JP6828639B2 - Analyzer, display method and display program of analyzer - Google Patents

Description

The present invention relates to an analyzer that analyzes an amplitude modulation component, a display method and a display program of the analyzer.

Conventionally, a method called DEMON (Dection of Modulation On Noise) has been used as a method for analyzing wideband noise amplitude-modulated by a narrow band component generated by the operation of a rotation mechanism. Information on the rotation mechanism can be obtained by analyzing the wideband noise amplitude-modulated by the DEMON process.

The object having a rotating mechanism has, for example, a propeller such as a ship, a helicopter, or a drone, and examples of information on the rotating mechanism of the object include the shaft rotation speed and the number of blades. Since the information about the rotation mechanism obtained by the DEMON process is the information unique to the object, it is important to acquire the information about the rotation mechanism when identifying the object.

The observation data handled by the DEMON process is that the amplitude of the broadband noise is periodically modulated by the rotation of the shaft or the like. The observation data includes a signal component caused by the rotation of the shaft and the like, a wideband noise component, and a noise component different from the wideband noise. In the DEMON process, signal components are extracted from the observation data, and frequency analysis is performed on the extracted signal components. As a result, information on the rotation mechanism can be obtained.

By the way, as a new DEMON method for effectively analyzing signal components, a method called CMC (Cyclic Modulation Coherence) has been proposed (see, for example, Non-Patent Documents 1 and 2). By analyzing the observation data using the CMC, it is possible to accurately analyze the signal component and the harmonic component caused by the rotation mechanism of the sound source.

D.Hanson et.al, “Cyclostationarity for ship detection using passive sonar”, Acoustics, 2009 J.Antoni et.al, “Detection of Surface Ships From Interception of Cyclostationary Signature With the Cyclic Modulation Coherence”, IEEE Journal of OCEANIC ENGINEERING Vol.37, No3, 2012

However, when the signal components and harmonic components caused by a plurality of sound sources are observed in a mixed manner, when the analysis results of the signal components and harmonic components are displayed using CMC, the respective signal components and harmonic components are displayed. The correspondence with and is unclear.

Therefore, an analyzer that displays the signal components and harmonic components so that the correspondence between the signal components and the harmonic components becomes clear when the signal components and harmonic components caused by a plurality of sound sources are mixed and observed. , Display methods and display programs for analyzers are desired.

The analyzer according to the present invention is an analyzer that analyzes amplitude-modulated observation data, and performs Fourier conversion processing on the observation data a plurality of times to show the relationship between acoustic frequency and periodic frequency. The FFT unit that calculates the above and the acoustic frequency direction of the periodic modulation coherence are divided into a plurality of acoustic frequency groups, and the periodic modulation coherence is band-added in the acoustic frequency direction for each band of the divided acoustic frequency groups to produce acoustics. The partial band addition unit that calculates the band addition output in the frequency direction, and the maximum value in the acoustic frequency direction and the index of the divided acoustic frequency group are detected for each periodic frequency of the band addition output in the acoustic frequency direction. The signal with the value detection unit, the color arrangement processing unit that determines the color arrangement when displaying the signal component and the harmonic component obtained from the band addition output in history based on the maximum value and the index, and the determined color arrangement. It is provided with a display generation unit that generates a history display of the component and the harmonic component.

Further, the display method according to the present invention is a display method of an analysis result by an analyzer that analyzes amplitude-modulated observation data, in which the observation data is subjected to Fourier conversion processing a plurality of times to obtain an acoustic frequency and a periodic frequency. The FFT step for calculating the periodic modulation coherence showing the relationship with, and the acoustic frequency direction of the periodic modulation coherence are divided into a plurality of acoustic frequency groups, and the periodic modulation coherence is divided into the acoustic frequencies for each band of the divided acoustic frequency groups. A partial band addition step that adds bands in the direction to calculate the band addition output in the acoustic frequency direction, and the maximum value in the acoustic frequency direction and the divided acoustic frequency for each periodic frequency of the band addition output in the acoustic frequency direction. A maximum value detection step for detecting the index of the group, and a color arrangement step for determining the color arrangement when historically displaying the signal component and the harmonic component obtained from the band addition output based on the maximum value and the index. It has a display generation step of generating a history display of the signal component and the harmonic component in the color arrangement.

