processing of an audio signal of a TV broadcast. An audio signal of a TV broadcast is usually transmitted by a main signal and a sub signal of a multiplexing unit. In the transmitted AM broadcast signal, the video signal and the audio signal are separately received by the video receiving unit 110 and the audio receiving unit 120. The received audio signal is processed and output by the signal processing circuit 210 including the signal correction unit 6.

When a TV broadcast is received by a mobile unit, a multipath is generated and a noise generation section becomes long. Therefore, by applying the present system, it is possible to remove noise accurately and without a sense of incongruity. (Embodiment 4-3) The embodiment shown in FIG. 26 shows a case where the noise elimination devices of the first and third embodiments of the present invention are applied to the processing of an audio signal of digital broadcasting. The received digital broadcast signal is transmitted to the video demodulation unit 111 and the audio demodulation unit 1.
At 21, video and audio are demodulated separately. The demodulated audio signal is introduced into the signal processing unit 211, and the signal correction unit 6 removes noise.

The audio signal of a digital broadcast is usually compressed by the MPEG technology or the like, but if the electric field condition is deteriorated, the signal is intermittent in a frame unit (24 ms or the like). In such an intermittent signal state, by applying the signal adjusting unit of the present invention, it is possible to output a sound that is more audible and comfortable even when the electric field state deteriorates. Embodiment 5 Hereinafter, various examples of the noise cut unit used in Embodiments 1, 2, and 3 of the present invention will be described.

(Embodiment 5-1) The embodiment shown in FIG.
The noise cut section 5 includes a register 5a and a 0 set section 5b.
It is introduced into the setting section 5b and operated, and the register 5a
Has a configuration in which a signal at the time of occurrence of noise is replaced with 0.

This has the advantage that the ADF effect is improved by removing abnormally large noise or the like when noise occurs. The zero set of the register can be easily realized by the clear instruction, and the development of the DSP processing program becomes easy and simple. FIG. 28A shows a signal waveform before processing including noise, and FIG. 28B shows a signal waveform in which a noise portion is replaced with 0, that is, an output waveform of the register 5a.
As is apparent from these figures, abnormally large noise is effectively removed in the noise cut unit 5, so that the ADF processing in the signal correction unit is performed more effectively.

(Embodiment 5-2) The embodiment shown in FIG.
The configuration of the noise cut unit has a configuration in which a value before the occurrence of noise is set in the noise portion. In the figure, 5c
Is a normal time, that is, a memory for storing a signal before noise occurrence,
5d stores the contents of the register 5a in the memory 5c when noise occurs.
Is a set unit that is replaced with the value stored in.

In this device, when a detection signal from the noise detection unit 4 is input to the set unit 5d, the set unit 5d operates to replace the noise portion of the signal with the contents of the memory. Since the contents of the memory are in a state before the signal contains noise, this makes it possible to effectively remove abnormally large noise and the like. As a result, the effect in the subsequent ADF processing is improved. Further, the occurrence of switching noise can be suppressed by replacing the noise portion of the signal with the signal value before the occurrence of the noise.

(Embodiment 5-3) The embodiment shown in FIG.
The noise cut section 5 has a peak limit section 5f for limiting the peak value of the noise, and a comparison section 5e for comparing the value of the input signal with the limit value. By setting the limit value to a value higher than that of a normal signal in the limit section 5f, when the input signal exceeds the limit value, the limit value can be set in the signal.

As a result, even if abnormally large noise enters when noise occurs, the value exceeding the limit is removed,
The effect of noise removal can be enhanced in the subsequent ADF. Further, in this embodiment, unlike the embodiments 5-1 and 5-2, since the apparatus can be operated without requiring a noise detection signal, there is also an advantage that omission of noise detection can be prevented. .

(Embodiment 5-4) The embodiment shown in FIG.
The digital filter 5g in the noise cut unit 5 and the LPF characteristic coefficient set / reset unit 5 for the filter 5g
h. The device of this embodiment has a coefficient for passing a signal in a normal state through, and the L is determined by a noise detection signal.
The coefficient of the PF characteristic is set. This makes it possible to suppress passage of high-frequency noise when noise occurs.

Embodiment 5-5 The embodiment shown in FIG.
The noise cut unit 5 includes a digital filter 5g, a coefficient update unit 5i for updating the coefficient of the filter, and a time constant setting unit 5j. In the noise cut section 5, the digital filter 5g is set with a coefficient so as to pass a signal in a normal state through, and by input of a noise detection signal,
The coefficient of the filter characteristic is set. At first, the filter characteristics are close to through, gradually become LPF characteristics, and finally operate to take a fixed value. The time constant setting unit 5j sets the transition time of the filter characteristic.

In this embodiment, the generation of high-frequency noise at the time of noise cut is suppressed by gradually changing the filter characteristics. Embodiment 6 Hereinafter, various examples of the signal adjustment unit 6 used in Embodiments 1, 2, and 3 of the present invention will be described.

(Embodiment 6-1) The embodiment shown in FIG.
In the signal correction unit 6 in the DSP processing unit, a memory amount control unit 7a is provided between the signal storage memory 7 and the noise detection unit 4 to increase or decrease the memory amount depending on the type of noise. . That is, based on information about the input signal and noise detection information from the noise detection unit 4, the amount of memory is increased or decreased according to the type of noise. This allows
For example, it is possible to perform processing such as increasing the amount of memory in the case of long noise and decreasing the amount of memory in the case of short noise.
The amount of memory is reduced and the accuracy of ADF prediction is improved.

(Embodiment 6-2) The embodiment shown in FIG.
A digital filter (LPF) 10 is inserted between the signal storage memory 7 and the prediction unit 8 including the ADF. In the noise removal system having this configuration, the signal from which the high frequency noise has been removed by the digital filter 10 is input to the prediction unit 8, and as a result, more accurate signal prediction can be performed for the low frequency signal.

