JP2021021815A - Voice grasping system of species - Google Patents

Abstract

To grasp the voice of species by easily and rapidly extracting only the voice of desired species.SOLUTION: When comparing actual data collected by recording means 5 with the frequency of basic voice data stored in basic voice data storage means 6, a voice grasping system performs filter processing to the frequency of actual voice by a filter function 11, eliminates noise etc., extracts the voice of a portion almost certainly including birdcalls, and compares data in which information on clear noise and sound other than the voice of species does not exist.SELECTED DRAWING: Figure 2

Description

The present invention relates to a species voice grasping system.

The importance of environmental assessment is increasing, and it is becoming important to understand the habitat status of wild animals (species) such as birds, mammals, amphibians, and insects. For this reason, the survey area is equipped with a sound source detection means that collects voices such as the barking of the species, and the voice information obtained by the sound source detection means is registered in the database via the communication network to grasp the habitat status of the species. A system (for example, a voice grasping system for a species) has been conventionally proposed (for example, Patent Document 1).

In the conventionally proposed voice grasping system for species, it is possible to create a database of behaviors based on the voice data of the species without holding a wireless terminal or the like in a specific species. For this reason, it is possible to collect audio data of species and grasp the habitat status by reducing the restrictions on the survey target without conducting on-site surveys or data collection.

In the conventionally proposed voice grasping system for species, voice data can be collected without holding a wireless terminal or the like in a specific species. In order to easily extract only the voice of a desired species, it is necessary to compare the recorded voice data with the voice data of the desired species and make an appropriate mechanical determination.

However, in the technique disclosed in Patent Document 1, the recorded voice data and the voice data of a desired biological species are mechanically and appropriately discriminated continuously for a long period of time or repeatedly for a long period of time. The point is not taken into consideration, and the current situation is that it has not been possible to easily extract only the voice of a desired species in a short time.

Japanese Patent Application No. 2018-99114

The present invention has been made in view of the above circumstances, and provides a voice grasping system for a species that can easily extract only the voice of a desired species in a short time and grasp the voice of the species. The purpose.

In particular, it is necessary to easily extract only the voice of a desired species in a short time based on the voice data recorded continuously for a long time or repeatedly for a long period of time to grasp the voice of the species. The purpose is to provide a voice grasping system for species that can be used.

The voice grasping system for a species of the present invention according to claim 1 for achieving the above object is a basic voice data storage means in which voice data of a target species is stored as basic voice data, and at a desired location. A recording means for recording audio to obtain actual data including actual audio data of a target organism, the basic audio data stored in the basic audio data storage means, and the recording means recorded by the recording means. The control means is provided with a control means for comparing the actual data and specifying a part of the actual voice data of the target organism from the actual data, and the control means compares the frequency of the basic voice data with the frequency of the actual data. By doing so, it has a comparison function for specifying a part of the actual audio data and a filter function for performing a filtering process for setting a frequency range to be compared with the actual data to be compared with the basic audio data. It is characterized by.

In the present invention according to claim 1, a filter process for setting a frequency range to be compared is performed on actual data including actual audio data of the target biological species (actual bark of the biological species and all surrounding sounds). The actual data that has been applied and filtered is compared with the basic audio data, and the actual audio data (the part of the actual bark) is extracted from the actual data in which the frequency range is set.

The time when the real voice data can be extracted from the recorded real data continuously for a long time or repeatedly for a long time at a desired place, and the bark of the target species is generated. By clarifying the (time) and frequency, it becomes possible to grasp the habitat status of species.

Therefore, it is possible to easily extract only the voice of a desired species in a short time and grasp the voice of the species.

The voice grasping system for the species of the present invention according to claim 2 is the voice grasping system for the species according to claim 1, wherein the control means sets the time range of the voice of the actual data into syllables of a predetermined division. It has a syllable division function for dividing, and the comparison function is characterized in that frequencies are compared in units of syllables divided by the syllable division function.

In the present invention according to claim 2, the voice of the species can be grasped by classifying the portion where the voice (squeal) of the species is generated as a syllable and comparing the frequencies of the divided syllables in units. ..

