A pig breeding health monitoring method, device, equipment and medium

Abstract

The present application relates to the technical field of pig breeding, and more particularly to a pig breeding health monitoring method, device, equipment and medium. By integrating thermal imaging data, millimeter wave radar data and audio data, three different types of sensing data are realized. Multi-dimensional, three-dimensional monitoring of pig health status. The thermal imaging data can accurately capture the temperature change of the pig, the millimeter wave radar data can obtain the vital sign parameters such as the respiration and heartbeat of the pig in real time, and the audio data can perceive the potential abnormal state through the analysis of the call sound. This multi-source data fusion method can more comprehensively and accurately reflect the health status of the pig, and improves the accuracy of pig health monitoring.

Description

Technical Field

[0001] This invention relates to the field of pig farming technology, and in particular to a method, device, equipment and medium for monitoring the health of pigs. Background Technology

[0002] Currently, health monitoring of pigs in pig farming relies heavily on manual inspections and experience-based judgment. This typically only provides limited information such as body temperature or activity level, failing to simultaneously obtain crucial vital signs like respiration and heartbeat. This leads to insufficient identification of heat stress, respiratory diseases, and sub-health conditions, resulting in the spread of disease and significant economic losses for farms. Furthermore, manual monitoring is highly subjective; different personnel may use different judgment standards, easily leading to missed or false positives. Therefore, improving the accuracy of health monitoring in pig farming is a pressing issue that needs to be addressed. Summary of the Invention

[0003] In view of this, the embodiments of this application provide a method, device, equipment and medium for monitoring the health of pigs, in order to solve the problem of low accuracy in monitoring the health of pigs during the pig farming process.

[0004] In a first aspect, embodiments of this application provide a method for monitoring the health of pig farming, the method comprising:

[0005] Acquire thermal imaging data, millimeter-wave radar data, and audio data within a pig farm per unit time period;

[0006] The thermal imaging data is segmented to obtain the target body temperature data of each pig in the pig farm;

[0007] The millimeter-wave radar data is processed to extract the respiratory rate and heart rate of each pig.

[0008] The audio data is subjected to abnormal vocalization detection to obtain the vocalization detection results for each pig.

[0009] Health monitoring is conducted on each pig based on its target body temperature, respiratory rate, heart rate, and vocalizations to obtain monitoring results.

[0010] Secondly, this application provides a pig farming health monitoring device, which includes a thermal imaging camera module, a millimeter-wave bio-radar module, a microphone module, and a signal processing module. The thermal imaging camera module, the millimeter-wave bio-radar module, and the microphone module are respectively communicatively connected to the signal processing module.

[0011] The thermal imaging camera module is used to acquire thermal imaging data in the pig farm and transmit the thermal imaging data to the signal processing module.

[0012] The millimeter-wave bio-radar module is used to acquire millimeter-wave radar data in the pig farm and transmit the millimeter-wave radar data to the signal processing module.

[0013] The microphone module is used to acquire audio data from the pig farm and transmit the audio data to the signal processing module;

[0014] The signal processing module is used to execute the above-described method for monitoring the health of pig farming.

[0015] Thirdly, embodiments of this application provide a computer device, the computer device including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the pig farming health monitoring method described above.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the pig farming health monitoring method described above.

[0017] The advantages of this application compared to the prior art are:

[0018] This application integrates three different types of sensor data—thermal imaging data, millimeter-wave radar data, and audio data—to achieve multi-dimensional and comprehensive monitoring of pig health status. Thermal imaging data can accurately capture changes in pig body temperature, millimeter-wave radar data can acquire vital signs such as respiration and heartbeat in real time, and audio data can detect potential abnormalities through analysis of vocalizations. This multi-source data fusion approach can more comprehensively and accurately reflect the health status of pigs, improving the accuracy of pig health monitoring. Furthermore, this intelligent monitoring method can promptly detect potential risks in the early stages of pig health abnormalities, providing a scientific basis for decision-making by farm managers, helping to reduce the risk of disease transmission, improve farming efficiency, and promote the digital and intelligent transformation and upgrading of the pig farming industry. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic flowchart of a method for monitoring the health of pig farming provided in one embodiment of this application;

[0021] Figure 2 This is a schematic diagram of the structure of a pig farming health monitoring device provided in one embodiment of this application;

[0022] Figure 3 This is another structural schematic diagram of a pig farming health monitoring device provided in one embodiment of this application;

[0023] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0026] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0027] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0028] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0029] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0030] References to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0031] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0032] To illustrate the technical solution of this application, specific embodiments are described below.

[0033] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a method for monitoring the health of pig farming according to an embodiment of this application, as shown below. Figure 1 As shown, the method for monitoring the health of pig farming may include the following steps.

[0034] S101: Acquire thermal imaging data, millimeter-wave radar data, and audio data within a pig farm per unit time.

[0035] In step S101, the thermal imaging data is infrared thermal radiation information data collected by the thermal imaging camera module deployed in the pig farm, the millimeter-wave radar data is data collected by the millimeter-wave bio-radar module deployed in the pig farm, and the audio data is data collected by the microphone module deployed in the pig farm.

[0036] In this embodiment, a thermal imaging camera module, a millimeter-wave bio-radar module, a microphone module, and a signal processing module are deployed within the pig farm. These modules are communicatively connected to the signal processing module. The thermal imaging camera module acquires thermal imaging data from the pig farm and transmits the data to the signal processing module. The millimeter-wave bio-radar module acquires millimeter-wave radar data from the pig farm and transmits the data to the signal processing module. The microphone module acquires audio data from the pig farm and transmits the audio data to the signal processing module. The signal processing module processes the thermal imaging data, millimeter-wave radar data, and audio data.

