Method and apparatus for detecting breathing patterns in animals

Abstract

The embodiment of the application provides a kind of animal respiratory mode detection method and device, by obtaining the image data corresponding to each target animal in target stall, processing image data, obtain the contour displacement feature and thermal radiation variation feature of target animal and generate fusion respiratory waveform signal, determine the animal species of target animal and corresponding preset unit time, determine the number of fusion respiratory waveform signal wave crest in preset unit time to determine the individual respiratory mode of target animal, determine all individual respiratory mode as group respiratory mode, determine the group coupling index of group respiratory mode, and determine the coupling similarity of group coupling index and normal coupling index, in the case where coupling similarity does not exceed preset similarity threshold, determine that there is sick animal in target stall.The method effectively solves the deficiency in real-time and continuous detection of animal respiratory frequency and the like, significantly improves the efficiency and accuracy of real-time and continuous detection of animal respiratory frequency.

Description

Technical Field

[0001] This application relates to the field of image processing, specifically to a method and apparatus for detecting animal breathing patterns. Background Technology

[0002] In the prior art, the respiratory rate and respiratory pattern of an animal are important physiological indicators that can reflect the animal's health status and potential diseases. The respiratory rate of an animal can be determined by visually observing the rise and fall of the animal's chest and abdomen or by using contact sensors, such as chest strap sensors or nasal airflow meters.

[0003] However, contact sensor devices can interfere with the animal's natural behavior, causing stress that affects the detected respiratory rate and pattern, and may also have a negative impact on the animal's health. Furthermore, it is difficult to visually assess the animal's movements in intensive farming. Summary of the Invention

[0004] To address the problems in the prior art, this application provides a method and device for detecting animal respiratory patterns, which can effectively solve the shortcomings of traditional technologies in that it is difficult to detect the respiratory rate of animals in real time and continuously, resulting in the failure to detect infected animals in a timely manner, and significantly improve the efficiency and accuracy of real-time and continuous detection of animal respiratory rate.

[0005] To solve at least one of the above problems, this application provides the following technical solution:

[0006] Firstly, this application provides a method for detecting animal respiratory patterns, including:

[0007] The image data corresponding to each target animal in the target field is acquired. The image data is processed by a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal. Based on the contour displacement features and thermal radiation change features, a fused respiratory waveform signal is generated. The image data includes the contour displacement image corresponding to the target animal acquired by the visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired by the infrared thermal imaging sensor.

[0008] The animal species of the target animal is determined based on the image data, and the preset unit time corresponding to the animal species is determined. The number of peaks of the fused respiratory waveform signal within the preset unit time is determined. The individual respiratory pattern of the target animal is determined based on the number of peaks. The individual respiratory pattern includes the individual respiratory rate, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0009] All individual breathing patterns are identified as group breathing patterns. The group coupling index of the group breathing pattern is determined, and the coupling similarity between the group coupling index and the normal coupling index is determined. If the coupling similarity does not exceed the preset similarity threshold, it is determined that there are diseased animals in the target field.

[0010] Furthermore, the pre-defined dual-stream neural network includes a visible photonic network and an infrared sub-network, which operate in parallel. It also includes:

[0011] Background subtraction processing is performed on continuous contour displacement images to obtain the initial animal contour region corresponding to the image data. Then, edge detection is performed on the initial animal contour region using an edge detection algorithm to obtain the contour edge information of the target animal.

[0012] The contour displacement features of the target animal are obtained by extracting contour edge information through convolutional layers in a visible photonic network.

[0013] Furthermore, the pre-defined dual-stream neural network includes a visible photonic network and an infrared sub-network, which operate in parallel. It also includes:

[0014] The thermal radiation region of the target animal's body surface is extracted from the thermal radiation image by a preset segmentation algorithm to obtain the initial thermal radiation region. The initial thermal radiation region is then analyzed in time series based on the image data to obtain the thermal radiation changes of the target animal's body surface.

[0015] The thermal radiation change characteristics of the target animal are obtained by extracting features of thermal radiation changes through convolutional layers in the infrared subnetwork.

[0016] Furthermore, it also includes: acquiring the time series corresponding to the image data, and aligning the contour displacement features and thermal radiation change features based on the time series;

[0017] The contour displacement features and thermal radiation change features after fusion and alignment are used to obtain comprehensive features, and the comprehensive features are then converted to obtain fused breathing waveform signals.

[0018] Furthermore, it also includes: determining the corresponding unit time threshold based on the animal species from a preset relationship mapping table, wherein the preset relationship mapping table is used to represent the mapping relationship between animal species and unit time threshold;

[0019] Based on the actual respiratory cycle corresponding to the animal species, a preset unit time is determined based on a unit time threshold.

[0020] Furthermore, it also includes: traversing the individual breathing waveforms corresponding to each individual breathing pattern in the group breathing pattern, and determining the morphological similarity between any two individual breathing waveforms within the same time window;

[0021] The average morphological similarity of the breathing waveforms of all individuals included in the group breathing pattern is determined as the group coupling index.

[0022] Furthermore, after confirming the presence of diseased animals in the target field, the following steps are also included:

[0023] In group breathing patterns, identify the zero-case animal with abnormal individual respiratory waveforms and determine the relative positions among target animals based on imaging data;

[0024] Based on the zero pathogen and its relative location, the respiratory similarity of the respiratory patterns of the target animals centered on the zero pathogen and adjacent to the zero pathogen is determined, and the decay gradient of the respiratory similarity is determined.

[0025] When the attenuation gradient spreads radially, it is determined that there is a group disease transmission in the current field.

[0026] Secondly, this application provides an animal respiratory pattern detection device, comprising:

[0027] The first processing module is used to acquire image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired based on the visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired based on the infrared thermal imaging sensor.

[0028] The second processing module is used to determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes the individual respiratory frequency, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0029] The third processing module is used to determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing pattern, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, it is determined that there are diseased animals in the target field.

