Wenchang chicken health detection method and device based on dual-spectrum imaging, processor and computer readable storage medium thereof

Abstract

The invention relates to a method for realizing Wenchang chicken health detection based on dual-spectrum imaging. The method comprises the following steps: acquiring a visible light image and a thermal imaging image of the Wenchang chicken under the same space-time reference; inputting the double-spectrum image into a key part positioning model trained by a labeled data set, and synchronously outputting bounding boxes of a cockscomb region, an eye region and a main body feather region; multi-dimensional health feature vectors are extracted in parallel based on the bounding box; and inputting the multi-dimensional health feature vector into a health assessment classification model trained by a labeled data set, and outputting a quantized health state score and a corresponding health level. According to the Wenchang chicken health detection method and device based on double-spectrum imaging, the processor and the computer readable storage medium thereof, key parts such as a cockscomb, feathers and eyes are automatically positioned by synchronously collecting visible light images and thermal imaging images, multi-dimensional health features are extracted, and the detection accuracy is improved. Based on the Wenchang chicken specific health assessment model, accurate health state judgment and early disease early warning are realized.

Description

Technical Field

[0001] This invention relates to the field of intelligent agricultural breeding technology, and more particularly to the field of livestock and poultry health monitoring technology. Specifically, it relates to a method, device, processor, and computer-readable storage medium for health detection of Wenchang chicken based on dual-spectral imaging. Background Technology

[0002] Currently, health monitoring in Wenchang chicken farming mainly relies on manual observation and experience-based judgment, which has the following technical shortcomings: 1. Highly subjective: Relies on the experience of the zookeepers and lacks objective, quantifiable standards; 2. Low efficiency: Manual inspection is time-consuming and labor-intensive, and cannot achieve all-weather monitoring; 3. Delayed early warning: It is often only discovered when the symptoms are obvious, thus missing the best time for treatment; 4. Inconsistent standards: Different keepers have different standards for judgment.

[0003] Traditional computer vision methods only use visible light images, which cannot obtain information about the internal physiological state of chickens, resulting in limited detection accuracy. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, device, processor and computer-readable storage medium for health detection of Wenchang chicken based on dual-spectral imaging that is automated, highly accurate and widely applicable.

[0005] To achieve the above objectives, the present invention provides a method, apparatus, processor, and computer-readable storage medium for health detection of Wenchang chickens based on dual-spectral imaging, as follows: The main feature of this method for health detection of Wenchang chickens based on dual-spectral imaging is that the method includes the following steps: (1) Visible light images and thermal images of Wenchang chickens were acquired under the same spatiotemporal reference by using a synchronously calibrated visible light camera and thermal imaging camera; (2) Input the bispectral image into the key part localization model trained on the labeled dataset, and output the bounding boxes of the comb, eyes and main feather regions simultaneously; (3) Extract multidimensional health feature vectors in parallel based on bounding boxes; (4) Input the multidimensional health feature vector into the health assessment classification model trained on the labeled dataset, and output the quantitative health status score and the corresponding health level; (5) If the health status score is lower than the preset threshold, generate and output health warning information containing specific abnormal feature dimensions.

[0006] Preferably, step (3) specifically includes the following steps: (3.1) Based on visible light images, extract the hue and saturation statistics of the HSV color space, and based on thermal imaging images, extract the mean and distribution variance of the temperature of the comb region to obtain the comb feature vector; among which, the normal range of body surface temperature of the comb region of healthy Wenchang chickens is 40.5℃~41.5℃. (3.2) Based on thermal imaging images, extract the surface temperature of the chicken's body, calculate the temperature gradient between the core area of ​​the torso and the surrounding feather-covered area, and obtain the feather feature vector; (3.3) Based on visible light images, color features of the eyelids and conjunctiva are extracted. Combined with the body surface temperature information provided by thermal imaging images, the recognition accuracy is improved by image fusion technology to obtain eye feature vectors.

[0007] Preferably, the key part localization model in step (2) is a convolutional neural network based on the attention mechanism. It is trained by centrally annotating the bounding box annotation data of the Wenchang chicken comb, eyes and trunk, and prioritizes the localization of the unique single comb or double comb structure of the Wenchang chicken.

[0008] Preferably, the health assessment classification model in step (4) adopts a weighted multi-feature fusion algorithm, and determines the weight of each feature vector based on the logistic regression feature importance analysis of the labeled dataset.