[0037] This project acquires thermal imaging data, millimeter-wave radar data, and audio data from a pig farm within a given time period. The thermal imaging data includes temperature distribution information for different areas of the farm, accurate to the temperature value of each pixel. The millimeter-wave radar data includes biological parameters such as the movement trajectory, respiratory rate, and heart rate of individual pigs. The audio data covers various sound information, including pig vocalizations, coughs, and breathing sounds. The thermal imaging data, millimeter-wave radar data, and audio data from the pig farm within a given time period are corresponding time-series data; that is, they are collected synchronously on the same time axis and have a one-to-one correspondence in the time dimension.

[0038] In this embodiment, acquiring thermal imaging data, millimeter-wave radar data, and audio data within a pig farm per unit time enables comprehensive and real-time capture of key information within the pig farm from multiple dimensions, providing a data foundation for accurate monitoring of pig health status.

[0039] S102: Segment the thermal imaging data to obtain the target body temperature data of each pig in the pig farm.

[0040] In step S102, the thermal imaging data is segmented to accurately identify and divide the individual pigs in the thermal imaging image into regions. The target body temperature data for each pig is the temperature value of that pig within the corresponding region in the thermal imaging image.

[0041] In this embodiment, when segmenting the thermal imaging data, a deep learning-based image segmentation model, such as Mask R-CNN, can be used. This model can accurately identify the contour boundaries of different pigs in the thermal imaging image, and can effectively distinguish individual regions even when pigs are close to each other or partially occluded. Other segmentation methods can also be used, and this embodiment does not limit them.

[0042] The thermal imaging data is segmented to obtain the outline boundary of each pig. Based on the temperature data within the outline boundary of each pig, the target body temperature data of each pig is calculated. This can be achieved by calculating the average temperature within the outline boundary of each pig and using this average as the target body temperature data, or by extracting corresponding key points and using the average temperature data at those key points as the target body temperature data. Other methods can also be used to calculate the target body temperature data of each pig; this embodiment does not limit the specific methods used.

[0043] Optionally, the thermal imaging data is segmented to obtain target body temperature data for each pig in the pig farm, including:

[0044] The thermal imaging data is segmented to obtain information about the head region of each pig;

[0045] Key regions are extracted from the information in the head region to obtain a set of key head regions;

[0046] Based on thermal imaging data, the temperature data of each key area of ​​the head in the set of key head areas is determined, and the reference temperature of each key head area is calculated based on the temperature data.

[0047] The reference temperatures of each key area of ​​the head are weighted and fused to obtain the target body temperature data for each pig.

[0048] In this embodiment, the thermal imaging data is segmented to obtain the head region information for each pig. A deep learning-based instance segmentation algorithm, such as the Mask R-CNN model, can be used. This model can accurately delineate the pixel-level boundaries of each pig's head while detecting the pig target. During the model training phase, a labeled dataset containing pig heads at different growth stages and poses needs to be constructed. The labeled information includes the bounding rectangle of the head and a pixel-level segmentation mask. By inputting the thermal imaging image into the trained Mask R-CNN model, the model outputs candidate regions for all pig heads in the image, classifies these candidate regions, performs bounding box regression, and generates an accurate head region segmentation mask, which serves as the head region information for each pig.

[0049] Key regions are extracted from the head region information to obtain a set of key head regions. These key head regions include the left ear root region, right ear root region, left eye region, and right eye region. When extracting key regions from the head region information, prior knowledge of pig head anatomy can be used to set the relative position and morphological feature parameters of each key region based on the obtained head region segmentation mask. For example, for the ear root region, based on its typical location on both sides of the head, near the lower part of the ear, and its specific connection to the head contour, edge detection and region growing algorithms are used to segment sub-regions conforming to the morphological features of the left and right ear roots from the head mask. For the eye region, based on its location below the forehead, its approximately elliptical contour, and the difference in grayscale from surrounding tissues, Hough circle transform or template matching methods are used to accurately locate the left and right eye regions.

[0050] Based on thermal imaging data, the temperature data of each key head region in the key head region set is determined. Then, a reference temperature for each key head region is calculated based on this temperature data. The formula for calculating the reference temperature is as follows:

[0051]

[0052] in, Let be the reference temperature of the k-th critical area of ​​the head of the i-th pig at time t. For time t Temperature data at the location, The temperature data for the k-th head region of the i-th pig at time t is... ,in, For the left eye area, The right eye area. The area around the left ear root. This refers to the area around the right ear root. This represents the p-quantile function.

[0053] The reference temperatures of each key head region are weighted and fused to obtain the target body temperature data for each pig. Different weight values ​​are assigned to each key head region, and these weights can be dynamically adjusted based on the sensitivity and stability of different key head regions in pig body temperature monitoring. For example, the eye region, due to its thinner skin and denser blood vessel distribution, is more sensitive to temperature changes and can be assigned a relatively higher weight. The ear root region, on the other hand, is more affected by external environmental temperature and is slightly less stable than the eye region, so it can be assigned a slightly lower weight. Based on the corresponding weight values, the reference temperatures of the key head regions are weighted to obtain the weighted temperature, and the average of the weighted temperatures is determined as the target body temperature data.