[0030] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the animal respiratory pattern detection method.

[0031] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the animal respiratory pattern detection method described above.

[0032] Fifthly, this application provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the animal respiratory pattern detection method.

[0033] As can be seen from the above technical solution, this application provides a method and device for detecting animal respiratory patterns. It innovatively acquires image data corresponding to each target animal in the target field, processes the image data using a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generates a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired by a visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. The animal species of the target animal is determined based on the image data, thereby determining the preset unit time corresponding to the animal species and the number of peaks in the fused respiratory waveform signal per unit time. The individual respiratory patterns of target animals are determined based on the number of peaks. These individual respiratory patterns include individual respiratory rate, individual respiratory depth waveform, and individual respiratory rhythm disorder. All individual respiratory models are then classified as group respiratory patterns, thereby determining the group coupling index and the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, the presence of diseased animals in the target area is confirmed. This method can determine the individual respiratory patterns of target animals using image data, avoiding errors associated with contact-based respiratory rate detection and improving the accuracy of individual respiratory patterns. Furthermore, it can analyze both individual and group respiratory patterns to determine the disease status of each animal in the target area. This method effectively addresses the shortcomings of real-time and continuous detection of animal respiratory rates, which can lead to delayed detection of diseased animals, significantly improving the efficiency and accuracy of real-time and continuous animal respiratory rate detection. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating the animal respiratory pattern detection method in the embodiments of this application;

[0036] Figure 2 This is a structural diagram of the animal respiratory pattern detection device in the embodiments of this application;

[0037] Figure 3 This is a schematic diagram of the structure of the electronic device in the embodiments of this application.

[0038] Figure label:

[0039] Electronic device 9600, central processing unit 9100, memory 9140, communication module 9110, input unit 9120, audio processor 9130, display 9160, power supply 9170, buffer memory 9141, application / function storage unit 9142, data storage unit 9143, driver storage unit 9144, antenna 9111, speaker 9131, microphone 9132. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0041] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0042] In view of the problems existing in the prior art, this application provides a method and device for detecting animal breathing patterns. It can continuously acquire video using a top-mounted camera or an underwater camera, separate the animal outline using background subtraction, and then extract key parts through edge detection. For example, in detecting the individual breathing patterns of pigs, the ventral curve fluctuations of the pig can be extracted from a top-down view, and the periodic movement from the pig's shoulder to the waist angle can be calculated to determine the individual breathing pattern of the pig. Simultaneously, by locating key parts of the target animal, such as the mouth of a fish or the abdomen of a pig, and tracking the target animal over a long period based on these key parts, long-term detection of the target animal's individual breathing patterns can be achieved.

[0043] To effectively address the shortcomings of traditional technologies in real-time and continuous detection of animal respiratory rate, which leads to delayed detection of infected animals, and to significantly improve the efficiency and accuracy of real-time and continuous detection of animal respiratory rate, this application provides an embodiment of an animal respiratory pattern detection method, see [link to embodiment]. Figure 1 The animal respiratory pattern detection method specifically includes the following:

[0044] Step S101: Obtain the image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features.

[0045] The image data includes contour displacement images of the target animal acquired using a visible light sensor and surface thermal radiation images of the target animal acquired using an infrared thermal imaging sensor.

[0046] Optionally, in this embodiment, image data corresponding to each target animal in the target field is obtained, that is, image data corresponding to all animals in the target field is obtained, and each animal is taken as a target animal to determine its corresponding individual respiratory pattern.

[0047] The image data includes contour displacement images of the target animal obtained by a visible light sensor and body surface thermal radiation images of the target animal obtained by an infrared thermal imaging sensor. The contour displacement images represent the physical movement of the target animal during respiration, and the body surface thermal radiation images represent the physiological changes of the target animal during respiration.

[0048] In addition, by processing image data using a preset dual-stream neural network model, the contour displacement features and thermal radiation change features of the target animal can be extracted. Before processing the image data using the preset dual-stream neural network model, noise removal and normalization can be performed on the image data to ensure that the obtained contour displacement features and thermal radiation change features are accurate and smooth.

[0049] Before processing the image data through the preset dual-stream neural network model, the image data can be upsampled and denoised. The image acquisition process requires, but is not limited to, the complete chest and abdomen of the target animal and other parts that can be affected by the respiratory process, as well as the clear outline of the target animal. The sensor needs to be kept stable so that the image is free from blurring, trailing, or other poor quality, and that the relative positions of different consecutive image data are the same.

[0050] Furthermore, a fused respiratory waveform signal is generated based on contour displacement features and thermal radiation change features. This can be achieved by fusing the contour displacement features and thermal radiation change features, or by splicing the contour displacement features and thermal radiation change features together. This allows the fused respiratory waveform signal to more comprehensively reflect the respiratory state of the target animal and improves the accuracy and reliability of the fused respiratory waveform signal.

[0051] In addition, contour displacement images can be processed by phase magnification algorithms or optical flow methods to convert the micron-level movement of respiratory organs into visual waveforms. For example, in the process of detecting the individual respiratory rate of cattle, optical flow methods can capture abdominal undulations of less than 1 millimeter.

[0052] Furthermore, image data can be acquired through continuous acquisition by a top-mounted visible light sensor and an infrared thermal imaging sensor, or through continuous acquisition by an underwater visible light sensor and an infrared thermal imaging sensor. When acquiring underwater image data, a turbidity compensation algorithm can be used to handle situations with low underwater visibility.

[0053] This embodiment enables the fusion of respiratory waveform signals corresponding to the target animal through image data of the target animal. The image data includes contour displacement images and body surface thermal radiation images, thereby improving the accuracy and reliability of the fusion of respiratory waveform signals.

[0054] Step S102: Determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks.

[0055] Individual breathing patterns include individual respiratory rate, individual respiratory depth waveform, and individual respiratory rhythm disorder.

