JPH09179579A - Retrieval device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To put the singing voices and names of animals in a data base and discriminate the animal name of a singing voice which is picked up. SOLUTION: A control part 6 extracts only the singing voice of a bird from a sound picked up by a microphone 1 by using spectrum analysis, linear predictive analysis, formant pitch extraction, voice analyzing method for correlation, etc., and stores this singing voice in a RAM 8 together with a bird name inputted from a key input part 5. Further, the control part 6 compares the singing voice which is picked up with singing voices stored in the RAM 8 to retrieve the bird name of a singing voice which matches or is similar to the picked-up singing voice, and displays the bird name at a display part 12. Further, the control part 6 outputs the collected singing voice from a speaker 11 as it is or reads a singing voice close to the collected singing voice out of the RAM8 and outputs it from a speaker 1.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a retrieval device for retrieving a corresponding animal name from an animal bark.

[0002]

2. Description of the Related Art Hitherto, so-called bird watching has been known, which involves listening to birds chirping in mountains or forests. In this bird watching, what kind of situation is there, such as identifying the type of bird by listening to the song of the bird, whether it is fighting the territory, or inviting hens during the breeding season, depending on how it calls Identify
I was observing the ecology of birds. Such interviews are conducted not only in birdwatching, but also in investigating bird habitats to monitor environmental damage and the like.

[0003]

However, since the above-mentioned bird identification and song recognition were all carried out based on human experience, the bird's song was correctly heard and the relevant bird was identified from among many birds. It took a lot of experience to do so. Therefore, there is a problem that skill is required and no one can easily perform it. Therefore, it is conceivable to record the sound of birds using a recording device such as a tape recorder.
In this case, it is necessary to investigate the name of the bird and the screaming condition under which the information cannot be obtained on the spot. Such a thing,
Not only birds but other animals as well.

It is therefore an object of the present invention to provide a retrieval device which can make a database of animal squeals and the names of the animals, and which allows an unskilled person to easily retrieve an animal from the screams.

[0005]

In order to achieve the above object, the search device according to the invention as defined in claim 1 is a storage device in which the calls of many animals are stored as sound data together with the animal name data, and the calls of the animals. A sound collecting means for collecting sound, a converting means for converting a squeal collected by the sound collecting means into sound data, the sound data converted by the converting means and the sound data stored in the storage means. It is provided with a search means for comparing and searching for animal name data corresponding to the sound data that is the same or is most similar.

According to the fourth aspect of the present invention, there is provided a search device, wherein a sounding means of a large number of animals is stored as sound data in association with animal name data, and an input means for inputting animal name data. Retrieval means for retrieving the sound data corresponding to the animal name data input by the input means from the sound data stored in the storage means, and converting the sound data retrieved by the search means into vocal data. It is characterized in that it is provided with a sound producing means.

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an embodiment of the present invention will be described with reference to the drawings as an embodiment applied to a bird name search device.

A. Configuration of Bird Name Search Device FIG. 1 shows a bird name search device according to an embodiment of the present invention (hereinafter, referred to as
It is a block diagram showing composition of a bird name retrieval device. In the figure, reference numeral 1 denotes a microphone, which collects a bark of a bird (hereinafter referred to as an audio signal) and supplies it to the amplifier 2.
The amplifier 2 amplifies the audio signal to a predetermined level and supplies it to the filter 3. The filter 3 is a bandpass filter, filters the amplified audio signal, removes noise and sounds other than squeaking, and supplies the A / D converter 4. The A / D converter 4 converts the audio signal into a digital signal (hereinafter referred to as waveform data) and supplies the digital signal to the control unit 6 described later.

Next, the key input unit 5 instructs the control unit 6 by the operator to change the operation mode (squealing sound collection mode, bird name search mode, call-in mode) of the bird name search device and each operation mode. Is used for inputting a bird name in a song collection mode, which will be described later. The control unit 6 controls each unit of the bird name search device by executing a predetermined program. In particular, in this embodiment, speech analysis (spectral analysis, linear prediction analysis,
Formant pitch extraction, correlation, etc.) is used to perform processing to obtain digital data in which only the necessary voice waveform (in this case, the bird's bark waveform) is extracted from the collected voice waveform data, and the extracted digital data of the bark (Hereinafter, referred to as waveform data) is stored in the RAM 8.

