Laying hen health condition monitoring method and system based on cockscomb form

Abstract

The invention relates to a computer vision technology, and discloses a laying hen health condition monitoring method and system based on a cockscomb form, and the method comprises the steps: collecting laying hen cockscomb images according to a preset time interval, and forming an image sequence; and carrying out gravity center identification on each image in the sequence, obtaining gravity center sequence data and analyzing gravity center offset. If the center-of-gravity offset exceeds a preset threshold value, color identification is directly carried out on the current image to obtain color data, and if the center-of-gravity offset does not exceed the preset threshold value, the form variation of the cockscomb edge area in the image sequence is analyzed. When the form variation is greater than a preset threshold value, judging that the laying hen is healthy; and if not, returning to execute the color identification step. And finally, performing disease judgment according to the color data to obtain a judgment result. The laying hen health condition monitoring efficiency can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image analysis, and in particular to a method and system for monitoring the health status of laying hens based on comb morphology. Background Art

[0002] With the increasing scale and intensification of egg-laying chicken farming in my country, stocking density and management complexity have significantly increased. Accurate monitoring of egg-laying chicken health has become a critical component in ensuring both farming profitability and food safety. Traditionally, egg-laying chicken health monitoring relies primarily on manual observation, with breeders regularly checking the chickens' mental state, diet, fecal form, and body characteristics (such as combs and feathers). The comb, a crucial surface feature of laying hens, exhibits significant correlations with changes in its color, shape, and texture. For example, a pale comb may indicate anemia or malnutrition, a cyanotic comb is often associated with respiratory disease, and a swollen comb may indicate inflammation or viral infection.

[0003] Existing methods for monitoring the health status of laying hens obtain physiological parameters such as body temperature and heart rate by implanting sensors or wearing devices such as smart collars. Although this method can achieve accurate data collection, it has problems such as cumbersome operation, easy stress response in laying hens, and high equipment maintenance costs, making it difficult to promote and apply in large-scale breeding environments. Summary of the Invention

[0004] The present invention provides a method and system for monitoring the health status of laying hens based on comb morphology, the main purpose of which is to solve the problems of low efficiency and high cost of existing methods for monitoring the health status of laying hens.

[0005] To achieve the above objectives, the present invention provides a method for monitoring the health status of laying hens based on comb morphology, comprising:

[0006] Using a camera preset on an inspection track of a laying hen pen to collect comb images of laying hens at preset time intervals to obtain a comb image sequence, performing comb center of gravity identification on each comb image in the comb image sequence to obtain comb center of gravity sequence data;

[0007] Performing center offset analysis based on the comb center of gravity sequence data to obtain a center of gravity offset;

[0008] Determining whether the center of gravity offset is greater than a preset offset threshold;

[0009] If the center of gravity offset is greater than the offset threshold, performing comb color recognition on the comb image to obtain color data;

[0010] If the center of gravity offset is less than or equal to the offset threshold, performing a morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount;

[0011] Determining whether the morphological change is greater than a preset change threshold;

[0012] If the morphological change is greater than the change threshold, it is confirmed that the current laying hen is healthy;

[0013] If the morphological change amount is less than or equal to the change amount threshold, returning to the step of performing comb color recognition on the comb image to obtain color data;

[0014] Diseases of laying hens are identified based on the color data to obtain disease identification results.

[0015] Optionally, the identifying the comb center of gravity of each comb image in the comb image sequence includes:

[0016] Extracting texture features of the cockscomb image;

[0017] extracting the infrared features of blood vessels of the cockscomb image;

[0018] Dynamically fusing the texture features and the vascular infrared features through a cross-attention mechanism to obtain a bimodal feature;

[0019] Dividing the cockscomb image into a plurality of image blocks of the same size;

[0020] generating a thermal distribution map of blood vessel distribution according to the dual-modal features;

[0021] Determining a local weight of each image block according to the heat distribution map;

[0022] Calculating a local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature;

[0023] The center of gravity position of the comb image is determined according to the image block with the largest congestion coefficient in each comb image.