[0056] Optionally, this embodiment can identify the species of the target animal based on the acquired image data. This can be achieved by inputting the contour displacement image from the image data into a deep learning model. The deep learning model can extract the feature information of the target animal from the image data and compare it with the animal species existing in a preset animal species database to determine the species of the target animal.

[0057] In addition, based on the identified animal species, a preset unit time is determined for that animal species. Within the preset unit time, the number of peaks in the fused respiratory waveform signal is analyzed to determine the individual respiratory pattern of the target animal. The individual respiratory pattern includes the individual respiratory rate, respiratory depth waveform, and respiratory rhythm disorder.

[0058] Individual respiratory rate represents the number of times the target animal breathes per unit time. Individual respiratory depth waveform represents the curve of the expansion and contraction amplitude of the respiratory organs during each breath of the target animal, indicating the amount of air entering and leaving the respiratory process. Respiratory rhythm disorder represents the degree of irregularity of the respiratory cycle of the target animal during the breathing process, and is used to measure the variability and uncertainty of the respiratory interval time.

[0059] Furthermore, adaptive windowing technology can be used to determine the preset unit time, that is, to dynamically adjust the preset unit time window based on the real-time monitored respiratory signals, so as to more flexibly adapt to the individual respiratory pattern changes of target animals of different animal species.

[0060] This embodiment enables the determination of the individual respiratory pattern of a target animal by fusing respiratory waveform signals, which can accurately analyze and evaluate the respiratory status of each target animal in detail, thereby improving accuracy.

[0061] Step S103: Determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing patterns, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, determine that there are infected animals in the target field.

[0062] Optionally, in this embodiment, all individual breathing patterns are defined as group breathing patterns, that is, the group breathing patterns include the individual breathing patterns corresponding to all target animals in the target field. The group coupling index of the group breathing patterns is determined, wherein the group coupling index is used to represent the consistency of the breathing behavior of target animals in the group.

[0063] In addition, the coupling similarity between the group coupling index and the normal coupling index is determined. If the coupling similarity does not exceed the preset similarity threshold, it is determined that there are diseased animals in the target column.

[0064] Furthermore, if the coupling similarity does not exceed a preset similarity threshold, an alarm message can be sent to indicate the presence of a diseased animal in the target field. Also, if the coupling similarity does not exceed a preset similarity threshold, the ventilation system of the target field can be driven to adjust ventilation data such as oxygen content or ventilation volume based on individual and group breathing patterns.

[0065] Furthermore, YOLOv7 or Mask R-CNN can be used to locate the breathing sites of target animals, such as the mouth of fish or the abdomen of mammals. Individual target animals can be tracked over a long period using ReID technology (e.g., based on pigmented markings). LSTM networks can also be used to process and fuse non-stationary signals in respiratory waveforms to identify long-term changes in an individual's breathing patterns, such as progressive shortness of breath caused by disease.

[0066] This embodiment enables the detection of respiratory patterns in the group behavior of target animals, thereby identifying potential health problems within the group and providing timely alerts.

[0067] This embodiment enables high-precision detection of individual respiratory patterns of target animals based on image data. It can also determine group respiratory patterns based on individual respiratory patterns and analyze group respiratory patterns to discover potential health problems of target animals in the target field, thereby improving the accuracy and reliability of respiratory rate detection.

[0068] In some embodiments, the preset dual-stream neural network includes a visible photonics subnetwork and an infrared subnetwork, and the visible photonics subnetwork and the infrared subnetwork are parallel.

[0069] By processing image data using a pre-defined dual-stream neural network model, the contour displacement features and thermal radiation change features of the target animal are obtained, including:

[0070] Background subtraction processing is performed on continuous contour displacement images to obtain the initial animal contour region corresponding to the image data. Then, edge detection is performed on the initial animal contour region using an edge detection algorithm to obtain the contour edge information of the target animal.

[0071] The contour displacement features of the target animal are obtained by extracting features of contour edge information through convolutional layers in a visible photonic network.

[0072] Optionally, the preset dual-stream neural network in this embodiment includes a visible light sub-network and an infrared sub-network, which are in parallel structure, thereby allowing simultaneous processing of image data from a visible light sensor and image data from an infrared thermal imaging sensor. The visible light sub-network is used to process contour displacement images, and the infrared sub-network is used to process body surface thermal radiation images, which improves data processing efficiency and enhances robustness.

[0073] In addition, continuous contour displacement images are acquired through image data, and background subtraction processing is performed on the continuous contour displacement images. The background subtraction processing can be carried out by comparing the differences between the continuous contour displacement images to separate the initial contour region of the target animal, thereby removing background interference and making the contour of the target animal clearer.

[0074] In addition, the initial animal contour region is processed by an edge detection algorithm to extract the contour edge information of the target animal. The contour edge information provides detailed data on the contour changes of the target animal.

[0075] Furthermore, by using the convolutional layer in the visible photonics network of the pre-defined dual-stream neural network, the features of the contour edge information are further extracted to obtain the contour displacement features of the target animal. The contour displacement features reflect the movement changes of the respiratory organs of the target animal during respiration.

[0076] Furthermore, an adaptive background subtraction algorithm can be used to automatically adjust the subtraction threshold according to changes in ambient light, thereby accurately extracting the contour edge information of the target animal under different lighting conditions. In addition, a multi-scale edge detection algorithm can be used to detect edge information at different scales, thereby improving the accuracy and completeness of contour edge information extraction.

[0077] This embodiment achieves high-precision detection of individual breathing patterns of target animals through a preset dual-stream neural network model. In particular, by using background subtraction and edge detection algorithms, the contour displacement features of the target animal can be accurately extracted, which improves the accuracy of individual breathing pattern detection and enhances robustness and adaptability. Thus, the contour displacement features of the target animal can be determined under different environmental conditions.

[0078] In some embodiments, the preset dual-stream neural network includes a visible photonics subnetwork and an infrared subnetwork, and the visible photonics subnetwork and the infrared subnetwork are parallel.

