Method and apparatus for determining the presence or absence of stomach upset in ruminants

Abstract

To determine presence or absence of stomach upset of ruminants.SOLUTION: A determination method comprises: (A) a step of executing a predetermined noise eliminating processing to abdominal sound data of a ruminant; (B) a step of calculating a time-varying sound pressure level from the sound data after having executed the predetermined eliminating noise processing; and (C) a determination step of determining the presence or absence of the stomach upset of the ruminant, on the basis of the frequency distribution of the calculated sound pressure level. Further, in the determination step, the presence or absence of the stomach upset of the ruminant may be determined, further on the basis of the value of a self-correlation coefficient of the calculated sound pressure level.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a technique for determining whether or not a ruminant has stomach upset. [Background technology]

[0002] Various information processing technologies to support livestock farming have been proposed in recent years. One document discloses a technology in which cows are fitted with activity status sensor modules and their health status is assessed by a management device. The activity status sensor module includes a triaxial acceleration sensor and a barometric pressure sensor. The management device identifies the activity status of the cow using a preset behavior identification model based on the cow's activity data received from the activity status sensor module, and calculates the cow's behavioral index, resting index, and rumination index corresponding to the identified activity status of the cow. Furthermore, the management device compares the temporal fluctuations of the cow's behavioral index, resting index, and rumination index with preset thresholds. If the fluctuations of any of the cow's behavioral index, resting index, and rumination index exceed the threshold, the management device assesses the presence or absence of an abnormality in the cow's health status using a predetermined judgment model.

[0003] The technology described in this document requires cows to be fitted with the above-mentioned activity sensor module, and additional communication devices are also required, which is costly. Furthermore, while the rumination behavior itself can be identified, the stomach condition cannot be directly observed. Furthermore, the judgment model is a model consisting of multiple pre-set judgment criteria based on past cases and recorded data that show that specific combinations of each indicator and the activity levels included in those indicators are associated with specific abnormalities (diseases or poor health) in cows. This model assumes that a large amount of data has been accumulated, but it is not easy to accumulate the data required to make accurate judgments even though the stomach condition cannot be directly observed. In other words, this technology is difficult to implement both in terms of cost and practicality. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-122368 Summary of the Invention [Problem to be solved by the invention]

[0005] According to one aspect, an object of the present invention is to provide a technique that enables easy determination of the presence or absence of stomach upset in a ruminant. [Means for solving the problem]

[0006] The determination method of the present invention includes: (A) a step of performing a predetermined noise removal process on sound data from the abdomen of a ruminant; (B) a step of calculating a time-varying sound pressure level from the sound data after the predetermined noise removal process has been performed; and (C) a determination step of determining whether or not the ruminant has stomach upset based on the frequency distribution of the calculated sound pressure level. [Effects of the Invention]

[0007] According to one aspect, it becomes possible to easily determine whether or not a ruminant has stomach upset. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing the configuration of a system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram showing a processing flow illustrating the processing content according to the first embodiment. [Figure 3A] FIG. 3A is a diagram showing an example of sound data of the abdomen of a healthy cow. [Figure 3B] FIG. 3B is a diagram showing an example of abdominal sound data of a cow with stomach upset. [Figure 4A] FIG. 4A shows an example of filtered acoustic data from the abdomen of a healthy cow. [Figure 4B] FIG. 4B shows an example of filtered sound data from the abdomen of a cow with stomach upset. [Figure 5A]FIG. 5A is a diagram showing an example of the change over time in sound pressure level in the abdomen of a healthy cow. [Figure 5B] FIG. 5B is a diagram showing an example of the change over time in sound pressure level in the abdomen of a cow with stomach upset. [Figure 6A] FIG. 6A is a diagram showing an example of threshold values ​​for sound pressure levels in the abdomen of a healthy cow. [Figure 6B] FIG. 6B shows an example of thresholds for sound pressure levels in the abdomen of a cow with stomach upset. [Figure 7A] FIG. 7A is a diagram showing an example of the range in which the sound pressure level in the abdomen of a healthy cow exceeds the threshold. [Figure 7B] FIG. 7B shows an example of the range in which the sound pressure level in the abdomen of a cow with stomach upset exceeds the threshold. [Figure 8] FIG. 8 is a diagram showing a processing flow illustrating processing contents according to the second embodiment. [Figure 9] FIG. 9 is a diagram showing a processing flow illustrating processing contents according to the second embodiment. [Figure 10A] FIG. 10A shows an example of filtered sound data from the abdomen of a cow with stomach upset. [Figure 10B] FIG. 10B is a diagram showing an example of the change over time in sound pressure level in the abdomen of a cow with stomach upset. [Figure 10C] FIG. 10C shows the change in autocorrelation coefficient for cows with stomach upset. [Figure 11A] FIG. 11A shows an example of filtered acoustic data of a healthy cow's abdomen. [Figure 11B] FIG. 11B is a diagram showing an example of the change over time in sound pressure level in the abdomen of a healthy cow. [Figure 11C] FIG. 11C shows the change in autocorrelation coefficients in healthy cows. DETAILED DESCRIPTION OF THE INVENTION