(Embodiment 6-3) The embodiment shown in FIG.
LPF section 8 between subtracter 8c and coefficient updating section 8b of prediction section 8
d is arranged to perform LPF processing on the error E of the ADF. 8a indicates a filter unit. In the system of this configuration, LPF is added to the error component.
By performing the processing, the high frequency component of the error is reduced, and A
The response of the DF to high-frequency components is slow. As a result, even when high-frequency noise is mixed, it is not adapted to the noise, and only the signal component is predicted, thereby enabling more accurate noise removal.

(Embodiment 6-4) The embodiment shown in FIG.
A limit unit 8e is arranged between the subtractor 8c and the coefficient update unit 8b of the prediction unit 8 to limit the error E of the ADF. In the system of this configuration, by limiting the error component, even if an abnormally large noise component is mixed in the error, the limit component 8e can reduce the abnormal component. Thus, abnormal operation can be reduced, and accurate noise removal can be performed.

(Embodiment 6-5) The embodiment shown in FIG.
The LPF 8 is provided between the subtractor 8c and the coefficient updating unit 8b of the prediction unit 8.
fa, HPF 8fb, first gain setting unit 8ga, second gain setting unit 8ga
The gain setting unit 8gb and the adder 8h are arranged. The LPF 8fa and the HPF 8fb form a band-pass filter.

In the system having this configuration, the error signal E output from the subtractor 8c is filtered by the LPF 8fa and the HPF 8fb, and each output is multiplied by the first and second gains, respectively. Here, by applying a small gain to the high frequency band, it is possible to configure a system that responds slowly to high frequency components. As a result, the system does not adapt to unusually large high-frequency noise, so that the signal component can be accurately predicted.

(Embodiment 6-6) The embodiment shown in FIG.
The LPF 8 is provided between the subtractor 8c and the coefficient updating unit 8b of the prediction unit 8.
fa, an HPF 8fb, a first limiter 8ia, a second limiter 8ib, and an adder 8h. The LPF 8fa and the HPF 8fb form a band-pass filter.

In the system having this configuration, the error signal E output from the subtracter 8c is added to the LPF 8fa and the HPF 8f.
b, and each limiter 8ia, 8ib
Sets the first and second limits for each output. Here, by setting the limit of the high-frequency band to be small, a system in which the reaction to the high-frequency component is further limited is configured. As a result, updating of the coefficient is limited for abnormally large high-frequency noise, and the prediction of the signal can be performed more accurately.

(Embodiment 6-7) The embodiment shown in FIG.
An oscillation preventing unit 8j is provided for the coefficient updating unit 8b of the prediction unit 8. When the filter coefficient is updated to a certain value or more when the filter coefficient is updated, the oscillation preventing unit 8j performs an operation of limiting the coefficient or setting the coefficient to 0.

In the system of the present configuration, even if abnormally large noise is mixed in, the malfunction of the ADF can be prevented by the operation of the oscillation preventing section 8j, and more accurate noise removal can be achieved. (Embodiment 6-8) In the embodiment shown in FIG. 40, a power detector 11 for detecting the power of the output signal b of the noise cut section 5 and the output signal Y of the filter section 8a of the prediction section 8; And an amplitude adjustment unit 12 that adjusts the amplitude of the output signal Y of the filter unit 8a under the control of the filter unit 8a.

In the system of this configuration, the power detector 11 detects the power of the b input signal and the power of the Y signal, and the amplitude adjustment unit 12 adjusts the amplitude of the Y signal so that these powers become approximately equal. Normally, since the output of the predictor 8 has a high-frequency component removed, the power of the Y signal is lower than that of the b input signal. Therefore, if the signal is frequently switched by the changeover switch unit 9 to remove noise, a sense of sound fluctuation will occur in the sense of hearing. The present system has the effect of reducing this sound fluctuation feeling by adjusting the power of both signals to the same extent.

(Embodiment 6-9) The embodiment shown in FIG.
This is a system characterized in that a band-pass filter (BPF) is arranged in front of a signal adjustment unit, and processing is changed according to a band. In this system, as shown
The LPF 13 and the HPF 14 are arranged in front of the signal adjusters 6a and 6b provided for each band. Note that the signal correction units 6a and 6b have the same configuration, and the signal storage memories 7 and 6, respectively.
It comprises a prediction unit 8 and a switching SW 9. Fifteen
Is an adder.

In the system of this configuration, each of the signal correction units 6a and 6b operates with different parameters.
The signal correction section on the side has a long ADF tap length,
The signal correction unit on the side shortens the tap length. This enables more accurate signal prediction. In addition, since the LPF side has a small amount of high-frequency noise components, it is possible to adopt a configuration in which all the signals pass through.
Thus, the program amount of the DSP processing unit can be reduced.

(Embodiment 6-10) In the embodiment shown in FIG. 42, the LPF 13 and the HPF 13 are provided before the signal adjusters 6a and 6b.
It is characterized by a structure in which a bandpass filter composed of F14 is provided, and an adjustment control unit 16 is provided in the signal adjustment units 6a and 6b. The adjustment control unit 16 is configured to execute the processing from the sixth embodiment to the sixth embodiment.
Each control mechanism described in -9, for example, a memory amount, a filter,
It incorporates all limit, oscillation prevention, amplitude and gain control mechanisms.

In the system having this configuration, appropriate processing can be performed according to the type of signal, the type of noise, and the like.
More optimal noise removal can be performed. (Embodiment 6-11) The embodiment shown in FIG.
Is characterized in that the adjustment control section 16 described in the embodiment 6-10 is provided in the apparatus of (1). Specifically, a system in which a control unit 16 is provided in the system of the embodiment 2-7 shown in FIG. 16 is shown.