Further, the voice grasping system for the biological species of the present invention according to claim 3 is the voice grasping system for the biological species according to claim 2, wherein the control means is the time lapse in the syllables divided by the syllable division function. It has a conversion function that converts a change in the frequency of sound into a frequency characteristic that is the relationship between the loudness of the sound and the frequency, and the comparison function is characterized in that the frequency characteristic converted by the conversion function is compared. To do.

In the present invention according to claim 3, the change with time of the sound in the syllable is converted into the frequency characteristic which is the relationship between the loudness (intensity) of the sound and the frequency, and the frequency characteristic is compared. Can be grasped.

Further, the sound grasping system for the species of the present invention according to claim 4 is the sound grasping system for the species according to claim 3, wherein the comparison function of the control means is a parameter when comparing frequency characteristics. The maximum peak value, which is the maximum value of the value at which the sound frequency changes from rising to falling, the minimum peak value, which is the minimum value of the value at which the sound frequency changes from falling to rising, and the difference between the maximum peak value and the minimum peak value. It is characterized in that at least one of a certain peak difference value, a value at which the sound frequency changes from an increase to a decrease, and an average value of a value at which a sound frequency changes from a decrease to an increase is used.

In the present invention according to claim 4, by using the value (peak value) at the position where the frequency of the sound changes, it is possible to compare the parts where the characteristics are likely to appear.

Further, the voice grasping system of the biological species of the present invention according to claim 5 is the voice grasping system of the biological species according to claim 4, and in the basic voice data storage means, the basic voice data is a voice time range. Is divided into basic tunes, which are the tunes of a predetermined division, and the time passage of the basic tunes is converted into the basic frequency characteristics, which is the relationship between the magnitude and the frequency, and the frequency of the sound rises and falls as a basic parameter from the basic frequency characteristics. The basic maximum peak value, which is the maximum value of the value that changes to, the basic minimum peak value, which is the minimum value of the value at which the sound frequency changes from decreasing to rising, and the basic peak difference value, which is the difference between the basic maximum peak value and the basic minimum peak value. , The basic average peak value, which is the average value of the value at which the frequency of the sound changes from rising to falling and the value at which the loudness changes from falling to rising, is defined, and a plurality of basic parameters including at least two of the basic parameters are extracted. Then, from the values of multiple basic parameters, a normal random number generated based on a multivariate variance matrix is generated, and a quadratic basic parameter in a state in which outliers are excluded based on the generated normal random number is specified. The control means is characterized in that the frequencies of the actual audio data and the basic audio data are compared by comparing the frequency characteristics of the secondary basic parameters and the actual data.

In the present invention according to claim 5, as the basic audio data on which the comparison is based, the syllables of the bark part, the state based on the basic parameters of the values at the positions where the frequency of the sound changes, and the average of the data The secondary basic parameters are specified in a state where many normal random numbers are generated from the variation and outliers are excluded, and the frequency characteristics of the secondary basic parameters and the actual data are compared, so that the real voice data and the basic voice data The frequencies are compared.

Therefore, the noise of the basic voice data can be eliminated, and the variation due to the regional difference of the basic voice data can be corrected. By using the secondary basic parameters as the comparison target, the mechanical comparison can be easily performed. Can be carried out.

The voice grasping system for a species of the present invention can easily extract only the voice of a desired species in a short time and grasp the voice of the species.

That is, the voice grasping system of the species of the present invention can easily obtain only the voice of a desired species in a short time based on the voice data recorded continuously for a long time or repeatedly for a long period of time. It is possible to extract and grasp the sound of the species, and confirm when (time, season) the desired species is present, how much it is, and whether it is not in that place.