[0079] By processing image data using a pre-defined dual-stream neural network model, the contour displacement features and thermal radiation change features of the target animal are obtained, including:

[0080] The thermal radiation region of the target animal's body surface is extracted from the thermal radiation image by a preset segmentation algorithm to obtain the initial thermal radiation region. The initial thermal radiation region is then analyzed in time series based on the image data to obtain the thermal radiation changes of the target animal's body surface.

[0081] The thermal radiation variation characteristics of the target animal are obtained by extracting features of thermal radiation variation through convolutional layers in the infrared subnetwork.

[0082] Optionally, the preset dual-stream neural network in this embodiment includes a visible light sub-network and an infrared sub-network, which are in parallel structure, thereby allowing simultaneous processing of image data from a visible light sensor and image data from an infrared thermal imaging sensor. The visible light sub-network is used to process contour displacement images, and the infrared sub-network is used to process body surface thermal radiation images, which improves data processing efficiency and enhances robustness.

[0083] In addition, during the processing of the surface thermal radiation image acquired by the infrared thermal imaging sensor, the thermal radiation region of the target animal's body surface is extracted by a preset segmentation algorithm to obtain the initial thermal radiation region. The extraction of the initial thermal radiation region of the target animal's body surface can be achieved by a region growing algorithm or a threshold segmentation algorithm, thereby effectively removing background interference and accurately extracting the initial thermal radiation region of the animal's body surface.

[0084] In addition, the time series of the image data is determined, and the initial thermal radiation area is analyzed based on the time series of the image data to identify the periodic fluctuations of the target animal's body surface temperature, thereby obtaining the changes in the target animal's body surface thermal radiation. Among them, the changes in body surface thermal radiation reflect the dynamic changes in body surface temperature during the target animal's respiration process, and can determine the physiological indicators of the target animal's respiratory state.

[0085] Furthermore, by extracting the features of thermal radiation changes through the convolutional layer in the infrared subnetwork of the pre-defined dual-stream neural network, the thermal radiation change features of the target animal are obtained, where the thermal radiation change features represent the local changes in the thermal radiation of the target animal.

[0086] Furthermore, in the process of extracting the thermal radiation region using a preset segmentation algorithm, an adaptive threshold segmentation algorithm can be used to adaptively adjust the segmentation threshold according to the background noise of the thermal radiation image, thereby enabling the extraction of the thermal radiation region under different environmental conditions.

[0087] Furthermore, deep convolutional layers can be added to the infrared subnetwork to extract deeper features of thermal radiation variation.

[0088] This embodiment achieves high-precision detection of individual respiratory patterns of target animals through a preset dual-stream neural network model. By extracting the thermal radiation region through a preset segmentation algorithm and performing time-series analysis, it can accurately extract the thermal radiation change characteristics of the target animal, improve the accuracy of detection, and enhance robustness and adaptability. Thus, it can determine the thermal radiation change characteristics of the target animal under different environmental conditions.

[0089] In some embodiments, a fused respiratory waveform signal is generated based on contour displacement features and thermal radiation change features, including:

[0090] Obtain the time series corresponding to the image data, and align the contour displacement features and thermal radiation change features based on the time series;

[0091] The contour displacement features and thermal radiation change features after fusion and alignment are used to obtain comprehensive features, and the comprehensive features are then converted to obtain fused breathing waveform signals.

[0092] Optionally, in this embodiment, the time series corresponding to the image data is obtained, and the contour displacement features and thermal radiation change features are aligned according to the time series.

[0093] One approach is to align contour displacement features and thermal radiation change features using time series alignment algorithms, such as Dynamic Time Warping (DTW). This algorithm adaptively adjusts the time series of contour displacement features and thermal radiation change features, and accurately aligns them to ensure that the two features remain synchronized in time. DTW can handle nonlinear changes in the time series, improve alignment accuracy, and reduce errors caused by time deviations.

[0094] In addition, the integrated features are obtained by fusing the aligned contour displacement features and thermal radiation change features. The integrated features include, but are not limited to, the physical motion information and physiological change information of the target animal's respiration.

[0095] Contour displacement features and thermal radiation change features can be fused using either weighted feature fusion or deep feature fusion. Weighted feature fusion can assign different weights to contour displacement features and thermal radiation change features based on their importance, while deep feature fusion can use a deep learning model to adaptively learn the fusion strategy to determine the complex relationship between contour displacement features and thermal radiation change features, thereby generating comprehensive features.

[0096] In addition, the comprehensive features are converted into a fused respiratory waveform signal, which can reflect the respiratory state of the target animal.

[0097] Furthermore, in the process of converting the comprehensive features into a fused respiratory waveform signal, signal processing techniques, such as wavelet transform or adaptive filters, can be used to remove noise and abnormal fluctuations, enhance the quality and stability of the fused respiratory waveform signal, and improve the clarity of the signal.

[0098] Furthermore, the composite signal can be treated as a physical signal and converted into a digital waveform signal, i.e., a fused respiratory waveform signal. This involves generating usable digital information from the composite signal, determining continuous position information based on a time series, obtaining multiple consecutive digital information based on the continuous position and digital information, and performing filtering and amplification operations on the digital information. Specifically, adjacent erroneous or jagged digital information is smoothed, and adjacent distant digital information is smoothed using limit values. Before processing erroneous digital information, it is first amplified to prevent correct digital information from being mishandled.

[0099] The fused respiratory waveform signal may also contain abnormal waveforms. These abnormal waveforms may originate from sensor jitter during image data recording or minor behaviors of the target animal other than breathing. Frame difference detection of the original image data can be performed by combining preset limit values ​​per unit time to identify the differences between normal and abnormal waveforms, thereby reducing abnormal waveforms in the fused respiratory waveform signal.

[0100] This embodiment achieves temporal synchronization of contour displacement features and thermal radiation change features through time series alignment, thereby improving the quality of fused respiratory waveform signals, enhancing the accuracy of individual respiratory pattern detection, and increasing robustness and adaptability.