Further, the display program according to the present invention is a display program that causes a computer to execute a display method of analysis results by an analyzer that analyzes amplitude-modulated observation data, and performs Fourier conversion processing on the observation data a plurality of times. The FFT step for calculating the periodic modulation coherence indicating the relationship between the acoustic frequency and the periodic frequency, and the acoustic frequency direction of the periodic modulation coherence are divided into a plurality of acoustic frequency groups, and each band of the divided acoustic frequency groups is divided. The partial band addition step for calculating the band addition output in the acoustic frequency direction by band-adding the periodic modulation coherence in the acoustic frequency direction, and the maximum value in the acoustic frequency direction for each periodic frequency of the band addition output in the acoustic frequency direction. And the maximum value detection step for detecting the index of the divided acoustic frequency group, and the color arrangement when historically displaying the signal component and the harmonic component obtained from the band addition output based on the maximum value and the index. The color arrangement step to be determined and the display generation step to generate the history display of the signal component and the harmonic component with the determined color arrangement are executed.

As described above, according to the present invention, the signal component and the harmonic component are determined based on the maximum value in the acoustic frequency direction detected for each periodic frequency of the band addition output in the acoustic frequency direction and the index of the acoustic frequency group. Determine the color arrangement when displaying the history. As a result, the signal component and the harmonic component can be displayed so that the correspondence between the signal component and the harmonic component becomes clear.

It is a graph which shows an example of the observation data input to an analyzer. It is a block diagram which shows an example of the structure of the conventional analyzer which performs DEMON processing using CMC. It is a graph which shows an example of the snapshot image of the periodic modulation coherence calculated by the division part of FIG. It is a graph which shows an example of the history display of the band addition output generated by the display generation part of FIG. It is a block diagram which shows an example of the structure of the analyzer which concerns on Embodiment 1. FIG. It is a graph for demonstrating the periodic modulation coherence when the whole acoustic frequency is divided into three. It is a schematic diagram for demonstrating the case of band addition of the periodic modulation coherence for each acoustic frequency group divided into a plurality of parts. It is the schematic for demonstrating the maximum value detection by the maximum value detection part of FIG. It is a graph which shows an example of the relationship between the acoustic frequency group number and a band addition output at a periodic frequency α ₀ . It is the schematic for demonstrating the method of determining the color scheme by the color scheme processing part of FIG. It is a schematic diagram which shows an example of the history display by Embodiment 1. FIG. It is a graph for demonstrating the case where the periodic modulation coherence shown in FIG. 6 is divided into 9 in the acoustic frequency direction. It is a graph for demonstrating the index when the periodic modulation coherence is divided into 9 in the acoustic frequency direction. It is the schematic which shows an example of the history display when the periodic modulation coherence is divided into 9 in the acoustic frequency direction. It is a block diagram which shows an example of the structure of the analyzer which concerns on Embodiment 2. FIG. It is a graph for demonstrating the band addition output which added the band in the periodic frequency direction. It is a graph which shows the relationship between the index of an acoustic frequency group, and the band addition output when the maximum value is band-added in the periodic frequency direction. It is a graph for demonstrating the maximum value detected by the maximum value detection part of FIG. It is the schematic which shows an example of the history display by Embodiment 2. FIG.

Embodiment 1.
Hereinafter, the analyzer according to the first embodiment of the present invention will be described. The analyzer extracts the signal component caused by the sound source from the input observation data, and displays the extracted signal component and harmonic component. First, before explaining the analyzer according to the first embodiment, the analyzer using CMC will be described.

(About observation data)
FIG. 1 is a graph showing an example of observation data input to the analyzer. In FIG. 1, the horizontal axis represents time and the vertical axis represents amplitude. As shown in FIG. 1, in the observation data, the amplitude of the broadband noise is amplitude-modulated by the signal component. Here, the signal component is a signal component caused by the rotation of the shaft or the like.

When the observation data is x [t], the observation data x [t] can be expressed by the equation (1). In the formula (1), s [t] represents a signal component and N [t] represents wideband noise. Further, v [t] indicates noise different from wideband noise. m indicates a coefficient representing the magnitude of amplitude modulation, and t indicates a time index.

In the DEMON process, the signal component s [t] is extracted from the observation data x [t] shown in FIG. Then, by performing frequency analysis on the extracted signal component s [t], information on rotation can be acquired.

(Construction of conventional analyzer)
FIG. 2 is a block diagram showing an example of the configuration of a conventional analyzer 100 that performs DEMON processing using CMC. The configuration of the analyzer 100 will be described with reference to FIG.