[0009] Preferably, the weights of the rooster comb feature in the aforementioned feature vectors are 0.28–0.32, the weights of the feather feature are 0.23–0.27, the weights of the eye feature are 0.18–0.22, the weights of the body temperature feature are 0.13–0.17, and the weights of the behavioral feature are 0.08–0.12.

[0010] The main feature of this system for health detection of Wenchang chickens based on dual-spectral imaging is that the system includes: The dual-spectral image acquisition module is used to acquire visible light and thermal images of Wenchang chickens under the same spatiotemporal reference. The data processing and analysis module, connected to the dual-spectral image acquisition module, is used to load and run the key part location model and the health assessment classification model; The health decision and output module, connected to the data processing and analysis module, is used to determine health status based on multi-feature fusion, generate health status scores, health levels and health warning information, and provide a human-computer interaction interface to generate health reports and abnormal warnings.

[0011] Preferably, the dual-spectrum image acquisition module includes a visible light camera, a thermal imaging camera, and an image registration unit. The image registration unit is connected to the visible light camera and the thermal imaging camera. The visible light camera and the thermal imaging camera are hardware synchronized. The image registration unit is used to ensure spatiotemporal consistency.

[0012] Preferably, the data processing and analysis module includes: The first processing unit is used to load and run the key part location model; The second processing unit is used to load and run the health assessment classification model; The database is used to store a dataset of labeled images of tens of thousands of chickens.

[0013] Preferably, the system further includes an intelligent gimbal control module and an automated gimbal. The dual-spectrum image acquisition module is installed on the automated gimbal, and the intelligent gimbal control module is connected to the automated gimbal. The intelligent gimbal control module tracks the chicken's movement in real time based on the DeepSORT multi-target tracking algorithm and controls the dual-spectrum image acquisition module to automatically adjust the shooting angle and capture images of key parts.

[0014] Preferably, the human-computer interaction interface of the health decision and output module includes a health status dashboard, which displays the distribution of health scores of the flock, a list of abnormal individuals, and historical health trends in the form of charts.

[0015] The main feature of this device for health detection of Wenchang chickens based on dual-spectral imaging is that the device includes: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the method for health detection of Wenchang chickens based on dual-spectral imaging.

[0016] The processor for health detection of Wenchang chicken based on dual-spectral imaging is characterized in that the processor is configured to execute computer-executable instructions, and when the computer-executable instructions are executed by the processor, the various steps of the above-mentioned method for health detection of Wenchang chicken based on dual-spectral imaging are implemented.

[0017] The main feature of this computer-readable storage medium is that it stores a computer program thereon, which can be executed by a processor to implement the various steps of the above-described method for health detection of Wenchang chickens based on dual-spectral imaging.

[0018] This invention employs a method, device, processor, and computer-readable storage medium for health detection of Wenchang chickens based on dual-spectral imaging. By simultaneously acquiring visible light and thermal images, it automatically locates key areas such as the comb, feathers, and eyes, extracts multi-dimensional health features, and achieves accurate health status judgment and early disease warning based on a Wenchang chicken-specific health assessment model. This invention solves the problems of low efficiency and poor accuracy of traditional manual detection, providing effective technical support for the scientific breeding of Wenchang chickens. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method for health detection of Wenchang chicken based on dual-spectral imaging according to the present invention. Detailed Implementation

[0020] To more clearly describe the technical content of the present invention, the following description is provided in conjunction with specific embodiments.

[0021] The method for health detection of Wenchang chickens based on dual-spectral imaging of the present invention includes the following steps: (1) Visible light images and thermal images of Wenchang chickens were acquired under the same spatiotemporal reference by using a synchronously calibrated visible light camera and thermal imaging camera; (2) Input the bispectral image into the key part localization model trained on the labeled dataset, and output the bounding boxes of the comb, eyes and main feather regions simultaneously; (3) Extract multidimensional health feature vectors in parallel based on bounding boxes; (4) Input the multidimensional health feature vector into the health assessment classification model trained on the labeled dataset, and output the quantitative health status score and the corresponding health level; (5) If the health status score is lower than the preset threshold, generate and output health warning information containing specific abnormal feature dimensions.

