Image recognition-based poultry health early warning system and method

Abstract

The present application provides a poultry health early warning system and method based on image recognition, and relates to the field of health early warning.The present application comprises a central processing unit, a communication module and a storage, the central processing unit is electrically connected with an outdoor information collection system, the outdoor information collection system comprises an outdoor camera module, an outdoor sound amplification module, a cage feeding execution module, a cage camera module and a far infrared thermal imager module;the poultry growth model changes in real time along with the growth of poultry, and the poultry growth model is stored in the report storage module at the same time, and the thermal imaging image data of the same day is intercepted at several time points, and the temperature data and the poultry growth model are stored and updated synchronously, so that the health archives of each poultry can be updated daily.

Description

Technical Field

[0001] This invention relates to a poultry health early warning system, specifically an image recognition-based poultry health early warning system and method, belonging to the field of health early warning technology. Background Technology

[0002] Poultry refers to birds raised in captivity. Humans raise them not only for their nutritious meat and eggs, but also for their highly economical feathers and droppings. Among the large family of poultry, chickens are one of the most commonly raised by humans.

[0003] With the continuous development of the economy and the continuous improvement of people's living standards, people's demand for chickens is increasing day by day, and the requirements for the quality of chicken meat are also getting higher and higher. Therefore, so-called ecological farming methods have emerged. For example, the existing patent application number: 202210321713.2 provides a method for ecological free-range poultry farming, which includes using a wheeled grazing vehicle as a mobile livestock house for outdoor poultry farming to carry out ecological free-range poultry farming; the area where the wheeled grazing vehicle moves is open grassland or relatively flat artificial forest; the poultry breed is a breed with excellent environmental resistance; the wheeled grazing vehicle is divided into upper and lower layers and is equipped with: fermentation bed, heat lamp, fan, ultraviolet lamp, air conditioner, disinfection facilities, and incineration.

[0004] However, when raising poultry and livestock outdoors in the mountains and forests, especially on a large scale, it is difficult for staff to detect whether the scattered poultry and livestock are sick, which can easily lead to the outbreak of diseases and cause a lot of economic losses. Summary of the Invention

[0005] To solve the above problems, the present invention is implemented through the following technical solution: a poultry health early warning system and method based on image recognition, including a central processing unit, a communication module and a storage device, wherein the central processing unit is electrically connected to an outdoor information collection system, and the outdoor information collection system includes an outdoor camera module, an outdoor loudspeaker module, a cage feeding execution module, a cage camera module and a far-infrared thermal imager module. The central processing unit stores a health early warning system, which includes an image processing module, a lesion locking module, a behavior judgment module, a cause locking module, an image analysis module, an animal locking module, a plant distribution recording module, a trajectory simulation module, a vulnerability locking module, and a summary judgment module.

[0006] Preferably, the communication module is used for data transmission between the outdoor information collection system and the central processing unit, and the communication module includes network communication equipment and cable module.

[0007] Preferably, the storage device is used to store various data generated by the outdoor information collection system and the health early warning system, and the storage device includes a disease model storage module, a report storage module, and an ecological information storage module.

[0008] Preferably, both the outdoor camera module and the cage camera module are equipped with a micro storage unit.

[0009] Preferably, the outdoor camera module consists of multiple outdoor cameras, and the outdoor loudspeaker module consists of multiple loudspeakers. The outdoor loudspeaker module plays voice messages to drive away birds, carnivores, or animals that cause significant damage to the breeding environment that are common near the outdoor farm.

[0010] Preferably, the cage feeding execution module includes multiple feeding devices and sound generation devices, the cage camera module consists of multiple indoor cameras, and the far-infrared thermal imager module consists of multiple far-infrared thermal imagers.

[0011] Preferably, the disease model storage module stores symptom models of common diseases in poultry and livestock, the report storage module stores the results and reports obtained during the operation of the health early warning system, and the ecological information storage module stores models of common animals or plants around outdoor farms.