The analyzer 100 is composed of software executed on an arithmetic unit such as a CPU (Central Processing Unit), hardware such as a circuit device that realizes various functions, and the like. As shown in FIG. 2, the analyzer 100 includes a first FFT (Fast Fourier Transform) unit 110, a second FFT unit 120, a division unit 130, a band addition unit 140, a display generation unit 150, and a display unit 160. Observation data x [t] is input to the analyzer 100.

The first FFT unit 110 performs the first Fourier transform process FFT on the observed data x [t], squares the amplitude, and calculates the spectrogram X [f, t]. The calculated spectrogram X [f, t] is supplied to the second FFT unit 120. Although f represents a frequency, it is referred to as an acoustic frequency here in order to distinguish it from a periodic frequency described later.

The second FFT unit 120 performs the second FFT with respect to the time axis direction of the spectrogram X [f, t] supplied from the first FFT unit 110, squares the amplitude, and squares the spectrogram X [f, α, t]. Is calculated. The calculated spectrogram X [f, α, t] is supplied to the division unit 130. α is a periodic frequency that serves as an index indicating the rotation frequency of the rotation mechanism.

The division unit 130 performs a process of dividing the spectrogram X [f, α, t] supplied from the second FFT unit 120 by a component having a periodic frequency α of 0 Hz for each acoustic frequency f, and the periodic modulation coherence C. Calculate [f, α, t]. The calculated periodic modulation coherence C [f, α, t] is supplied to the band addition unit 140.

The band addition unit 140 adds bands to the periodic modulation coherence C [f, α, t] supplied from the division unit 130 over the entire band in the acoustic frequency direction, and the band addition output IC [α, t]. Is calculated. The calculated band addition output IC [α, t] is supplied to the display generation unit 150.

The display generation unit 150 generates a history display based on the band addition output IC [α, t] supplied from the band addition unit 140. In the history display, the output level of the band addition output IC [α, t] is displayed in shades when the horizontal axis is the periodic frequency α and the vertical axis is the time. The display unit 160 displays the history display generated by the display generation unit 150.

(History display)
FIG. 3 is a graph showing an example of a snapshot image of the periodic modulation coherence C [f, α, t] calculated by the division unit 130 of FIG. In FIG. 3, the horizontal axis represents the periodic frequency α, and the vertical axis represents the acoustic frequency f. In the example shown in FIG. 3, it is shown that there are three signal components, each of which modulates the amplitude of wideband noise in different acoustic frequency bands.

In general, when the amplitude modulation component of wideband noise amplitude-modulated by the signal component of the rotation mechanism is analyzed, the harmonic component of the fundamental period frequency is also observed in addition to the signal component of the basic period frequency. Here, how the harmonic component is generated is determined according to the source of the signal component. For example, when the source of the signal component is a propeller, how the harmonic component is generated is determined by the number of blades of the propeller.

FIG. 3 shows that a harmonic component for the signal component of the basic periodic frequency is observed, and the component in the region surrounded by the alternate long and short dash line A is the harmonic component for the signal component of the basic periodic frequency α _0. Is. Here, the output level of the harmonic component with respect to the signal component of the fundamental periodic frequency α ₀ decreases as the order increases with the second harmonic and the third harmonic, but in the fourth harmonic, the harmonic component decreases. The output level is rising. In this way, since the output level of the harmonic component changes from decreasing to increasing in the fourth harmonic, it can be estimated that the number of blades of the propeller is four.

FIG. 4 is a graph showing an example of history display of the band addition output IC [α, t] generated by the display generation unit 150 of FIG. In FIG. 4, the horizontal axis represents the periodic frequency α and the vertical axis represents the time t. FIG. 4 shows a band addition output IC [α, t] in which the periodic modulation coherence C [f, α, t] of FIG. 3 is band-added over the entire band in the acoustic frequency direction.

Here, the signal component and the harmonic component of the rotation mechanism do not always have a high output level. Therefore, it is not possible to reliably display the signal component and the harmonic component only by displaying the periodic modulation coherence C [f, α, t] shown in FIG. Therefore, in the DEMON processing using the CMC, even if the signal component is weak, it can be recognized by calculating the band addition output IC [α, t] having an integral effect on the display.

(Problems of history display)
As shown in FIG. 4, when the history display is generated by using the CMC, the signal component and the harmonic component caused by a plurality of sound sources are observed in a mixed manner. Therefore, even if the history display is confirmed, it is difficult to analyze the basic periodic frequency and the harmonic component of each signal component. Specifically, it becomes difficult to analyze how many signal components there are and which signal component the harmonic component corresponds to.