[0022] In a preferred embodiment of the present invention, step (3) specifically includes the following steps: (3.1) Based on visible light images, extract the hue and saturation statistics of the HSV color space, and based on thermal imaging images, extract the mean and distribution variance of the temperature of the comb region to obtain the comb feature vector; among which, the normal range of body surface temperature of the comb region of healthy Wenchang chickens is 40.5℃~41.5℃. (3.2) Based on thermal imaging images, extract the surface temperature of the chicken's body, calculate the temperature gradient between the core area of ​​the torso and the surrounding feather-covered area, and obtain the feather feature vector; (3.3) Based on visible light images, color features of the eyelids and conjunctiva are extracted. Combined with the body surface temperature information provided by thermal imaging images, the recognition accuracy is improved by image fusion technology to obtain eye feature vectors.

[0023] As a preferred embodiment of the present invention, the key part localization model in step (2) is a convolutional neural network based on the attention mechanism. The model is trained by centrally annotating the bounding box annotation data of the Wenchang chicken comb, eyes and trunk, and prioritizes the localization of the unique single comb or double comb structure of the Wenchang chicken.

[0024] As a preferred embodiment of the present invention, the health assessment classification model in step (4) adopts a weighted multi-feature fusion algorithm, and determines the weight of each feature vector based on the logistic regression feature importance analysis of the labeled dataset.

[0025] In a preferred embodiment of the present invention, the weights of the rooster comb feature in each feature vector are 0.28 to 0.32, the weight of the feather feature is 0.23 to 0.27, the weight of the eye feature is 0.18 to 0.22, the weight of the body temperature feature is 0.13 to 0.17, and the weight of the behavior feature is 0.08 to 0.12.

[0026] The present invention discloses a system for health detection of Wenchang chickens based on dual-spectral imaging, wherein the system comprises: The dual-spectral image acquisition module is used to acquire visible light and thermal images of Wenchang chickens under the same spatiotemporal reference. The data processing and analysis module, connected to the dual-spectral image acquisition module, is used to load and run the key part location model and the health assessment classification model; The health decision and output module, connected to the data processing and analysis module, is used to determine health status based on multi-feature fusion, generate health status scores, health levels and health warning information, and provide a human-computer interaction interface to generate health reports and abnormal warnings.

[0027] In a preferred embodiment of the present invention, the dual-spectrum image acquisition module includes a visible light camera, a thermal imaging camera, and an image registration unit. The image registration unit is connected to the visible light camera and the thermal imaging camera. The visible light camera and the thermal imaging camera are hardware synchronized. The image registration unit is used to ensure spatiotemporal consistency.

[0028] In a preferred embodiment of the present invention, the data processing and analysis module includes: The first processing unit is used to load and run the key part location model; The second processing unit is used to load and run the health assessment classification model; The database is used to store a dataset of labeled images of tens of thousands of chickens.

[0029] In a preferred embodiment of the present invention, the system further includes an intelligent gimbal control module and an automated gimbal. The dual-spectrum image acquisition module is installed on the automated gimbal, and the intelligent gimbal control module is connected to the automated gimbal. The intelligent gimbal control module tracks the movement of the chicken in real time based on the DeepSORT multi-target tracking algorithm and controls the dual-spectrum image acquisition module to automatically adjust the shooting angle and capture images of key parts.

[0030] In a preferred embodiment of the present invention, the human-computer interaction interface of the health decision and output module includes a health status dashboard, which displays the health score distribution of the flock, a list of abnormal individuals, and historical health trends in the form of charts.

[0031] The device for health detection of Wenchang chicken based on dual-spectral imaging of the present invention includes: A processor is configured to execute computer-executable instructions; The memory stores one or more computer-executable instructions, which, when executed by the processor, implement the various steps of the method for health detection of Wenchang chickens based on dual-spectral imaging.

[0032] The processor of the present invention for health detection of Wenchang chicken based on dual-spectral imaging is configured to execute computer-executable instructions, which, when executed by the processor, implement the various steps of the above-described method for health detection of Wenchang chicken based on dual-spectral imaging.

[0033] The computer-readable storage medium of the present invention stores a computer program thereon, which can be executed by a processor to implement the various steps of the above-described method for health detection of Wenchang chicken based on dual-spectral imaging.

[0034] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for health detection of Wenchang chicken based on dual-spectral imaging. Through the synergistic analysis of visible light and thermal imaging, the health status of Wenchang chicken can be detected automatically and accurately.

[0035] The two-stage detection logic of this invention is "first identify the chicken, then determine its health status": 1. Identification stage: First, thermal imaging is used to quickly locate the chicken's position, and then visible light images are used to confirm the chicken's outline and key parts; 2. Judgment stage: Based on multimodal feature fusion, combined with temperature threshold and visual feature weights, the health status is comprehensively judged.

