Fish respiration monitoring method and system

Abstract

The embodiment of the invention provides a fish respiration monitoring method and system, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring fish body video data; intercepting an image of the fish body video data; performing real-time reasoning on each frame of video by adopting a hybrid cascade network model to obtain a motion bounding box of each frame of fish mouth and gill motion; performing spatio-temporal feature extraction operation on the motion bounding box to obtain time sequence motion signals of the fish mouth and the fish gills; determining a fish body breathing frequency value according to the time sequence motion signal; and determining the fish body pressure grade according to the fish body breathing frequency value. Compared with the prior art, the method and the system provided by the embodiment of the invention realize accurate monitoring of fish respiration under the influence of water waves through combination of the hybrid cascade network model and the time sequence motion signal.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a fish respiration monitoring method and system. Background Art

[0002] In current aquaculture, the assessment of fish health status usually relies on manual observation or traditional biosensors. These methods are often inefficient, inaccurate, and easily affected by environmental factors. With the development of computer vision technology, non-contact monitoring using video images has become possible. As an effective motion estimation method, optical flow has been used in various biological behavior analysis scenarios. However, traditional optical flow (such as Farneback) relies on image gradient changes and is prone to losing motion information in low-texture areas or when the fish swims rapidly. Summary of the Invention

[0003] The purpose of the embodiments of the present invention is to provide a method and system for monitoring fish respiration, which improves the monitoring accuracy of fish respiration.

[0004] To achieve the above objectives, an embodiment of the present invention provides a fish respiration monitoring method, comprising:

[0005] Obtain fish video data;

[0006] intercepting an image of the fish body video data;

[0007] A hybrid cascade network model is used to perform real-time inference on each frame of video to obtain the motion bounding box of the fish's mouth and gills in each frame;

[0008] Performing a spatiotemporal feature extraction operation on the motion bounding box to obtain a time series motion signal of the fish mouth and gills;

[0009] Determine the respiratory frequency value of the fish body according to the time series motion signal;

[0010] The fish body stress level is determined according to the fish body breathing frequency value.

[0011] Optionally, obtaining fish video data includes:

[0012] Use Gaussian filtering to perform low-frequency filtering on the acquired original video stream;

[0013] Use median filtering to process the original video stream after low-frequency filtering;

[0014] The multi-spectral channel equalization method is used to adjust the processed original video stream;

[0015] The resized original video stream is enhanced using contrast-limited adaptive histogram equalization.

[0016] Optionally, it also includes:

[0017] Split the fish body video frame by frame to obtain the fish body image;

[0018] Marking a fish head region in the fish body image;

[0019] Performing at least one of random rotation, dynamic scaling, atomization simulation, and motion blur on the annotated fish image;

[0020] Combining the processed fish body images into a training data set;

[0021] The ConvNeXt network structure is used as the backbone network to build the initial model;

[0022] Determine the loss function of the initial model according to formula (1) to formula (3):

[0023] FL(p)=-α(1-p) γ log(p), (1)

[0024] BCE(p)=-log(p), (2)

[0025] Loss(p)=FL(p)+BCE(p), (3)

[0026] Wherein, α is the balancing factor, p is the predicted probability of the initial model, γ is the focus factor, and Loss(p) is the loss function;

[0027] The data set is used to train an initial model to obtain the hybrid cascade network model.

[0028] Optionally, performing a spatiotemporal feature extraction operation on the motion bounding box to obtain a time series motion signal of the fish mouth and gills includes:

[0029] A multi-scale feature pyramid is used to extract spatiotemporal motion features between the motion bounding boxes of every two frames;

[0030] Calculating optical flow field components based on the spatiotemporal motion characteristics using an iterative update mechanism;

[0031] Calculating a motion intensity representation based on the optical flow field components;

[0032] generating a time series signal based on the motion intensity representation at a preset sampling rate;

[0033] A time alignment mechanism is used to compensate for the phase deviation of the time series signal using a model;

[0034] The time series signals are synchronized according to the result of the phase deviation inference to obtain the time series motion signal.

[0035] Optionally, an iterative update mechanism is used to calculate the optical flow field components according to the spatiotemporal motion characteristics, including:

[0036] A median filter is used to eliminate abnormal optical flow components in the optical flow field components.

[0037] Optionally, calculating the motion intensity representation based on the optical flow field component includes:

[0038] The exercise intensity representation is calculated using formula (4):

[0039]

[0040] in, is the motion intensity representation of the i-th sampling point, u i 、v i are different sub-components in the optical flow field component of the i-th sampling point, and λ is the statistical quantile.