[0012] The usage of the poultry health early warning system includes the following steps: Step 1: Conduct reflex conditioning training when poultry and livestock are young to reduce the physical and mental effort required for later breeding. Step 2: During the day, poultry and livestock move around inside the outdoor farm. The outdoor camera module captures outdoor image data of the poultry and livestock's activities. The behavior judgment module identifies and judges the images. Based on the images, the behavior judgment module judges the poultry and livestock to be abnormal. Based on the outdoor image data collected by the outdoor information collection system, the staff locates the poultry and livestock that have lost their ability to move or have even died, and quickly retrieves them. Step 3: The central processing unit controls the sound-generating equipment to generate sound at regular intervals to attract poultry and livestock into the nearby cages. The central processing unit controls the cage camera module and far-infrared thermal imager module to work. Multiple indoor cameras and multiple far-infrared thermal imagers work to collect image data and thermal image data of poultry and livestock in the cages. Step 4: The image processing module establishes a poultry growth model based on the external characteristics of poultry at multiple time points from outdoor image data and poultry and livestock image data inside cages. The poultry and livestock growth model changes in real time with the growth of poultry and livestock, and is synchronized to the report storage module for storage, updating the health record of each poultry and livestock daily. Step 5: The lesion locking module determines whether the poultry or livestock is sick based on the external symptom models of common poultry and livestock diseases stored in the disease model storage module, the external characteristics of poultry and livestock in the poultry and livestock growth model stored in the report storage module, and the temperature data at several time points captured from the thermal imaging image data. Step Six: After determining that the poultry and livestock are sick, the image processing module can quickly locate the sick poultry and livestock based on the poultry and livestock growth model, and then locate their location in real time based on the image. Staff can cooperate with the personnel in charge of the central processor to accurately capture the sick poultry and livestock.

[0013] This invention provides a poultry health early warning system and method based on image recognition, which has the following beneficial effects: 1. This image recognition-based poultry health early warning system and method enables the poultry growth model to change in real time according to the growth status of poultry. The poultry growth model is synchronously stored in the report storage module. At the same time, the thermal imaging image data of the day is extracted and the temperature data at several time points are synchronously stored and updated with the poultry growth model, which can update the health record of each poultry daily. In addition, before the outdoor image data that the outdoor information collection system transmits to the central processor in real time through the communication module enters the storage, the behavior judgment module recognizes and judges the image. When poultry is accidentally injured, sick, or loses its ability to move or even dies, the behavior judgment module judges the poultry abnormality based on the image. Then, the staff locates the incapacitated or dead poultry based on the outdoor image data collected by the outdoor information collection system and quickly retrieves it, avoiding the occurrence of odor, bacterial growth or even the spread of disease due to the accidental death of poultry.

[0014] 2. This image recognition-based poultry health early warning system and method generates reports with information such as poultry and livestock disease determination, external symptoms, body temperature, and causes of illness through a summary judgment module. While the report is updated and stored internally in the report storage module, it is also displayed on the central processing unit's screen, facilitating targeted prevention and control by staff based on the poultry and livestock's condition and reducing economic losses. The disease cause identification module identifies ingested plants and animals from the disease model storage module, ecological information storage module, and online data. The summary judgment module generates potential disease reports based on the specific effects and components of the ingested plants and animals, providing staff with a reference for disease cause identification and reducing the likelihood of undetermined causes in outdoor forest farming, thus quickly mitigating losses.

[0015] 3. The image recognition-based poultry health early warning system and method, through a summary judgment module, generates a treatment method for the plant based on the plant distribution map and the ecological information storage module's information on whether the plant is a protected plant and the corresponding management method. This avoids situations where staff, due to a lack of understanding of the plant, inadvertently kill large numbers of protected plants, thus reducing unexpected losses. Attached Figure Description

[0016] Figure 1 This is an overall schematic diagram of the present invention; Figure 2This is a schematic diagram of the summary judgment module of the present invention; Figure 3 This is a schematic diagram of the storage device of the present invention; Figure 4 This is a schematic diagram of the communication module of the present invention; Figure 5 This is a schematic diagram of the outdoor camera module of the present invention.

[0017] Explanation of reference numerals in the attached diagram: 1. Central Processing Unit; 2. Outdoor Information Collection System; 201. Outdoor Camera Module; 202. Outdoor Amplification Module; 203. Cage Feeding Execution Module; 204. Cage Camera Module; 205. Far-Infrared Thermal Imaging Module; 3. Communication Module; 301. Network Communication Equipment; 302. Cable Module; 4. Health Early Warning System; 401. Image Processing Module; 402. Lesion Locking Module; 403. Behavior Judgment Module; 404. Cause Locking Module; 405. Image Analysis Module; 406. Animal Locking Module; 407. Plant Distribution Recording Module; 408. Trajectory Simulation Module; 409. Vulnerability Locking Module; 410. Summary Judgment Module; 5. Storage Module; 501. Disease Model Storage Module; 502. Report Storage Module; 503. Ecological Information Storage Module. Detailed Implementation

[0018] This invention provides a poultry health early warning system and method based on image recognition.