[0054] In this embodiment, by accurately extracting and weighting temperature data from key areas of the pig's head, non-contact, high-precision monitoring of the pig's body temperature is achieved. This effectively reduces the impact of external environmental factors on the temperature measurement results, making the target body temperature data more accurately reflect the pig's true physiological state.

[0055] Optionally, the reference temperatures of each key area of ​​the head are weighted and fused to obtain the target body temperature data for each pig, including:

[0056] Based on the temperature data of each key area of ​​the head, the temperature variance of the corresponding key area of ​​the head is calculated.

[0057] Based on thermal imaging data, the corresponding background temperature data in the pig farm is determined. Based on the background temperature data and the temperature data of each key area of ​​the head, the regional visibility evaluation value of the corresponding key area of ​​the head is calculated.

[0058] The weight values ​​of the corresponding key head regions are calculated based on the temperature variance and the regional visibility evaluation value.

[0059] The reference temperature of each key area of ​​the head is weighted and fused with the corresponding weight value of the key area of ​​the head to obtain the target body temperature data for each pig.

[0060] In this embodiment, the temperature variance of each key head region is calculated based on its temperature data. Specifically, the average temperature of each key head region is calculated using the temperature data, and the temperature variance is then calculated using the variance formula and the average temperature.

[0061] Based on thermal imaging data, the background temperature data within the pig farm is determined. This background temperature data refers to the ambient temperature outside the pig farming area. The thermal imaging data is segmented to extract individual pig regions, and the temperature data for these regions is defined as the background temperature data. Based on the background temperature data and the temperature data for each key head region, the visibility evaluation value for that key head region is calculated. Specifically, a first average temperature for the corresponding key head region and a second average temperature from the background temperature data are calculated. The average temperature difference is then calculated based on these two average temperatures. Finally, the normalized thermal contrast ratio is calculated using the average temperature difference, the maximum temperature value, and the minimum temperature value from the thermal imaging data. The calculation formula is as follows:

[0062]

[0063] in, for , The average temperature difference This represents the maximum temperature. This is the minimum temperature.

[0064] Based on the background temperature data, the temperature standard deviation of the background region is calculated. The thermal signal-to-noise ratio is then calculated based on the ratio of the average temperature difference to the temperature standard deviation. The spatial contrast of the region is calculated based on the spatial gradient of the thermal imaging data, using the following formula:

[0065]

[0066] in, For regional spatial contrast. Let j be the temperature gradient of the j-th pixel at the edge of the key head region. M represents the average temperature corresponding to the key area of ​​the head, and M is the number of pixels at the edge of the ROI.

[0067] The interference coefficient is calculated based on the ratio of the edge area of ​​the critical head region to the total area of ​​the critical head region. The interference coefficient represents the degree of influence of edge interference on the overall temperature detection result.

[0068] The regional visibility evaluation value is calculated based on the normalized thermal contrast ratio, thermal signal-to-noise ratio, regional spatial contrast ratio, and interference coefficient. The calculation formula is as follows:

[0069]

[0070] in, Let be the regional visibility evaluation value of the k-th key head region of the i-th pig. To normalize thermal contrast, For thermal signal-to-noise ratio, For regional spatial contrast. This represents the interference coefficient.

[0071] Based on temperature variance and regional visibility evaluation values, the weight values ​​of corresponding key head regions are calculated. The calculation formula is as follows:

[0072]

[0073]

[0074] in, Let be the weight value of the k-th key head region of the i-th pig. It is a non-negative function. To correspond to the temperature variance in key areas of the head, This refers to the visibility evaluation value of the corresponding key head region.

[0075] The target body temperature data for each pig is obtained by weighting and fusing the reference temperature of each key head region with the corresponding weight value of that key head region, using the following formula:

[0076]

[0077]

[0078] in, For the left eye area, The right eye area. The area around the left ear root. The area around the right ear root. Let be the weight value of the k-th key head region of the i-th pig. Let be the reference temperature of the k-th critical area of ​​the head of the i-th pig at time t. To prevent the denominator from being zero, To assess the temperature, , is the smoothing coefficient. For target body temperature data.

[0079] S103: Process millimeter-wave radar data to extract the respiratory and heart rates of each pig.

[0080] In step S103, the millimeter-wave radar data is the millimeter-wave radar echo signal, and the respiratory rate is the number of times the pig breathes per unit time. The heart rate is the number of times the pig's heart beats per unit time. By analyzing and processing the millimeter-wave radar echo signal, these two physiological parameters can be obtained non-contactly.

[0081] In this embodiment, time-frequency analysis is performed on millimeter-wave radar data. For example, short-time Fourier transform or wavelet transform is used to convert the time-domain signal to the time-frequency domain to clearly show the changes of different frequency components in the signal over time. Then, based on the physiological characteristic frequency range of pig respiration and heartbeat, the frequency interval of interest is determined in the time-frequency graph, and the characteristic frequency peaks of corresponding respiration and heartbeat are identified. The peak frequency peaks of corresponding respiration and heartbeat are identified within this interval using a peak detection algorithm.

[0082] In this embodiment, the respiratory and heart rates of each pig are extracted through the processing of millimeter-wave radar data described above. This non-contact measurement method avoids direct contact with the pigs, reduces interference with their normal physiological state, and ensures that the acquired respiratory and heart rate data more accurately reflect the pigs' natural physiological condition, thereby improving the accuracy and reliability of health monitoring. Furthermore, millimeter-wave radar technology can penetrate certain obstacles for detection and is not significantly affected by environmental factors such as light and weather, enabling continuous monitoring around the clock and improving the stability and adaptability of the monitoring.