[0101] In some embodiments, determining a preset unit of time corresponding to an animal species includes:

[0102] The corresponding unit time threshold is determined from the preset relationship mapping table based on the animal species. The preset relationship mapping table is used to represent the mapping relationship between animal species and unit time threshold.

[0103] Based on the actual respiratory cycle corresponding to the animal species, a preset unit time is determined based on a unit time threshold.

[0104] Optionally, in this embodiment, the unit time threshold corresponding to the animal species corresponding to the current target animal is obtained from a preset relationship mapping table. The preset relationship mapping table is a data structure used to record the mapping relationship between different animal species and their corresponding unit time thresholds.

[0105] The mapping relationship between different animal species and their corresponding unit time thresholds is based on the physiological characteristics and respiratory cycles of different target animals, so that each animal species has a suitable unit time threshold.

[0106] In addition, based on the actual respiratory cycle of the animal species and combined with the unit time threshold, the preset unit time corresponding to the current target animal is adjusted and determined, so as to combine the respiratory changes of the target animal under different environmental conditions and make the preset unit time more accurately reflect the actual respiratory state of the target animal.

[0107] The actual respiratory cycle can be obtained from real-time monitoring data. That is, a suitable preset unit time can be determined within a unit time threshold based on the actual respiratory cycle of the target animal and environmental changes. Furthermore, the preset unit time can be updated in real time according to changes in the environment of the target animal, thereby improving the adaptability and accuracy of the preset unit time. For example, if the respiratory rate of the target animal is detected to increase due to changes in environmental temperature, the unit time threshold can be adaptively shortened within a unit time threshold.

[0108] Furthermore, machine learning algorithms can be used to periodically update the preset relationship mapping table. This involves analyzing a large amount of monitoring data to identify and update the mapping relationship between animal species and unit time thresholds, thereby maintaining the accuracy and timeliness of the preset relationship mapping table.

[0109] This embodiment achieves the detection of individual respiratory patterns of target animals by obtaining the unit time threshold corresponding to animal species from a preset relationship mapping table and determining the preset unit time based on the actual respiratory cycle, thereby improving accuracy and enhancing adaptability and reliability.

[0110] In some embodiments, determining the group coupling index of group breathing patterns includes:

[0111] Traverse the individual respiratory waveforms corresponding to each individual's respiratory pattern in the group's respiratory patterns, and determine the morphological similarity between any two individual respiratory waveforms within the same time window;

[0112] The average morphological similarity of the breathing waveforms of all individuals included in the group breathing pattern is determined as the group coupling index.

[0113] Optionally, in this embodiment, the individual breathing waveforms corresponding to each individual breathing pattern in the group breathing pattern are traversed, and any two individual breathing waveforms are determined as a pair of individual breathing waveforms. For each pair of individual breathing waveforms, their morphological similarity is evaluated within the same time window.

[0114] Among them, the morphological similarity calculation can be based on the shape and frequency characteristics of an individual's respiratory waveform, and the morphological similarity can be quantified by comparing the degree of matching between two waveforms within a specific time window.

[0115] In addition, the average morphological similarity of all individual respiratory waveforms is determined, which is the group coupling index. The group coupling index can provide a quantitative indicator for monitoring the health status of a group, making it easier to detect abnormal breathing patterns in the group in a timely manner.

[0116] Furthermore, the size of the time window can be dynamically adjusted according to the characteristics of the individual respiratory waveform through an adaptive time window method to adapt to the changes in the individual respiratory patterns of target animals of different species, thereby improving the calculation accuracy of the population coupling index. For example, for target animals with high respiratory rates, a shorter time window can be adaptively selected to capture the details of the respiratory waveform more accurately.

[0117] Furthermore, in addition to morphological similarity, similarity can be assessed through other dimensions, such as frequency similarity and amplitude similarity. By integrating similarity across multiple dimensions, the consistency between individual respiratory waveforms can be assessed more comprehensively, thereby improving the reliability and accuracy of the population coupling index. For example, frequency similarity can be determined by calculating the frequency spectrum similarity of two respiratory waveforms using Fourier transform, while amplitude similarity can be determined by comparing the amplitudes of the respiratory waveforms.

[0118] This embodiment realizes the consistency of the group's breathing patterns by traversing the individual breathing waveforms corresponding to each individual breathing pattern in the group breathing pattern and determining the morphological similarity between any two individual breathing waveforms within the same time window. This provides a quantitative indicator for monitoring the health status of the group and facilitates the timely detection of abnormal breathing patterns in the group.

[0119] In some embodiments, after determining that an infected animal exists in the target field, the method further includes:

[0120] In group breathing patterns, identify the zero-case animal with abnormal individual respiratory waveforms and determine the relative positions among target animals based on imaging data;

[0121] Based on the zero pathogen and its relative location, the respiratory similarity of the respiratory patterns of the target animals centered on the zero pathogen and adjacent to the zero pathogen is determined, and the decay gradient of the respiratory similarity is determined.

[0122] When the attenuation gradient spreads radially, it is determined that there is a group disease transmission in the current field.

[0123] Optionally, if this embodiment determines that there are infected animals in the target field, the group's respiratory pattern can be further analyzed to identify and locate the zero pathogen. The zero pathogen refers to the individual in the group that first shows abnormal respiratory waveforms, and whose individual respiratory pattern differs from that of healthy animals.

[0124] By analyzing image data, the relative positions of each target animal in the target field can be determined. The image data provides spatial distribution information of each target animal in the target field, thereby enabling accurate identification of the location of the zero pathogen and its relative positional relationship with the surrounding target animals.

[0125] In addition, based on the zero pathogen and relative location information, the respiratory similarity of the respiratory patterns of the target animals adjacent to the zero pathogen can be determined. The morphological similarity of the respiratory waveforms of adjacent target animals can be analyzed, and their respiratory similarity can be quantified by comparing the shape and frequency characteristics of the respiratory waveforms.