[0009] In the embodiment of the present invention, a cow will be described as an example of a ruminant animal, but sheep and other ruminant animals may also be used.

[0010] [Embodiment 1] Ruminants such as cows stir up the food they eat in their stomachs to make it easier to digest. This stomach motility moves food and air through the digestive tract, producing a gurgling sound in healthy cows. These sounds are called gastric sounds, which is the general term for sounds produced by gastric motility. On the other hand, in cows whose gastric motility is weakened due to illness or other reasons, gastric sounds are not observed or are produced for a shorter period of time. In other words, it is thought that there is a certain correlation between the presence or absence of gastric sounds, the duration of their production, and the health of individual cows.

[0011] In this embodiment, a system constructed based on such knowledge will be described.

[0012] An example of a system according to this embodiment is shown in Figure 1. In this embodiment, an electronic stethoscope 100 is placed against the abdomen of a cow 1000 to acquire abdominal sound data. The electronic stethoscope 100 has the function of storing digital sound data, and this digital sound data is imported into an information processing device 200 that executes the main processing in this embodiment for processing. Alternatively, the sound data of the cow may be measured using an analog device other than the electronic stethoscope 100, and then A / D (Analog-to-Digital) converted and imported into the information processing device 200. Communication between the electronic stethoscope 100 and the information processing device 200 may be wireless or wired.

[0013] The information processing device 200 is, for example, a personal computer, a tablet, a smartphone, or the like, and in some cases may be implemented by a combination of a terminal device such as a personal computer and a server or other computer (which may be a virtual computer).

[0014] The information processing device 200 includes a measurement data storage unit 201 , a band-limiting filter unit 203 , a first data storage unit 205 , a sound pressure level calculation unit 207 , a second data storage unit 209 , a determination unit 211 , and an output unit 213 .

[0015] The measurement data storage unit 201 stores digital sound data captured from, for example, the electronic stethoscope 100. The band-limiting filter unit 203 performs filtering to extract only signals in a frequency band related to stomach sounds, and stores the processing results in the first data storage unit 205. The sound pressure level calculation unit 207 calculates a sound pressure level for each predetermined frame from the sound data stored in the first data storage unit 205, and stores the calculated sound pressure level data in the second data storage unit 209. The determination unit 211 determines whether or not the target cow has stomach upset based on the sound pressure level data stored in the second data storage unit 209, and the output unit 213 outputs the determination result. The output unit 213 displays the determination result on the display unit of the information processing device 200, for example, or outputs the determination result to a printer or another information processing device.

[0016] Next, the processing contents of the information processing device 200 will be described with reference to FIGS. 2 to 7B.

[0017] First, the information processing device 200 acquires sound data of the abdomen of a ruminant from the electronic stethoscope 100 or the like, and stores the data in the measurement data storage unit 201 (FIG. 2: step S1).