[0024] Optionally, calculating the local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature includes:

[0025] Obtaining a feature vector for each image block according to the bimodal feature;

[0026] Calculating a spatial correlation term of each image block according to the feature vector;

[0027] Perform Fourier operation on the feature vector of each image block to obtain the frequency domain features of each image block;

[0028] The local congestion coefficient of each image block is calculated according to the spatial correlation term and the frequency domain feature.

[0029] Optionally, the spatial correlation term of each image block is calculated using the following formula:

[0030]

[0031] in, is the first The spatial correlation term of image blocks, represents the number of image block rows of the cockscomb image, represents the number of image block columns of the cockscomb image, Indicates that in the cockscomb image, Rank The feature vector of the image block.

[0032] Optionally, the local congestion coefficient is calculated using the following formula:

[0033]

[0034] in, Indicates the first The local congestion coefficient of the image block, For the The local weight of the image patch, is the dimension of the feature vector, Indicates the first The image block in The feature vector of dimension is the preset feature vector importance weight, is the preset first adjustment parameter, is the first The spatial correlation term of image blocks, is the preset second adjustment parameter, is the number of components of the frequency domain feature, Indicates the first The frequency domain features of the image block The real part of the component, Indicates the first The frequency domain features of the image block A portion.

[0035] Optionally, performing center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset includes:

[0036] Calculating the image clarity of each cockscomb image in the cockscomb image sequence;

[0037] Determine the image weight of each cockscomb image according to the image clarity of each cockscomb image, and obtain an image weight sequence;

[0038] Convert the comb barycenter sequence data into coordinate data to obtain a comb barycenter coordinate sequence;

[0039] The center of gravity offset is calculated according to the image weight sequence and the cockscomb center of gravity coordinate sequence.

[0040] Optionally, the center of gravity offset is calculated using the following formula:

[0041]

[0042] in, is the center of gravity offset, is the number of samples of the cockscomb barycenter coordinate sequence, Indicates the image weight sequence image weights, Indicates the first The horizontal coordinate of the coordinate data, Indicates the first The vertical coordinate of the coordinate data.

[0043] Optionally, performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain the morphological change amount includes:

[0044] Obtaining visible light edge features of each cockscomb image in the cockscomb image sequence;

[0045] Acquire infrared edge features of each comb image in the comb image sequence;

[0046] Performing feature fusion on the visible light edge feature and the infrared edge feature to obtain a multimodal edge feature;

[0047] Calculating the contour area of ​​each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour area sequence;

[0048] Calculating the contour perimeter of each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour perimeter sequence;

[0049] The morphological change amount is calculated according to the contour area sequence and the contour perimeter sequence.

[0050] Optionally, the calculation formula of the morphological change amount is as follows:

[0051]

[0052] in, is the morphological change amount, is the preset variation fusion weight, is the number of samples in the cockscomb image sequence, Indicates the first The contour perimeter, Indicates the first A contour area.

[0053] In order to solve the above problems, the present invention also provides a laying hen health monitoring system based on comb morphology, characterized in that the system includes a center of gravity recognition module, an offset analysis module, a first judgment module, a second judgment module and a disease discrimination module, wherein:

[0054] The center of gravity recognition module is used to use a camera preset on the inspection track of the laying hen pen to collect the comb images of the laying hen according to the preset time interval to obtain a comb image sequence, and perform comb center of gravity recognition on each comb image in the comb image sequence to obtain comb center of gravity sequence data;

[0055] The offset analysis module is used to perform center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset;

[0056] The first judgment module is used to determine whether the center of gravity offset is greater than a preset offset threshold; if the center of gravity offset is greater than the offset threshold, performing comb color recognition on the comb image to obtain color data; if the center of gravity offset is less than or equal to the offset threshold, performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount;

[0057] The second judgment module is used to judge whether the morphological change is greater than a preset change threshold. If the morphological change is greater than the change threshold, it is confirmed that the current laying hen is healthy. If the morphological change is less than or equal to the change threshold, it returns to the first judgment module and performs comb color recognition on the comb image to obtain color data.

[0058] The disease identification module is used to identify diseases of laying hens based on the color data to obtain disease identification results.