[0126] In addition, determining the decay gradient of respiratory similarity, that is, the trend of respiratory similarity changes from the zero pathogen outwards, if the decay gradient spreads radially, it indicates that the spread of the disease may start from the zero pathogen and spread to the surrounding animals. Radial decay gradient is a typical feature of the spread of group diseases and can be used to indicate that there may be the spread of group diseases in the current target field.

[0127] If the respiratory similarity is consistent and no decay gradient is observed, it indicates that the patient zero is infected and the disease is not a group disease.

[0128] Furthermore, disease transmission models can be used to predict the transmission path and scope of a disease based on the decay gradient of respiratory similarity. For example, the SIR (susceptible-infected-recovered) model can simulate the disease transmission process and predict the target animals that may be affected in the future, thus providing a scientific basis for prevention and control measures.

[0129] This embodiment enables the analysis of group respiratory patterns after identifying the presence of infected animals in the target field, determining the respiratory similarity and attenuation gradient of target animals adjacent to the zero-case animal, thereby judging whether there is a spread of group diseases, improving positioning accuracy and response speed, enhancing predictive ability, and facilitating disease prevention and control.

[0130] To effectively address the shortcomings of traditional technologies in real-time and continuous detection of animal respiratory rate, which leads to delays in detecting diseased animals, and to significantly improve the efficiency and accuracy of real-time and continuous detection of animal respiratory rate, this application provides an embodiment of an animal respiratory pattern detection device for realizing all or part of the aforementioned animal respiratory pattern detection. See [link to embodiment]. Figure 2 The animal respiratory pattern detection device specifically includes the following components:

[0131] The first processing module 10 is used to acquire image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired based on the visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired based on the infrared thermal imaging sensor.

[0132] The second processing module 20 is used to determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes the individual respiratory frequency, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0133] The third processing module 30 is used to determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing pattern, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, it is determined that there are diseased animals in the target field.

[0134] As described above, the animal respiratory pattern detection device provided in this application can innovatively acquire image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired by a visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. The animal species of the target animal is determined based on the image data, thereby determining the preset unit time corresponding to the animal species and the number of peaks in the fused respiratory waveform signal per unit time. The individual respiratory patterns of target animals are determined based on the number of peaks. These individual respiratory patterns include individual respiratory rate, individual respiratory depth waveform, and individual respiratory rhythm disorder. All individual respiratory models are then classified as group respiratory patterns, thereby determining the group coupling index and the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, the presence of diseased animals in the target area is confirmed. This method can determine the individual respiratory patterns of target animals using image data, avoiding errors associated with contact-based respiratory rate detection and improving the accuracy of individual respiratory patterns. Furthermore, it can analyze both individual and group respiratory patterns to determine the disease status of each animal in the target area. This method effectively addresses the shortcomings of real-time and continuous detection of animal respiratory rates, which can lead to delayed detection of diseased animals, significantly improving the efficiency and accuracy of real-time and continuous animal respiratory rate detection.

[0135] From a hardware perspective, in order to effectively address the shortcomings of traditional technologies in real-time and continuous detection of animal respiratory rates, which leads to the failure to detect diseased animals in a timely manner, and to significantly improve the efficiency and accuracy of real-time and continuous detection of animal respiratory rates, this application provides an embodiment of an electronic device for implementing all or part of the aforementioned animal respiratory pattern detection method. The electronic device specifically includes the following components:

[0136] The system comprises a processor, memory, a communications interface, and a bus; wherein the processor, memory, and communications interface communicate with each other via the bus; the communications interface is used to realize information transmission between the animal respiratory pattern detection device and core business systems, user terminals, and related databases and other related devices; the logic controller can be a desktop computer, tablet computer, or mobile terminal, etc., and this embodiment is not limited to these. In this embodiment, the logic controller can be implemented with reference to the embodiments of the animal respiratory pattern detection method and the animal respiratory pattern detection device in the embodiments, the contents of which are incorporated herein, and repeated details will not be described again.

[0137] It is understood that the user terminal may include smartphones, tablet computers, network set-top boxes, portable computers, desktop computers, personal digital assistants (PDAs), in-vehicle devices, smart wearable devices, etc. Among these, the smart wearable devices may include smart glasses, smartwatches, smart bracelets, etc.

[0138] In practical applications, the animal breathing pattern detection method can be partially executed on the electronic device side as described above, or all operations can be completed in the client device. The choice can be made based on the processing power of the client device and the limitations of the user's usage scenario. This application does not impose any limitations on this. If all operations are completed in the client device, the client device may further include a processor.

[0139] The aforementioned client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission. The server may include a server on the task scheduling center side; in other implementation scenarios, it may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a distributed server structure.

[0140] Figure 3 This is a schematic block diagram illustrating the system configuration of the electronic device 9600 according to an embodiment of this application. Figure 3 As shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It is worth noting that... Figure 3 This is an example; other types of structures can also be used to supplement or replace this structure to achieve telecommunications functions or other functions.

[0141] In one embodiment, the animal respiratory pattern detection method function can be integrated into the central processing unit 9100. The central processing unit 9100 can be configured to perform the following control:

[0142] Step S101: Obtain image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal obtained based on the visible light sensor and the body surface thermal radiation image corresponding to the target animal obtained based on the infrared thermal imaging sensor.

[0143] Step S102: Determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes the individual respiratory rate, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0144] Step S103: Determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing patterns, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, determine that there are infected animals in the target field.