Further, the RAM 8 is used as a work area for storing various data in accordance with the execution of the program by the control unit 6, and sequentially stores the collected crying (waveform data) and sings the bird name. It is stored in correspondence with.

Here, FIG. 2 is a conceptual diagram showing a data structure of the RAM 8. In the figure, the display register is
It is used to display a bird name and the like on the display unit 12 described later, and the search result is stored in the process described later. M
Is a register indicating the above-mentioned operation mode, which is “0” in the bark collection mode and “1” in the bird name search mode.
It becomes "2" in the call-in mode. The contents of each mode will be described later. Next, the buffer is an area for temporarily storing the waveform data of the bark of the bird collected from the microphone 1 and the bird name input from the key input unit 5.

Next, the bird name is a register for storing the bird name input from the key input unit 5 for the bark (waveform data) once stored in the buffer, and the bark is temporarily stored in the buffer. This is a register for storing the waveform data of the collected cry (waveform data). It is possible to store a plurality of calls (waveform data) in one bird name. This is usually due to the cry of birds, the cry of calling the opposite sex in the breeding season, the cry of turf war, etc.
This is because there are multiple patterns. These bird names and their corresponding squeals can be stored for several hundred types of birds, depending on the capacity of the RAM 8, as shown in the figure. And the type is entered by the operator,
This is a register that stores the type of bird indicated by the bird name. In addition, in addition to this, it is possible to store information indicating, in correspondence with a call, what kind of situation the call is in, that is, when calling the opposite sex, when it is a territory battle, etc. May be.

Next, returning to FIG. 1, the ROM 7 stores in advance a program for executing the processing in each of the predetermined modes. In addition, the ROM 7 stores a large number of bird name data indicating the types of birds in advance, and a plurality of different types of bird name data, such as a cry when calling the opposite sex, a cry during a territory battle, etc. The waveform data of the bark is stored. In addition to storing the waveform data of different types of bark corresponding to each bird name data, image data of each bird may be further stored. This case will be described later.

The D / A converter 9 converts the waveform data supplied from the control unit 6 into an analog signal (hereinafter referred to as a voice signal).
And is supplied to the amplifier 10. The amplifier 10 amplifies the supplied audio signal to a predetermined level and supplies it to the speaker 11. The speaker 11 produces the amplified audio signal. This output system outputs, from the ROM 7 or the RAM 8, the sound of a bird of the same kind as the sound of the collected sound, the sound of the same kind of the sound of the collected sound, or the sound of a different kind of bird in the call mode described later. It is read out and pronounced, and so-called bird call is realized in which birds are gathered by hearing a squealing sound produced by the bird name search device.

Next, the display section 12 is composed of a liquid crystal display or the like provided on the front surface of the main body, and displays various data (bird name, operation mode, etc.) supplied from the control section 23.

B. Next, an operation of the above-described embodiment will be described. Here, FIG. 3 to FIG. 6 are flowcharts for explaining the operation of the bird name search device according to the present embodiment. Hereinafter, each operation mode will be described.

(A) Main routine Steps S10 to S16 shown in FIG. 3, which will be described below, are processes for switching the operation mode in accordance with an operator's instruction supplied via the key input unit 5. Control unit 9
First, in step S10, it is determined whether or not the mode switch (not shown) of the key input unit 5 is pressed. When the mode switch is pressed, the register M (mode M) is incremented (step S12) ( Then, in step S14, it is determined whether or not the mode M is "3" or more. In this embodiment, as described above, there are three types of operation modes, and the mode M corresponds to each operation mode.
It takes a value of "0", "1", or "2". Therefore, when the mode M becomes "3" or more, the mode M is returned to "0" in step S16, and the process returns to step S10. On the other hand, if the mode M is not "3" or more, the process directly returns to step S10. That is, each time the mode switch is pressed, the mode M is “0” → “1”.
->"2"->"0"->"1"->"2"-> ...