[0059] The embodiment of the present invention uses a camera preset on an inspection track of a laying hen pen to collect comb images of laying hens according to preset time intervals to obtain a comb image sequence, performs comb center of gravity identification on each comb image in the comb image sequence to obtain comb center of gravity sequence data, so as to improve the efficiency of subsequent center of gravity offset analysis; performs center offset analysis on the comb center of gravity sequence data to obtain a center of gravity offset; determines whether the center of gravity offset is greater than a preset offset threshold; if the center of gravity offset is greater than the offset threshold, performs comb color identification on the comb image to obtain color data, which can improve the efficiency of subsequent disease identification and Accuracy; if the center of gravity offset is less than or equal to the offset threshold, then performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change; judging whether the morphological change is greater than a preset change threshold; if the morphological change is greater than the change threshold, confirming that the current laying hen is healthy; if the morphological change is less than or equal to the change threshold, returning to the step of performing comb color recognition on the comb image to obtain color data can improve the accuracy of laying hen health monitoring; performing disease identification on the laying hen according to the color data to obtain a disease identification result can improve the efficiency of laying hen disease identification. Therefore, the laying hen health status monitoring method and system based on comb morphology proposed in the present invention can solve the problems of low efficiency and high cost of existing laying hen health status monitoring methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A schematic flow chart of a method for monitoring the health status of laying hens based on comb morphology provided by one embodiment of the present invention;

[0061] Figure 2 A schematic structural diagram of an inspection track provided by one embodiment of the present invention;

[0062] Figure 3 This is a functional module diagram of a laying hen health monitoring system based on comb morphology provided by one embodiment of the present invention.

[0063] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0064] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0065] The embodiment of the present application provides a method for monitoring the health status of laying hens based on the morphology of the comb. The execution subject of the method for monitoring the health status of laying hens based on the morphology of the comb includes but is not limited to at least one of the electronic devices such as the server and the terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for monitoring the health status of laying hens based on the morphology of the comb can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0066] Reference Figure 1 FIG. 1 is a flow chart of a method for monitoring the health status of laying hens based on comb morphology according to an embodiment of the present invention. In this embodiment, the method for monitoring the health status of laying hens based on comb morphology includes:

[0067] S1. Use a camera preset on the inspection track of the laying hen pen to collect the comb images of the laying hen according to the preset time interval to obtain a comb image sequence, and perform comb center of gravity identification on each comb image in the comb image sequence to obtain comb center of gravity sequence data.

[0068] In an embodiment of the present invention, the inspection track is a mechanical track specially designed for automated inspection, which can provide a fixed moving path for inspection equipment such as cameras, thereby achieving full coverage monitoring of a group of laying hens.

[0069] In the embodiment of the present invention, referring to Figure 2 FIG. 1 is a schematic structural diagram of an inspection track provided by an embodiment of the present invention.

[0070] In detail, the inspection track 2 is installed in the laying hen pen 1, and a slidable camera 3 is provided on the inspection track 2; the inspection track 2 is electrically connected to the camera 3, and the camera 3 can be controlled by electrical signals to slide on the inspection track 2, and the camera 3 can be controlled by electrical signals to take pictures of laying hens.

[0071] In the embodiment of the present invention, the preset time interval may be one day.

[0072] In the embodiment of the present invention, the collecting of the comb images of laying hens according to the preset time intervals is the collection of images of the combs of the same laying hen.

[0073] Specifically, a leg ring with a unique number printed on it can be worn on the feet of each laying hen. After collecting images of all laying hens, the comb image of the same laying hen can be collected by identifying the number of the leg ring.

[0074] In the embodiment of the present invention, the step of identifying the comb center of gravity of each comb image in the comb image sequence includes:

[0075] Extracting texture features of the cockscomb image;

[0076] extracting the infrared features of blood vessels of the cockscomb image;

[0077] Dynamically fusing the texture features and the vascular infrared features through a cross-attention mechanism to obtain a bimodal feature;

[0078] Dividing the cockscomb image into a plurality of image blocks of the same size;

[0079] generating a thermal distribution map of blood vessel distribution according to the dual-modal features;

[0080] Determining a local weight of each image block according to the heat distribution map;

[0081] Calculating a local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature;

[0082] The center of gravity position of the comb image is determined according to the image block with the largest congestion coefficient in each comb image.

