JP3784431B2 - Tone detection device and tone detection method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention belongs to the field of telecommunications, and particularly relates to a technique for detecting dial tone in telephone communications.
[0002]
[Prior art]
As a conventional tone detection device, for example, a configuration shown in FIG. 6 is known. FIG. 6 is a block diagram showing a basic configuration of a conventional tone detection apparatus, in which 11 is tone band filtering means, 12 is out-of-band filtering means, 16 and 17 are amplitude averaging means, tone / out-of-band level comparison judging means, 19 is a judgment means.
[0003]
Next, the operation of the apparatus having such a configuration will be described with reference to FIG. FIG. 7 is a flowchart showing the operation of the conventional apparatus. First, when a data signal is input (S21), the data signal is input to the tone band filtering means 11 and the out-of-band filtering means 12, respectively, and separated into a tone band signal and an out-of-band signal (S22).
[0004]
For the separated tone band signal and out-of-band signal, respective average amplitude levels are obtained by the first and second amplitude averaging means 16 and 17 (S23). The tone signal level and the out-of-band signal level obtained by the first and second amplitude averaging means 16 and 17 are compared with each other by the tone / out-of-band level comparing and judging means 18 (S24, S25). If the difference between the two is smaller than the specified value, it is determined that the input is a voice input (S26). If the difference is larger than the specified value, it is determined that the signal is a tone signal, and the subsequent processing is performed (S27).
[0005]
[Problems to be solved by the invention]
However, the quality of human voice (mainly frequency bands) varies, and depending on the voice quality, the difference between the tone signal level and the out-of-band signal level may be larger than the specified value. In some cases, it was erroneously determined as tone detection even with voice input.
[0006]
The present invention has been made in view of the above points, and an object of the present invention is to provide a tone detection apparatus and a tone detection method that do not erroneously detect as a tone signal no matter what quality of speech is input. There is.
[0007]
[Means for Solving the Problems]
In order to solve the above problems, the present invention includes a tone band filtering unit 1 for filtering a tone band signal, and a first amplitude averaging unit 6 for obtaining an average amplitude level of the tone band signal filtered by the tone band filtering unit 1. An out-of-band filtering unit 2 for filtering a signal other than the tone band signal, a second amplitude averaging unit 7 for obtaining an average amplitude level of the out-of-band signal filtered by the out-of-band filtering unit 2, and the first Tone / out-of-band level for comparing whether the tone signal level obtained by the first amplitude averaging means 6 and the out-of-band signal level obtained by the second amplitude averaging means 7 are larger than the specified value In the tone detection apparatus having the comparison judgment means 8, the voice pitch extraction means 3 for extracting the voice pitch component from the tone band signal, and the voice pitch extraction means 3 extract the voice pitch component. A third amplitude averaging means 4 for obtaining an average amplitude level of the voice pitch signal, and a pitch for judging whether or not the voice pitch level averaged by the third amplitude averaging means 4 is larger than a threshold value. The apparatus includes a determination unit 5 and a determination unit 5 that determines whether the tone signal is output from the output of the pitch determination unit 5 and the output 8 of the tone / out-of-band level comparison determination unit.
[0008]
Alternatively, the tone band signal and the out-of-band signal are separated (S1), the respective average amplitude levels of both signals are obtained (S4), and the obtained tone signal level and the out-of-band signal level are compared (S5). In the tone detection method, when a tone signal is determined to be a predetermined value or more (S6, S7), an audio pitch component is extracted from the separated tone signal (S3), and an average amplitude level of the obtained audio pitch signal (S4), the voice pitch level thus obtained is compared with a threshold value (S8), and the tone signal level and the out-of-band signal level are compared with each other and become equal to or more than a prescribed value. A method of determining that the signal is a tone signal when it is below the threshold (S9, S10, S11) is used.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Next, an embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a block diagram showing the basic configuration of a tone detection apparatus according to the present invention. In the figure, 1 is tone band filtering means for filtering the frequency band of the tone signal, 2 is out-of-band filtering means for filtering the frequency band other than the tone signal band, and 6 is the tone signal filtered by the tone band filtering means 1. A first amplitude averaging means 7 that averages amplitudes and obtains an average level value thereof, and 7 averages signals other than the tone band filtered by the out-of-band filtering means 2 and obtains an average level value thereof. 2 amplitude averaging means, 8 compares the tone signal level and the out-of-band level, which is the average value of the level of the tone signal and the tone out-of-band signal, averaged by the first and second amplitude averaging means, It is a tone / out-of-band level comparison judging means for judging whether or not it is larger than a prescribed value.