[0018] Sound data measured by the electronic stethoscope 100 on a healthy cow has a waveform such as that shown in FIG. 3A. In FIG. 3A, the vertical axis represents amplitude [V] and the horizontal axis represents time [seconds]. This example includes the first 0.5 seconds or so where the electronic stethoscope 100 is not in contact with the cow's abdomen. On the other hand, sound data measured by the electronic stethoscope 100 on a cow with stomach upset has a waveform such as that shown in FIG. 3B. In FIG. 3B, the vertical axis represents amplitude [V] and the horizontal axis represents time [seconds]. This example also includes the first 0.5 seconds or so where the electronic stethoscope 100 is not in contact with the cow's abdomen. This waveform includes noise from the contact between the electronic stethoscope 100 and the cow, the cow's heartbeat, and other noises. The sampling frequency is, for example, 4000 Hz.

[0019] Next, the band-limiting filter unit 203 applies a predetermined band-limiting filter (for example, an IIR (Infinite Impulse Response) filter) to the sound data stored in the measurement data storage unit 201, and stores the processing result in the first data storage unit 205 (step S3). In the case of a cow, the frequency band related to stomach sounds is 200 Hz to 600 Hz, and contact noise and heartbeat sounds below 200 Hz and sounds above 600 Hz are removed. Note that instead of using such a band-limiting filter, noise removal may also be performed using a spectral subtraction method or the like.

[0020] When the above-described band-limiting filter is applied to the sound data shown in Fig. 3A, sound data with a waveform as shown in Fig. 4A is obtained. On the other hand, when the above-described band-limiting filter is applied to the sound data shown in Fig. 3B, sound data with a waveform as shown in Fig. 4B is obtained.

[0021] Then, the sound pressure level calculation unit 207 calculates the sound pressure level for each frame from the sound data stored in the first data storage unit 205, and stores the calculated sound pressure level in the second data storage unit 209 (step S5).

[0022] The sound pressure level calculation unit 207 first normalizes the sound data stored in the first data storage unit 205 so that the range is −1 to +1. Then, for example, the sound pressure level [dB] is calculated using the following formula with a frame width of 100 ms and a frame shift width of 50 ms.

number

[0023] For example, the sound pressure level of sound data with a waveform such as that shown in Figure 4A changes over time as shown in Figure 5A. In Figure 5A, the vertical axis represents sound pressure level [dB], and the horizontal axis represents time. On the other hand, the sound pressure level of sound data with a waveform such as that shown in Figure 4B changes over time as shown in Figure 5B. In this example, the sound pressure level changes more significantly for healthy cows, while the sound pressure level for cows with stomach upset is higher overall, but the change is smaller.

[0024] Then, the determination unit 211 determines a threshold value based on the sound pressure level data stored in the second data storage unit 209 (step S7). In this embodiment, the threshold value for the sound pressure level is dynamically calculated. For example, the determination unit 160 generates a frequency distribution of the sound pressure levels by sorting the sound pressure level values ​​in ascending order, and identifies the q (e.g., q=5) percentile in the frequency distribution. More specifically, it identifies the number of frames×q / 100th sound pressure level value. Then, it determines the q percentile+α (e.g., 10) as the threshold value.

[0025] In the case of Fig. 5A, the sound pressure levels represented by the dotted lines represent the 5th percentile, and the thresholds determined based on this are represented by the solid lines, as shown in Fig. 6A. In the case of Fig. 5B, the sound pressure levels represented by the dotted lines represent the 5th percentile, and the thresholds determined based on this are represented by the solid lines, as shown in Fig. 6B.

[0026] The overall level of sound pressure varies from measurement to measurement, and a fixed threshold value cannot be used to accurately determine the level. Therefore, in this embodiment, sound pressure levels of statistically noteworthy portions of the calculated sound pressure levels are identified, and a threshold value is determined from the identified sound pressure levels. This allows portions with high sound pressure levels to be identified in relation to the threshold value.