It is a conceptual diagram explaining the whole situation of the voice grasping system of the species which concerns on one Example of this invention. It is a schematic block block diagram of the voice grasping system of the biological species which concerns on one Example of this invention. It is a figure (basic frequency characteristic) of the voice level explaining the basic syllable of the basic voice data which is the basis of comparison. It is a figure which shows the time-dependent change (basic parameter) of the peak frequency in a basic syllable. It is a conceptual diagram of a normal random number generated from the basic frequency characteristic. It is a figure (frequency characteristic) of the voice level explaining the syllable of the real data. It is a figure which shows the time-dependent change (parameter) of the peak frequency in a syllable.

When it is necessary to grasp the habitat status of a specific bird (simply referred to as a bird) of the species to be surveyed in the environmental assessment, the song of the bird (voice data) is prepared as the basic voice data. Then, at a desired place, the voice (actual data) including the actual sounds of the birds and all the surrounding sounds is continuously or repeatedly recorded for a predetermined period of time, and the actual data and the basic voice data are recorded. By comparing, the actual bark (actual voice data) of birds is extracted (identified) from the actual data.

By extracting (identifying) the actual song of birds (actual voice data), the time, season, frequency, etc. of the song of birds are clarified at the desired location, and the habitat status of the species can be grasped. can do. In other words, it is possible to confirm when (time, season) the desired bird is, how much it is, and whether it is not in that place.

In the present invention, when the actual song of a bird (actual voice data) is extracted, the actual song of the bird is specified by comparing the frequencies of the actual data and the basic voice data. Then, when making a comparison, by applying a filter process, noise and the like are eliminated, and the sound of the part that is almost certain to be the song of a bird is extracted, and the range of frequencies to be compared is specified. It is a feature to do.

Specifically, the bandpass filter sets the upper and lower limits of the frequency, and excludes the audio information of the frequency larger than the upper limit and the frequency smaller than the lower limit, and the frequencies of the actual data and the basic audio data. To compare.

Therefore, it is possible to compare data in which there is no clear noise or voice information other than the bird's bark, and it is possible to easily extract only the desired bird's bark in a short time and grasp the bird's voice. become.

The present invention will be specifically described with reference to the drawings.

FIG. 1 shows a concept for explaining the overall situation of a voice grasping system for a species according to an embodiment of the present invention.

As shown in the figure, in order to investigate the habitat status of a specific bird (target species), which is a desired species, the survey target location, for example, agricultural land 1 adjacent to the city area or out of the city area. Recording means 5 is installed in the mountainous area 2 and the forest group 3 extending along the sea. The recording means 5 records audio at a desired location (farmland 1, mountainous area 2, forest group 3) and includes actual audio data of a specific bird (sound of a specific bird) (all recorded data). Is obtained.

A basic voice data storage means 6 for storing the actual bark data of a desired bird as basic voice data (voice data) is provided, and the information of the basic voice data storage means 6 is stored in the control means 7. The control means 7 is input with real data (all recorded data) including real voice data of a specific bird (song of a specific bird) recorded by the recording means 5.

In the control means 7, the frequency of the basic audio data (audio data) and the frequency of the actual data (all recorded data) are compared, and the part of the actual audio data (song of a specific bird) of a specific bird is specified. (Comparison function). Then, the actual data to be compared with the basic voice data (voice data) is subjected to a filter process for setting a frequency range to be compared (filter function).

The configuration of the function in the control means 7 will be described with reference to FIG. FIG. 2 shows a schematic block configuration of a function in the control means 7 of the voice grasping system according to the embodiment of the present invention.

As shown in the figure, the actual data recorded by the recording means 5 is input to the control means 7. The control means 7 has a filter function 11 for setting a frequency range of actual data to be compared. It also has a syllable division function 12 that divides the time range of the voice of the actual data into syllables (a set of phonemes) of a predetermined division.

Then, the control means 7 has a conversion function 13 that converts a change in the loudness of the sound in the syllables divided by the syllable division function 12 into a frequency characteristic that is a relationship between the loudness and the frequency. There is. Then, the frequency characteristics converted by the conversion function 13 are sent to the comparison function 14, and the comparison function 14 compares the frequency characteristics converted by the conversion function 13.

Therefore, the voice of a bird can be grasped by classifying the portion where the voice of the bird (screaming) is generated as a syllable and comparing the frequencies of the divided syllables in units.