[0083] Optionally, the millimeter-wave radar data is processed to extract the respiratory and heart rates of each pig, including:

[0084] The thermal imaging data is segmented to obtain the trunk region information of each pig, and the trunk region information of each pig is converted into a set of voxels in the central space of the trunk.

[0085] For any central spatial voxel of the torso, radio frequency features of the central spatial voxel of the torso are extracted. Based on the radio frequency features, a phase sequence is constructed. The phase sequence is subjected to phase expansion processing to obtain the phase change amount. The phase change amounts of all central spatial voxels of the torso are weighted and fused to obtain the temporal information.

[0086] The timing information is filtered by the respiratory frequency band to obtain the respiratory frequency, and the timing information is filtered by the heart rate band to obtain the heart rate.

[0087] In this embodiment, the thermal imaging data is segmented to obtain the trunk region information of each pig, and the trunk region information of each pig is converted into a set of spatial voxels at the trunk center. The millimeter-wave radar data uses the same coordinate system as the radar coordinate system, and the spatial coordinates of any trunk center voxel include range, azimuth, and elevation information in the radar coordinate system.

[0088] For any central spatial voxel of the torso, the radio frequency (RF) features of the central spatial voxel of the torso are extracted. The RF features of the set of central spatial voxels of the torso are represented as follows:

[0089]

[0090] in, Let t be the set of radio frequency features of the set of voxels in the center space of the trunk corresponding to the i-th pig. The central space voxel of the torso is Radio frequency characteristics at that time Let t be the set of voxels in the center space of the trunk of the i-th pig. For distance, The pitch angle, It is the azimuth angle.

[0091] Based on radio frequency characteristics, a phase sequence is constructed. First, a time window of length L is set. Used for extracting temporal respiratory and heart rate data, for any central spatial voxel of the torso. Its complex echo information at time t is A phase sequence is constructed for this, represented as:

[0092]

[0093] in, This indicates the phase angle calculation. Voxel of the central space of the torso The phase sequence.

[0094] Phase expansion is performed on the phase sequence to obtain the phase change. The phase change is expressed as:

[0095]

[0096]

[0097] in, For phase expansion operators, Voxel of the central space of the torso phase sequence, For time t, the central spatial voxel of the torso The phase after being processed by the phase expansion operator, For the central spatial voxel of the torso at time t-1 The phase after being processed by the phase expansion operator, Voxel of the central space of the torso The amount of phase change between adjacent frames suppresses static phase offset and drift.

[0098] The phase changes of all trunk center space voxels are weighted and fused to obtain temporal information. First, the weight value of each trunk center space voxel is determined based on amplitude stability and phase stability. The formulas for calculating amplitude stability and phase stability are as follows:

[0099]

[0100]

[0101] in, To correspond to the amplitude stability of the voxels in the central space of the torso, The length of the time window, Radio frequency characteristics of the central spatial voxel of the torso. For time window, To ensure the phase stability of the corresponding voxel in the central space of the torso, This indicates variance calculation. Voxel of the central space of the torso Phase change between adjacent frames.

[0102] Based on amplitude stability and phase stability, the weight values ​​of the corresponding voxels in the center space of the torso are calculated, and the total calculations are as follows:

[0103]

[0104] in, For the corresponding central space voxel of the torso The weight value, For the corresponding central space voxel of the torso amplitude stability, For the corresponding central space voxel of the torso Phase stability, The amplitude threshold should be determined based on the actual situation. For the corresponding central space voxel of the torso amplitude stability, A value greater than 0 indicates a phase stability adjustment coefficient. Used to prevent the denominator from being zero.

[0105] The temporal information is obtained by weighted summation based on the phase change of each central voxel of the torso and its corresponding weight value. The calculation formula is as follows:

[0106]

[0107] in, Let be the temporal information of the i-th pig at time t. For the corresponding central space voxel of the torso The weight value, Voxel of the central space of the torso Phase change between adjacent frames For time windows.

[0108] The timing information is filtered by the respiratory frequency band to obtain the respiratory rate, and the timing information is filtered by the heart rate band to obtain the heart rate. Specifically, the respiratory frequency band filtering of the timing information extracts the respiratory waveform and the heart rate waveform. The calculation formulas for the respiratory waveform and the heart rate waveform are as follows:

[0109]

[0110]

[0111] in, This is a respiratory waveform. This is a heartbeat waveform. For time series information, For the bandpass filter operator in the breathing frequency band, This is a bandpass filter operator for the heartbeat frequency band.

[0112] Calculate the power spectral density of the respiratory and heartbeat waveforms. Based on the power spectral density of the respiratory and heartbeat waveforms, extract the respiratory and heartbeat frequencies. The calculation formulas are as follows:

[0113]

[0114]

[0115] in, Respiratory rate, Heart rate The power spectral density of the respiratory waveform. The power spectral density of the heartbeat waveform. This is a preset range for the respiratory rate of pigs. This is the preset range of pig heart rate.

[0116] Optionally, the torso region information of each pig is converted into a set of torso center spatial voxels, including:

[0117] Based on the trunk region information of each pig, determine the coordinates of the trunk geometric center of the corresponding pig;

[0118] For any live pig, the geometric center coordinates of the trunk are transformed to obtain the target trunk center space voxel corresponding to the geometric center coordinates of the trunk.

[0119] The target torso's central space voxels are expanded in three dimensions to obtain a set of torso central space voxels.