In the system having the above configuration, appropriate processing can be performed according to the type of signal and the type of noise, and more optimal noise removal can be performed. Further, the signal adjusting units 6, 6L, 6
It is possible to set a memory amount corresponding to each of R. For example, in the case of the FM signal, by setting the memory amount to be short in the signal adjusting unit 6 and long in the signal adjusting units 6L and 6R, the high frequency is set in the signal adjusting unit 6 and the low frequency is set in the signal adjusting unit 6L. , 6R. As a result, optimal adjustment becomes possible, the effect of noise removal is further improved, and the DSP processing amount is reduced.

Seventh Embodiment In the noise elimination apparatuses according to the first, second, and third embodiments of the present invention, various examples will be described in which a noise detector has a function of determining the type of noise. (Embodiment 7-1) The embodiment shown in FIG. 44 is characterized in that the noise detection unit 4 of the noise elimination device shown in FIG. 1 is provided with a type determination means 43 for determining the type of noise. For example, in radio broadcasting, the type of noise is determined by extracting the noise component from the AC component of the S meter. In the figure, 1
s is an S meter signal input means, 2s is an AD converter, and the type discriminating means 43 applies a narrow band BPF to the S meter to extract noise characteristics.

In the drawing, reference numeral 41 denotes noise detection means, and reference numeral 42 denotes flag generation means. In the system of the present embodiment, since it is possible to cope with the characteristics of the extraneous noise, finer control of the noise is possible, and the audibility is improved. (Embodiment 7-2) In the embodiment shown in FIG. 45, in place of the noise type discrimination using the S meter shown in FIG. 44, the noise detector 4 having the configuration shown in FIG. Type discriminating means 4 using a signal from the ADF
In g, the type of noise is determined.
The input signal a 'is introduced to one input of the subtractor 4e, and the output Y of the filter unit 4c of the ADF is introduced to the other input. Therefore, when pulse noise is mixed as an unsteady signal into the input signal a ', the output Y of the filter unit 4c is output.
, A non-stationary signal, that is, pulse noise appears at the output E of the subtractor 4e. Therefore, from the noise extracted in this way, the noise type discriminating means 4
g to determine the type, for example, noise level detection,
Detection of noise width, frequency detection, and the like are performed.

In the system of this configuration, if a large value is detected in the level detection, it can be understood that the amplitude difference between the stationary signal and the noise is large. Also, if the width detection has a large value and the frequency is high, it can be understood that the noise is a multipath noise. As a result, a more accurate noise type can be determined, and more accurate noise removal can be performed. (Embodiment 7-3) The embodiment shown in FIG. 46 shows details of the noise detector 4 of the system shown in FIG.
The detection unit 4 includes a band-pass filter (BPF) 4h for a composite signal, a comparator 4i, a signal processing control signal generation circuit 4j for noise cut, a signal processing control signal generation circuit 4k for ADF, and S A meter signal band-pass filter (BPF) 41, a comparator 4m, and a noise type determination signal generation circuit 4n are provided.

In the system having this configuration, in order to switch the noise detection sensitivity, the sensitivity of the comparator 4g in the noise detection system (composite signal system) is controlled as shown in the figure. For example, the noise detection sensitivity at the time of noise A (for example, pulse system noise) is set as a default, and when the noise B (for example, multipath noise) is detected in the type discrimination signal generation circuit 4n, the reference of the comparator 4i is set to the noise. Switch to B (detection sensitivity control 4o). Accordingly, the noise sensitivity can be set according to the type of noise, so that the noise removing effect is improved.

(Embodiment 7-4) The embodiment shown in FIG.
44 shows details of the noise detection unit 4 of the system shown in FIG. 44. In particular, the type of noise is determined, and the result is introduced to a signal processing control signal generation circuit 4j for noise cut, thereby cutting the noise cut. The width is controlled to correspond to the type of noise (cut width control 4p).

In this embodiment, the pulse noise cut width is cut only by the amount detected by the detection unit, while the multipath noise cut width is set to have a cut width longer by a certain time than the detected amount. For example, in the external noise of FM broadcasting, the noise width that appears in the composite
Multipath noise becomes very long compared to pulse noise. However, multipath noise is small as the noise frequency. Further, when viewed microscopically, the frequency of occurrence of noise is higher in multipath noise in the same time width. From such a feature, it is preferable to cut off only the detected noise with respect to the pulse noise in consideration of the noise occurrence frequency (noise repetition frequency). On the other hand, if the multipath noise is cut only for the detected amount, the period between cuts is short, so that the signal after cut has a waveform with a considerable offset,
It is not preferable because harmonics are included therein.

As described above, the multipath noise cut is microscopically detected by the detection unit. Therefore, the correction is performed by the signal generation unit, and as a result, the macro cut is performed. I have. As an effect of the present embodiment, there is an improvement in audibility when multipath noise occurs. (Embodiment 7-5) The embodiment shown in FIG. 48 shows a system in which the ADF flag width is controlled (ADF flag width control 4q) according to the type of the detected noise. Specifically, according to the type of noise detected by the noise type determination signal generation circuit 4j, A
The width of the DF flag is controlled. Also in the present embodiment, as in the case of the embodiment 7-4, by setting the ADF flag width 4q of the multipath noise to be long, it is possible to improve the audibility characteristics at the time of noise removal.

(Embodiment 7-6) The embodiment shown in FIG.
Control is performed to change the time constant of the ADF flag according to the type of the detected noise (ADF flag time constant control 4
r) The configuration is characterized in that: In this embodiment, the ADF flag time constant (release time) when a multipath occurs
Is designed to be longer than the pulse noise system.