[0027] Furthermore, the determining unit 211 counts the number of frames in which a sound pressure level exceeding the determined threshold value was obtained (step S9).

[0028] In the case of Fig. 6A, as shown in Fig. 7A, sound pressure levels roughly above the threshold are obtained in ranges A to D, and in this example, 1273 frames out of a total of 2999 frames are above the threshold. On the other hand, in the case of Fig. 6B, as shown in Fig. 7B, sound pressure levels roughly above the threshold are obtained in ranges E and F, and in this example, 160 frames out of a total of 2999 frames are above the threshold.

[0029] In this way, in the case of a healthy cow, there are more frames in which stomach sounds are considered to be present, and in the case of a cow with stomach upset, there are fewer frames in which stomach sounds are considered to be present.

[0030] Then, the determination unit 160 calculates the percentage of frames whose sound pressure level is equal to or greater than the threshold by dividing the counted number of frames by the total number of frames (step S11). Since a frame lasts for a certain period of time, this is also called a time percentage. In this way, the determination unit 160 sets an appropriate threshold based on the frequency distribution of the sound pressure level, and calculates the number of frames whose sound pressure level is equal to or greater than the threshold (or, if the total number of frames changes, the number of frames whose sound pressure level is equal to or greater than the threshold / total number of frames) as an index value to determine whether or not stomach upset is present.

[0031] The determination unit 211 determines whether the calculated ratio exceeds a reference value (step S13). The determination unit 211 outputs the result of this determination to the output unit 213. The reference value is, for example, a value determined experimentally, and in the example described above, 0.2 is set. In the case of FIG. 7A, the ratio is 0.424, which satisfies the condition of step S13. On the other hand, in the case of FIG. 7B, the ratio is 0.053, which does not satisfy the condition of step S13.

[0032] If the condition of step S13 is met, the output unit 213 outputs that the ruminant being judged is "healthy" (step S15), and the process then ends.

[0033] On the other hand, if the condition of step S13 is not satisfied, the output unit 213 outputs a message indicating that the ruminant being judged has "upset stomach" (step S17), and the process then ends.

[0034] Until now, veterinarians have judged whether a ruminant has stomach upset by placing a stethoscope on the animal's abdomen and listening directly to the sounds transmitted through the stethoscope. However, the sounds transmitted through the stethoscope contain noises other than stomach sounds, such as those mentioned above, which make it difficult to distinguish between them. There was also the problem of inconsistent judgment depending on the experience and physical condition of the individual veterinarian.

[0035] According to this embodiment, it is possible to reliably determine the presence or absence of stomach upset with a simple configuration without attaching a sensor to the ruminant animal to be evaluated or collecting the amount of data required for machine learning. Note that screening can also be performed by a person who is not a veterinarian before making the final veterinarian decision.

[0036] [Embodiment 2] In the first embodiment, the judgment was based on the ratio of the number of frames in which a sound pressure level exceeding the threshold was obtained, but it was found that the judgment accuracy could be further improved by additionally using the autocorrelation coefficient as shown below. This is because a new finding was obtained that the stomach sounds of healthy ruminants have a certain periodicity.

[0037] The processing flow according to this embodiment is shown in Figures 8 and 9. Note that some of the processing contents of determination unit 211 of information processing device 200 shown in Figure 1 are different from those in the first embodiment, and this section will mainly explain those parts. That is, steps S1 to S11 shown in Figure 8 are the same as steps S1 to S11 in Figure 2 according to the first embodiment, and therefore explanations thereof will be omitted.

[0038] Moving on to the explanation of the processing shown in FIG. 9 after transition via terminal A in FIG. 8, the determination unit 211 determines whether the calculated ratio exceeds the reference value (step S13). This step is the same as in the first embodiment. If the condition of step S13 is satisfied, the determination unit 211 notifies the output unit 213 that the condition of step S13 is satisfied. Therefore, the output unit 213 outputs that the ruminant being determined is "healthy" (step S15). Then, the processing ends.