When comparing the frequency characteristics, the comparison function 14 of the control means 7 has, as parameters, a maximum peak value which is the maximum value of the value at which the sound frequency changes from an increase to a decrease, and a value at which the sound frequency changes from a decrease to an increase. The average value of the minimum peak value, which is the minimum value, the peak difference value, which is the difference between the maximum peak value and the minimum peak value, the value at which the sound frequency changes from rising to falling, and the value at which the sound frequency changes from falling to rising. A certain average peak value is obtained.

Therefore, by using the value (peak value) at the position where the frequency of the sound changes, it is possible to compare the parts where the features are likely to appear.

On the other hand, in the basic voice data storage means 6, the basic voice data is divided into basic syllables whose voice time range is a syllable of a predetermined division, and the time passage of the basic syllable is a basic frequency characteristic in which the relationship between the magnitude and the frequency. Is converted to.

Then, from the basic frequency characteristics, as basic parameters, the basic maximum peak value, which is the maximum value of the value at which the sound frequency changes from rising to falling, and the basic minimum peak value, which is the minimum value of the value at which the sound frequency changes from falling to rising, The basic peak difference value, which is the difference between the basic maximum peak value and the basic minimum peak value, the value at which the sound frequency changes from rising to falling, and the basic average peak value, which is the average value of the values at which the sound frequency changes from falling to rising. Is defined and multiple basic parameters including the basic parameters are extracted.

Furthermore, from the values of multiple basic parameters, a normal random number generated based on a multivariate variance matrix is generated, and outliers are excluded based on the generated normal random number. To.

In the comparison function 14 of the control means 7, the frequencies of the actual voice data and the basic voice data are compared by comparing the secondary basic parameters of the basic voice data with the frequency characteristics of the actual data.

For this reason, as basic audio data that is the basis of comparison, a large number of normal random numbers are used based on the basic parameters of the syllable part of the bark and the value at the position where the frequency of the sound changes, and from the average and variation of the data. The frequencies of the actual audio data and the basic audio data are compared by specifying the secondary basic parameters and comparing the frequency characteristics of the secondary basic parameters and the actual data in a state of generating and eliminating outliers.

The basic voice data storage means 6 in which voice data is stored as basic voice data will be specifically described with reference to FIGS. 3 to 5. That is, the situation until the secondary basic parameters are obtained will be specifically described.

FIG. 3 shows the passage of time of the voice level explaining the basic syllables of the basic voice data that is the basis of comparison, FIG. 4 shows the time course of the peak frequency in the basic syllables (basic frequency characteristics, basic parameters), and FIG. 5 shows. Shows the concept of normal syllables generated from the basic frequency characteristics.

As shown in FIG. 3, the basic voice data is divided into basic syllables whose voice time range is the syllables of the predetermined divisions (1), (2), (3), (4), ... (N). For example, in the case of a bird's bark, the time range of the voice is divided into syllable units such as "pee" and "pee". The time course of many fundamental syllables is transformed into the fundamental frequency characteristics, which is the relationship between magnitude and frequency. The basic syllable of the basic frequency characteristic is a collection of many phonemes.

Then, for each of the basic frequency characteristics, as basic parameters, the basic maximum peak value, which is the maximum value of the value at which the sound frequency changes from rising to falling, and the basic value, which is the minimum value of the value at which the sound frequency changes from falling to rising. The minimum peak value, the basic peak difference value which is the difference between the basic maximum peak value and the basic minimum peak value, the value at which the sound frequency changes from rising to falling, and the average value of the values at which the sound frequency changes from falling to rising. The basal average peak value is defined and the basal parameters are extracted.

With respect to the basic syllables (1), (2), (3), (4), ... (N), the peak frequency in the basic syllable changes with time (1a) and (2a) based on the extracted basic parameters. ) (3a) (4a) ... (na) are shown. For example, taking the time-dependent change (2a) of the peak frequency in the basic syllable of the basic syllable (2) as an example, the relationship between the magnitude of the frequency and the time is shown as the time-dependent change of the peak as shown in FIG. ..