[0145] As described above, the electronic device provided in this application innovatively acquires image data corresponding to each target animal in the target field, processes the image data through a preset dual-stream neural network model, obtains the contour displacement features and thermal radiation change features of the target animal, and generates a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes a contour displacement image corresponding to the target animal acquired by a visible light sensor and a body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. The animal species of the target animal is determined based on the image data, thereby determining a preset unit time corresponding to the animal species, and determining the number of peaks in the fused respiratory waveform signal per unit time. This method determines the individual respiratory patterns of target animals, including individual respiratory rate, respiratory depth waveform, and respiratory rhythm disorder. All individual respiratory patterns are then classified as group respiratory patterns, and a group coupling index is determined. The coupling similarity between the group coupling index and the normal coupling index is also determined. If the coupling similarity does not exceed a preset similarity threshold, the presence of diseased animals in the target area is confirmed. This method can determine the individual respiratory patterns of target animals using image data, avoiding errors associated with contact-based respiratory rate detection and improving the accuracy of individual respiratory patterns. Furthermore, it can analyze both individual and group respiratory patterns to determine the disease status of each animal in the target area. This method effectively addresses the shortcomings of real-time and continuous detection of animal respiratory rates, which can lead to delayed detection of diseased animals, significantly improving the efficiency and accuracy of real-time and continuous animal respiratory rate detection.

[0146] In another embodiment, the animal respiratory pattern detection device can be configured separately from the central processing unit 9100. For example, the animal respiratory pattern detection device can be configured as a chip connected to the central processing unit 9100, and the animal respiratory pattern detection method function can be implemented through the control of the central processing unit.

[0147] like Figure 3 As shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily need to include these components. Figure 3 All components shown; in addition, the electronic device 9600 may also include Figure 3 For components not shown, please refer to existing technology.

[0148] like Figure 3 As shown, the central processing unit 9100, sometimes also referred to as a controller or operating control, may include a microprocessor or other processor device and / or logic device, which receives inputs and controls the operation of various components of the electronic device 9600.

[0149] The memory 9140 may be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store the aforementioned failure-related information, and also store a program for executing that information. The central processing unit 9100 may execute the program stored in the memory 9140 to perform information storage or processing, etc.

[0150] Input unit 9120 provides input to central processing unit 9100. Input unit 9120 may be, for example, a keypad or touch input device. Power supply 9170 provides power to electronic device 9600. Display 9160 displays images and text. Display may be, for example, an LCD display, but is not limited thereto.

[0151] The memory 9140 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs. The memory 9140 can also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142 for storing application programs and function programs or processes for executing the operation of the electronic device 9600 via the central processing unit 9100.

[0152] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various drivers for the electronic device's communication functions and / or for performing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0153] The communication module 9110 is a transmitter / receiver that sends and receives signals via the antenna 9111. The communication module 9110 (transmitter / receiver) is coupled to the central processing unit 9100 to provide input signals and receive output signals, which is the same as in a conventional mobile communication terminal.

[0154] Based on different communication technologies, multiple communication modules 9110 can be configured in the same electronic device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module 9110 (transmitter / receiver) is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the microphone 9132, thereby realizing typical telecommunications functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. Additionally, the audio processor 9130 is coupled to a central processing unit 9100, enabling on-device recording via the microphone 9132 and on-device playback of stored audio via the speaker 9131.

[0155] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the animal respiratory pattern detection method with a server or client execution subject in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the animal respiratory pattern detection method with a server or client execution subject in the above embodiments. For example, when the processor executes the computer program, it implements the following steps:

[0156] Step S101: Obtain image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal obtained based on the visible light sensor and the body surface thermal radiation image corresponding to the target animal obtained based on the infrared thermal imaging sensor.

[0157] Step S102: Determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes the individual respiratory rate, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0158] Step S103: Determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing patterns, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, determine that there are infected animals in the target field.

[0159] As described above, the computer-readable storage medium provided in this application innovatively acquires image data corresponding to each target animal in the target field, processes the image data through a preset dual-stream neural network model, obtains the contour displacement features and thermal radiation change features of the target animal, and generates a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes a contour displacement image corresponding to the target animal acquired by a visible light sensor and a body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. The animal species of the target animal is determined based on the image data, thereby determining a preset unit time corresponding to the animal species, and determining the number of peaks in the fused respiratory waveform signal per unit time. The number of peaks determines the individual respiratory pattern of the target animal. This individual respiratory pattern includes individual respiratory rate, individual respiratory depth waveform, and individual respiratory rhythm disorder. All individual respiratory models are defined as a group respiratory pattern, thereby determining the group coupling index and the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, the presence of a diseased animal in the target area is confirmed. This method can determine the individual respiratory pattern of the target animal using its image data, avoiding errors associated with contact-based respiratory rate detection and improving the accuracy of individual respiratory patterns. Furthermore, it can analyze both individual and group respiratory patterns to determine the disease status of each animal in the target area. This method effectively addresses the shortcomings of real-time and continuous detection of animal respiratory rates, which can lead to delayed detection of diseased animals, significantly improving the efficiency and accuracy of real-time and continuous animal respiratory rate detection.

[0160] Embodiments of this application also provide a computer program product capable of implementing all steps of the animal respiratory pattern detection method described above, where the execution subject is a server or client. When executed by a processor, this computer program / instruction implements the steps of the animal respiratory pattern detection method. For example, the computer program / instruction implements the following steps:

[0161] Step S101: Obtain image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal obtained based on the visible light sensor and the body surface thermal radiation image corresponding to the target animal obtained based on the infrared thermal imaging sensor.

[0162] Step S102: Determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes the individual respiratory rate, the individual respiratory depth waveform, and the degree of individual respiratory rhythm disorder.

[0163] Step S103: Determine all individual breathing patterns as group breathing patterns, determine the group coupling index of the group breathing patterns, and determine the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed the preset similarity threshold, determine that there are infected animals in the target field.