[0039] On the other hand, if the condition of step S13 is not satisfied, the determining unit 211 calculates the autocorrelation coefficients with a lag equal to or greater than a predetermined lag X (step S21).

[0040] For example, the i-th sound pressure level is x i and the average sound pressure level is expressed as x ave where h is the lag, and the autocorrelation coefficient r h is expressed as follows:

number

[0041] Then, the determination unit 211 identifies the maximum value of the autocorrelation coefficients calculated for lags X or more (step S23). As will be described later, the autocorrelation coefficients change depending on the lag, so the maximum value is identified for lags X or more. Then, the determination unit 3100 determines whether the maximum value of the autocorrelation coefficients is equal to or greater than a predetermined threshold (step S25). If the condition of step S25 is met, the determination unit 211 notifies the output unit 213 that the condition of step S25 is met. Then, the processing proceeds to step S15. That is, the output unit 213 outputs a message that the ruminant to be determined is "healthy".

[0042] On the other hand, if the condition of step S25 is not satisfied, the determination unit 211 notifies the output unit 213 that the conditions of steps S13 and S25 are not satisfied. In response, the output unit 213 outputs that the ruminant being determined has "stomach upset" (step S17). Then, the process ends.

[0043] In this way, even if the percentage of frames in which sound pressure levels exceed the threshold is below the threshold, if a certain periodicity is found in the stomach sounds based on the autocorrelation coefficient of the sound pressure levels, it is determined that the animal is healthy.

[0044] An example of a cow with stomach upset will be described using Figures 10A to 10C. Figure 10A shows the results of step S3. Specifically, the vertical axis represents amplitude [V], and the horizontal axis represents time [seconds]. Performing step S5 on the waveform of this sound data results in the time variation of the sound pressure level [dB], as shown in Figure 10B. Note that the figure also shows a smoothed line a of the sound pressure level, which is not actually calculated because it is suitable for obtaining an overview of the periodicity of the sound pressure level. Thus, for a cow with stomach upset, no clear periodicity is evident in the time variation of the sound pressure level. Performing step S21 then results in the values ​​of the autocorrelation coefficient for each lag, as shown in Figure 10C. In Figure 10C, the vertical axis represents the autocorrelation coefficient, and the horizontal axis represents the lag h [seconds]. In the example of Figure 10C, the autocorrelation coefficient is calculated for lags less than h = 30 seconds, but in practice, it is sufficient to perform the calculation for lags greater than h = 30 seconds. In this example, a maximum value of approximately 0.16 is detected around h = 47 seconds. If the threshold value is set to 0.2, when the condition in step S13 is not satisfied, the condition in step S25 is also not satisfied, and thus stomach upset is detected.

[0045] In contrast, an example of a healthy cow will be described using Figures 11A to 11C. Figure 11A shows the results of step S3. Specifically, the vertical axis represents amplitude [V], and the horizontal axis represents time [seconds]. Performing step S5 on this sound data waveform results in the time variation of sound pressure level [dB], as shown in Figure 11B. Similar to Figure 10B, the figure also shows a smoothed line b for sound pressure level, which is not actually calculated. Thus, for healthy cows, a clear periodicity appears in the time variation of sound pressure level. Performing step S21 then yields the autocorrelation coefficient values ​​for each lag, as shown in Figure 11C. While Figure 11C also shows calculations for lags less than h = 30 seconds, calculations can actually be performed for lags greater than h = 30 seconds. In this example, a maximum value of approximately 0.48 is detected around h = 62 seconds. If the threshold value is set to 0.2, even if the condition of step S13 is not met, the condition of step S25 is met, and the cow is therefore determined to be healthy.

[0046] Although the embodiments of the present invention have been described above, the present invention is not limited to these. For example, although numerical examples for cows have been shown, for other ruminants, appropriate numerical values ​​will be used. Furthermore, although a ratio is calculated above, if the total number of frames is set to a fixed value, the judgment may be made based on the number of frames. Furthermore, in the above example, a threshold value is determined from the frequency distribution of sound pressure levels and used to judge stomach upset, but the judgment may also be made using a method such as a Gaussian Mixture Model (GMM).