[0083] In an embodiment of the present invention, the texture features of the comb image are extracted by using lightweight ConvNeXt-Tiny to extract the RGB texture features of the comb image.

[0084] ConvNeXt is a convolutional neural network (CNN)-based architecture that combines the advantages of the modern Transformer architecture with the efficiency of traditional CNNs. Tiny ConvNeXt is a lightweight version of the ConvNeXt architecture. By reducing the number of model parameters and computational complexity, it reduces computing resource requirements while maintaining performance, making it more suitable for running on resource-constrained devices.

[0085] Specifically, the cross-attention mechanism is a commonly used mechanism in deep learning to capture the interrelationships between features from different modalities. In this solution, it is used to dynamically fuse texture features and vascular infrared features. The cross-attention mechanism allows the model to focus on important parts of one feature while processing another, thereby better integrating the information of the two features.

[0086] Specifically, a heatmap is a visualization method that uses color to represent data density or intensity. In this solution, a heatmap of vascular distribution generated based on bimodal features uses color depth to represent the density or activity of vascular distribution in different areas of the comb. Darker areas may indicate denser vascular distribution or higher levels of congestion.

[0087] In this embodiment of the present invention, the infrared features of the blood vessels in the comb image are extracted by performing denoising and enhancement on the comb image and setting an appropriate threshold based on the grayscale features of the comb blood vessels in the infrared image. Pixels with grayscale values ​​above the threshold are considered to be blood vessels, while pixels below the threshold are considered to be background or other tissue.

[0088] In the embodiment of the present invention, the texture features and the vascular infrared features are dynamically fused using the following formula:

[0089]

[0090] in, is the multimodal feature, is the preset fusion weight, is the texture feature, is the infrared characteristic of the blood vessel, is the Sigmoid function, is the channel number of the vascular infrared feature, is the number of channels of the multimodal feature.

[0091] In detail, the fusion weight may be 0.35.

[0092] In detail, the number of channels of the vascular infrared feature and the multimodal feature may be 128.

[0093] In detail, the calculating of the local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature includes:

[0094] Obtaining a feature vector for each image block according to the bimodal feature;

[0095] Calculating a spatial correlation term of each image block according to the feature vector;

[0096] Perform Fourier operation on the feature vector of each image block to obtain the frequency domain features of each image block;

[0097] The local congestion coefficient of each image block is calculated according to the spatial correlation term and the frequency domain feature.

[0098] Specifically, the local congestion coefficient is a comprehensive indicator used to measure the degree of congestion in each image block within a cockscomb image. It is calculated using a specific formula, combining information such as the local weight of the image block, bimodal features, spatial correlation terms, and frequency domain characteristics. A larger congestion coefficient indicates a higher degree of congestion in that image block.

[0099] Specifically, the spatial correlation term of each image block is calculated using the following formula:

[0100]

[0101] in, is the first The spatial correlation term of image blocks, represents the number of image block rows of the cockscomb image, represents the number of image block columns of the cockscomb image, Indicates that in the cockscomb image, Rank The feature vector of the image block.

[0102] In an embodiment of the present invention, the core logic of the calculation formula of the spatial correlation term of each image block is to quantify the degree of local change of the image block in space by measuring the difference in the eigenvectors of adjacent image blocks, thereby providing a spatial correlation basis for subsequent analysis of the congestion state and center of gravity position of the cockscomb.

[0103] Specifically, the local hyperemia coefficient is calculated using the following formula:

[0104]

[0105] in, Indicates the first The local congestion coefficient of the image block, For the The local weight of the image patch, is the dimension of the feature vector, Indicates the first The image block in The feature vector of dimension is the preset feature vector importance weight, is the preset first adjustment parameter, is the first The spatial correlation term of image blocks, is the preset second adjustment parameter, is the number of components of the frequency domain feature, Indicates the first The frequency domain features of the image block The real part of the component, Indicates the first The frequency domain features of the image block A portion.