[0120] In this embodiment, the geometric center coordinates of the corresponding pig's torso are determined based on the torso region information of each pig. Specifically, the centroid of the area of ​​all regions included in the torso region information is used as the geometric center coordinates. The geometric center coordinates of the torso are coordinates in the temperature image coordinate system corresponding to the thermal imaging data, and the formula for the geometric center coordinates of the torso is as follows:

[0121]

[0122] in, Let be the coordinates of the geometric center of the trunk of the i-th pig at time t. This represents the number of sub-regions in the torso region information. Let be the trunk region of the i-th pig at time t. This represents the two-dimensional location of the sub-region.

[0123] For any live pig, the geometric center coordinates of the torso are transformed to obtain the target torso center spatial voxel corresponding to these coordinates. First, the geometric center coordinates of the torso are converted to the corresponding two-dimensional camera coordinates in the camera plane coordinate system. Specifically, the geometric center coordinates of the torso are distorted and converted to homogeneous coordinates, as follows:

[0124]

[0125]

[0126] in, Let be the coordinates of the geometric center of the trunk of the i-th pig at time t. , Let be the camera distortion correction function, where the camera is a thermal imaging camera. These correspond to homogeneous coordinates.

[0127] It should be noted that the camera's intrinsic parameters are... The camera is installed at a height of [height] above the ground. The distortion-free torso geometric center coordinates in the camera plane coordinate system are calculated as the two-dimensional coordinates, i.e., the two-dimensional camera coordinates, which are expressed as:

[0128]

[0129]

[0130] in, Let x be the x-coordinate in the camera plane coordinate system. Let be the ordinate in the camera plane coordinate system.

[0131] Based on the preset offset and the 2D camera coordinates, the radar plane coordinates of the torso's geometric center in the radar plane coordinate system are calculated. Here, the radar plane coordinate system is the plane coordinate system corresponding to the millimeter-wave bio-radar acquiring millimeter-wave radar data. The preset offset is the offset of the millimeter-wave bio-radar relative to the camera on the reference horizontal plane. The coordinates of the torso's geometric center in the millimeter-wave bio-radar plane coordinate system, calculated based on the preset offset and the 2D camera coordinates, are expressed as follows:

[0132]

[0133] in, This is the preset offset. The coordinates of the torso's geometric center are in the millimeter-wave bioradar plane coordinate system.

[0134] Further converted to angular domain form in the millimeter-wave bio-radar coordinate system, it can be expressed as:

[0135]

[0136] in, The pitch angle, Let be the azimuth angle. Using the coordinates corresponding to the azimuth domain as the center point, construct an extended set of the azimuth domain, expanding the center point to a central range, represented as:

[0137]

[0138] in, , This is the preset corner domain expansion amount.

[0139] The target distance center is obtained by retrieving amplitude peaks along the distance cells within the angular domain extension set, denoted as:

[0140]

[0141] in, For distance, The preset target height range for pigs, This represents the amplitude value.

[0142] Based on the corresponding pitch angle, azimuth angle, and distance, determine the spatial voxel of the target torso center. .

[0143] The target torso center space voxels are three-dimensionally expanded to construct a torso center space voxel set based on millimeter-wave bio-radar, represented as:

[0144]

[0145] in, The set of voxels in the central space of the torso. This is the distance extension amount. , This is the preset angular domain expansion amount.

[0146] S104: Perform abnormal sound perception detection on the audio data to obtain the sound perception results for each pig.

[0147] In step S104, abnormal vocalization perception detection involves identifying anomalies in the audio data of pigs, and the vocalization perception results include normal vocalizations and abnormal vocalizations.

[0148] In this embodiment, abnormal vocalization detection is performed on the audio data, extracting acoustic features such as Mel-frequency cepstral coefficients (MFCC), short-time energy, zero-crossing rate, and spectral centroid. These features effectively reflect the time and frequency domain characteristics of the vocalization. The extracted acoustic features are input into a pre-trained abnormal vocalization detection model. This model, built based on deep learning algorithms, learns the feature differences between normal and abnormal vocalizations to classify and judge the input audio features, thereby obtaining the perception result of whether the vocalization of each pig is abnormal at the corresponding time point, i.e., the vocalization perception result.

[0149] In this embodiment, through precise analysis of pig audio data, automated and non-contact monitoring of abnormal pig vocalizations is achieved. This significantly reduces labor costs and subjective judgment errors, and can also capture subtle changes in pig vocalizations in the early stages of disease or under stress, providing timely warnings to farmers so that intervention measures can be taken as early as possible, effectively reducing the risk of disease transmission and economic losses.

[0150] Optionally, the vocalization perception results include normal vocalizations and abnormal vocalizations;

[0151] Anomaly detection processing is performed on the audio data to obtain the sound perception results for each pig, including:

[0152] The audio data is segmented to obtain a multi-frame sub-audio data sequence, and the short-time energy, spectral power and noise energy of each frame sub-audio data are extracted.

[0153] The noise anomaly level of the corresponding frame of sub-audio data is calculated based on the short-time energy and noise energy of each frame of sub-audio data.

[0154] The proportion of high-frequency energy in the corresponding frame of audio data is calculated based on the spectral power of each frame of sub-audio data.

[0155] Based on the noise abnormality level and high-frequency energy ratio of each frame of sub-audio data, the corresponding frame of sub-audio data is subjected to call abnormality detection. If the noise abnormality level and high-frequency energy ratio meet the preset conditions, the corresponding frame of sub-audio data is determined to be call abnormal.