[0047] Although the maximum value of the autocorrelation coefficient is used as the feature amount, a plurality of peak values ​​may be extracted and the feature amount may be calculated from the peak values.

[0048] Also, the functional block configuration in Figure 1 is an example and may not match the program module configuration. The processing flow shown in Figure 2 may also include steps that can be executed in parallel or in a different order as long as the processing result is the same.

[0049] Furthermore, the information processing device 200 may be implemented as a single device or as multiple devices. In this way, the information processing device 200 may be constructed as an information processing system realized by one or multiple devices, and the case where it is a single device may also be referred to as an information processing system. On the other hand, the electronic stethoscope 100 itself may be configured to have the functions of the information processing device 200. In this case, the electronic stethoscope 100 also becomes an information processing system.

[0050] The information processing device 200 described above is a computer device, and includes a memory, a CPU (Central Processing Unit), a hard disk drive (HDD), a display control unit connected to a display device, a drive device for a removable disk, an input device, and a communication control unit for connecting to a network, all connected via a bus. The operating system (OS) and application programs for implementing the processes described in this embodiment are stored on the HDD and are read from the HDD to memory when executed by the CPU. The CPU controls the display control unit, communication control unit, and drive device according to the processing content of the application program to perform predetermined operations. Data during processing is primarily stored in memory, but may also be stored on the HDD. In an embodiment of the present invention, the application program for implementing the processes described above is distributed on a computer-readable removable disk and installed from the drive device to the HDD. It may also be installed on the HDD via a network such as the Internet and a communication control unit. Such a computer device realizes the various functions described above through the organic cooperation of the hardware, such as the CPU and memory, and programs, such as the OS and application programs.

[0051] The present embodiment described above can be summarized as follows.

[0052] The determination method according to this embodiment includes: (A) a step of performing a predetermined noise removal process on sound data from the abdomen of a ruminant; (B) a step of calculating a time-varying sound pressure level from the sound data after the predetermined noise removal process has been performed; and (C) a determination step of determining whether or not the ruminant has stomach upset based on the frequency distribution of the calculated sound pressure level.

[0053] The presence or absence of stomach upset in ruminants is reflected in the frequency distribution of sound pressure levels as described above. In this embodiment, the presence or absence of stomach upset in ruminants such as cows can be determined with such a simple configuration.

[0054] The above-mentioned determination step may also include (c1) a calculation step of calculating the number of frames in which the sound pressure level is equal to or greater than a threshold value, or the ratio of the number of frames to the total number of frames, based on the frequency distribution of the calculated sound pressure level, and (c2) a step of determining whether or not the number of frames or the ratio exceeds a reference value. For example, if the sound pressure level is always calculated for a fixed number of frames, the determination can be made based on the number of frames, and if the total number of frames fluctuates, the determination can be made by calculating the ratio as described above.

[0055] The calculation step may include a step of determining the threshold value based on the frequency distribution of the calculated sound pressure level, thereby enabling appropriate threshold values ​​to be set for various environments and ruminants.

[0056] Alternatively, the threshold value may be determined based on a sound pressure level of a predetermined percentile among the sound pressure levels for each frame. By statistically specifying a reference sound pressure level, an appropriate threshold value can be set.

[0057] Furthermore, in the determination step described above, the presence or absence of stomach upset in a ruminant may be determined based on the value of the autocorrelation coefficient of the calculated sound pressure level. This is based on the novel finding that healthy ruminants exhibit periodic changes in the sound pressure level of their stomach sounds. By adopting such a configuration, stomach upset in a ruminant can be detected with greater accuracy.

[0058] The aforementioned determination step may include a calculation step of calculating the number of frames whose sound pressure level is equal to or greater than a threshold value or the ratio of the number of frames to the total number of frames based on the frequency distribution of the calculated sound pressure level, a step of determining whether the value of the autocorrelation coefficient satisfies a predetermined condition if the number of frames or the ratio is equal to or less than a reference value, and a step of outputting a signal indicating that the ruminant has a stomach upset if the value of the autocorrelation coefficient does not satisfy the predetermined condition. By using the autocorrelation coefficient, it is possible to accurately detect stomach upset in ruminants.