[0041] Optionally, determining the fish respiratory frequency value according to the time series motion signal includes:

[0042] Preprocessing the time series motion signal in sequence using a fourth-order Butterworth low-pass filter and a Symlet wavelet function;

[0043] Determine respiratory cycle feature points based on the preprocessed time series motion signal using a sliding window dynamic peak detection algorithm;

[0044] Determining a weighted average of adjacent peak intervals based on the respiratory cycle characteristic points;

[0045] The fish respiratory frequency value is calculated according to the weighted average value.

[0046] Optionally, determining the fish body stress level according to the fish body breathing frequency value includes:

[0047] Determining the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio based on the fish respiratory rate value;

[0048] Constructing a characteristic matrix based on the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio;

[0049] The feature matrix is ​​weighted using three-level pressure labels, and the degree of incoordination between the fish mouth and gill movements is determined by combining the SHAP value interpretability analysis method;

[0050] The fish body pressure level is determined according to the incoordination degree.

[0051] On the other hand, the present invention further provides a fish respiration monitoring system, the system comprising a processor, and the processor is configured to execute any of the above methods.

[0052] Through the above technical solution, embodiments of the present invention provide a method and system for monitoring fish respiration. This method and system uses a hybrid cascade network model to extract a motion bounding box representing the movement of the fish's mouth and gills. Based on this motion bounding box, the system then determines a time series motion signal of the fish's mouth and gills. Finally, based on this time series motion signal, the system determines the fish's respiratory rate. Compared to the prior art, the method and system provided by embodiments of the present invention, through the combination of a hybrid cascade network model and time series motion signals, achieves accurate monitoring of fish respiration under the influence of water ripples.

[0053] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation of the embodiments of the present invention. In the accompanying drawings:

[0055] Figure 1 is a flow chart of a fish respiration monitoring method according to one embodiment of the present invention;

[0056] Figure 2 is an example diagram of a hybrid cascade network model according to one embodiment of the present invention;

[0057] Figure 3 is a flowchart of a method for training and acquiring a hybrid machine-online network model according to one embodiment of the present invention;

[0058] Figure 4 is a flowchart of a method for acquiring a time series motion signal according to one embodiment of the present invention;

[0059] Figure 5 is a flow chart of a method for determining a respiratory frequency value of a fish according to one embodiment of the present invention;

[0060] Figure 6 is a flow chart of a method for determining a fish body stress level according to one embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following describes the specific implementation of the embodiment of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiment of the present invention and is not used to limit the embodiment of the present invention.

[0062] In the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0063] like Figure 1 The figure shows a flow chart of a fish respiration monitoring method according to one embodiment of the present invention. Figure 1 In the method, the method may include the following steps:

[0064] In step S10, fish video data is obtained;

[0065] In step S11, an image of the fish body video data is captured;

[0066] In step S12, a hybrid cascade network model is used to perform real-time inference on each frame of the video to obtain a motion bounding box of the fish mouth and gills in each frame;

[0067] In step S13, a spatiotemporal feature extraction operation is performed on the motion bounding box to obtain a time series motion signal of the fish mouth and gills;

[0068] In step S14, the respiratory frequency value of the fish is determined according to the time series motion signal;

[0069] In step S15, the fish body stress level is determined according to the fish body breathing frequency value.

[0070] In this Figure 1In the method shown, step S10 can be used to obtain fish video data. The specific method for obtaining the fish video data can be various forms known to those skilled in the art, such as setting an image acquisition device underwater or in the surrounding area. Further, in order to improve the acquisition quality of the fish video data, step S10 can be to use an underwater camera with a high frame rate (greater than or equal to 120Hz) and a wide-angle lens and an LED cold light source fill light system to collect the fish video data in real time. Furthermore, considering the influence of water ripples, bubbles, etc. of the water body itself, step S10 can also be to first use Gaussian filtering to perform low-frequency filtering on the acquired original video stream, thereby eliminating the low-frequency noise generated by water ripples in the fish video data. In one example of the present invention, the Gaussian filtering operation can be performed using a convolution kernel of 5*5 and a σ value of 1.5. Then, for the original video stream after low-frequency filtering, step S10 can also be processed by a median filtering method to remove the high-frequency salt and pepper noise caused by bubbles and suspended particles in the water body. In one example of the present invention, the sliding window size of the median filter can be set to 7*7. Finally, to address the color flake problem caused by underwater light absorption, step S10 can also use a multispectral channel equalization method to adjust the processed original video stream, and use a contrast-limited adaptive histogram equalization method to enhance the adjusted original video stream, thereby obtaining the fish video data.