[0019] Please see Figure 1 , Figure 2 , Figure 3 , Figure 4 and Figure 5 The system includes a central processing unit (1), a communication module (3), and a storage unit (5). The central processing unit (1) is electrically connected to an outdoor information collection system (2). The outdoor information collection system (2) includes an outdoor camera module (201), an outdoor loudspeaker module (202), a cage feeding execution module (203), a cage camera module (204), and a far-infrared thermal imager module (205). The central processing unit (1) stores a health early warning system (4). The health early warning system (4) includes an image processing module (401), a lesion locking module (402), a behavior judgment module (403), a cause locking module (404), an image analysis module (405), an animal locking module (406), a plant distribution recording module (407), a trajectory simulation module (408), a vulnerability locking module (409), and a summary judgment module (410).

[0020] Specifically, the communication module (3) is used for data transmission between the outdoor information collection system (2) and the central processing unit (1). The communication module (3) consists of a network communication device (301) and a cable module (302). The network communication device (301) meets the needs of wireless communication and realizes real-time data transmission. The cable module (302) meets the needs of wired communication. Miniature storage elements are provided on the outdoor camera module (201) and the cage camera module (204) used to collect information in the outdoor information collection system (2). The miniature storage elements are used to temporarily store the data collected by the outdoor camera module (201) and the cage camera module (204). After each unit of time, the communication module (3) transmits the data inside the micro storage element to the storage (5) in the central processing unit (1) via the cable module (302). The data transmitted by the cable module (302) has a higher priority in the storage (5). The data transmitted by the cable module (302) covers the same data already stored in the storage (5). The cable module (302) and the network communication device (301) cooperate with each other to avoid the loss of data collected by the outdoor camera module (201) installed in the outdoor mountains and forests, reduce the possibility of data loss, adapt to the situation where the network signal is poor due to sudden rain, strong wind, heavy snow, etc., and ensure that the data collected by the outdoor information collection system (2) can be completely transmitted into the central processing unit (1) to avoid data loss due to sudden situations and ensure the stable operation of the outdoor information collection system (2).

[0021] The outdoor information collection system (2) installed outdoors consists of an outdoor camera module (201), an outdoor loudspeaker module (202), a cage feeding execution module (203), a cage camera module (204), and a far-infrared thermal imager module (205). The outdoor camera module (201) consists of multiple outdoor cameras, which are installed in various parts of the outdoor farm to meet the health monitoring data collection work of free-range poultry and livestock, and are used to collect environmental data and poultry and livestock activity image data of the outdoor farm. The outdoor loudspeaker module (202) consists of multiple loudspeakers, which are installed in multiple locations in the outdoor farm. The loudspeakers are used to play multiple common voices stored in the storage unit (5). The outdoor loudspeaker module (202) plays voices to drive away common birds, carnivores or animals that cause great damage to the farm environment near the outdoor farm, thereby reducing the possibility of livestock being harmed by animals. The cage feeding execution module (203) includes multiple feeding devices and sound-making devices. The feeding devices and sound-making devices correspond one-to-one. Cages are built in multiple locations in the outdoor farm. Each cage is equipped with a feeding device and a sound-making device. Reflex condition training can be carried out when poultry and livestock are young, reducing the physical and mental effort required for later poultry and livestock breeding. The sound produced by the sound-making device is the sound for reflex training of poultry and livestock. The central processor (1) controls the sound-making device to work and produce sound at regular intervals. After attracting poultry and livestock into the nearby cages, the central processor (1) controls the feeding device to work and feed a certain amount of feed to the poultry and livestock. Alternatively, the central processor (1) controls the cages to work and close, so that the poultry and livestock can stay in the cages overnight or avoid rain, reducing the safety hazards of poultry and livestock breeding. The cage camera module (204) consists of multiple indoor cameras, which are installed inside multiple cages. The far-infrared thermal imaging module (205) consists of multiple far-infrared thermal imaging cameras, which are installed inside multiple cages. Before diseases cause changes in tissue structure and morphology, abnormal cell metabolism occurs, and the temperature of the organism changes. The temperature, the shape of the temperature field, and the size of the temperature difference can reflect the location, nature, and degree of the disease. Far-infrared thermal imaging technology is based on this phenomenon to detect various diseases such as inflammation, tumors, and vascular diseases. Since far-infrared thermal imaging technology does not harm the organism, does not cause any pollution to the environment, and is simple and economical, far-infrared thermal imaging cameras can be fully applied to large-scale poultry and livestock breeding production.