[0106] In the embodiment of the present invention, the core logic of the calculation formula of the local congestion coefficient is to quantify the congestion degree of the comb tissue by integrating multimodal features (visible light texture, infrared blood vessels), spatial correlation and frequency domain characteristics, providing a key indicator for subsequent center of gravity identification and health judgment. It calculates the Euclidean norm square of the feature vector, normalizes the numerator, avoids the deviation caused by the difference in numerical scale in high-dimensional features, converts the congestion coefficient into the response value under unit feature energy, and improves the comparability between different image blocks; It represents the enhancement of frequency domain features. Reflects the amplitude of the frequency domain component. High amplitude corresponds to significant frequency features (such as high-frequency texture of dense blood vessels). The role of is to act as a frequency weight, giving weights to different frequency components (such as the high-frequency components corresponding to the focused microvessels). The vascular distribution of a healthy comb is usually regular (dominated by a specific frequency), and the diseased area may destroy this regularity. By setting different frequency weights, abnormalities can be captured.

[0107] In detail, the first adjustment parameter may be 1.25, and the second adjustment parameter may be 0.88.

[0108] In detail, after Fourier transform is performed on the bimodal feature vector, the obtained frequency domain feature is a complex vector, in which each element has a real part and an imaginary part.

[0109] S2. Perform center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset.

[0110] In the embodiment of the present invention, the center of gravity offset is a coefficient representing the degree of center of gravity offset of the laying hen's comb.

[0111] Specifically, under normal circumstances, the center of gravity of a laying hen's comb is relatively stable. This is due to the comb's normal physiological structure and growth. The comb's internal tissues are evenly distributed, and the blood vessels, muscles, and connective tissues coordinate with each other, allowing the comb to maintain a relatively balanced position on the head. The comb's center of gravity also depends on factors such as the laying hen's breed and head posture. Generally speaking, when laying hens are healthy and in normal physiological condition, the comb's center of gravity fluctuates within a narrow range.

[0112] Furthermore, when the center of gravity of the comb of laying hens shifts significantly in a short period of time, it indicates that the laying hens may have related diseases. For example, when laying hens suffer from neurological diseases, such as brain infections, nerve damage, etc., it will affect the balance perception and body coordination ability of laying hens. Laying hens may not be able to control their head posture correctly, which will cause the center of gravity of the comb to shift; if there is a tumor on the comb, the growth of the tumor will increase the weight of one side of the comb, causing the center of gravity of the comb to shift toward the side where the tumor grows. Or if there is severe inflammation and swelling in the local area of ​​the comb, the balance of the comb will be broken and the center of gravity will change; when a part of the blood vessels in the comb is embolized or narrowed, the tissue in that area may atrophy due to ischemia, while other parts are relatively normal, which will change the mass distribution of the comb and cause the center of gravity to shift.

[0113] In the embodiment of the present invention, performing center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset includes:

[0114] Calculating the image clarity of each cockscomb image in the cockscomb image sequence;

[0115] Determine the image weight of each cockscomb image according to the image clarity of each cockscomb image, and obtain an image weight sequence;

[0116] Convert the comb barycenter sequence data into coordinate data to obtain a comb barycenter coordinate sequence;

[0117] The center of gravity offset is calculated according to the image weight sequence and the cockscomb center of gravity coordinate sequence.

[0118] In the embodiment of the present invention, the center of gravity offset is calculated using the following formula:

[0119]

[0120] in, is the center of gravity offset, is the number of samples of the cockscomb barycenter coordinate sequence, Indicates the image weight sequence image weights, Indicates the first The horizontal coordinate of the coordinate data, Indicates the first The vertical coordinate of the coordinate data.

[0121] In detail, the image clarity of each cockscomb image in the cockscomb image sequence can be calculated by converting the image into a grayscale image to simplify the calculation, and using the Laplace operator to perform a convolution operation on the grayscale image. The Laplace operator is a second-order derivative operator that can highlight the edges and detail information in the image. Areas with rich edges and details will have larger response values ​​in the convolution result. After the convolution is completed, the variance of the Laplace convolution result image is calculated. The larger the variance, the richer the edges and details in the image, and the clearer the image; conversely, the smaller the variance, the blurrier the image.