On the other hand, when the display shown in FIG. 3 is displayed as a history display, the observation data at a certain time can be analyzed. However, in the display of FIG. 3, since the integral effect on the display cannot be obtained, there is a possibility that a weak signal component and a harmonic component cannot be detected.

Therefore, in the display shown in FIGS. 3 and 4, it is not possible to reliably detect the signal component and the harmonic component caused by the plurality of sound sources. Therefore, in the first embodiment, the signal component and the harmonic component caused by a plurality of sound sources can be reliably detected.

[Analyzer configuration]
FIG. 5 is a block diagram showing an example of the configuration of the analyzer 1 according to the first embodiment. The configuration of the analyzer 1 will be described with reference to FIG. In the following description, the parts common to the analyzer 100 shown in FIG. 2 are designated by the same reference numerals, and detailed description thereof will be omitted.

As shown in FIG. 5, the analyzer 1 includes a first FFT unit 110, a second FFT unit 120, a division unit 130, a partial band addition unit 40, a maximum value detection unit 70, a color scheme processing unit 80, a display generation unit 50, and a display unit. It is composed of 160. Observation data x [t] is input to the analyzer 1.

The partial band addition unit 40 divides the acoustic frequency direction of the periodic modulation coherence C [f, α, t] supplied from the division unit 130 into a set number of divisions such as K division, and the periodic modulation coherence C [f, α, t] t] is grouped by the acoustic frequency of each F / K in the acoustic frequency direction. The number F of the acoustic frequency is preferably an integral multiple of the number of divisions K, but in other cases, the entire acoustic frequency may be divided into K by performing processing such as rounding the F / K. ..

When the output of the k-th acoustic frequency group is C _k [f, α, t] (k = 0, 1, ..., K), the partial band addition unit 40 has a periodic modulation coherence C _k [f, For α, t], bands are added in the acoustic frequency direction in each acoustic frequency group. Then, the partial band addition unit 40 outputs the band addition output IC _k [α, t] as a result of the band addition.

FIG. 6 is a graph for explaining the periodic modulation coherence when the entire acoustic frequency is divided into three. In the example shown in FIG. 6, the entire acoustic frequency is divided into three, and the periodic modulation coherences C ₁ [f, α, t], C ₂ [f, α, t] and C corresponding to each of the three acoustic frequency groups. _It can be seen that ₃ [f, α, t] is obtained.

FIG. 7 is a schematic diagram for explaining a case where the periodic modulation coherence for each acoustic frequency group divided into a plurality of units is band-added. In the example shown in FIG. 7, the band addition outputs IC ₁ [α, t], IC ₂ [α, t] and IC ₃ [α, t] are derived for each of the acoustic frequency group numbers k 1-3. Has been done. In this way, by band-adding the periodic modulation coherence C _k [f, α, t] for each acoustic frequency group in the acoustic frequency direction, the band addition output IC _k [α, corresponding to each acoustic frequency group number k] t] is output from the partial band addition unit 40.

The maximum value detection unit 70 of FIG. 5 sets the maximum value peak [α, t] in the k direction for each periodic frequency α with respect to the band addition output IC _k [α, t] supplied from the partial band addition unit 40. , The index k [α, t] that takes the maximum value peak [α, t] is detected. The detected maximum value peak [α, t] and index k [α, t] are supplied to the color scheme processing unit 80.

FIG. 8 is a schematic view for explaining the maximum value detection by the maximum value detection unit 70 of FIG. In the example shown in FIG. 8, similarly to FIG. 7, the band addition output IC ₁ [α, t], IC ₂ [α, t] and IC ₃ derived when the acoustic frequency group numbers k are 1 to _{3 are obtained.} Indicates [α, t]. The maximum value detection unit 70 derives the output level of the band addition output IC _k [α, t] in each acoustic frequency group k for a certain periodic frequency α.

FIG. 9 is a graph showing an example of the relationship between the acoustic frequency group number and the band addition output at the periodic frequency α ₀ . The example of FIG. 9 shows the maximum value peak [α ₀ , t] when the acoustic frequency group k is 1 to 3 for the periodic frequency α ₀ .