That is, as shown in FIG. 4, as the change with time of the peak frequency, the basic maximum peak value M, the basic minimum peak value m, the basic peak difference value which is the difference between the basic maximum peak value M and the basic minimum peak value m, and the peak frequency. The basic average peak value a, which is the average value of, is defined.

Further, as shown in FIG. 5, a normal random number is generated based on a multivariate variance matrix from the values of two (plurality) basic parameters that correlate with each other (for example, the basic maximum peak value and the basic minimum peak value). .. For example, there are outliers that are not within the dotted line in the figure, as shown by ●.

It is determined that the outliers not within the middle dotted line in FIG. 5 correspond to, for example, the temporal changes (1a) and (7a) of the peak frequency in the basic syllable in FIG. That is, it is determined that the time-dependent changes (1a) and (7a) of the peak frequency in the basic syllable do not approximate the state shown in FIG. 4, and the time-dependent change of the peak frequency in the basic syllable corresponding to the outlier is determined. Changes (1a) and (7a).

Then, the basic frequency characteristics (1) and (7) of the outliers that are not contained in the middle dotted line in FIG. 5 are excluded, and the basic frequency characteristics (the basic syllables that are contained in the dotted line in the figure) are excluded. Basic frequency characteristics) is specified as a secondary basic parameter. That is, the frequency characteristics of the basic syllables (2), (3), (4), (5), and (6) in the state shown in FIG. 3 are specified as secondary basic parameters.

That is, in the state shown in FIG. 3, for example, the change with time of the peak frequency in the basic syllable (2a) with respect to the basic syllables (1), (2), (3), (4), ... (N). It is determined that (3a), (4a), (5a), and (6a) are close to the state shown in FIG. 4, and they are specified as secondary basic parameters.

Therefore, the noise of the basic voice data can be eliminated, and the variation due to the regional difference of the basic voice data can be corrected. By using the secondary basic parameters as the comparison target, the mechanical comparison can be easily performed. Can be carried out.

It should be noted that the values of a plurality of basic parameters that generate normal random numbers based on the multivariate variance matrix also generate normal random numbers in combination with other parameters.

The secondary basic parameters of the basic voice data constructed as described above are stored in the basic voice data storage means 6, and are compared with the frequency characteristics of the actual data by the comparison function 14 of the control means 7.

The construction of the secondary basic parameters of the basic voice data is not limited to the method of the above-described embodiment, and a large number of data are stored in a database and variously selected from the database according to the purpose of use to obtain the desired secondary basic parameters. It is also possible to.

Filter function 11 (FIG. 2) that sets the frequency characteristics to be compared with the secondary basic parameters from the actual data (all recorded data) including the actual voice data (songs of specific birds) based on FIGS. 6 and 7. , The situation of the syllable division function (FIG. 2) and the conversion function 13 (FIG. 2) will be specifically described.

FIG. 6 shows the passage of time of the voice level explaining the syllables of the real data (all recorded data) including the real voice data (songs of specific birds), and FIG. 7 shows the time course of the peak frequency in the syllables (frequency). Characteristics (characteristics, parameters) are shown.

Since the actual data includes noise, living sounds of other surrounding creatures, and the like, the frequency range is set by the filter function 11. That is, a bandpass filter in a predetermined frequency range sets an upper limit value and a lower limit value of the frequency, and excludes voice information having a frequency larger than the upper limit value and a frequency smaller than the lower limit value. For this reason, obvious noise is eliminated, and a frequency band within a range that is not mistaken for the frequency of the song of a specific bird is set.

The actual data including the actual voice data of the target bird (the actual bark of the bird and all the surrounding sounds) was filtered by applying a filter process to set the frequency range to be compared. It is possible to compare the actual data with the basic voice data and extract the actual voice data (the part of the actual bark) from the actual data in which the frequency range is set.