[0122] In detail, the image weight of each rooster's comb image is determined according to the image clarity of each rooster's comb image, by normalizing the image clarity of each rooster's comb image and using the normalized result as the weight.

[0123] In detail, the coordinate data conversion of the comb center of gravity sequence data is performed by establishing a two-dimensional coordinate system with a preset position in the comb image as the coordinate origin, and converting the pixel position in the comb image into a coordinate position.

[0124] S3. Determine whether the center of gravity offset is greater than a preset offset threshold.

[0125] If the center of gravity offset is greater than the offset threshold, step S4 is executed to perform comb color recognition on the comb image to obtain color data.

[0126] If the center of gravity offset is less than or equal to the offset threshold, executing S5, performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount.

[0127] In the embodiment of the present invention, the step of performing morphological change analysis on the edge region of the comb according to the comb image sequence to obtain the morphological change amount includes:

[0128] Obtaining visible light edge features of each cockscomb image in the cockscomb image sequence;

[0129] Acquire infrared edge features of each comb image in the comb image sequence;

[0130] Performing feature fusion on the visible light edge feature and the infrared edge feature to obtain a multimodal edge feature;

[0131] Calculating the contour area of ​​each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour area sequence;

[0132] Calculating the contour perimeter of each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour perimeter sequence;

[0133] The morphological change amount is calculated according to the contour area sequence and the contour perimeter sequence.

[0134] Specifically, the visible light edge feature refers to the characteristic information of the comb edge in the visible light image. By performing edge detection and other operations on the visible light image of the comb, the position, shape and other features of the comb edge can be extracted.

[0135] In detail, the infrared edge feature refers to the characteristic information of the comb edge in the infrared image, which can reflect the thermal radiation information of the comb.

[0136] In detail, the calculation of the morphological change amount based on the contour area sequence and the contour perimeter sequence can be to calculate the change rate of the contour area and contour perimeter between adjacent cockscomb images, and then fuse the two with Android preset weights to obtain the morphological change amount.

[0137] In detail, the calculation formula of the morphological change amount is as follows:

[0138]

[0139] in, is the morphological change amount, is the preset variation fusion weight, is the number of samples in the cockscomb image sequence, Indicates the first The contour perimeter, Indicates the first A contour area.

[0140] In an embodiment of the present invention, the calculation formula of the morphological change amount is to quantify the dynamic fluctuation of the comb morphology in the time series by fusing the contour perimeter change rate and the contour area change rate, thereby judging the health status of the laying hen.

[0141] S6. Determine whether the morphological change is greater than a preset change threshold.

[0142] If the morphological change is greater than the change threshold, S7 is executed to confirm the current laying hen is healthy.

[0143] If the morphological change is less than or equal to the change threshold, the process returns to S4 to perform comb color recognition on the comb image to obtain color data.

[0144] S8. Perform disease identification on the laying hens according to the color data to obtain a disease identification result.

[0145] In an embodiment of the present invention, the disease identification of laying hens based on the color data is performed according to a preset color-disease correspondence rule. For example, if the color data is white, it indicates that the laying hen suffers from anemia-related diseases (such as coccidiosis and leukocytozoonosis) or nutritional deficiencies. If the color data is purple, it indicates that the laying hen may suffer from respiratory diseases (such as infectious bronchitis and mycoplasma infection) or circulatory diseases (such as pericardial effusion syndrome). If the color data is yellowish, it indicates that the laying hen may suffer from infectious diseases (such as fowl pox or fowl cholera) or have an allergic reaction.

[0146] like Figure 3 1 is a functional module diagram of a laying hen health status monitoring system based on comb morphology provided by an embodiment of the present invention.

[0147] The laying hen health monitoring system 100 based on comb morphology described in the present invention can be installed in an electronic device. Depending on the functionality to be implemented, the laying hen health monitoring system 100 based on comb morphology can include a center of gravity identification module 101, an offset analysis module 102, a first judgment module 103, a second judgment module 104, and a disease identification module 105. A module, also referred to as a unit, is a series of computer program segments that can be executed by a processor in an electronic device and perform a fixed function. These modules are stored in the memory of the electronic device.