As shown in FIG. 9, it can be seen that the maximum value peak [α ₀ , t] of the band addition output at the periodic frequency α ₀ is when the acoustic frequency group k is 1. That is, at the periodic frequency α ₀ , when the index k [α ₀ , t] is 1, the maximum value peak [α ₀ , t] is taken. The maximum value detection unit 70 performs such processing for all periodic frequencies α, and supplies the maximum value peak [α, t] and the index k [α, t] obtained by the processing to the color arrangement processing unit 80. ..

The color scheme 80 of FIG. 5 determines the color scheme when displaying the signal components in history based on the maximum value peak [α, t] and the index k [α, t] supplied from the maximum value detection unit 70. Generate color information for. Here, as an example, an HSV color space in which a color is determined by three parameters of hue H, saturation S, and lightness V is used as the color space, and the color arrangement processing unit 80 uses hue H, saturation S, and color information as color information. Calculate the brightness V. The color information generated in this way is supplied to the display generation unit 50.

FIG. 10 is a schematic view for explaining a method of determining a color scheme by the color scheme processing unit 80 of FIG. As shown in FIG. 10, the hue H is a parameter indicating the type of color of the signal component, and is determined using the equation (2) based on the index k [α, t]. In the formula (2), H _max and H _min indicate the maximum value and the minimum value that the hue H can take at the time of display, respectively. Generally, the range that hue H can take is 0 to 360. Therefore, the maximum value H _max of the hue H is 360 at the maximum, and the minimum value H _min is 0 at the minimum.

The saturation S is a parameter indicating the vividness of the color of the signal component, and is determined by using the equation (3) based on the maximum value peak [α, t]. In the formula (3), S _min and S _max indicate the maximum value and the minimum value that the saturation S can take at the time of display, respectively. Bias _S indicates a set value to be subtracted from the maximum value peak [α, t]. G _S indicates a coefficient used when converting the maximum value peak [α, t] into saturation S. Further, max (x, y) is a function that returns the larger value of x and y. min (x, y) is a function that returns the smaller value of x and y. Generally, the range that the saturation S can take is 0 to 1. Therefore, the maximum value S _max of the saturation S is 1 at the maximum, and the minimum value S _min is 0 at the minimum.

The brightness V is a parameter indicating the brightness of the color of the signal component, and is determined by using the equation (4) based on the maximum value peak [α, t]. In the formula (4), V _min and V _max indicate the maximum value and the minimum value that the brightness V can take at the time of display, respectively. Bias _V indicates a set value to be subtracted from the maximum value peak [α, t]. G _V indicates a coefficient used when converting the maximum value peak [α, t] into lightness V. Generally, the range that the brightness V can take is 0 to 1. Therefore, the maximum value V _max of the lightness V is 1 at the maximum, and the minimum value V _min is 0 at the minimum.

The display generation unit 50 of FIG. 5 generates a history display by arranging colors at the maximum value peak [α, t] of each time based on the color information supplied from the color arrangement processing unit 80. Then, the display generation unit 50 causes the display unit 160 to display the generated history display.

FIG. 11 is a schematic view showing an example of history display according to the first embodiment. In FIG. 11, reference numeral a indicates “green”, reference numeral b indicates “yellow”, and reference numeral c indicates “red”. Further, it is assumed that the saturation S and the lightness V decrease in the order of a1 to a3, b1 to b4, and c1 to c4, respectively.

As shown in FIG. 11, when there are three acoustic frequency groups, different colors are set for each acoustic frequency group. For example, the first acoustic frequency group is set to red, the second acoustic frequency group is set to yellow, and the third acoustic frequency group is set to green. Further, in each acoustic frequency group, the saturation S and the brightness V of the harmonic components are displayed so as to be different from the colors of the basic signal components. As described above, in the history display according to the first embodiment, the signal components and harmonic components caused by the plurality of sound sources are displayed in different colors.

As described above, the analyzer 1 according to the first embodiment divides the acoustic frequency direction of the periodic modulation coherence into a plurality of acoustic frequency groups, and divides the periodic modulation coherence into the acoustic frequency direction for each band of the divided acoustic frequency groups. Band is added to and the band addition output is calculated. Then, the maximum value in the acoustic frequency direction and the index of the divided acoustic frequency group are detected for each periodic frequency of the band addition output in the acoustic frequency direction, and the signal component and the harmonic component are historically displayed based on the maximum value and the index. Determine the color arrangement when doing.

As a result, each signal component and harmonic component are displayed in different colors. Therefore, the signal component and the harmonic component can be displayed so that the correspondence between the signal component and the harmonic component becomes clear.