[0071] Step S11 can be used to capture an image of the fish video data. Specifically, in one example of the present invention, step S11 can be, for example, by presetting a time period and capturing the fish video data at at least one time point in each time period, thereby obtaining the captured image.

[0072] Step S12 can be used to perform real-time reasoning on each frame of video using a hybrid cascade network model to obtain a motion bounding box for the fish mouth and gills in each frame. The specific structure of the hybrid cascade network model can be various forms known to those skilled in the art. In one example of the present invention, the hybrid cascade network model can include: Figure 2 The structure shown in Figure 2In the example, the hybrid cascade network model may include a first convolutional layer (Conv0), a second convolutional layer (Conv_stage1), a third convolutional layer (Conv_stage2), a fourth convolutional layer (Conv_stage3), a fifth convolutional layer (Conv_stage3), a PPM unit, an FPN unit, a feature fusion unit (FUSE), and a feature refinement unit (Refined Feature) connected in sequence. In this example, the specific parameters of the first convolutional layer (Conv0), the second convolutional layer (Conv_stage1), the third convolutional layer (Conv_stage2), the fourth convolutional layer (Conv_stage3), the fifth convolutional layer (Conv_stage3), the PPM unit, the FPN unit, the feature fusion unit (FUSE), and the feature refinement unit (Refined Feature) may be, for example, as shown in the following Table 1:

[0073] Table 1

[0074]

[0075] In this Figure 2 In the network, the first convolutional layer (Conv0), the second convolutional layer (Conv_stage1), the third convolutional layer (Conv_stage2), the fourth convolutional layer (Conv_stage3), and the fifth convolutional layer (Conv_stage3) are connected in sequence, the PPM unit is connected to the output of the fifth convolutional layer (Conv_stage3), and is connected to the starting input of the FPN unit. The FPN unit includes a first upsampling layer C1, a second upsampling layer C2, a third upsampling layer C3, and a fourth upsampling layer C4 connected in sequence. Among them, the first upsampling layer C1 is used to receive the output of the PPM unit, the second upsampling unit C2 is used to receive the output of the fourth convolutional layer and the first upsampling unit C1, the third upsampling unit C3 is used to receive the output of the third convolutional layer and the second upsampling unit C2, and the fourth upsampling unit C4 is used to receive the output of the second convolutional layer and the third upsampling unit C3. The feature fusion unit is used to perform feature fusion on the outputs of the first upsampling layer C1, the second upsampling layer C2, the third upsampling layer C3 and the fourth upsampling layer C4. The feature refining unit is used to refine the feature map output by the feature fusion unit to obtain a feature map with a motion bounding box.

[0076] The specific method for training the hybrid cascade network model can be various forms known to those skilled in the art. In one example of the present invention, considering the characteristics of the image of the fish video data, the method for training the hybrid cascade network model can include the following: Figure 3 The steps shown in Figure 3In the method of training and obtaining the hybrid machine-connected network model, the method may include the following steps:

[0077] In step S20, the fish body video is split frame by frame to obtain a fish body image;

[0078] In step S21, the fish head area in the fish body image is marked;

[0079] In step S22, the annotated fish image is subjected to at least one of random rotation, dynamic scaling, fog simulation, and motion blur. In one example of the present invention, step S22 may be performed by performing at least one of random rotation, dynamic scaling, fog simulation, and motion blur, respectively, thereby increasing the data volume of the processed fish image by 10 times.

[0080] In step S23, the processed fish images are combined into a training data set;

[0081] In step S24, the ConvNeXt network structure is used as the backbone network to construct the initial model. The schematic diagram of the constructed initial model is as follows Figure 2 As shown;

[0082] In step S25, the loss function of the initial model is determined according to formulas (1) to (3):

[0083] FL(p)=-α(1-p) γ log(p), (1)

[0084] VCE(p)=-log(p), (2)

[0085] Loss(p)=FL(p)+BCE(p), (3)

[0086] Where α is the balancing factor, p is the predicted probability of the initial model, γ is the focus factor, and Loss(p) is the loss function. In one example of the present invention, to adjust the influence between positive and negative samples, the balancing factor α can be set to 0.8, and to adjust the weight between difficult and easy samples, the initial value of the focus factor γ can be set to 2.