[0022] In this application, the storage device (5) used for storing data is divided into three storage areas: a disease model storage module (501), a report storage module (502), and an ecological information storage module (503). The disease model storage module (501) stores the external symptom models of common diseases of poultry and livestock, which are used to determine whether poultry and livestock are sick. The report storage module (502) stores the results and reports obtained during the operation of the health early warning system (4) for easy retrieval later. The ecological information storage module (503) is used to store common animal or plant models around outdoor farms, especially models of protected animals or protected plants. It also stores models of plants planted and animals raised inside the farm for the purpose of raising poultry and livestock.

[0023] When the poultry and livestock health early warning system, which consists of a central processor (1), an outdoor information collection system (2), a communication module (3), a health early warning system (4), and a storage device (5), is in operation, during the daytime, poultry and livestock are active in the outdoor farm, and the outdoor camera module (201) works to capture outdoor image data of poultry and livestock activities; when poultry and livestock return to the cages to spend the night, the central processor (1) controls the cage camera module (204) and the far-infrared thermal imager module (205) to work, and multiple indoor cameras and multiple far-infrared thermal imagers work to collect image data of poultry and livestock in the cages and thermal image data of poultry and livestock; Outdoor image data, caged livestock image data, and thermal imaging image data collected by the outdoor information collection system (2) are transmitted to the central processing unit (1) through the communication module (3). The central processing unit (1) stores and processes the outdoor image data, caged livestock image data, and thermal imaging image data. The image processing module (401) establishes a livestock growth model based on the external characteristics of livestock at multiple time points from outdoor image data and livestock images inside cages. This work is performed once a day, so that the livestock growth model changes in real time with the growth of livestock. The livestock growth model is also stored in the report storage module (502) in sync. At the same time, the thermal imaging image data of the day is extracted and the temperature data at several time points are stored and updated in sync with the livestock growth model, so that the health record of each livestock can be updated daily. The outdoor information collection system (2) transmits the data in real time through the communication module (3). Before the outdoor image data in the central processing unit (1) is stored in the storage unit (5), the behavior judgment module (403) identifies and judges the image. When livestock are injured, sick, or lose their mobility or even die, the behavior judgment module (403) judges the livestock as abnormal based on the image. Then, the staff locks the location of the livestock that has lost its mobility or even died based on the outdoor image data collected by the outdoor information collection system (2) and quickly retrieves them to avoid the occurrence of odor, germ growth, or even disease spread caused by the accidental death of livestock.

[0024] Every one to three days, the lesion locking module (402) in the health early warning system (4) works once. The lesion locking module (402) determines whether the poultry or livestock is sick based on the external symptom model of common poultry and livestock diseases stored in the disease model storage module (501) and the external characteristics of poultry and livestock in the poultry and livestock growth model stored in the report storage module (502) and the temperature data at several time points when the thermal imaging image data is captured. When it is determined that the poultry or livestock is sick, the outdoor camera module (201) and the cage camera module (204) in the outdoor information collection system (2) are set so that the image processing module (401) can quickly lock the location of the sick poultry or livestock based on the poultry and livestock growth model, and then lock its location in real time based on the image. The staff can cooperate with the staff in charge of the central processor (1) to accurately capture the sick poultry or livestock, and avoid the situation of slow or wrong capture of the sick free-range poultry or livestock when the external symptoms of poultry and livestock are not obvious, so as to facilitate the staff to quickly deal with the sick poultry or livestock.

[0025] After the lesion locking module (402) identifies the disease of poultry and livestock, the cause locking module (404) determines the cause of the disease based on the disease information stored in the disease model storage module (501). The summary judgment module (410) receives the judgment results from the lesion locking module (402) and the cause locking module (404), as well as the poultry and livestock external features and body temperature data captured in the image processing module (401). The summary judgment module (410) then generates a report containing information such as the poultry and livestock disease determination, external symptoms, body temperature, and cause of disease. While the report is updated and stored in the report storage module (502), it is also displayed on the display interface of the central processing unit (1), which makes it convenient for staff to carry out targeted prevention and control based on the poultry and livestock disease situation and reduce economic losses.