[0036] The hardware configuration for dual-spectral image acquisition and preprocessing is as follows: Visible light camera: resolution ≥1080P, used to capture the appearance features of chickens; Thermal imaging camera: resolution ≥384×288, temperature sensitivity ≤0.05℃; Synchronization control unit: Ensures synchronized acquisition time between the two cameras; Image registration module: Enables spatial alignment between visible light and thermal imaging images.

[0037] The dual-spectral image acquisition process is as follows: 1. Set up fixed collection points or mobile collection devices inside the chicken coop; 2. Simultaneously trigger dual cameras to capture images of the same scene; 3. Perform camera parameter calibration and image registration using a calibration board; 4. Establish a spatiotemporal consistent mapping relationship.

[0038] II. The identification strategy for chicken identification and key body part location is as follows: 1. Thermal imaging-based priority positioning: By utilizing the difference between the chicken's body temperature (normal 40.5~41.5℃) and the ambient temperature, the chicken's area can be quickly located by segmenting the temperature threshold, eliminating interference from environmental heat sources.

[0039] 2. Visible light profile confirmation: Based on the thermal imaging localization results, the corresponding region is extracted from the visible light image, and a deep learning model is used to finely segment the chicken outline and identify key parts such as the comb, eyes, and torso.

[0040] 3. Key component location model: The CNN network is based on an attention mechanism and is trained on tens of thousands of labeled images of Wenchang chickens. The focus is on optimizing the recognition of the unique single / double comb structure of Wenchang chickens.

[0041] III. The process for extracting multidimensional health features is as follows: 1. Health characteristics of rooster combs: Visible light characteristics are the statistics of hue (H) and saturation (S) in the HSV color space; the health threshold is redness of 0.72 to 1.0; abnormal manifestations are paleness, purplish discoloration, and swelling; thermal imaging characteristics are the mean and variance of temperature distribution in the comb area; the normal range is 38.5 to 40.5℃; abnormal manifestations are abnormal increases or decreases in temperature.

[0042] 2. Characteristics of healthy feathers: Thermal imaging features include the temperature gradient between the core and outer feathers of the torso. Excellent is >2.0℃ (good insulation of feathers); good is 1.0~2.0℃; average is 0.5~1.0℃; poor is <0.5℃ (sparse or damp feathers). Visible light features include analysis of feather luster and uniformity.

[0043] 3. Characteristics of eye health: Visible light characteristics include the color characteristics of the eyelids and conjunctiva. Healthy eyelids are pink and the conjunctiva is moist; abnormalities include redness, swelling, and increased secretions. Thermal imaging characteristics include the abnormal temperature index of the eye periphery, reflecting inflammation or infection.

[0044] 4. Behavioral health characteristics: Activity level is the distance traveled based on continuous frame analysis; abnormal postures include head-down, drooping wings, crouching, and other unusual postures; feeding and drinking behaviors are analyzed by frequency and duration.

[0045] V. Early warning and decision support are as follows: The early warning mechanism is as follows: 1. Threshold warning: Automatically triggered when health score falls below 60; 2. Trend Warning: A continuous downward trend in scores; 3. Feature Anomaly Warning: A single key feature is severely abnormal.

[0046] The warning information includes an abnormal individual identifier, specific abnormal characteristics, severity assessment, and recommended handling measures.

[0047] The hardware system of this invention is as follows: The dual-spectrum acquisition module includes a visible light camera array, a thermal imaging camera array, a synchronous trigger controller, and an intelligent gimbal system based on the DeepSORT tracking algorithm.

[0048] The data processing unit includes edge computing devices, graphics processing units (GPUs), and data storage devices.

[0049] The output display unit includes a large monitoring screen, a mobile terminal, and an audible and visual alarm device.

[0050] The software system of this invention is as follows: The image processing layer includes an image registration module, a feature extraction module, and a target tracking module.

[0051] The intelligent analysis layer includes a key component location model, a health assessment classification model, and a multi-feature fusion algorithm.

[0052] The application service layer includes a health status dashboard, an early warning management module, a report generation module, and a historical data tracing module.

[0053] In specific embodiments of the present invention, the following examples are provided: Example 1 illustrates a large-scale farm application, deployed at the entrance passage of the Wenchang chicken shed. Testing was conducted twice daily, morning and evening, with a daily testing volume exceeding 1000 birds. The effect was early detection of respiratory infection cases, with an early warning accuracy rate of 92.3%.