[0010]
3 is a voice pitch extracting means for extracting a pitch component of the voice from the tone signal filtered by the tone band filtering means 1, and 4 is a third for obtaining an average level of the pitch signal obtained from the voice pitch extracting means 3. Amplitude averaging means 5 for comparing the averaged voice pitch level with a predetermined threshold value to determine whether or not the voice pitch level is greater than the threshold value, 9 for the tone / From the output of the determination results of the out-of-band level comparison determination means and the pitch determination means and the output of the first amplitude averaging means 6, it is determined whether it is a voice input or a tone signal, and the determination of the tone signal level and the subsequent processing are performed. It is a judgment means.
[0011]
Next, the main operation of the apparatus having such a configuration will be described. In addition to the conventional apparatus or method, the present invention is characterized in that an audio pitch component is extracted, an average level of the audio pitch signal is compared with a predetermined threshold value, and this is used as a judgment element for tone detection. Have. Thus, paying attention to the presence of pitch in human voiced sound, the degree of pitch included in the input signal is detected to distinguish between voice and tone signals.
[0012]
FIG. 2 conceptually shows the operation of the tone signal extraction means 3, the third amplitude averaging means 4, and the pitch judgment means 5 which are characteristic parts of the present invention, and other parts are omitted. is there. As shown in FIG. 1, the tone signal extracting means 3 is composed of a half-wave rectifying means 31, an envelope detecting means 32, and a pitch band filtering means 33. Note that the detailed configuration for extracting the voice pitch is not limited to this, and it is only necessary that the same function can be obtained by other technical means.
[0013]
The tone band signal filtered by the tone band filtering means 1 is half-wave rectified by the half-wave rectifying means 31 of the tone signal extracting means 3, and a positive waveform is extracted. Next, envelope detection is performed by the envelope detection means 32, and a pitch component signal is extracted. Only the pitch component of the voice frequency band is extracted from the extracted pitch signal by the pitch band filtering means 33 (for example, a band pass filter of fc = 80-500 HZ) that filters only the voice band. The extracted voice pitch signal is averaged by the third amplitude averaging means 4 to obtain an average level. The obtained voice pitch level and a predetermined threshold value are compared by the pitch determination means 5 to determine whether or not it is greater than the threshold value.
[0014]
If it is larger than the threshold value, it is recognized that there is a voice pitch component, and it is determined that the voice is input. The threshold value is obtained by an empirical rule. Thus, by extracting a pitch component peculiar to voice, it is possible to reliably discriminate between voice input and tone signal input, and to prevent malfunction. Further, here, the operation is described by analog expression means for convenience of explanation, but in reality, it handles a digitized PCM signal, and the above-described elements are constituted by well-known digital signal processing means.
[0015]
Next, the tone detection operation will be described. FIG. 3 is a flowchart showing the tone detection operation. First, a data signal such as a tone signal or an audio signal is input (S1). The tone band signal and the out-of-band signal are separated from the input data signal by the tone band filtering unit 1 and the out-of-band filtering unit 2 (S2). Then, the audio pitch signal is extracted from the separated tone band signal by the audio pitch extraction means 3 (S3).
[0016]
Average levels of the tone band signal, the out-of-band signal, and the voice pitch signal are obtained by the first amplitude averaging means 6, the second amplitude averaging means 7, and the third amplitude averaging means 4, respectively. (S4). The tone signal level obtained from the first amplitude averaging means and the second amplitude averaging means is compared with the out-of-band level (S5). If the difference between the two is not greater than the specified value (S6), it is determined that the voice is input (S7). If the difference between the two is equal to or greater than the specified value (S6), the voice pitch level obtained by the third amplitude averaging means is compared with a predetermined threshold value (S8). If the voice pitch level is equal to or higher than the threshold value (S9), it is determined that the voice is input (S10). If the voice pitch level is equal to or lower than the threshold value (S9), it is determined that the tone signal is detected, and the subsequent processing is performed. Proceed (S11).
[0017]
Thus, by determining the voice pitch in addition to the conventional method, it is possible to reliably detect tone signals for voices of any voice quality, and in various audible sounds, DTMF signals, and data communications on a communication line. Answer tones can be detected reliably. Further, although signals having two different frequencies are used in the DTMF signal, even in this case, the voice and the tone signal can be reliably detected without being confused.
[0018]
【Example】
Next, preferred embodiments of the present invention will be described. FIG. 4 shows a first embodiment of the present invention. The boot ROM 22 stores an algorithm (program) for realizing each means or method shown in FIG. 1 or 3, and the DSP ( 1 is read into the internal memory of the Digital Signal Processor), and the DSP 21 operates as a device having the means shown in FIG. 1 or a device for realizing the method shown in FIG. The input data is determined by the DSP 21, and the determination result is output as 1-bit data from the tone detection result output port of the DSP 21.