In the ADF switching, an input signal and a prediction signal are switched, so that it is conceivable that a feeling of strangeness occurs at the time of switching or noise occurs at the time of switching. Therefore, a time constant is provided in the ADF flag. It is preferable that this time constant is somewhat long. Looking at this by external noise, multipath noise has a low frequency of occurrence of macro-type noise (the generation interval is long), and there is no problem in taking a long ADF flag time constant. However, in the case of pulse noise, when considering vehicle ignition noise, the frequency increases, and if a time constant similar to that of multipath noise is provided, a prediction signal may be output forever. Finally, no sound is output from the speaker.

For this reason, it is desirable to switch the ADF flag time constant according to the noise type, thereby improving the audibility. (Embodiment 7-7) The embodiment shown in FIG.
A system having all the parameter controls in the noise type determination shown in 7, 48 and 49 is shown. Therefore, in the system of this configuration, all the effects are included, and control suitable for pulse noise can be performed for pulse noise and control suitable for multipath noise can be performed for multipath noise.

FIG. 51 shows a waveform (a) for each type of noise.
And a cut flag waveform (b) and an ADF flag waveform (c) suitable for each type of noise waveform. As shown in the waveform diagram (a), the multi-path noises 1, 2, and 3 have a lower macro-frequency (the interval between the multi-path noises is narrower) than the pulse noise, but the micro-pulse shows The frequency of occurrence is high. For example, looking at the multipath noise 1 in FIG. 9A, it can be seen that the frequency between the multipath noises 1, 2, and 3 is low, although the interval between the pulses is narrow. Therefore, the noise type determination signal generation circuit 4n performs control such that the multipath noise is macroscopically regarded as one noise. That is, as shown in FIG. 51 (b), the cut width of the noise in the case of multi-path noise is made longer, and as shown in FIG. 51 (c), the pulse width and flag time constant of the ADF flag are made longer. Note that the detection sensitivity of the multipath noise is low, and for example, the multipath noise 1 is not detected.

Eighth Embodiment Hereinafter, various examples in which the system shown in the first, second, and third embodiments is provided with a means for detecting the reception electric field intensity will be described. In the noise removal of FM broadcasting, the noise detection is strongly affected by the received radio wave condition. For example,
When comparing the detection state of the same pulse noise between a place where the electric field is strong and a place where the electric field is weak, if the noise detection filter can be properly set in the place where the electric field is strong, only noise (external noise in this case) can be normally detected. However, in a place where the electric field is weak, the frequency components of the white noise and the external noise become close to each other, so that only the external noise cannot be properly detected.

As a result, the noise canceling system is not so preferable because it may cause deterioration of the signal component or cause no sound (by continuing to operate the noise canceling). In particular, the system of the present invention has an ADF, and it is important to understand the reception status even from the background that the range of signal processing is wider than that of the conventional system. Therefore, detecting the electric field strength and controlling the signal processing is an effective means. As a result, noise can be removed according to the reception environment, so that the audibility is improved.

(Embodiment 8-1) The embodiment shown in FIG.
The basic system of the present invention shown in the embodiment (1-1) (see FIG. 1) is characterized in that an electric field intensity detecting means is provided. The intensity of the received electric field is detected by taking the S meter signal into the electric field intensity detection unit 44 provided in the noise detection unit 4 via the input unit 1s and the AD converter (ADC) 2s. When taking in the S meter signal,
The signal may be smoothed by applying an LPF.

In this system, the electric field intensity detecting means 44
Based on the electric field strength of the S-meter signal detected in the above, the signal processing operation of the noise detecting means 41 and the flag generating means 42 is controlled. This makes it possible to remove noise corresponding to the reception environment, and the audibility is improved. (Embodiment 8-2) The embodiment shown in FIG.
1) shows a system for varying the noise detection sensitivity according to the received electric field strength (sensitivity detection control 44b).

In this system, the S meter signal is
The signal is input to the electric field intensity signal generation circuit 44a via 4s to obtain an electric field intensity detection signal. As the strength of the detection signal decreases, that is, as the reception electric field becomes weaker, the detection sensitivity decreases linearly (or may be fixed). As a result, erroneous detection due to white noise can be prevented. As a result, it becomes possible to detect noise corresponding to the reception environment, to prevent malfunction of the ADF, and to further improve the audibility.

(Embodiment 8-3) The embodiment shown in FIG.
In the embodiment shown in FIGS. 46 to 49, a system is shown in which the strength of the received electric field is detected, and the noise detection sensitivity and various flags can be arbitrarily set based on the reception environment. In this system, the noise detection sensitivity and various flags can be arbitrarily set according to the state of the external noise and the state of the received electric field, so that the audibility is improved.

Ninth Embodiment Hereinafter, various examples in which the system shown in the first, second and third embodiments is provided with a means for detecting the frequency of noise will be described. (Embodiment 9-1) The embodiment shown in FIG. 55 shows a system in which a means 45 for detecting the frequency of noise is provided in the noise detection unit 4 in the noise removal apparatus shown in FIG. In the present embodiment, the noise frequency is detected by measuring the interval between the flags (noise cut flag or ADF flag) generated by the flag generation unit 42.

It is difficult to predict how often external noise will enter the system. Particularly in an on-vehicle environment, the situation is complicated, and there is a possibility that the noise removal device may rather deteriorate the signal. For example, when the environment is very bad and noise is generated for a long time, the noise elimination device must work in the direction of turning off or reducing the function, but the above-described device is so. Absent. In this embodiment, in consideration of this point, the noise frequency detecting means 45 is provided so that the noise removing operation can be controlled in accordance with the frequency of detecting the external noise.