From step S18 shown in FIG.
S20 is a process of executing a predetermined process according to the operation mode set to the mode M. Step S1 described above
At 0, if the mode switch is not pressed,
In step S18, it is determined whether or not the start switch (not shown) of the key input unit 5 has been pressed. here,
If the start switch is not pressed, the process proceeds to other processing. On the other hand, when the start switch is pressed, the process proceeds to step S18 and thereafter, and the process in the operation mode according to the value of the mode M at that time is executed.

That is, in step S20, the mode M
Is "0", in step S24, the mode M
Is determined to be “1”. Here, if the mode M is "0", the determination result in step S20 is "YES", and the flow advances to step S22 to perform a bark collecting process (bark collecting mode). On the other hand, mode M
Is "1", the determination result in step S24 is "YES", the process proceeds to step S26, and the bird name search process (bird name search mode) is executed. Further, if the mode M is "2", the determination result in steps S20 and S24 is "NO", and the process proceeds to step S28 to execute the call-in process (call-in mode). After each process is completed, the process returns to step S10 and steps S10 to S2.
Repeat step 8.

(B) Singing voice collection mode In steps S40 to S48 shown in FIG. 4, which will be described below, the sounding of birds (waveform data) collected by the microphone 1 is stored in the RAM 8.
Is stored in the RAM 8, and the bird name is input and stored in the RAM 8 in association with the bark. The control unit 9
First, in step S40 shown in FIG. 4, the call of a bird collected by the microphone 1 is sampled. The bark (voice signal) collected by the microphone 1 is amplified by the amplifier 2 and then filtered by the filter 3 to obtain the A / D converter 4
Is converted into a digital signal (waveform data) and supplied to the control unit 6. Next, in step S42, only the waveform data of the required sound (in this case, the bark of the bird) is extracted from the waveform data by using voice analysis (spectral analysis, linear prediction analysis, formant pitch extraction, correlation, etc.). .

Next, in step S44, the extracted song of the bird is stored in the RAM 8 as waveform data, and in step S44.
At 46, the bird name is input from the key input unit 5. At this time, a message for inputting the bird name (please input the bird name, etc.) may be displayed on the display unit 12 to notify the operator of that fact. When the operator inputs a bird name from the key input unit 5, the bird name is stored in the RAM 8 in step S48 in association with the bark. Then, the process returns to the main routine shown in FIG.

(C) Bird Name Retrieval Mode Steps S60 to S68 shown in FIG. 5 are processes for retrieving and displaying the bird name corresponding to the bark of the collected bird. Steps S70 to S80 collect the sound. This is a process of storing, in the RAM, the bark of a bird that has been collected when the bark of the bird is not registered and the bird name cannot be retrieved. First,
In step S60 shown in FIG. 5, the control unit 9 samples the bird squeals collected by the microphone 1. The bark (voice signal) collected by the microphone 1 is amplified by the amplifier 2, filtered by the filter 3, converted into a digital signal (waveform data) by the A / D converter 4, and supplied to the control unit 6. And is once stored in the buffer of the RAM 8. Next, in step S62, only the necessary sound (in this case, the bark of the bird) is extracted from the waveform data by using voice analysis (spectral analysis, linear prediction analysis, formant pitch extraction, correlation, etc.).

Next, in step S64, the waveform data of the bark and the RAM 8 stored in the ROM 7 are sequentially read out and compared with the collected bark in sequence. Then, in step S66, it is determined whether or not the two match. Here, if they do not match, the process proceeds to step S70 and R
It is determined whether or not all the calls that are already registered in the OM 7 and the RAM 8 are compared. If the comparison is not completed, the process returns to step S64. In this way, ROM7
And the squeals stored in the RAM 8 are sequentially read one by one and compared with the collected squeals. If they match, that is, if the collected voice is registered in the ROM 7 or the RAM 8, the determination result in step S66 is "YES", and the process proceeds to step S68. In step S68, the bird name corresponding to the matched call is read from the ROM 7 or RAM 8 and displayed on the display unit 12.
Is displayed, the process ends, and the process returns to the main routine shown in FIG. As described above, when the image data of the bird's figure is stored in association with the bird name data in the ROM 7, this image data is also displayed on the display unit 12. By doing this, you can see the sounding birds,
It helps to find out.