[0019]
FIG. 5 shows a second embodiment of the present invention. In this example, instead of the boot ROM 22 of FIG. 4, a CPU 23 which is another processor is provided, and a program similar to the above is loaded from the CPU 23 to the internal memory of the DSP 21. Read in. The determination result is sent to the CPU 23 through the internal bus. As described above, tone detection is performed by using the DSP 21 operating at high speed and digital signal processing by software, so that tone detection can be performed with high speed and reliability with a small and simple configuration. In addition, it is possible to obtain an equivalent function by analog signal processing if a known signal processing means is used. However, as described above, digital signal processing requires a small and simple configuration, and has a predetermined configuration. Since the number can be easily changed, it can be said that the configuration of the above embodiment is excellent in this respect.
[0020]
【The invention's effect】
As described above, according to the apparatus of the present invention, in addition to the result from the tone / out-of-band level comparison determination unit, the determination of the audio pitch by the audio pitch extraction unit, the third amplitude averaging unit, and the pitch determination unit is also determined. Because you are typing in the means
Tone signals can be reliably detected even for voices of any quality, various audible sounds on communication lines, DTMF signals, answer tones in data communication, etc. can be reliably detected, and there are two different types of DTMF signals. Even in this case, it is possible to provide a device that can reliably detect the sound and the tone signal without confusion.
[0021]
According to the method of the present invention, in addition to the conventional method, the voice pitch level is extracted and compared with a predetermined threshold value.
Tone signals can be reliably detected even for voices of any quality, various audible sounds on communication lines, DTMF signals, answer tones in data communication, etc. can be reliably detected, and there are two different types of DTMF signals. Even in this case, it is possible to provide a method capable of reliably detecting the voice and the tone signal without confusion.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a basic configuration of an apparatus according to the present invention.
FIG. 2 is a diagram conceptually showing main actions of an audio pitch extraction means, a third amplitude averaging means, and a pitch determination means.
FIG. 3 is a flowchart showing the operation or method of the apparatus according to the present invention.
FIG. 4 is a diagram showing a first embodiment of the present invention.
FIG. 5 is a diagram showing a second embodiment of the present invention.
FIG. 6 is a block diagram showing a schematic configuration of a conventional tone detection apparatus.
FIG. 7 is a flowchart showing an operation of a conventional tone detection apparatus or a tone detection method.
[Explanation of symbols]
1 Tone band filtering means 2 Out-of-band filtering means 3 Voice pitch extracting means 4 Third amplitude averaging means 5 Pitch judging means 6 First amplitude averaging means 7 Second amplitude averaging means 8 Tone / out-of-band level comparison Determination means 9 Determination means 31 Half-wave rectification means 32 Envelope detection means 33 Pitch band filtering means

Claims (2)

Tone band filtering means for filtering the tone band signal, first amplitude averaging means for obtaining an average amplitude level of the tone band signal filtered by the tone band filtering means, and a band for filtering signals other than the tone band signal An outer filtering means, a second amplitude averaging means for obtaining an average amplitude level of the out-of-band signal filtered by the out-of-band filtering means, a tone signal level obtained by the first amplitude averaging means, and a second A tone detection apparatus having a tone / out-of-band level comparing / determining unit for comparing with an out-of-band signal level obtained by the amplitude averaging unit of
Voice pitch extraction means for extracting a voice pitch component from the tone band signal, third amplitude averaging means for obtaining an average amplitude level of the voice pitch signal extracted by the voice pitch extraction means, and the third amplitude average A pitch determination means for determining whether or not the voice pitch level averaged by the conversion means is greater than a threshold value, and whether the tone signal is obtained from the output of the pitch determination means and the output of the tone / out-of-band level comparison determination means A tone detection apparatus comprising: a determination means for determining whether or not. The tone band signal and the out-of-band signal are separated, the average amplitude level of both signals is obtained, the obtained tone signal level is compared with the out-of-band signal level, and it is determined that the signal is a tone signal when the level is equal to or greater than a specified value. In the tone detection method,
A voice pitch component is extracted from the separated tone signal, an average amplitude level of the obtained voice pitch signal is obtained, a voice pitch level obtained thereby is compared with a threshold value, and the tone signal level is out of band. A tone detection method characterized in that a tone signal is determined when the signal level is equal to or greater than a predetermined value by comparison with a signal level and when the signal level is less than or equal to the threshold value.

Images