[0148] In this embodiment, the functions of each module / unit are as follows:

[0149] The center of gravity recognition module 101 is used to use a camera preset on the inspection track of the laying hen pen to collect the comb images of the laying hen according to the preset time interval to obtain a comb image sequence, and perform comb center of gravity recognition on each comb image in the comb image sequence to obtain comb center of gravity sequence data;

[0150] The offset analysis module 102 is used to perform center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset;

[0151] The first judgment module 103 is used to judge whether the center of gravity offset is greater than a preset offset threshold; if the center of gravity offset is greater than the offset threshold, performing comb color recognition on the comb image to obtain color data; if the center of gravity offset is less than or equal to the offset threshold, performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount;

[0152] The second judgment module 104 is used to judge whether the morphological change is greater than a preset change threshold. If the morphological change is greater than the change threshold, it is confirmed that the current laying hen is healthy. If the morphological change is less than or equal to the change threshold, the process returns to the first judgment module 103, and the process of performing comb color recognition on the comb image to obtain color data is performed.

[0153] The disease identification module 105 is used to identify diseases of laying hens based on the color data to obtain disease identification results.

[0154] In detail, each module in the laying hen health monitoring system 100 based on comb morphology according to the embodiment of the present invention is used in the same manner as above. Figure 1 The same technical means are used as the method for monitoring the health status of laying hens based on comb morphology described in , and can produce the same technical effects, so they will not be repeated here.

[0155] In the embodiments provided herein, it should be understood that the disclosed devices, systems, and methods may be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the module division is merely a logical functional division, and actual implementation may employ other division methods.

[0156] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected to achieve the purpose of the solution of this embodiment according to actual needs.

[0157] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0158] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0159] Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the claims are intended to be embraced therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0160] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, techniques, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results.

[0161] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems recited in a system claim may also be implemented by a single unit or system through software or hardware. Terms such as "first" and "second" are used to indicate names and do not imply any particular order.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for monitoring the health status of laying hens based on comb morphology, characterized in that: The method comprises: Using a camera preset on an inspection track of a laying hen pen to collect comb images of laying hens at preset time intervals to obtain a comb image sequence, performing comb center of gravity identification on each comb image in the comb image sequence to obtain comb center of gravity sequence data; Performing center offset analysis based on the comb center of gravity sequence data to obtain a center of gravity offset; Determining whether the center of gravity offset is greater than a preset offset threshold; If the center of gravity offset is greater than the offset threshold, performing comb color recognition on the comb image to obtain color data; If the center of gravity offset is less than or equal to the offset threshold, performing a morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount; Determining whether the morphological change is greater than a preset change threshold; If the morphological change is greater than the change threshold, it is confirmed that the current laying hen is healthy; If the morphological change amount is less than or equal to the change amount threshold, returning to the step of performing comb color recognition on the comb image to obtain color data; Diseases of laying hens are identified based on the color data to obtain disease identification results.

2. The method for monitoring the health status of laying hens based on comb morphology according to claim 1, wherein: The step of identifying the center of gravity of each comb image in the comb image sequence includes: Extracting texture features of the cockscomb image; extracting the infrared features of blood vessels of the cockscomb image; Dynamically fusing the texture features and the vascular infrared features through a cross-attention mechanism to obtain a bimodal feature; Dividing the cockscomb image into a plurality of image blocks of the same size; generating a thermal distribution map of blood vessel distribution according to the dual-modal features; Determining a local weight of each image block according to the heat distribution map; Calculating a local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature; The center of gravity position of the comb image is determined according to the image block with the largest congestion coefficient in each comb image.

3. The method for monitoring the health status of laying hens based on comb morphology according to claim 2, wherein: Calculating the local congestion coefficient of each image block according to the local weight of each image block and the bimodal feature includes: Obtaining a feature vector for each image block according to the bimodal feature; Calculating a spatial correlation term of each image block according to the feature vector; Perform Fourier operation on the feature vector of each image block to obtain the frequency domain features of each image block; The local congestion coefficient of each image block is calculated according to the spatial correlation term and the frequency domain feature.