On the other hand, if there is no corresponding squealing even when compared with all the squeakings stored in the ROM 7 and the RAM 8, the determination result in step S70 becomes "YES", and the process proceeds to step S72. In step S72, the display unit 12 displays a message not applicable. Next, the process proceeds to step S74, and it is determined whether or not the collected bark of the bird is registered, that is, whether or not to be stored in the RAM 8. It should be noted that the operator is instructed by key operation through the key input unit 5 whether or not to register. Then, when the registration is not made, the processing is ended and the process returns to the main routine shown in FIG.

On the other hand, if the operator gives an instruction to register, the result of the determination in step S74 is "YE
S ”, and the process proceeds to step S76. In step S76, the extracted song of the bird is stored in the RAM 8 as waveform data, and in step S78, the bird name is input from the key input unit 5. When the operator inputs a bird name from the key input unit 5, the bird name is stored in the RAM 8 in step S80 in association with the bark. Then, the process returns to the main routine shown in FIG.

(D) Calling-up mode Steps S90 to S98 shown in FIG. 6 which will be described below are processings of calling out a bird by reproducing the sound of a bird that has been collected, and steps S99 to S110 are input or selected. This is a process of calling out and calling out the sound of the birds. First, in step S90 shown in FIG. 6, the control unit 9 samples the bark of the bird collected by the microphone 1. The bark (voice signal) collected by the microphone 1 is amplified by the amplifier 2, filtered by the filter 3, converted into a digital signal (waveform data) by the A / D converter 4, and supplied to the control unit 6. To be done. Next, in step S92, only a necessary sound (in this case, a bark of a bird) is extracted from the waveform data by using a voice analysis (spectral analysis, linear prediction analysis, formant pitch extraction, correlation, etc.).

Next, in step S94, the extracted squeals are temporarily stored in the buffer of the RAM 8 and then in step S96.
Then, it is determined whether or not there is an instruction to reproduce the collected calls. Here, when the operator gives an instruction for reproduction from the key input unit 5, the process proceeds to step S98, the bark stored in the buffer of the RAM 8 is output to the D / A converter 9, and the speaker 11 is output from the speaker 11 via the amplifier 10. Pronounce. Then, the processing is ended, and the process returns to the main routine shown in FIG. On the other hand, if reproduction is not instructed, this flow ends.

In step S99, the name of the bird to be called is input from the key input unit 5, or the name of the bird to be called is selected from among a plurality of bird names displayed in advance. When the key input in step S99 is completed, the process proceeds to step S100. Step S100
Then, it is determined whether or not the operator has operated the key for starting the search. Here, if the operator does not instruct the search, the processing is ended as it is, while if the operator instructs the search, the process proceeds to step S102 and the ROM
The bird name data stored in 7 is read from the beginning and compared with the bird name data keyed in. And step S
At 104, it is determined whether the two match. here,
If they do not match, the process proceeds to step S106, where RO
It is determined whether or not all the bird name data stored in M7 have been compared, and if the comparison is complete, step S102.
Return to and read the next bird name data. The bird name data stored in the ROM 7 are sequentially read and compared.

If they match, that is, if the bird name data input by the key matches the bird name data stored in the ROM 7, steps S104 to S104 are performed.
In step 108, the waveform data of the bark stored corresponding to the matched bird name data is read out and read into the work area of the RAM 8. If there is no matching bird name data, the waveform data of the bird name of a similar squeal may be read. Next, in step S110, the bark read in the RAM 8 is output to the D / A converter 9, and the amplifier 10
The speaker 11 produces a sound through. Then, the processing is ended, and the process returns to the main routine shown in FIG.

As described above, in the above-described embodiment, the bird's squealing from the sound collected by the microphone 1 is stored in the RAM 8 together with the bird name input from the key input unit 5. Can be stored in association with the calls of birds, and the calls of various birds can be collected. In addition, by comparing the bark of the collected bird with the bark stored in the RAM 8, the bird name of the bark that matches or approximates the bark of the collected voice is searched and the bird name is displayed on the display unit 12. As a result, the unknown cry can be identified.
In addition, when a sound of a collected bird is uttered from the speaker 11 or a sound that matches or approximates to the collected sound of a bird is stored in the RAM 8, the squeal of the sound is output to the speaker 11.
Since it is pronounced from, it is possible to call for birds of the same or different species. Further, since the singing voice of the bird name input by the key is read out from the ROM 7 and sounded, it is possible to call in birds of the same type or different types.