[0156] The number of consecutive frames in the statistical sub-audio data that indicate abnormal vocalizations is determined. If the number of consecutive frames meets the preset abnormality condition, the vocalization perception result is determined to be abnormal; otherwise, the vocalization perception result is determined to be normal.

[0157] In this embodiment, the audio data is segmented according to frame length N to obtain a multi-frame sub-audio data sequence, which is represented as follows:

[0158]

[0159] Extract the short-time energy, spectral power, and noise energy of each frame of sub-audio data. The short-time energy, spectral power, and noise energy are expressed as follows:

[0160]

[0161]

[0162]

[0163] in, Let be the short-time energy of the m-th frame sub-audio data. The audio signal of the m-th frame sub-audio data. Let be the spectral power of the m-th frame of sub-audio data. The frequency domain signal obtained by performing a discrete Fourier transform on the m-th frame of sub-audio data. For frequency point index, Let be the noise energy of the m-th frame of sub-audio data. This describes the process of calculating the median.

[0164] The noise anomaly level of the corresponding frame of audio data is calculated based on the short-time energy and noise energy of each frame. First, the background noise energy scale is calculated based on the short-time energy and noise energy, using the following formula:

[0165]

[0166] in, As a background noise energy scale, Let be the noise energy of the m-th frame of sub-audio data. Let be the short-time energy of the j-th frame sub-audio data. This is the process of calculating the median. Used to prevent zero values. This is the scaling correction factor.

[0167] Energy normalization is performed based on short-time energy, noise energy, and background noise energy scales to obtain the noise anomaly level of the corresponding frame sub-audio data. The calculation formula is as follows:

[0168]

[0169] in, Noise anomaly level, which is the degree of anomaly in the energy of the m-th frame sub-audio data relative to the background noise.

[0170] The proportion of high-frequency energy in the corresponding frame of sub-audio data is calculated based on the spectral power of each frame. The calculation formula is as follows:

[0171]

[0172] in, For the proportion of high-frequency energy, Spectral power of sub-audio data Used to prevent the denominator from being zero. This is a preset high-frequency index range.

[0173] Based on the noise anomaly level and high-frequency energy ratio of each frame of sub-audio data, the corresponding frame of sub-audio data is subjected to call anomaly detection. If the noise anomaly level and high-frequency energy ratio meet preset conditions, the corresponding frame of sub-audio data is determined to have a call anomaly. Specifically, to determine whether the noise anomaly level and high-frequency energy ratio are met, firstly, based on preset constraints, the frame of sub-audio data with call anomalies in the multi-frame sub-audio data sequence is determined. The preset constraint formula is as follows:

[0174]

[0175] in, For indicator functions, when When the condition is true, that is Greater than ,and Greater than When the function value is 1, that is... The value is 1; when the condition is false, the function value is 0, that is... The value is 0. The threshold for the degree of noise anomaly. This is the threshold for the proportion of high-frequency energy.

[0176] Based on the output of each frame of sub-audio data, determine the frame of audio data corresponding to an output result of 1. If the noise abnormality level and the proportion of high-frequency energy meet the preset conditions, the corresponding frame sub-audio data is identified as having abnormal vocalizations.

[0177] The statistical sub-audio data is used to count the number of consecutive frames with abnormal vocalizations. If the number of consecutive frames meets a preset abnormality condition, the vocalization perception result is determined to be abnormal; otherwise, the vocalization perception result is determined to be normal. The determination of whether the number of consecutive frames meets the preset abnormality condition is based on the following formula.

[0178] First, continuously satisfy The sub-audio data frames are merged into an abnormal cry event, represented as:

[0179]

[0180] in, As the starting frame, To terminate the frame, The sampling rate of the audio data. This is the frame shift length, i.e., the number of sampling points between two adjacent frames. A threshold greater than or equal to a preset threshold is used as a filtering condition for valid abnormal vocalization events. The probability of an event set within a statistical window of length T is calculated and expressed as:

[0181]

[0182] in, For a set of events, i.e. The number of abnormal vocalizations that are greater than or equal to a preset threshold. For the corresponding probability. When If the probability exceeds the preset probability threshold, the sound perception result is determined to be abnormal; otherwise, the sound perception result is determined to be normal.

[0183] S105: Health monitoring is conducted based on the target body temperature data, respiratory rate, heart rate, and vocalization perception results of each pig to obtain monitoring results.

[0184] In step S105, health monitoring involves converting the pig's body temperature data, respiratory rate, heart rate, and vocalization perception results into a quantitative health status assessment value.

[0185] In this embodiment, normal range thresholds are set for target body temperature data, respiratory rate, and heart rate, and status indicators are set for the vocalization perception results. For a single pig, its real-time collected target body temperature data is compared with the preset normal body temperature range; if it exceeds the range, it is marked as abnormal body temperature. Similarly, respiratory rate and heart rate are compared with their respective preset normal ranges; if they exceed the range, they are marked as abnormal breathing and abnormal heart rate, respectively. A multi-dimensional health status assessment is performed by combining the abnormal markings of body temperature, breathing, and heart rate with the vocalization perception results. For example, when abnormal body temperature is accompanied by an abnormal vocalization perception result, or when respiratory rate and heart rate are both abnormal and the vocalization perception result is abnormal, the pig's health monitoring result can be determined to be abnormal. When the target body temperature data, respiratory rate, and heart rate are all within the normal range and the vocalization perception result is normal, the health monitoring result is determined to be normal. This multi-parameter fusion judgment method can improve the accuracy and reliability of pig health monitoring and promptly detect potential health problems.