Therefore, in this system, noise can be sufficiently removed in response to the situation of external noise, so that the audibility and the stability of the noise removal system (ADF) are improved. (Embodiment 9-2) The embodiment shown in FIG. 56 shows a system in which the noise frequency detected by the apparatus shown in FIG. 55 is applied to the noise cut width control 45a. Normally, in this system, when the frequency of noise increases, the signal processing control signal generation circuit 4j narrows the noise cut width to prevent the signal from being excessively cut. As a result, in the present system, it is possible to effectively cope with the situation of the external noise and improve the audibility.

In particular, this system is suitable for use in an on-vehicle radio AM.
Suitable for removing noise during broadcast reception. This is for the following reasons. In general, the pulse response appearing in the detection output of AM broadcasting has a characteristic that the high frequency range is sufficiently dropped due to the necessity of its frequency characteristic, and the pulse response is greatly affected by this, and the pulse response is greatly reduced. . As a result, the pulse width at the detection output becomes very long, about 1 to 2 ms. On the other hand, the pulse noise that occurs repeatedly is proportional to the number of revolutions of the engine when the vehicle is mounted on a vehicle, and the maximum repetition frequency may be 100 Hz or more. When the pulse width is 2 ms, 2
An attempt is made to have a cut width of m seconds. In this case, however, the time of 1/5 of the original signal is cut, and there is a possibility that a sense of incongruity may appear due to excessive cutting of the signal. Apply this system to prevent this and make sure to have an appropriate cut in the noise situation.

(Embodiment 9-3) The embodiment shown in FIG.
The noise frequency detected by the device shown in FIG.
The system used for F flag width control 45b is shown. A
Since it is necessary to form a flag longer than the noise cut width for the DF from the beginning, the effect is great when the apparatus of the present embodiment is applied.

(Embodiment 9-4) The embodiment shown in FIG.
The noise frequency detected by the device shown in FIG.
The system used for control 45c of the time constant of the F flag is shown. The effect of this embodiment is almost the same as that of the embodiment 9-2. In particular, since the ADF originally needs to have a longer time constant than the noise cut, the necessity of the apparatus of this embodiment is high.

(Embodiment 9-5) The embodiment shown in FIG.
55 shows a system in which the noise frequency detected by the device shown in FIG. 55 is used for control 45d when the detection of noise is stopped. The effect of this embodiment is the same as that of the embodiment 9-1, but is particularly suitable for turning off the system when noise continues for a long time.

(Embodiment 9-6) The embodiment shown in FIG.
55 shows a system in which the noise frequency detected by the device shown in FIG. 55 is used for noise detection sensitivity control 45e. In a general noise elimination device, audio AGC is performed to prevent a sound malfunction (erroneous detection of a sound).
Is provided to change the detection sensitivity depending on the power of the signal component. However, the AGC amount is affected by noise, and even if the audio power is the same, if the noise frequency increases, the AGC amount changes accordingly and the noise detection sensitivity decreases (this is correct as an operation). The apparatus according to the present embodiment is designed to correct a decrease in detection sensitivity in such a case.

As a result, an improvement in the audibility is seen. (Embodiment 9-7) In the embodiment shown in FIG. 61, the noise frequency detected by the apparatus shown in FIG. 55 is used for noise cut width control 45a, ADF flag time constant width control 45c, and detection sensitivity control 54e. The following shows the system.

Since the apparatus of this embodiment can effectively cope with external noise, it has effects in terms of improvement in audibility and stability of the noise elimination system. (Embodiment 9-8) The embodiment shown in FIG. 62 shows a system for detecting a noise frequency by using a flag generated by a signal processing control signal generation (for ADF) circuit 4k.

Means for detecting the noise frequency include a method of directly viewing the waveform after the HPF of the detection block, a method of viewing the noise cut flag, a method of viewing the ADF flag, and the like.
Although there are various methods, in this embodiment, noise frequency detection is performed using the ADF flag having the widest pulse width. This is to achieve compatibility between the frequency detection of pulse noise and the frequency of multipath noise when receiving FM broadcasting.

The characteristics of the pulse noise and the multipath noise are as shown in FIG. 51. In order to detect both, the noise frequency needs to be detected macroscopically. If noise frequency detection is performed microscopically, in the case of multipath noise, the noise frequency becomes abnormally higher than that of pulse noise, and the requirements of the system cannot be satisfied.

Therefore, in this embodiment, noise is detected by using the ADF flag having the widest noise. (Embodiment 9-9) In the embodiment shown in FIG. 63, in the system of Embodiment 8-3 shown in FIG. 54, a frequency detecting means 45 is further provided, and the detected noise frequency is controlled by the noise cut width control, ADF The system used for control of the flag time constant width is shown.

In the apparatus of this embodiment, the processing can be performed while monitoring both the noise state and the radio wave state, so that an ideal characteristic can be realized as a noise removing apparatus. In addition, since the characteristics can be arbitrarily changed according to the user's preference, the system is easy to use. Embodiment 10 In the systems shown in Embodiments 1, 2 and 3, a delay unit 50 is arranged in the signal system, and the noise detection unit 4 is used.
A description will be given of an embodiment in which a time delay for creating a flag is eliminated.

(Embodiment 10-1) In the embodiment shown in FIG. 64, a delay unit 50 is provided between the ADC 2 and the noise cut unit 5 to eliminate the time delay for flag creation in the detection unit 4. Shows the system. In the normal noise detection unit 4, filter processing or the like (ADF
(Including detection), a slight time delay occurs. Therefore, when the noise is cut or the ADF is input, the noise is left for the delay. In the apparatus according to the present embodiment, a delay unit 50 is provided for the purpose of preventing this.

Therefore, in the apparatus of this embodiment, it is possible to prevent the noise removal effect from deteriorating due to the detection delay as described above. Embodiment 11 In the systems shown in Embodiments 1, 2 and 3 below, delay units 50a and 50b are arranged in a signal system, and a prediction unit 8 and a changeover switch 9 are provided before noise is cut by a noise cut unit 5. An embodiment of a system in which is operated will be described.