In the above-described embodiment, the bird name corresponding to the bark of the collected bird is retrieved and displayed, but the present invention is not limited to this, and the bark corresponds to the bark, and the bark is the bark in any situation. It is also possible to store information indicating whether it is the time to call the opposite sex, the time to fight a territory, or the like, and display in what kind of situation the call is. Further, in addition to the bird name, image data of the bird may be stored and displayed together with the bird name. By doing so, the operator can obtain more detailed information.

Further, normally, a bird sings one continuous squeal which has some meaning or is suitable for a situation by combining monosyllables. Therefore, instead of or in addition to the singing voice, a single sound in which the singing of the bird is separated by syllables is registered, and the singing of the collected bird is separated into single syllables by the time division. You may make it compare and perform a more detailed analysis. In this case, by combining the above monosyllabic voices,
It becomes possible to have a pseudo conversation with the bird.

Furthermore, in the above embodiment, the bird was described, but it is of course applicable to the calls of other animals. For example, if the calls of various animals are stored in the mountains, Also, it has an effect of being able to know what kind of animal is nearby in a forest or the like.

[0034]

As described above, according to the present invention,
The following effects can be obtained. (1) Since the sound of animals collected by the sound collecting means is stored in the storage means in association with the animal name input from the input means, it is possible to collect the sounds of various animals. Names and can be stored in a database. (2) In addition, by comparing the sound of the collected sound and the sound stored in the storage means, the animal name of the sound closest to the collected sound is searched, and the animal name is displayed on the display means. By doing so, even an unskilled person can easily identify the animal name from the bark. (3) Also, the sound of the collected animal is pronounced from the sounding means, or the sound of sound that matches or approximates to the sound of the animal that is key-in is read out from the storage means, and the sound of the sound is pronounced from the sounding means. So you can call in the same or different birds. (4) Since the sound data collected by the sound collecting means is analyzed by using a sound analysis method such as spectrum analysis, linear prediction analysis, formant pitch extraction, and correlation, only the singing of the desired animal is extracted. can do.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a bird name search device according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram showing a data configuration of a RAM of the bird name search device.

FIG. 3 is a flowchart for explaining a main operation of the bird name search device.

FIG. 4 is a flowchart for explaining a bark sound collection process of the bird name search device.

FIG. 5 is a flowchart for explaining a bird name search process of the bird name search device.

FIG. 6 is a flowchart for explaining a calling process of the bird name search device.

[Explanation of symbols]

 1 Microphone (sound collecting means) 2 Amplifier 3 Filter 4 A / D converter 5 Key input section (input means) 6 Control section (control means) 7 ROM 8 RAM (storage means) 9 D / A converter 10 Amplifier 11 Speaker (Sounding means) 12 Display unit (Display means)

Claims (5)

[Claims] 1. A storage means in which the sounds of a large number of animals are stored as sound data together with the animal name data, a sound collecting means for collecting the sounds of the animals, and a sound collected by the sounds collecting means. Conversion means for converting to data and search means for comparing the sound data converted by this conversion means with the sound data stored in the storage means, and searching for animal name data corresponding to the matching or most similar sound data And a search device. 2. The sound data of the bark of a large number of animals stored in the storage means is the sound data of bark of different types of birds, and the animal name data is the type name data of birds. The search device according to claim 1, wherein: 3. The search device according to claim 2, wherein the sound data of the bark of the bird stored in the storage means stores a plurality of different sound data for one type of bird. 4. A storage means in which the sounds of many animals are stored as sound data in association with the animal name data, an input means for inputting the animal name data, and the sound data stored in the storage means. From the input means to search for sound data corresponding to the animal name data, and sounding means for converting the sound data searched by the search means into vocal data and producing sound. And search device. 5. The sound data of the barks of a large number of animals stored in the storage means are the sound data of bark of different types of birds, and the animal name data input by the input means is bird sounds. The search device according to claim 4, wherein the search device is type name data.