4. The method for monitoring the health status of laying hens based on comb morphology according to claim 3, wherein: The spatial correlation term of each image block is calculated using the following formula: ； in, is the first The spatial correlation term of image blocks, represents the number of image block rows of the cockscomb image, represents the number of image block columns of the cockscomb image, Indicates that in the cockscomb image, Rank The feature vector of the image block.

5. The method for monitoring the health status of laying hens based on comb morphology according to claim 3, wherein: The local hyperemia coefficient is calculated using the following formula: ； in, Indicates the first The local congestion coefficient of the image block, For the The local weight of the image patch, is the dimension of the feature vector, Indicates the first The image block in The feature vector of dimension is the preset feature vector importance weight, is the preset first adjustment parameter, is the first The spatial correlation term of image blocks, is the preset second adjustment parameter, is the number of components of the frequency domain feature, Indicates the first The frequency domain features of the image block The real part of the component, Indicates the first The frequency domain features of the image block A portion.

6. The method for monitoring the health status of laying hens based on comb morphology according to claim 1, wherein: The center offset analysis is performed based on the comb center of gravity sequence data to obtain the center of gravity offset, including: Calculating the image clarity of each cockscomb image in the cockscomb image sequence; Determine the image weight of each cockscomb image according to the image clarity of each cockscomb image, and obtain an image weight sequence; Convert the comb barycenter sequence data into coordinate data to obtain a comb barycenter coordinate sequence; The center of gravity offset is calculated according to the image weight sequence and the cockscomb center of gravity coordinate sequence.

7. The method for monitoring the health status of laying hens based on comb morphology according to claim 6, wherein: The center of gravity offset is calculated using the following formula: ； in, is the center of gravity offset, is the number of samples of the cockscomb barycenter coordinate sequence, Indicates the image weight sequence image weights, Indicates the first The horizontal coordinate of the coordinate data, Indicates the first The vertical coordinate of the coordinate data.

8. The method for monitoring the health status of laying hens based on comb morphology according to claim 1, wherein: The step of performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain the morphological change amount includes: Obtaining visible light edge features of each cockscomb image in the cockscomb image sequence; Acquire infrared edge features of each comb image in the comb image sequence; Performing feature fusion on the visible light edge feature and the infrared edge feature to obtain a multimodal edge feature; Calculating the contour area of ​​each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour area sequence; Calculating the contour perimeter of each cockscomb image in the cockscomb image sequence according to the multimodal edge feature to obtain a contour perimeter sequence; The morphological change amount is calculated according to the contour area sequence and the contour perimeter sequence.

9. The method for monitoring the health status of laying hens based on comb morphology according to claim 8, wherein: The calculation formula of the morphological change is as follows: ； in, is the morphological change amount, is the preset variation fusion weight, is the number of samples in the cockscomb image sequence, Indicates the first The contour perimeter, Indicates the first A contour area.

10. A laying hen health monitoring system based on comb morphology, characterized in that: The system includes a center of gravity identification module, an offset analysis module, a first judgment module, a second judgment module and a disease discrimination module, wherein: The center of gravity recognition module is used to use a camera preset on the inspection track of the laying hen pen to collect the comb images of the laying hen according to the preset time interval to obtain a comb image sequence, and perform comb center of gravity recognition on each comb image in the comb image sequence to obtain comb center of gravity sequence data; The offset analysis module is used to perform center offset analysis based on the comb center of gravity sequence data to obtain the center of gravity offset; The first judgment module is used to determine whether the center of gravity offset is greater than a preset offset threshold; if the center of gravity offset is greater than the offset threshold, performing comb color recognition on the comb image to obtain color data; if the center of gravity offset is less than or equal to the offset threshold, performing morphological change analysis on the edge area of ​​the comb according to the comb image sequence to obtain a morphological change amount; The second judgment module is used to judge whether the morphological change is greater than a preset change threshold. If the morphological change is greater than the change threshold, it is confirmed that the current laying hen is healthy. If the morphological change is less than or equal to the change threshold, it returns to the first judgment module and performs comb color recognition on the comb image to obtain color data. The disease identification module is used to identify diseases of laying hens based on the color data to obtain disease identification results.

Images