[0026] When the disease cause locking module (404) determines that the cause of illness in more than 10 poultry and livestock is due to eating a certain plant, the image analysis module (405) retrieves the model of each time period in the growth cycle of the plant from the ecological information storage module (503). The image analysis module (405) locks the growth location of the plant from the outdoor image data captured by the outdoor camera module (201) based on the plant model. The plant growth location is recorded by the plant distribution recording module (407). Finally, the distribution map of plants harmful to the growth of poultry and livestock in the farm generated by the plant distribution recording module (407) is obtained. The distribution map is sent to the summary judgment module (410). The summary judgment module (410) generates a treatment method for the plant based on the plant distribution map and whether the plant is a protected plant and the treatment method stored in the ecological information storage module (503). This avoids the situation where staff kill a large number of protected plants because they do not understand the plants, thus reducing unexpected losses.

[0027] When the cause-identification module (404) fails to identify the cause of illness and more than 30 poultry or livestock die or become seriously ill due to the same disease, the behavior judgment module (403) determines the date of illness based on the daily changes in the poultry or livestock growth model. Subsequently, the image analysis module (405) identifies images from the outdoor image data stored in the storage unit (5) within two days prior to the date of illness of the poultry or livestock. The trajectory simulation module (408) simulates the movement trajectory of the poultry or livestock within two days based on the images, obtaining a reference trajectory. Similarly, 10 reference trajectories are obtained based on the movement trajectories of 10 poultry or livestock with the same disease. Subsequently, the behavior judgment module (403) The system identifies the animals and plants consumed by the livestock (excluding feed, common plants, and common insects) from 10 reference trajectories, and identifies the animals and plants that the 10 livestock have consumed together. The system obtains ingestion data. The disease cause identification module (404) then identifies the animals and plants in the ingestion data from the disease model storage module (501), the ecological information storage module (503), and the network. The summary judgment module (410) generates a possible disease report based on the specific effects and components of the animals and plants in the ingestion data. The possible disease report provides staff with a reference for finding the cause of the disease, reducing the occurrence of situations where the cause of the disease cannot be determined in outdoor mountain forest breeding, and quickly stopping losses.

[0028] Furthermore, the ecological information storage module (503) stores the local animal distribution and the living habits of common carnivores and omnivores. The animal locking module (406) works during the most active time of day for carnivores and omnivores. The animal locking module (406) judges whether carnivores and omnivores have entered the breeding site from the outdoor image data captured and transmitted by the outdoor camera module (201), and locks the location of the carnivores and omnivores. The presence of carnivores and omnivores and their location are displayed on the central processing unit (1). While the central processing unit (1) issues an alarm to remind the staff, the central processing unit (1) controls the storage device (5) to transmit relevant sounds to the loudspeakers in the outdoor loudspeaker module (202) near the location of the carnivores and omnivores. The outdoor loudspeaker module (202) amplifies and plays the sound. At the same time, it calculates and controls the loudspeakers in the relevant locations based on the location of the carnivores and omnivores to drive the carnivores and omnivores out of the breeding site. When the outdoor loudspeaker module (202) fails to affect carnivores and omnivores, the summary judgment module (410) in the health warning system (4) judges the movement of carnivores and omnivores based on the behavior judgment module (403), generates a report of failure to automatically drive away carnivores and omnivores, and the summary judgment module (410) judges the animal's aggressiveness, attack method, and whether it is a protected animal based on the information stored in the ecological information storage module (503) to generate treatment suggestions, so as to avoid the situation where staff rashly drive away carnivores and omnivores and are harmed, and also violate the law, thereby reducing the accidental losses of poultry and livestock raised in outdoor mountain forests.

[0029] When the animal locking module (406) locks the carnivorous and omnivorous animals, the central processing unit (1) controls the outdoor camera module (201) and the cage camera module (204) in the outdoor information collection system (2) to lock and photograph the animal. The animal image is acquired by the trajectory simulation module (408), which generates an image of the animal's movement trajectory. After the animal is processed, the vulnerability locking module (409) locks the location of the nearby building where the animal may enter based on its movement trajectory and the breeding site construction data stored in the storage (5), so as to ensure that the staff can quickly repair the building of the breeding site and reduce losses.