[0059] Furthermore, the predetermined condition may be that the maximum value of the autocorrelation coefficients for lags equal to or greater than a certain value is equal to or greater than a second threshold. This is based on new findings obtained about ruminants.

[0060] Furthermore, the above-described determination method may further include a step of outputting that the ruminant does not have stomach upset if the number of frames or the percentage exceeds a reference value or if the value of the autocorrelation coefficient satisfies a predetermined condition.

[0061] A program for executing the above processing can be created, and the program is stored in a computer-readable storage medium or storage device, such as a flexible disk, optical disk (CD-ROM, DVD-ROM, etc.), magneto-optical disk, semiconductor memory, hard disk, etc. Intermediate processing results are temporarily stored in a storage device such as main memory. [Explanation of symbols]

[0062] 100 Electronic Stethoscope 200 Information processing device 201 Measurement data storage unit 203 Band limiting filter unit 205 First data storage unit 207 Sound pressure level calculation unit 209 Second data storage unit 211 Determination unit 213 Output section

Claims

1. performing a predetermined noise removal process on the abdominal sound data of the ruminant; calculating a time-varying sound pressure level from the sound data after the predetermined noise removal processing; a determining step of determining whether or not the ruminant has stomach upset based on the calculated frequency distribution of the sound pressure levels and the value of the autocorrelation coefficient of the sound pressure levels calculated for the sound data measured this time; A program that causes a computer to execute the above.

2. The determining step a calculation step of calculating the number of frames in which the sound pressure level is equal to or greater than a threshold value, or the ratio of the number of frames to the total number of frames, based on the calculated frequency distribution of the sound pressure level; determining whether or not the ruminant has stomach upset based on whether or not the number of frames or the ratio exceeds a reference value; The program of claim 1 .

3. The calculation step determining the threshold value based on the frequency distribution of the calculated sound pressure level; The program according to claim 2, comprising:

4. The threshold value is determined by identifying a sound pressure level of a predetermined percentile among the sound pressure levels for each frame and determining the threshold value based on the identified sound pressure level. The program according to claim 3.

5. The determining step a calculation step of calculating the number of frames in which the sound pressure level is equal to or greater than a threshold value, or the ratio of the number of frames to the total number of frames, based on the calculated frequency distribution of the sound pressure level; If the number of frames or the ratio is equal to or less than a reference value, determining whether the value of the autocorrelation coefficient satisfies a predetermined condition; if the value of the autocorrelation coefficient does not satisfy a predetermined condition, outputting that the ruminant has a stomach upset; The program of claim 1 .

6. The predetermined condition is that the maximum value of the autocorrelation coefficients with a lag of a certain value or more is equal to or greater than a second threshold value. The program according to claim 5.

7. If the number of frames or the ratio exceeds a reference value or if the value of the autocorrelation coefficient satisfies the predetermined condition, outputting that the ruminant does not have stomach upset. The program of claim 5, further comprising:

8. performing a predetermined noise removal process on the abdominal sound data of the ruminant; calculating a sound pressure level for each frame from the sound data after the predetermined noise removal processing; a determining step of determining whether or not the ruminant has stomach upset based on the calculated frequency distribution of the sound pressure levels and the value of the autocorrelation coefficient of the sound pressure levels calculated for the sound data measured this time; and a computer-implemented determination method.

9. means for performing a predetermined noise removal process on the abdominal sound data of the ruminant; means for calculating a sound pressure level for each frame from the sound data after the predetermined noise removal processing; a determination means for determining whether or not the ruminant has stomach upset based on the frequency distribution of the calculated sound pressure levels and the value of the autocorrelation coefficient of the sound pressure levels calculated for the sound data measured this time; An information processing system having the above.

Images