[0186] Optionally, health monitoring is conducted based on the target body temperature data, respiratory rate, heart rate, and vocalization perception results of each pig to obtain monitoring results, including:

[0187] Obtain a preset pig health prediction model, input the target body temperature data, respiratory rate, heart rate and vocalization perception results into the pig health prediction model, and output the health prediction results.

[0188] In this embodiment, the swine health prediction model is a machine learning model trained based on historical swine health data. This historical swine health data includes a large amount of body temperature data, respiratory rate, heart rate, vocalization perception results, and corresponding health status labels, such as healthy, slightly abnormal, and severely abnormal, for swine with known health status. The target body temperature data, respiratory rate, heart rate, and vocalization perception results are input into the swine health prediction model, and the model outputs a health prediction result.

[0189] In this embodiment, the pig health monitoring method introduces a pig health prediction model, comprehensively analyzing multi-dimensional physiological indicators and vocalization perception results to achieve intelligent and precise monitoring of pig health status. This effectively improves the accuracy and timeliness of health status identification, providing strong technical support for the refined management of modern pig farming.

[0190] This application integrates three different types of sensor data—thermal imaging data, millimeter-wave radar data, and audio data—to achieve multi-dimensional and comprehensive monitoring of pig health status. Thermal imaging data can accurately capture changes in pig body temperature, millimeter-wave radar data can acquire vital signs such as respiration and heartbeat in real time, and audio data can detect potential abnormalities through analysis of vocalizations. This multi-source data fusion approach can more comprehensively and accurately reflect the health status of pigs, improving the accuracy of pig health monitoring. Furthermore, this intelligent monitoring method can promptly detect potential risks in the early stages of pig health abnormalities, providing a scientific basis for decision-making by farm managers, helping to reduce the risk of disease transmission, improve farming efficiency, and promote the digital and intelligent transformation and upgrading of the pig farming industry.

[0191] Please see Figure 2 , Figure 2 This is a schematic diagram of a pig farming health monitoring device according to an embodiment of this application. This pig farming health monitoring device corresponds one-to-one with the pig farming health monitoring methods described in the above embodiments. Please refer to [link / reference] for details. Figure 1 as well as Figure 1 The relevant descriptions in the corresponding embodiments are shown below. For ease of explanation, only the parts relevant to this embodiment are shown. See also... Figure 2 The pig farming health monitoring device 20 includes: a thermal imaging camera module, a millimeter-wave bioradar module, a microphone module, and a signal processing module. The thermal imaging camera module, millimeter-wave bioradar module, and microphone module are all communicatively connected to the signal processing module. Specifically, the thermal imaging camera module includes a thermal imaging camera device, the millimeter-wave bioradar module includes a millimeter-wave bioradar device, and the microphone module includes a microphone device.

[0192] The thermal imaging camera module is used to acquire thermal imaging data within the pig farm and transmit the thermal imaging data to the signal processing module.

[0193] The millimeter-wave bio-radar module is used to acquire millimeter-wave radar data in pig farms and transmit the millimeter-wave radar data to the signal processing module.

[0194] The microphone module is used to acquire audio data within the pig farm and transmit the audio data to the signal processing module.

[0195] The signal processing module is used to execute the above-mentioned methods for monitoring the health of pig farming.

[0196] Please see Figure 3 , Figure 3 This is another structural schematic diagram of a pig farming health monitoring device provided in one embodiment of this application. The pig farming health monitoring device also includes a battery module and a WIFI / Bluetooth module. The battery module is used to power the entire pig farming health monitoring device. The WIFI / Bluetooth module is used to transmit the health detection results from the signal processing module to the server.

[0197] It should be noted that the swine health monitoring device is installed within the swine farm, and its specific installation location can be flexibly chosen based on the farm's layout and monitoring needs. For example, the device can be fixedly installed at a high position such as on a wall or pillar within the pens to ensure that the thermal imaging camera module, millimeter-wave bio-radar module, and microphone module can cover a large monitoring area without obstruction, thereby comprehensively capturing the swine's activity status and physiological information. Simultaneously, the installation height and angle of the device should be reasonably adjusted to avoid damage to the device due to the swine's daily activities and to prevent the device itself from interfering with the swine's normal life. Furthermore, for some large-scale farms, the monitoring device can be installed separately in different farming units according to the division of farming areas, enabling independent monitoring and management of the swine health status in each area.

[0198] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0199] Figure 4 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application. For example... Figure 4 As shown, the computer device of this embodiment includes: at least one processor ( Figure 4 Only one is shown in the diagram), a memory, and a computer program stored in the memory and capable of running on at least one processor, wherein the processor executes the computer program to implement the steps in any of the above embodiments of the pig farming health monitoring method.

[0200] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 4 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown in the illustration, or combinations of certain components, or different components, such as network interfaces, displays, and input devices.

[0201] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0202] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of a computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal storage units and external storage devices of the computer device. Memory is used to store the operating system, applications, bootloader, data, and other programs, such as program code for computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0203] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable medium can include at least: any entity or device capable of carrying computer program code, a recording medium, a computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0204] The implementation of all or part of the processes in the methods of the above embodiments can also be accomplished by a computer program product. When the computer program product is run on a computer device, it enables the computer device to execute the steps in the above method embodiments.