Claims

1. A poultry health early warning system based on image recognition, comprising a central processing unit (1), a communication module (3), and a storage device (5), characterized in that: The central processing unit (1) is electrically connected to an outdoor information collection system (2), which includes an outdoor camera module (201), an outdoor loudspeaker module (202), a cage feeding execution module (203), a cage camera module (204), and a far-infrared thermal imager module (205). The central processing unit (1) stores a health early warning system (4), which includes an image processing module (401), a lesion locking module (402), a behavior judgment module (403), a cause locking module (404), an image analysis module (405), an animal locking module (406), a plant distribution recording module (407), a trajectory simulation module (408), a vulnerability locking module (409), and a summary judgment module (410). The image processing module (401) establishes a livestock growth model; the lesion locking module (402) determines whether livestock are sick; the behavior judgment module (403) determines livestock abnormalities; the cause locking module (404) determines the cause of livestock illness; the image analysis module (405) locks the location of plant growth; the animal locking module (406) determines whether carnivores and omnivores have entered the breeding area and locks the location of carnivores and omnivores; the plant distribution recording module (407) generates a distribution map of plants harmful to livestock growth in the breeding area; the trajectory simulation module (408) generates animal movement trajectory images; the vulnerability locking module (409) locks the location of nearby buildings where animals may enter; the summary judgment module (410) generates plant treatment methods and animal treatment suggestions. The storage unit (5) is used to store various data generated by the outdoor information collection system (2) and the health early warning system (4). The storage unit (5) includes a disease model storage module (501), a report storage module (502) and an ecological information storage module (503). Both the outdoor camera module (201) and the cage camera module (204) are equipped with a micro storage unit; The outdoor camera module (201) consists of multiple outdoor cameras, and the outdoor loudspeaker module (202) consists of multiple loudspeakers. The outdoor loudspeaker module (202) plays voice messages to drive away birds and carnivores that are common near the outdoor farm.

2. The poultry health early warning system based on image recognition according to claim 1, characterized in that: The communication module (3) is used for data transmission between the outdoor information collection system (2) and the central processing unit (1). The communication module (3) includes a network communication device (301) and a cable module (302).

3. The poultry health early warning system based on image recognition according to claim 1, characterized in that: The cage feeding execution module (203) includes multiple feeding devices and sound generation devices, the cage camera module (204) consists of multiple indoor cameras, and the far-infrared thermal imager module (205) consists of multiple far-infrared thermal imagers.

4. The poultry health early warning system based on image recognition according to claim 1, characterized in that: The disease model storage module (501) stores the external symptom models of common diseases of poultry and livestock. The report storage module (502) is used to store the results and reports obtained during the operation of the health early warning system (4). The ecological information storage module (503) is used to store common animal or plant models around outdoor farms.

5. A method for using a poultry health early warning system, based on the image recognition-based poultry health early warning system according to any one of claims 1-4, characterized in that: Includes the following steps: Step 1: Conduct reflex conditioning training when poultry and livestock are young to reduce the physical and mental effort required for later poultry and livestock farming. Step 2: During the day, poultry and livestock are active in the outdoor farm. The outdoor camera module (201) works to capture outdoor image data of poultry and livestock activities. The behavior judgment module (403) identifies and judges the images. The behavior judgment module (403) judges the poultry and livestock to be abnormal based on the images. The staff locks the location of poultry and livestock that have lost the ability to move or even died based on the outdoor image data collected by the outdoor information collection system (2) and quickly retrieves them. Step 3: The central processing unit (1) controls the sound-generating equipment to produce sound at regular intervals to attract poultry and livestock into the nearby cages. The central processing unit (1) controls the cage camera module (204) and the far-infrared thermal imager module (205) to work. Multiple indoor cameras and multiple far-infrared thermal imagers work to collect image data of poultry and livestock in the cages and thermal image data of poultry and livestock. Step 4: The image processing module (401) establishes a poultry growth model based on the external features of poultry at multiple time points from outdoor image data and poultry and livestock image data in cages. The poultry growth model changes in real time with the growth of poultry and livestock, and the poultry growth model is synchronized to the report storage module (502) for storage. The health record of each poultry and livestock is updated daily. Step 5: The lesion locking module (402) determines whether the poultry or livestock is sick based on the symptom models of common poultry and livestock diseases stored in the disease model storage module (501), the symptom models of poultry and livestock growth models stored in the report storage module (502), and the temperature data at several time points where thermal imaging data is extracted. Step 6: After determining that the poultry and livestock are in a sick state, the image processing module (401) can quickly locate the sick poultry and livestock based on the poultry and livestock growth model, and then locate their location in real time based on the image. The staff and the personnel in charge of the central processor (1) cooperate to accurately capture the sick poultry and livestock.

Images