Embodiment 2.
Next, the analyzer according to the second embodiment will be described. In the first embodiment, a case where the signal component and the harmonic component caused by a certain sound source are included in the same acoustic frequency group when the periodic modulation coherence is divided in the acoustic frequency direction has been described. However, when the signal component and the harmonic component caused by a certain sound source are included in different acoustic frequency groups and the history display is performed, the signal component and the harmonic component may not be displayed in the same color.

FIG. 12 is a graph for explaining a case where the periodic modulation coherence shown in FIG. 6 is divided into 9 in the acoustic frequency direction. As shown in FIG. 12, the periodic modulation coherences C ₁ [f, α, t] to C ₃ [f, α, t] of the groups having acoustic frequency group numbers k 1 to 3 include a certain signal component # 1 and The harmonic component is included. Further, the periodic modulation coherences C ₄ [f, α, t] to C ₆ [f, α, t] of the group having the acoustic frequency group number k 4 to 6 have a signal component # 2 different from the signal component # 1. And its harmonic components are included. Further, the periodic modulation coherences C ₇ [f, α, t] to C ₇ [f, α, t] of the group having the acoustic frequency group number k 7 to 9 have further different signal components # 3 and their harmonic components. It is included. In such a case, if the maximum value peak [α, t] and the index k [α, t] are derived as described in the first embodiment, the signal components caused by the same sound source are displayed when the history is displayed. And harmonic components may not be displayed in the same color.

FIG. 13 is a graph for explaining the index k [α, t] when the periodic modulation coherence is divided into 9 in the acoustic frequency direction. In FIG. 13, only the indexes k [α, t] corresponding to the signal component and the harmonic component are displayed for easy display.

As shown in FIG. 13, when the index k [α, t] is derived when the periodic modulation coherence is divided into 9 in the acoustic frequency direction, the index k [α, t] of the signal component and the harmonic component caused by the same sound source is derived. ] May be different. Therefore, when the history is displayed, different colors are assigned to the signal components and harmonic components caused by the same sound source.

FIG. 14 is a schematic view showing an example of history display when the periodic modulation coherence is divided into 9 in the acoustic frequency direction. In FIG. 14, reference numerals a to i indicate different colors. By increasing the number of divisions in the acoustic frequency direction with respect to the periodic modulation coherence, the acoustic frequency on which the signal component and the harmonic component are superimposed can be grasped in detail. However, as shown in FIG. 14, the signal component and the harmonic that are originally desired to be displayed in the same color can be grasped. Wave components are displayed in different colors.

In this way, when the signal component and the harmonic component caused by the same sound source are displayed in different colors in the history display, it is necessary to provide information so that the relationship between the signal component and the harmonic component can be easily grasped. I can't. Therefore, in the second embodiment, even when different acoustic frequency groups include signal components and harmonic components caused by the same sound source, the signal components and harmonic components are displayed in the same color.

[Analyzer configuration]
FIG. 15 is a block diagram showing an example of the configuration of the analyzer 2 according to the second embodiment. The configuration of the analyzer 2 will be described with reference to FIG. In the following description, the parts common to the analyzer 100 shown in FIG. 2 and the analyzer 1 shown in FIG. 5 are designated by the same reference numerals, and detailed description thereof will be omitted.

As shown in FIG. 15, the analyzer 2 includes a first FFT unit 110, a second FFT unit 120, a division unit 130, a partial band addition unit 40, a maximum value detection unit 70, a second band addition unit 91, and a maximum value detection unit. It is composed of 92, an index correction unit 93, a color arrangement processing unit 80, a display generation unit 50, and a display unit 160. Observation data x [t] is input to the analyzer 2.

The second band addition unit 91 adds the maximum value peak [α, t] supplied from the maximum value detection unit 70 in the periodic frequency direction, and calculates the band addition output sum_peak [k, t] in the periodic frequency direction. To do. The calculated band addition output sum_peak [k, t] is supplied to the maximum value detection unit 92.

FIG. 16 is a graph for explaining the band addition output sum_peak [k, t] in which the band is added in the periodic frequency direction. Further, FIG. 17 shows a band addition in which the index k of the acoustic frequency group and the maximum value peak [α, t] are band-added in the periodic frequency direction in consideration of the band addition output sum_peak [k, t] shown in FIG. It is a graph which shows the relationship with the output sum_peak [k, t].