[0054] Example 2 illustrates quality control in a breeder chicken farm, with the application scenario being the assessment of reproductive health status as a key indicator for breeder chicken health screening. The result was a 95.8% accuracy rate in screening high-quality breeder chickens.

[0055] The present invention also has the following embodiments: Example 3: Fixed-point detection Dual-spectrum acquisition equipment is installed at key locations in the chicken house to achieve timed automatic data collection, individual chicken identification, health status assessment, and automatic alarm for abnormalities.

[0056] Example 4: Mobile Inspection and Detection The inspection robot, equipped with a dual-spectrum camera, enables full-coverage inspection, tracking of key individuals, dynamic behavior analysis, and real-time health monitoring.

[0057] Example 5: Cloud-based collaborative detection The edge device + cloud analytics architecture enables rapid local detection, in-depth cloud analysis, multi-farm data aggregation, and health trend prediction.

[0058] This invention focuses on data and standards to protect data assets. The image annotation dataset of tens of thousands of chickens includes more than 5,000 specific annotations for Wenchang chickens. All annotations have undergone a three-level quality review process, including initial annotation by farmers, review by veterinary experts, and algorithm consistency verification, with an annotation consistency rate of more than 95%.

[0059] The training data for the health assessment classification model includes more than 1,000 clinically diagnosed unhealthy Wenchang chicken samples, enabling the model to identify characteristic patterns associated with anemia, hypoxia, fever, and early respiratory infections.

[0060] The core innovations of this invention are as follows: Dual-spectral collaborative detection, leveraging the complementary advantages of visible light and thermal imaging; The logic of identifying before judging improves detection accuracy and efficiency; Multi-feature weight fusion, scientific weight allocation based on a large amount of experimental data; A model specifically designed for Wenchang chickens, featuring optimized algorithms tailored to their physiological characteristics.

[0061] This invention improves detection accuracy by over 25% compared to single-spectrum detection, provides timely early warnings (detecting health abnormalities 24-48 hours in advance), automates the process (reducing manual inspections by 80%), and standardizes output by unifying health assessment standards and eliminating subjective differences. Through the innovative application of dual-spectrum imaging technology, this invention solves key technical challenges in the health detection of Wenchang chickens, providing a new technical approach and implementation plan for livestock and poultry health monitoring.

[0062] For the specific implementation scheme of this embodiment, please refer to the relevant descriptions in the above embodiments, which will not be repeated here.

[0063] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0064] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0065] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0066] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution device. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0067] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The corresponding program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0068] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0069] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0070] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0071] This invention employs a method, device, processor, and computer-readable storage medium for health detection of Wenchang chickens based on dual-spectral imaging. By simultaneously acquiring visible light and thermal images, it automatically locates key areas such as the comb, feathers, and eyes, extracts multi-dimensional health features, and achieves accurate health status judgment and early disease warning based on a Wenchang chicken-specific health assessment model. This invention solves the problems of low efficiency and poor accuracy of traditional manual detection, providing effective technical support for the scientific breeding of Wenchang chickens.

[0072] In this specification, the invention has been described with reference to specific embodiments thereof. However, it will be apparent that various modifications and variations can be made without departing from the spirit and scope of the invention. Therefore, the specification and drawings should be considered illustrative rather than restrictive.

Claims

1. A method for detecting the health of Wenchang chicken based on dual-spectrum imaging, characterized in that, The method comprises the following steps: (1) Collecting the visible light image and the thermal imaging image of the Wenchang chicken under the same space-time reference by synchronously calibrating the visible light camera and the thermal imaging camera; (2) Inputting the dual-spectrum image into the key part positioning model trained by the labeled data set, and synchronously outputting the boundary box of the comb, the eye and the main feather area; (3) Extracting the multi-dimensional health feature vector in parallel based on the boundary box; (4) Inputting the multi-dimensional health feature vector into the health assessment classification model trained by the labeled data set, and outputting the quantitative health status score and the corresponding health grade; (5) If the health status score is lower than the preset threshold, generating and outputting the health warning information containing the specific abnormal feature dimension.

2. The method for detecting the health of Wenchang chicken based on dual-spectrum imaging according to claim 1, characterized in that, The step (3) specifically comprises the following steps: (3.1) Extracting the hue and saturation statistics of the HSV color space based on the visible light image, and extracting the temperature mean and distribution variance of the comb area based on the thermal imaging image, to obtain the comb feature vector; wherein the normal range of the body surface temperature of the comb area of the healthy Wenchang chicken is 40.5℃-41.5℃; (3.2) Extracting the body surface temperature of the chicken based on the thermal imaging image, calculating the temperature gradient value of the core area of the trunk and the peripheral feather covering area, to obtain the feather feature vector; (3.3) Extracting the color features of the eyelid and conjunctiva area based on the visible light image, combining the body surface temperature information provided by the thermal imaging image, and improving the recognition accuracy through image fusion technology, to obtain the eye feature vector.