[0164] As described above, the computer program product provided in this application innovatively acquires image data corresponding to each target animal in the target field, processes the image data through a preset dual-stream neural network model, obtains the contour displacement features and thermal radiation change features of the target animal, and generates a fused respiratory waveform signal based on the contour displacement features and thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired by a visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. The animal species of the target animal is determined based on the image data, thereby determining the preset unit time corresponding to the animal species, and determining the number of peaks in the fused respiratory waveform signal per unit time. The number of peaks determines the individual respiratory patterns of the target animal. These individual respiratory patterns include individual respiratory rate, individual respiratory depth waveform, and individual respiratory rhythm disorder. All individual respiratory models are then classified as a group respiratory pattern, thereby determining the group coupling index and the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, the presence of a diseased animal in the target area is confirmed. This method can determine the individual respiratory patterns of the target animal using its image data, avoiding errors associated with contact-based respiratory rate detection and improving the accuracy of individual respiratory patterns. Furthermore, it can analyze both individual and group respiratory patterns to determine the disease status of each animal in the target area. This method effectively addresses the shortcomings of real-time and continuous detection of animal respiratory rates, which can lead to delayed detection of diseased animals, significantly improving the efficiency and accuracy of real-time and continuous animal respiratory rate detection.

[0165] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0166] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0167] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0169] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for detecting animal respiratory patterns, characterized in that, The method includes: Image data corresponding to each target animal in the target field is acquired. The image data is processed by a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal. A fused respiratory waveform signal is generated based on the contour displacement features and the thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired by a visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired by an infrared thermal imaging sensor. Based on the image data, the animal species of the target animal is determined, and the preset unit time corresponding to the animal species is determined. The number of peaks of the fused respiratory waveform signal within the preset unit time is determined. Based on the number of peaks, the individual respiratory pattern of the target animal is determined. The individual respiratory pattern includes individual respiratory frequency, individual respiratory depth waveform, and individual respiratory rhythm disorder. All individual breathing patterns are identified as a group breathing pattern. The group coupling index of the group breathing pattern is determined, including: traversing the individual breathing waveforms corresponding to each individual breathing pattern in the group breathing pattern, determining the morphological similarity of any two individual breathing waveforms within the same time window; determining the average morphological similarity of all individual breathing waveforms included in the group breathing pattern as the group coupling index; and determining the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, it is determined that there is a diseased animal in the target field.

2. The method according to claim 1, characterized in that, The preset dual-stream neural network includes a visible light sub-network and an infrared sub-network, and the visible light sub-network and the infrared sub-network are parallel. The process of processing the image data using a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal includes: Background subtraction processing is performed on the continuous contour displacement images to obtain the initial animal contour region corresponding to the image data, and edge detection is performed on the initial animal contour region using an edge detection algorithm to obtain the contour edge information of the target animal. The contour edge information is extracted by the convolutional layer in the visible photonic network to obtain the contour displacement features of the target animal.

3. The method according to claim 1, characterized in that, The preset dual-stream neural network includes a visible light sub-network and an infrared sub-network, and the visible light sub-network and the infrared sub-network are parallel. The process of processing the image data using a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal includes: The thermal radiation region of the target animal's body surface is extracted from the thermal radiation image by a preset segmentation algorithm to obtain the initial thermal radiation region. The initial thermal radiation region is then analyzed based on the time series of the image data to obtain the thermal radiation changes of the target animal's body surface. The thermal radiation change features of the target animal are obtained by extracting the features of the thermal radiation change through the convolutional layer in the infrared subnetwork.

4. The method according to claim 1, characterized in that, The generation of the fused breathing waveform signal based on the contour displacement features and the thermal radiation change features includes: Obtain the time series corresponding to the image data, and align the contour displacement feature and the thermal radiation change feature based on the time series; The contour displacement features and thermal radiation change features after fusion and alignment are used to obtain a comprehensive feature, and the comprehensive feature is then converted into a fused breathing waveform signal.

5. The method according to claim 1, characterized in that, The step of determining the preset unit time corresponding to the animal species includes: The corresponding unit time threshold is determined from the animal species from a preset relationship mapping table, wherein the preset relationship mapping table is used to represent the mapping relationship between the animal species and the unit time threshold; The preset unit time is determined based on the actual respiratory cycle corresponding to the animal species and the unit time threshold.

6. The method according to claim 1, characterized in that, After determining that an infected animal exists in the target field, the method further includes: In the group breathing pattern, identify the zero-case animal whose individual breathing waveform is abnormal, and determine the relative position between the target animals based on the image data; Based on the zero pathogen and the relative position, determine the respiratory similarity of the respiratory patterns of the target animals corresponding to the zero pathogen as the center and adjacent to the zero pathogen, and determine the decay gradient of the respiratory similarity; When the attenuation gradient spreads radially, it is determined that there is a spread of a group disease in the current column.

7. An animal respiratory pattern detection device, characterized in that, The device includes: The first processing module is used to acquire image data corresponding to each target animal in the target field, process the image data through a preset dual-stream neural network model to obtain the contour displacement features and thermal radiation change features of the target animal, and generate a fused respiratory waveform signal based on the contour displacement features and the thermal radiation change features. The image data includes the contour displacement image corresponding to the target animal acquired based on a visible light sensor and the body surface thermal radiation image corresponding to the target animal acquired based on an infrared thermal imaging sensor. The second processing module is used to determine the animal species of the target animal based on the image data, determine the preset unit time corresponding to the animal species, determine the number of peaks of the fused respiratory waveform signal within the preset unit time, and determine the individual respiratory pattern of the target animal based on the number of peaks. The individual respiratory pattern includes individual respiratory frequency, individual respiratory depth waveform, and individual respiratory rhythm disorder. The third processing module is used to determine all the individual breathing patterns as a group breathing pattern and to determine the group coupling index of the group breathing pattern, including: traversing the individual breathing waveforms corresponding to each individual breathing pattern in the group breathing pattern, determining the morphological similarity of any two individual breathing waveforms within the same time window; determining the average value of the morphological similarity corresponding to all the individual breathing waveforms included in the group breathing pattern as the group coupling index; and determining the coupling similarity between the group coupling index and the normal coupling index. If the coupling similarity does not exceed a preset similarity threshold, it is determined that there is a diseased animal in the target field.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the animal respiratory pattern detection method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the animal respiratory pattern detection method according to any one of claims 1 to 6.

Images