(Embodiment 11-1) In the embodiment shown in FIG. 65, the delay unit 50b is arranged between the noise detection unit 4 and the noise cut unit 5 so that the prediction unit 8 and the switching unit can be switched before the noise cut. The figure shows a system in which the ADF flag is advanced to the switch unit 9. According to this system, the ADF coefficient can be reliably updated before the noise cut, and the changeover switch 9 can be reliably switched and set, so that the ADF processing in the noise section can be performed normally.

[0103]

As described above with reference to each embodiment, the noise elimination apparatus of the present invention replaces the signal in the noise period with the signal predicted from the signal in the period in which no noise occurs. Due to noise removal,
Even for a wide range of noise, the noise can be effectively removed without giving the listener an uncomfortable feeling.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first example of a noise elimination device according to a first embodiment of the present invention.

FIG. 2 is a waveform chart for explaining the operation of the apparatus shown in FIG. 1;

FIG. 3 is a waveform chart for explaining the operation of the apparatus of FIG. 1 together with the waveform chart of FIG. 2;

FIG. 4 is a block diagram of a second example of the noise elimination device according to the first embodiment of the present invention.

FIG. 5 is a block diagram of a third example of the noise elimination device according to the first embodiment of the present invention.

FIG. 6 is a block diagram of a fourth example of the noise elimination device according to the first embodiment of the present invention;

FIG. 7 is a block diagram of a fifth example of the noise elimination device according to the first embodiment of the present invention;

FIG. 8 is a block diagram of a first example of the receiving system according to the second embodiment of the present invention;

FIG. 9 is a block diagram of a second example of the receiving system according to the second embodiment of the present invention;

FIG. 10 is a block diagram of a third example of the receiving system according to the second embodiment of the present invention;

FIG. 11 is a block diagram of a fourth example of the receiving system according to the second embodiment of the present invention;

FIG. 12 is a block diagram of a fifth example of the receiving system according to the second embodiment of the present invention;

FIG. 13 is a block diagram of a sixth example of the receiving system according to the second embodiment of the present invention;

14 is a block diagram showing a part of the device shown in FIG. 13 in detail.

FIG. 15 is a block diagram showing a part of the apparatus shown in FIG. 13 in detail.

FIG. 16 is a block diagram of a seventh example of the receiving system according to the second embodiment of the present invention;

FIG. 17 is a block diagram of a first example of a noise elimination device according to a third embodiment of the present invention.

FIG. 18 is a comparison table showing the effect of the system shown in FIG. 17;

FIG. 19 is a block diagram of a second example of the noise elimination device according to the third embodiment of the present invention.

FIG. 20 is a main part block diagram of a third example of the noise elimination device according to the third embodiment of the present invention;

FIG. 21 is a main part block diagram of a fourth example of the noise elimination device according to the third embodiment of the present invention;

FIG. 22 is a main part block diagram of a fifth example of the noise elimination device according to the third embodiment of the present invention;

FIG. 23 is a main part block diagram of a sixth example of the noise elimination device according to the third embodiment of the present invention;

FIG. 24 is a block diagram of a first example of a receiving system including a noise removing device of the present invention as a fourth embodiment of the present invention.

FIG. 25 is a block diagram of a second example of the receiving system according to the fourth embodiment of the present invention;

FIG. 26 is a block diagram of a third example of the receiving system according to the fourth embodiment of the present invention;

FIG. 27 is a block diagram of a first example showing details of a noise cut unit shown in FIG. 1 as a fifth embodiment of the present invention.

28 is a waveform chart used for describing the operation of the device shown in FIG. 27.

FIG. 29 is a block diagram of a second example showing details of the noise cut unit shown in FIG. 1 as a fifth embodiment of the present invention;

FIG. 30 is a block diagram of a third example showing details of the noise cut unit shown in FIG. 1 as a fifth embodiment of the present invention.

FIG. 31 is a block diagram of a fourth example showing details of the noise cut unit shown in FIG. 1 as a fifth embodiment of the present invention.

FIG. 32 is a block diagram of a fifth example showing details of the noise cut unit shown in FIG. 1 as a fifth embodiment of the present invention.

FIG. 33 is a block diagram of a first example of a noise elimination device according to a sixth embodiment of the present invention.

FIG. 34 is a block diagram of a second example of the noise elimination device according to the sixth embodiment of the present invention.

FIG. 35 is a main part block diagram of a third example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 36 is a main part block diagram of a fourth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 37 is a main part block diagram of a fifth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 38 is a main part block diagram of a sixth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 39 is a main part block diagram of a seventh example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 40 is a main part block diagram of an eighth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 41 is a main part block diagram of a ninth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 42 is a main part block diagram of a tenth example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 43 is a main part block diagram of an eleventh example of the noise elimination device according to the sixth embodiment of the present invention;

FIG. 44 is a block diagram of a first example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 45 is a main part block diagram of a second example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 46 is a block diagram of a third example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 47 is a block diagram of a fourth example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 48 is a block diagram of a fifth example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 49 is a block diagram of a sixth example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 50 is a block diagram of a seventh example of the noise elimination device according to the seventh embodiment of the present invention;

FIG. 51 is a waveform chart used for describing the operation of the device shown in FIG. 50.

FIG. 52 is a block diagram of a first example of the noise elimination device according to the eighth embodiment of the present invention;

FIG. 53 is a block diagram of a second example of the noise elimination device according to the eighth embodiment of the present invention;

FIG. 54 is a block diagram of a third example of the noise elimination device according to the eighth embodiment of the present invention;

FIG. 55 is a block diagram of a first example of a noise elimination device according to a ninth embodiment of the present invention.