3. The method for detecting the health of Wenchang chicken based on dual-spectrum imaging according to claim 1, characterized in that, The key part positioning model of the step (2) is a convolutional neural network based on attention mechanism, which is trained by the boundary box labeling data of the comb, the eye and the trunk in the labeled data set, and preferentially focuses on and locates the single comb or double comb structure unique to the Wenchang chicken.

4. The method for realizing healthy detection of Wenchang chicken based on dual-spectrum imaging according to claim 1, characterized in that, The health assessment classification model of the step (4) adopts a weighted multi-feature fusion algorithm, and the weights of the feature vectors are determined based on the logical regression feature importance analysis of the labeled data set.

5. The method for realizing Wenchong chicken health detection based on dual-spectrum imaging according to claim 4, characterized in that, The comb feature weight in the weight of each feature vector is 0.28-0.32, the feather feature weight is 0.23-0.27, the eye feature weight is 0.18-0.22, the body temperature feature weight is 0.13-0.17, and the behavior feature weight is 0.08-0.

12. The system comprises:

6. A system for realizing health detection of Wenchang chicken based on dual-spectrum imaging, characterized in that, A dual-spectrum image acquisition module for acquiring the visible light image and the thermal imaging image of the Wenchang chicken under the same space-time reference; A data processing and analysis module connected with the dual-spectrum image acquisition module, for loading and running the key part positioning model and the health assessment classification model; A health decision and output module connected with the data processing and analysis module, for judging the health status based on multi-feature fusion, generating the health status score, the health grade and the health warning information, and providing a human-computer interaction interface to generate a health report and an abnormal warning. The dual-spectrum image acquisition module comprises a visible light camera, a thermal imaging camera and an image registration unit, the image registration unit is connected with the visible light camera and the thermal imaging camera, the visible light camera and the thermal imaging camera are hardware-synchronized, and the image registration unit is used for ensuring the space-time consistency.

7. The system for realizing healthy detection of Wenchang chicken based on dual-spectrum imaging according to claim 6, characterized in that, ​ 8. The system for realizing healthy detection of Wenchang chicken based on dual-spectrum imaging according to claim 6, characterized in that, The data processing and analysis module comprises: a first processing unit for loading and running a key position positioning model; a second processing unit for loading and running a health assessment classification model; a database for storing image annotation data sets of tens of thousands of chickens.

9. The system for realizing healthy detection of Wenchang chicken based on dual-spectrum imaging according to claim 6, characterized in that, The system further comprises an intelligent pan-tilt control module and an automatic pan-tilt, the dual-spectrum image acquisition module is installed on the automatic pan-tilt, the intelligent pan-tilt control module is connected with the automatic pan-tilt, the intelligent pan-tilt control module tracks the movement of the chickens in real time based on a DeepSORT multi-target tracking algorithm, and controls the dual-spectrum image acquisition module to automatically adjust a shooting angle and collect images of key positions.

10. The system for realizing healthy detection of Wenchang chicken based on dual-spectrum imaging according to claim 6, characterized in that, The human-computer interaction interface of the health decision and output module comprises a health status board, which displays the health score distribution of the chicken population, the list of abnormal individuals and the historical health trend in the form of a chart.

11. A device for realizing health detection of Wenchang chicken based on dual-spectrum imaging, characterized in that, The device comprises: a processor configured to execute computer executable instructions; a memory storing one or more computer executable instructions, which, when executed by the processor, implement the steps of the method for realizing health detection of Wenchang chicken based on dual-spectrum imaging according to any one of claims 1 to 5. 12.A processor for implementing healthy detection of Wenchang chicken based on dual-spectrum imaging, characterized in that, The processor is configured to execute computer executable instructions, which, when executed by the processor, implement the steps of the method for realizing health detection of Wenchang chicken based on dual-spectrum imaging according to any one of claims 1 to 5.

13. A computer-readable storage medium, characterized in that, A computer program is stored thereon, which can be executed by the processor to implement the steps of the method for realizing health detection of Wenchang chicken based on dual-spectrum imaging according to any one of claims 1 to 5.

Images