[0087] In step S26, the data set is used to train the initial model to obtain a hybrid cascade network model.

[0088] Step S13 can be used to perform spatiotemporal feature extraction on the motion bounding box to obtain a time series motion signal of the fish mouth and gills. Specifically, in one example of the present invention, the method for obtaining the time series motion signal can include: Figure 4 The steps shown in Figure 4The method for obtaining the time series motion signal may include the following steps:

[0089] In step S30 , a multi-scale feature pyramid is used to extract spatiotemporal motion features between motion bounding boxes of every two frames.

[0090] In step S31, an iterative update mechanism is used to calculate the optical flow field components based on the spatiotemporal motion features. Specifically, in this example, step S31 may be to use a median filter to eliminate abnormal optical flow components in the optical flow field components.

[0091] In step S32, the motion intensity representation is calculated based on the optical flow field component. Specifically, step S32 may be to calculate the motion intensity representation using formula (4):

[0092]

[0093] in, is the motion intensity representation of the i-th sampling point, u i 、v i are different sub-components in the optical flow field component of the i-th sampling point, and λ is the statistical quantile.

[0094] In step S33, a time series signal is generated based on the motion intensity representation at a preset sampling rate. In one example of the present invention, the sampling rate may be set to, for example, 120 Hz.

[0095] In step S34, a time alignment mechanism is used to compensate for the phase deviation of the time series signal using a model (≤30ms).

[0096] In step S35 , the time series signal is synchronized according to the result of the phase deviation inference to obtain a time series motion signal.

[0097] Step S14 can be used to determine the fish's respiratory frequency value based on the time series motion signal. Specifically, in one example of the present invention, step S14 may include: Figure 5 The method shown in Figure 5 The method for determining the respiratory frequency value of the fish may include the following steps:

[0098] In step S40, the time series motion signal is preprocessed using a fourth-order Butterworth low-pass filter and a Symlet wavelet function (Sym5). Specifically, in this example, the cutoff frequency of the fourth-order Butterworth low-pass filter can be 2 Hz. The Symlet wavelet function can be decomposed into a five-layer soft threshold to eliminate baseline drift caused by water turbulence.

[0099] In step S41, a sliding window dynamic peak detection algorithm is used to determine respiratory cycle feature points based on the preprocessed time series motion signal. The window length of the sliding window dynamic peak detection algorithm can be set to 10 seconds and the step length can be 1 second.

[0100] In step S42, a weighted average of the time intervals between adjacent peaks is determined based on the respiratory cycle characteristic points. The specific calculation method for the weighted average can be various methods known to those skilled in the art. In one example of the present invention, when calculating the weighted average, the weight of the operculum signal can be set to 0.6, and the weight of the fish mouth signal can be set to 0.4.

[0101] In step S43, the fish's respiratory frequency value is calculated based on the weighted average value.

[0102] Step S15 can be used to determine the fish body stress level based on the fish body breathing frequency value. Among them, the specific method for determining the fish body stress level can be various forms known to those skilled in the art, such as presetting a threshold to grade the fish body breathing frequency. In one example of the present invention, considering that the present invention uses fish body video data, the fish body breathing frequency obtained based on the fish body video data is a continuous data. Therefore, in one example of the present invention, it can be adopted as follows Figure 6 The method shown is used to determine the fish body pressure level. Figure 6 In the method for determining the stress level of the fish body, the method may include the following steps:

[0103] In step S50, the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio are determined based on the fish's respiratory rate value. The energy entropy ratio can be calculated based on the 0.1-0.5 Hz (low-frequency stress response) and 0.5-2 Hz (high-frequency respiratory rhythm) sub-bands.

[0104] In step S51, a feature matrix is ​​constructed based on the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio;

[0105] In step S52, the three-level pressure label weighted feature matrix is ​​used in combination with the SHAP value interpretability analysis method to determine the degree of incoordination between the fish mouth and gill movements. Specifically, in one example of the present invention, step S52 can be first inputting the weighted feature matrix into a trained XGBoost classifier, and then determining the degree of incoordination based on the SHAP value output by the XGBoost classifier.

[0106] In step S53, the fish body pressure level is determined according to the degree of incoordination.

[0107] On the other hand, the present invention further provides a fish respiration monitoring system, the system comprising a processor, and the processor is configured to execute any of the above methods.