Therefore, the real voice data can be extracted from the recorded real data continuously for a long time or repeatedly for a long time at a desired place, and the song of the target bird is generated. It is possible to clarify the time (time) and frequency of the bird, and to grasp the habitat of birds, and it is possible to easily extract only the desired bird's voice in a short time and grasp the bird's voice. Become.

As shown in FIG. 6, the actual voice data in which a predetermined frequency band is set has a syllable in which the voice time range is divided into predetermined divisions (1), (2), (3), (4), ... (N). It is divided (syllable division function 12). Similar to the basic syllables, for example, in the case of a bird song, the time range of the voice is divided into syllable units such as "pee" and "pee".

The time course of a large number of syllables is converted into frequency characteristics, which is the relationship between magnitude and frequency (conversion function 13). A syllable with frequency characteristics is a collection of many phonemes. Then, for each of the frequency characteristics, as parameters, the maximum peak value, which is the maximum value of the value at which the sound frequency changes from rising to falling, and the minimum peak value, which is the minimum value of the value at which the sound frequency changes from falling to rising, The peak difference value, which is the difference between the maximum peak value and the minimum peak value, the value at which the sound frequency changes from rising to falling, and the average peak value, which is the average value of the values at which the sound frequency changes from falling to rising, are defined. At least one parameter is extracted.

With respect to each of the syllables (1), (2), (3), (4), ... (N), the peak frequency in the syllable changes with time (1b), (2b), and (3b) based on the extracted parameters. ) (4b) ... (nb) is shown. For example, taking the time-dependent change (2b) of the peak frequency in the basic syllable of the syllable (2) as an example, the relationship between the magnitude of the frequency and the time is shown as the time-dependent change of the peak as shown in FIG.

That is, as shown in FIG. 7, as changes with time of the peak frequency, the maximum peak value M, the minimum peak value m, the peak difference value which is the difference between the maximum peak value M and the minimum peak value m, and the average value of the peak frequencies. The average peak value a is defined.

That is, based on the extracted parameters, for each of the syllables (1), (2), (3), (4), ... (N), the syllables (1), (2), (3), (4) ... The time course of the peak frequency in (n) (1b), (2b), (3b), (4b), ... (Nb) is shown (conversion function 13). The time course of the peak frequency in the syllable is used to compare the frequency characteristics. By using the value (peak value) at the position where the frequency of the sound changes, it is possible to compare the parts where the characteristics are likely to appear.

In the comparison function 14 of the control means 7, the basic parameter in which the value (peak value) at the position where the loudness changes is added, and the parameter are taken into consideration, and the secondary basic parameter and the frequency characteristic of the actual data are taken into consideration. By comparing, the frequencies of the actual audio data and the basic audio data are compared.

In the comparison function 14, the secondary basic parameter, which is the frequency characteristic of the basic syllable, and the frequency characteristic of the actual data, which is the frequency characteristic of the syllable, are compared, and as a result of the comparison, for example, the degree of matching (recent degree). When is high, the audio of the actual data is regarded as the audio of the basic audio data, and it is determined that the audio of the birds recorded in the basic audio data matches the audio of the birds extracted from the actual data. As a result, it is possible to specifically confirm the existence status of desired birds.

For example, as a result of comparison, the frequency changes (2a) (3a) (4a) (5a) (6a) (7a) in the basic syllables (2) (3) (4) (5) (6) (7) The frequency characteristics (see FIG. 6), which are the temporal changes (2b) (4b) (6b) of the peak frequencies in the syllables (2), (4), and (6), were approximated to a certain secondary basic parameter (see FIG. 3). It is judged as a result. Therefore, it is determined that the bird's voice in the syllables (2), (4), and (6) extracted from the actual data matches the bird's voice recorded in the basic voice data.

As a result, the existence status of the desired bird can be specifically confirmed at the time corresponding to the syllables (2), (4), and (6). In other words, it can be confirmed that the desired bird does not exist except for the time corresponding to the syllables (2), (4) and (6).