The maximum value detection unit 92 of FIG. 15 detects the maximum value in the acoustic frequency group number direction of the band addition output sum_peak [k, t] in the periodic frequency direction supplied from the second band addition unit 91. Here, if the band addition output sum_peak [k, t] varies due to the influence of noise, an unnecessary maximum value is also detected. Therefore, in this case, the maximum value detection unit 92 suppresses an unnecessary maximum value caused by noise by applying a moving average in the acoustic frequency group number direction to the band addition output sum_peak [k, t].

FIG. 18 is a graph for explaining the maximum value detected by the maximum value detecting unit 92 of FIG. As shown in FIG. 18, the maximum value detection unit 92 smoothes the band addition output sum_peak [k, t] in the direction of the acoustic frequency group number, and has P maximum values local_Max [p] (p = 1, 2, 2, ..., P) is detected. In the example of FIG. 18, it is shown that three maximum values local_Max [1], local_Max [2] and local_Max [3] were detected.

The index correction unit 93 of FIG. 15 corrects the index k [α, t] based on the P maximum values local_Max [p] detected by the maximum value detection unit 92. The modified index k [α, t] is supplied to the color scheme processing unit 80.

The index correction unit 93 sets the correction width δ, and first searches for the index k [α, t] in which the acoustic frequency group number is included in the range of “local_Max [p] −δ to local_Max [p] + δ”. When the index k [α, t] is found, the index correction unit 93 replaces the index k [α, t] with the maximum value local_Max [p]. The index correction unit 93 performs this process on all the maximum values local_Max [p] and corrects the index k [α, t]. Here, it is assumed that the maximum value local_Max [p] has a relationship of "local_Max [p] <local_Max [p + 1]".

If the range of "local_Max [p] -δ to local_Max [p] + δ" and the range of "local_Max [p + 1] -δ to local_Max [p + 1] + δ" overlap, the index correction unit 93 has two. Reset the boundary to the middle position of the maximum value. For example, in this example, the range of "local_Max [p] -δ to local_Max [p] + δ" is reset to the range of "local_Max [p] -δ to (local_Max [p + 1] + local_Max [p]) / 2". Will be done. Further, the range of "local_Max [p + 1] -δ to local_Max [p + 1] + δ" is reset to the range of "(local_Max [p + 1] + local_Max [p]) / 2-local_Max [p + 1] + δ".

The color arrangement processing unit 80 is a color for history display based on the maximum value peak [α, t] supplied from the maximum value detection unit 70 and the index k [α, t] supplied from the index correction unit 93. Generate information. The generated color information is supplied to the display generation unit 50.

FIG. 19 is a schematic view showing an example of history display according to the second embodiment. In FIG. 19, reference numerals b, e and h indicate different colors. Further, it is assumed that the saturation S and the lightness V decrease in the order of b1 to b3, e1 to e4, and h1 to h4, respectively. In the example of FIG. 14, since the indexes k [α, t] were uneven, the colors of the signal components and the harmonic components caused by the same sound source were different. However, the analyzer 2 according to the second embodiment is used. By using the history display, the same color is set as shown in FIG.

As described above, the analyzer 2 according to the second embodiment detects and detects the maximum value in the index direction of the band addition output obtained by band-adding the maximum value detected by the maximum value detection unit 70 in the periodic frequency direction. Correct the index of the acoustic frequency group based on the maximum value. The color scheme processing unit 80 determines the color scheme of the signal component and the harmonic component based on the maximum value and the corrected index. As a result, the variation in the index is suppressed, so that the signal component and the harmonic component caused by the same sound source can be set to the same color, and the correspondence between the signal component and the harmonic component becomes clear. The signal component and harmonic component can be displayed.

Although embodiments 1 and 2 of the present invention have been described above, the present invention is not limited to the above-described embodiments 1 and 2 of the present invention, and varies within the scope of the gist of the present invention. Can be transformed and applied. For example, in embodiments 1 and 2, hue H, saturation S, and lightness V are used based on maximum values peak [α, t] and index k [α, t] using equations (2) to (4). Was calculated linearly, but the color scheme is not limited to this, and the color scheme may be performed using the equations (5) to (7). In equations (5) to (7), f _H (x, y), f _S (x, y) and f _V (x, y) are arbitrary functions with respect to the arguments x and y.

Further, the color scheme processing unit 80 does not calculate the hue H, the saturation S, and the lightness V each time by using the equations (5) to (7), and the maximum value peak [α, t] and the index k [α, The relationship between [t] and hue H, saturation S, and lightness V may be previously associated and stored in a table. This makes it possible to express more complicated relationships that cannot be expressed by mathematical formulas. Further, it is not necessary to sequentially calculate the hue H, the saturation S and the brightness V by using the maximum value peak [α, t] and the index k [α, t], and the processing amount required for the color arrangement processing can be suppressed. ..