FIG. 56 is a block diagram of a second example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 57 is a block diagram of a third example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 58 is a block diagram of a fourth example of the noise elimination device according to the ninth embodiment of the present invention;

FIG. 59 is a block diagram of a fifth example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 60 is a block diagram of a sixth example of the noise elimination device according to the ninth embodiment of the present invention;

FIG. 61 is a block diagram of a seventh example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 62 is a block diagram of an eighth example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 63 is a block diagram of a ninth example of the noise elimination device according to the ninth embodiment of the present invention.

FIG. 64 is a block diagram of a first example of a noise elimination device according to a tenth embodiment of the present invention;

FIG. 65 is a block diagram of a first example of a noise elimination device according to an eleventh embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Input means 2 ... Analog-digital converter 3 ... Digital signal processing apparatus 3a ... High frequency band processing part 3b ... Downsampling part 3c ... Low frequency band processing part 4 ... Noise detection part 4a ... Signal storage memory 4b ... Prediction part 5 … Noise cut section 5a… Register section 5b… 0 set section 5c… Memory 5d… Set section 5e… Comparison section 5f… Limit section 6… Signal adjustment section 7… Signal storage memory 7a… Memory amount control section 8… Prediction section 8a ... Filter section 8b ... Coefficient update section 8c ... Subtractor 8d ... LPF 8e ... Limit section 9 ... Switching switch section 10 ... Digital filter 11 ... Power detection section 12 ... Amplitude adjustment section 13 ... LPF 14 ... HPF 15 ... Adder 16 ... Adjustment control unit 20 Digital-analog converter 41 Noise detection unit 42 Flag generation unit 43 Type Another means 44 ... field intensity detecting means 45 ... frequency detecting unit 200 ... signal processing unit 400 ... noise removal device 500 ... MPX section

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5K046 AA05 BB03 EE06 EE23 EF24 5K052 AA01 BB03 BB04 CC04 DD03 DD04 DD21 EE11 FF05 FF24 FF27 FF33 GG03 GG57

Claims (42)