[0205] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0206] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0207] In the embodiments provided in this application, it should be understood that the disclosed apparatus / computer devices and methods can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0208] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0209] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for monitoring the health of pigs in breeding, characterized in that, The methods for monitoring the health of pig farming include: Acquire thermal imaging data, millimeter-wave radar data, and audio data within a pig farm per unit time period; The thermal imaging data is segmented to obtain the target body temperature data of each pig in the pig farm; The millimeter-wave radar data is processed to extract the respiratory rate and heart rate of each pig. The audio data is subjected to abnormal vocalization detection to obtain the vocalization detection results for each pig. Health monitoring is conducted on each pig based on the target body temperature data, respiratory rate, heart rate, and vocalization perception results to obtain monitoring results; The step of segmenting the thermal imaging data to obtain the target body temperature data for each pig in the pig farm includes: The thermal imaging data is segmented to obtain head region information for each pig; The key regions of the head region are extracted to obtain a set of key head regions; Based on the thermal imaging data, the temperature data of each key head region in the set of key head regions is determined, and the reference temperature of each key head region is calculated based on the temperature data. The reference temperatures of each key area of ​​the head are weighted and fused to obtain the target body temperature data for each pig. The process involves weighted fusion of reference temperatures from key areas of the head to obtain target body temperature data for each pig, including: Based on the temperature data of each key area of ​​the head, the temperature variance of the corresponding key area of ​​the head is calculated. Based on the thermal imaging data, the corresponding background temperature data in the pig farm is determined, and the regional visibility evaluation value of the corresponding key head area is calculated based on the background temperature data and the temperature data of each key head area. The weight values ​​of the corresponding key head regions are calculated based on the temperature variance and the regional visibility evaluation value. The reference temperature of each key area of ​​the head is weighted and fused with the corresponding weight value of the key area of ​​the head to obtain the target body temperature data for each pig.

2. The method for monitoring the health of pig farming as described in claim 1, characterized in that, The process of processing the millimeter-wave radar data to extract the respiratory and heart rates of each pig includes: The thermal imaging data is segmented to obtain the trunk region information of each pig, and the trunk region information of each pig is converted into a set of voxels in the central space of the trunk. For any central spatial voxel of the torso, the radio frequency features of the central spatial voxel of the torso are extracted. Based on the radio frequency features, a phase sequence is constructed. The phase sequence is then subjected to phase expansion processing to obtain the phase change amount. The phase change values ​​of all central spatial voxels of the torso are weighted and fused to obtain temporal information; The timing information is subjected to respiratory frequency band filtering to obtain respiratory frequency, and the timing information is subjected to heartbeat frequency band filtering to obtain heartbeat frequency.

3. The method for monitoring the health of pig farming as described in claim 2, characterized in that, The process of converting the trunk region information of each pig into a set of trunk center spatial voxels includes: Based on the trunk region information of each pig, determine the coordinates of the trunk geometric center of the corresponding pig; For any live pig, the geometric center coordinates of the trunk are transformed to obtain the target trunk center space voxel corresponding to the geometric center coordinates of the trunk. The target torso center space voxels are three-dimensionally expanded to obtain a torso center space voxel set.

4. The method for monitoring the health of pig farming as described in claim 1, characterized in that, The vocalization perception results include normal vocalizations and abnormal vocalizations. Anomaly detection processing is performed on the audio data to obtain the sound perception results for each pig, including: The audio data is segmented to obtain a multi-frame sub-audio data sequence, and the short-time energy, spectral power and noise energy of each frame sub-audio data are extracted. The noise anomaly level of the corresponding frame of audio data is calculated based on the short-time energy and the noise energy of each frame of sub-audio data. The proportion of high-frequency energy in the corresponding frame of audio data is calculated based on the spectral power of each frame of sub-audio data. Based on the noise abnormality level and the high-frequency energy ratio of each frame of sub-audio data, the corresponding frame of sub-audio data is subjected to call abnormality detection. If the noise abnormality level and the high-frequency energy ratio meet the preset conditions, the corresponding frame of audio data is determined to be call abnormal. The number of consecutive frames in the statistical sub-audio data that indicate abnormal vocalizations is determined. If the number of consecutive frames meets a preset abnormality condition, the vocalization perception result is determined to be abnormal; otherwise, the vocalization perception result is determined to be normal.

5. The method for monitoring the health of pig farming as described in claim 1, characterized in that, The health monitoring is performed based on the target body temperature data, respiratory rate, heart rate, and vocalization perception results of each pig, and the monitoring results include: Obtain a preset pig health prediction model, input the target body temperature data, respiratory rate, heart rate and vocalization perception results into the pig health prediction model, and output the health prediction result.

6. A health monitoring device for pig farming, characterized in that, The pig farming health monitoring device includes a thermal imaging camera module, a millimeter-wave bio-radar module, a microphone module, and a signal processing module. The thermal imaging camera module, the millimeter-wave bio-radar module, and the microphone module are respectively communicatively connected to the signal processing module. The thermal imaging camera module is used to acquire thermal imaging data in the pig farm and transmit the thermal imaging data to the signal processing module. The millimeter-wave bio-radar module is used to acquire millimeter-wave radar data in the pig farm and transmit the millimeter-wave radar data to the signal processing module. The microphone module is used to acquire audio data within the pig farm and transmit the audio data to the signal processing module. The signal processing module is used to execute the pig farming health monitoring method according to any one of claims 1 to 5.

7. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the pig farming health monitoring method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the pig farming health monitoring method as described in any one of claims 1 to 5.

Images