[0108] Through the above technical solution, embodiments of the present invention provide a method and system for monitoring fish respiration. This method and system uses a hybrid cascade network model to extract a motion bounding box representing the movement of the fish's mouth and gills. Based on this motion bounding box, the system then determines a time series motion signal of the fish's mouth and gills. Finally, based on this time series motion signal, the system determines the fish's respiratory rate. Compared to the prior art, the method and system provided by embodiments of the present invention, through the combination of a hybrid cascade network model and time series motion signals, achieves accurate monitoring of fish respiration under the influence of water ripples.

[0109] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0110] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0111] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0113] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0114] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0115] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0116] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0117] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for monitoring fish respiration, characterized in that: include: Obtain fish video data; intercepting an image of the fish body video data; A hybrid cascade network model is used to perform real-time inference on each frame of video to obtain the motion bounding box of the fish's mouth and gills in each frame; Performing a spatiotemporal feature extraction operation on the motion bounding box to obtain a time series motion signal of the fish mouth and gills; Determine the respiratory frequency value of the fish body according to the time series motion signal; The fish body stress level is determined according to the fish body breathing frequency value.

2. The method according to claim 1, characterized in that Obtain fish video data, including: Use Gaussian filtering to perform low-frequency filtering on the acquired original video stream; Use median filtering to process the original video stream after low-frequency filtering; The multi-spectral channel equalization method is used to adjust the processed original video stream; The resized original video stream is enhanced using contrast-limited adaptive histogram equalization.

3. The method according to claim 1, characterized in that Also includes: Split the fish body video frame by frame to obtain the fish body image; Marking a fish head region in the fish body image; Performing at least one of random rotation, dynamic scaling, atomization simulation, and motion blur on the annotated fish image; Combining the processed fish body images into a training data set; The ConvNeXt network structure is used as the backbone network to build the initial model; Determine the loss function of the initial model according to formula (1) to formula (3): FL(p)=-α(1-p) γ log(p), (1) BCE(p)=-log(p), (2) Loss(p)=FL(p)+BCE(p), (3) Wherein, α is the balance factor, p is the predicted probability of the initial model, γ is the focus factor, and Loss(p) is the loss function; The data set is used to train an initial model to obtain the hybrid cascade network model.

4. The method according to claim 1, wherein Performing a spatiotemporal feature extraction operation on the motion bounding box to obtain a time series motion signal of the fish mouth and gills, including: A multi-scale feature pyramid is used to extract spatiotemporal motion features between the motion bounding boxes of every two frames; Calculating optical flow field components based on the spatiotemporal motion characteristics using an iterative update mechanism; Calculating a motion intensity representation based on the optical flow field components; generating a time series signal based on the motion intensity representation at a preset sampling rate; A time alignment mechanism is used to compensate for the phase deviation of the time series signal using a model; The time series signals are synchronized according to the result of the phase deviation inference to obtain the time series motion signal.

5. The method according to claim 4, characterized in that An iterative update mechanism is used to calculate the optical flow field components based on the spatiotemporal motion characteristics, including: A median filter is used to eliminate abnormal optical flow components in the optical flow field components.

6. The method according to claim 4, characterized in that Calculating a motion intensity representation based on the optical flow field components includes: The exercise intensity representation is calculated using formula (4): in, is the motion intensity representation of the i-th sampling point, u i 、v i are different sub-components in the optical flow field component of the i-th sampling point, and λ is the statistical quantile.

7. The method according to claim 1, characterized in that Determining the respiratory frequency value of the fish according to the time series motion signal includes: Preprocessing the time series motion signal in sequence using a fourth-order Butterworth low-pass filter and a Symlet wavelet function; Determine respiratory cycle feature points based on the preprocessed time series motion signal using a sliding window dynamic peak detection algorithm; Determining a weighted average of adjacent peak intervals based on the respiratory cycle characteristic points; The fish respiratory frequency value is calculated according to the weighted average value.

8. The method according to claim 1, characterized in that Determining the fish body stress level according to the fish body breathing frequency value includes: Determining the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio based on the fish respiratory rate value; Constructing a characteristic matrix based on the respiratory rate variation coefficient, peak period standard deviation, and energy entropy ratio; The feature matrix is ​​weighted using three-level pressure labels, and the degree of incoordination between the fish mouth and gill movements is determined by combining the SHAP value interpretability analysis method; The fish body pressure level is determined according to the incoordination degree.

9. A fish respiration monitoring system, characterized in that: The system comprises a processor, and the processor is configured to execute the method according to any one of claims 1 to 8.

Images