[Claims] A noise detection unit that detects noise in an input signal; a noise cut unit that cuts a high frequency component of noise from the input signal based on a noise detection result of the noise detection unit; A signal storage unit for temporarily storing the output of the noise cut unit; and a signal storage memory for temporarily storing the output of the noise cut unit, and forming a prediction signal from the output signal of the noise cut unit in a normal state. A signal prediction unit that forms a prediction signal from an output signal of the signal storage memory when noise is detected by the noise detection unit; and a noise cut unit output when noise is detected by the noise detection unit. A noise elimination device including a changeover switch unit that outputs an output of the signal prediction unit. 2. The noise detecting unit includes a signal storage memory into which an input signal is introduced, and a signal prediction unit into which the signal storage memory output is introduced, wherein the signal prediction unit includes a filter unit and the filter unit. A coefficient updating unit for updating a filter coefficient of the unit, and a subtractor for detecting an error between the output of the filter unit and the input signal and inputting the error output to the coefficient updating unit. The noise removing device according to claim 1, wherein the noise removing device is a signal. 3. An FM signal receiving means, an MPX circuit,
A receiving system comprising the noise removing device according to claim 1. 4. The receiving system according to claim 3, wherein the noise elimination device and the MPX circuit are configured as a digital signal processing device and formed as one chip. 5. The signal correction unit according to claim 1, further comprising a pilot signal in the FM signal as an input signal, and introducing an output of the signal correction unit to the MPX circuit. 3. The receiving system according to 3. 6. A downsampling unit is further provided between a high-frequency band processing unit including the noise cut unit and the detection unit and a low-frequency band processing unit including the signal adjustment unit. Noise removal device. 7. The receiving system according to claim 3, wherein the noise removing device is configured by the noise removing device according to claim 6. 8. The receiving system according to claim 3, wherein a downsampling unit is provided between said noise removing device and said MPX circuit. 9. The noise removing apparatus according to claim 1, wherein the input signal is an AM detection wave. 10. The noise removing device according to claim 1, wherein the input signal is a television broadcast audio signal. 11. The receiving system according to claim 3, wherein the FM signal is an audio signal of a television broadcast. 12. The noise removing apparatus according to claim 1, wherein the input signal is a digital broadcast audio signal. 13. The noise cut section includes a register into which an input signal is introduced and a 0 set section that sets a value of the register to 0 based on a noise detection signal from the detection circuit. 7. The noise removing device according to any one of claims 2 and 6. 14. The noise cut unit includes a register into which an input signal is introduced, a memory for temporarily storing an output of the register, and a noise detection signal from the detection circuit to which the content of the memory is input. 7. The noise removing device according to claim 1, further comprising: a setting unit configured to set the input memory content in the register. 8. 15. The noise cut section includes a register into which an input signal is introduced, a limit value setting section for determining a maximum allowable value of the register, and a value in the register compared with a value in the limit value setting section. And a setting unit that sets the value of the limit value setting unit to the register when the value in the register is larger than the value of the limit value setting unit in the comparison. 7. The noise removing device according to any one of 6. 16. The noise cut section includes a digital filter section into which an input signal is introduced, and an LP of the digital filter section based on a noise detection signal from the detection section.
7. The noise removing device according to claim 1, further comprising a coefficient setting unit that sets or resets the F characteristic coefficient. 17. The noise removing device according to claim 16, further comprising a time constant setting unit that sets a time constant of the coefficient setting unit. 18. A memory amount control unit is further provided between the detection unit and the signal storage memory, and determines the type of noise based on information of an input signal and noise detection information from the detection unit. The noise eliminator according to claim 1, wherein the increase / decrease of the memory amount of the signal storage memory is controlled. 19. The noise removing device according to claim 1, wherein a digital filter serving as a low-pass filter (LPF) is further inserted between the signal storage memory and the prediction unit. 20. The prediction unit, comprising: a filter unit, a coefficient update unit, a subtractor that receives an output of the filter unit as one input, and an output of the noise cut unit as another input, and detects an error between the two. 7. The noise removing apparatus according to claim 1, further comprising a low-pass filter section that cuts a high-frequency component of a filter output and inputs the cut-off signal to the coefficient updating section. 21. The prediction unit, comprising: a filter unit; a coefficient update unit; a subtractor that receives the output of the filter unit as one input and detects an error between the two with the output of the noise cut unit as the other input; 7. The noise removal apparatus according to claim 1, further comprising: a limit unit that limits the value of the error output from the filter and inputs the value to the coefficient update unit. 22. The prediction unit, comprising: a filter unit; a coefficient update unit; a subtractor that receives an output of the filter unit as one input, and an output of the noise cut unit as another input to detect an error between the two; 7. The apparatus according to claim 1, further comprising: a band-pass filter configured to filter the error output from the filter for each band, and a gain setting unit configured to set a different gain to the filter output filtered for each band. The noise removing device as described in the above. 23. The prediction unit, comprising: a filter unit; a coefficient update unit; a subtractor that receives an output of the filter unit as one input and an output of the noise cut unit as another input to detect an error between the two; 7. A band pass filter for filtering the error output from the filter for each band, and a limit value setting unit for setting different limit values for the filter output filtered for each band. The noise eliminator according to Item. 24. The prediction unit, wherein a filter unit, a coefficient update unit, and an output of the filter unit are used as one input, an output of the noise cut unit is used as another input, and an error between the two is detected and input to the update unit. The noise elimination device according to claim 1, further comprising a subtractor and an oscillation prevention unit configured to prevent oscillation of the coefficient update unit. 25. The prediction unit, comprising: a filter unit; a coefficient update unit; an output of the filter unit as one input; an output of the noise cut unit as the other input; And a power detection unit that detects the power of the output of the noise cut unit and the output of the filter unit, and further adjusts the output of the filter unit such that the detected powers are approximately the same. The noise removal device according to claim 1, further comprising an amplitude adjustment unit introduced into the changeover switch unit. 26. The detection section, comprising: noise detection means;
The noise according to claim 1, further comprising: a type determination unit configured to determine a type of the noise based on the S meter signal; and a flag generation unit configured to generate a flag based on the output of the noise detection unit and the output of the type determination unit. Removal device. 27. The detection unit, comprising: a signal storage memory connected to an input signal; a signal prediction unit connected to the memory; a filter unit; and a coefficient update unit for updating a filter coefficient of the filter unit. A subtractor for detecting an error between a filter unit output and the noise cut unit output and inputting the error output to the coefficient update unit; and a type of noise based on the subtractor output of the signal prediction unit. The noise elimination device according to claim 1, further comprising: a type determination unit configured to determine the type. 28. The noise removing apparatus according to claim 26, further comprising means for switching a noise detection sensitivity of said noise detecting means based on an output of said noise type determining means. 29. The flag generation means includes a signal processing control signal generation means for determining a cut width in the noise cut section, and the cut width is changed by an output of the noise type determination means. , Claim 2
7. The noise removing device according to 6. 30. The flag generation means includes signal processing control signal generation means for forming a flag signal to be input to the prediction section, and the width of the flag signal is determined based on an output of the noise type determination means. 27. The noise removing device according to claim 26, wherein the device is changed. 31. The flag generating means includes a signal processing control signal generating means for forming a flag signal to be input to the prediction section, and a time constant of the flag signal based on an output of the noise type determining means. Is to change the
The noise removing device according to claim 26. 32. The noise detection unit, comprising: a noise detection unit; an electric field intensity detection unit that detects an electric field intensity of an input signal from an S meter signal; and an output of the noise detection unit and an output of the electric field intensity detection unit. 7. The noise removing apparatus according to claim 1, further comprising: a flag generating unit that forms a flag. 33. The noise removing apparatus according to claim 32, further comprising: means for switching a noise detection sensitivity of said noise detecting means based on an output of said electric field strength detecting means. 34. The noise detection unit, wherein: a noise detection unit; a frequency detection unit for detecting a frequency of the noise; and a flag generation unit for forming a flag based on the output of the noise detection unit and the output of the noise frequency detection unit. The noise removal device according to claim 1, comprising: 35. The flag generating means includes a signal processing control signal generating means for determining a cut width in the noise cut section, and changing the cut width based on an output of the noise frequency detecting means. 35. The noise removing device according to claim 34, wherein: 36. The flag generation means includes a signal processing control signal generation means for forming a flag signal to be input to the prediction unit, and the width of the flag signal is determined based on an output of the noise frequency detection means. 35. The noise removing device according to claim 34, wherein the device is changed. 37. The flag generating means includes a signal processing control signal generating means for forming a flag signal to be input to the prediction section, and a time constant of the flag signal based on an output of the noise frequency detecting means. Is to change the
35. The noise removing device according to claim 34. 38. The noise eliminator according to claim 34, further comprising means for turning off the operation of the noise eliminator in response to the output of the frequency detection means. 39. The noise removing apparatus according to claim 34, further comprising means for varying the noise detection sensitivity in accordance with the frequency detection means output. 40. The flag generation means includes signal processing control signal generation means for forming a flag signal to be input to the prediction unit, and the frequency detection means includes a rising edge of an ADF flag as an output of the signal generation means. The noise removal device according to claim 34, wherein the frequency is detected by detecting an interval. 41. The noise removing device according to claim 1, wherein a delay unit is provided at a stage preceding the noise cut unit. 42. The apparatus according to claim 41, wherein a second delay unit is arranged between the detection unit and the noise cut unit.
The noise eliminator according to 1.