1, 2, 100 analyzer, 40 partial band addition unit, 50 display generation unit, 70 maximum value detection unit, 80 color scheme processing unit, 91 second band addition unit, 92 maximum value detection unit, 93 index correction unit, 110 1st FFT unit, 120 2nd FFT unit, 130 division unit, 140 band addition unit, 150 display generation unit, 160 display unit.

Claims (6)

An analyzer that analyzes amplitude-modulated observation data.
An FFT unit that performs Fourier transform processing on the observed data multiple times and calculates periodic modulation coherence indicating the relationship between the acoustic frequency and the periodic frequency.
The acoustic frequency direction of the periodic modulation coherence is divided into a plurality of acoustic frequency groups, and the periodic modulation coherence is band-added in the acoustic frequency direction for each band of the divided acoustic frequency groups to obtain a band addition output in the acoustic frequency direction. Partial band addition part to calculate and
A maximum value detection unit that detects the maximum value in the acoustic frequency direction and the index of the divided acoustic frequency group for each periodic frequency of the band addition output in the acoustic frequency direction.
A color scheme processing unit that determines the color scheme when displaying the signal component and the harmonic component obtained from the band addition output in history based on the maximum value and the index.
An analyzer comprising a display generation unit that generates a history display of the signal component and the harmonic component with a determined color scheme. A band addition unit that adds bands to the maximum value detected by the maximum value detection unit in the periodic frequency direction and calculates a band addition output in the periodic frequency direction.
A maximum value detection unit that detects the maximum value in the index direction of the acoustic frequency group of the band addition output in the periodic frequency direction, and
Further provided with an index correction unit that corrects the index of the acoustic frequency group based on the maximum value.
The color scheme processing unit
The analyzer according to claim 1, wherein the color scheme of the signal component and the harmonic component is determined based on the maximum value and the corrected index. The color scheme processing unit
Hue is calculated based on the index,
The analyzer according to claim 1 or 2, wherein the saturation and the brightness are calculated based on the maximum value. The color scheme processing unit
Holds a table in which the maximum value and the index are associated with hue, saturation and lightness.
The analyzer according to claim 1 or 2, wherein the color scheme of the signal component and the harmonic component is determined based on the table. It is a method of displaying the analysis result by an analyzer that analyzes the amplitude-modulated observation data.
An FFT step in which the Fourier transform process is performed a plurality of times on the observed data to calculate the periodic modulation coherence indicating the relationship between the acoustic frequency and the periodic frequency.
The acoustic frequency direction of the periodic modulation coherence is divided into a plurality of acoustic frequency groups, and the periodic modulation coherence is band-added in the acoustic frequency direction for each band of the divided acoustic frequency groups to obtain a band addition output in the acoustic frequency direction. Partial band addition step to calculate and
A maximum value detection step for detecting the maximum value in the acoustic frequency direction and the index of the divided acoustic frequency group for each periodic frequency of the band addition output in the acoustic frequency direction.
A color scheme step for determining a color scheme for historically displaying the signal component and harmonic component obtained from the band addition output based on the maximum value and the index.
A display method comprising: a display generation step of generating a historical display of the signal component and the harmonic component with a determined color scheme. A display program that causes a computer to execute a method of displaying analysis results by an analyzer that analyzes amplitude-modulated observation data.
An FFT step in which the Fourier transform process is performed a plurality of times on the observed data to calculate the periodic modulation coherence indicating the relationship between the acoustic frequency and the periodic frequency.
The acoustic frequency direction of the periodic modulation coherence is divided into a plurality of acoustic frequency groups, and the periodic modulation coherence is band-added in the acoustic frequency direction for each band of the divided acoustic frequency groups to obtain a band addition output in the acoustic frequency direction. Partial band addition step to calculate and
A maximum value detection step for detecting the maximum value in the acoustic frequency direction and the index of the divided acoustic frequency group for each periodic frequency of the band addition output in the acoustic frequency direction.
A color scheme step for determining a color scheme for historically displaying the signal component and harmonic component obtained from the band addition output based on the maximum value and the index.
A display program comprising executing a display generation step of generating a historical display of the signal component and the harmonic component with a determined color scheme.

Images