The voice grasping system described above compares the filtered real data with the basic voice data, and extracts the real voice data (the part of the actual bark) from the real data in which the frequency range is set. Then, since the parameter of the value (peak value) at the position where the loudness changes is taken into consideration, the actual data and the basic voice data can be compared in the part where the feature is likely to appear.

Therefore, it is possible to easily extract (identify) the actual song of birds (actual voice data) based on the recorded voice data for a long period of time continuously or repeatedly for a long period of time, which is desired. It is possible to grasp the habitat status of birds by clarifying the time, season, frequency, etc. of the birds singing at the place.

In other words, it is possible to easily confirm when (time, season) the desired bird is present, how much it is, and whether it is not in that place. That is, it is possible to easily extract only the desired bird's voice in a short time and grasp the bird's voice.

In the above-mentioned examples, birds have been taken as an example as a biological species, but the present invention can target wild animals such as mammals, amphibians, and insects as biological species. By applying the invention, it becomes possible to grasp the habitat status of various species.

The present invention can be used in the industrial field of voice grasping systems for species.

1 Agricultural land 2 Mountains 3 Forests 5 Recording means 6 Basic voice data storage means 7 Control means 11 Filter function 12 Syllable division function 13 Conversion function 14 Comparison function

Claims (5)

Basic voice data storage means that stores voice data of the target species as basic voice data,
A recording means for recording audio at a desired location to obtain actual data including actual audio data of the target species, and
The basic voice data stored in the basic voice data storage means and the actual data recorded by the recording means are compared, and a part of the real voice data of the target organism species is selected from the actual data. With control means to identify
The control means
A comparison function that identifies a part of the actual voice data by comparing the frequencies of the basic voice data and the actual data, and
A voice grasping system for a species having a filter function of performing a filtering process for setting a frequency range to be compared with the actual data to be compared with the basic voice data. In the voice grasping system for the species according to claim 1,
The control means
It has a syllable division function that divides the time range of the actual data voice into syllables of a predetermined division.
The comparison function
A voice grasping system for species, characterized in that frequencies are compared in units of syllables divided by the syllable division function. In the voice grasping system for the species according to claim 2,
The control means
It has a conversion function that converts the change in the frequency of the sound over time in the syllables divided by the syllable division function into the frequency characteristic that is the relationship between the loudness and frequency of the sound.
The comparison function
A voice grasping system for a species, characterized in that the frequency characteristics converted by the conversion function are compared. In the voice grasping system for the species according to claim 3,
The comparison function of the control means
As a parameter when comparing frequency characteristics
The maximum peak value, which is the maximum value of the value at which the frequency of sound changes from rising to falling,
The minimum peak value, which is the minimum value of the value at which the frequency of sound changes from decreasing to increasing,
Peak difference value, which is the difference between the maximum peak value and the minimum peak value,
A voice grasping system for a species, characterized in that it uses at least one of an average peak value, which is the average value of the value at which the sound frequency changes from rising to falling and the value at which the sound frequency changes from falling to rising. In the voice grasping system for the species according to claim 4,
In the basic voice data storage means,
The basic voice data is
The time range of the voice is divided into basic syllables, which are syllables of a predetermined division.
The time course of the basic syllable is converted into the basic frequency characteristic, which is the relationship between magnitude and frequency.
From the basic frequency characteristics, as basic parameters, the basic maximum peak value, which is the maximum value of the value at which the sound frequency changes from rising to falling, the basic minimum peak value, which is the minimum value of the value at which the sound frequency changes from falling to rising, and the basic maximum. The basic peak difference value, which is the difference between the peak value and the basic minimum peak value, the value at which the sound frequency changes from rising to falling, and the basic average peak value, which is the average value of the values at which the sound frequency changes from falling to rising, are defined. Being done
Multiple basic parameters are extracted, including at least two of the basic parameters
From the values of multiple basic parameters, a normal random number generated based on a multivariate variance matrix is generated, and outliers are excluded based on the generated normal random number.
In the control means
A voice grasping system for species, characterized in that the frequencies of real voice data and basic voice data are compared by comparing the frequency characteristics of secondary basic parameters and real data.

Images