Prediction method and system of silkworm diseases based on silkworm rearing environment

Abstract

The present invention discloses a method and system for predicting silkworm diseases based on a silkworm rearing environment. The present invention relates to the technical field of prediction methods. The system determines the silkworm rearing environment based on environmental detection of the silkworm environmental area, determines abnormal areas in the silkworm rearing room based on the silkworm rearing environment and the silkworm's activity trajectory, and marks abnormal silkworm images, thereby ensuring accurate detection of abnormal areas in the silkworm rearing room and the accuracy of abnormal silkworm images. Therefore, abnormal skin features and abnormal posture features of the silkworm are determined based on the abnormal silkworm images, and the morphology of the silkworm's excrement is determined along the detection of the silkworm's activity trajectory. The type of silkworm disease is predicted based on the abnormal skin features, abnormal posture features, and morphology of the silkworm's excrement, and the environmental control logic of the silkworm rearing room is triggered based on the type of silkworm disease and the silkworm rearing environment, thereby achieving accurate prediction of the type of silkworm disease and timely optimization of the silkworm rearing environment in the silkworm rearing room.

Description

Technical Field

[0001] The present invention relates to the technical field of prediction methods, and in particular to a method and system for predicting silkworm diseases based on a silkworm breeding environment. Background Art

[0002] With the development of science and technology, silkworms are a kind of animal and are cultivated in silkworm rearing rooms. At this time, the silkworms move in the silkworm rearing rooms and are affected by the silkworm rearing environment in real time. In the existing technology, the silkworm rearing environment and the morphology of the silkworms are collected, and the types of silkworm diseases of the silkworms are predicted based on the silkworm rearing environment and the morphology of the silkworms. However, the influence of the abnormal characteristics of the silkworms is not taken into account, and the accurate prediction of the types of silkworm diseases of the silkworms cannot be achieved. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art. The present invention provides a method and system for predicting silkworm diseases based on a silkworm rearing environment.

[0004] The embodiment of the present invention provides a method for predicting silkworm diseases based on a silkworm rearing environment, comprising:

[0005] Real-time monitoring of the silkworm rearing room and collection of the silkworm rearing distribution map of the silkworm rearing room;

[0006] Determine the silkworm activity trajectory and silkworm environment area based on the identification of silkworm distribution map;

[0007] Determine the silkworm rearing environment based on the environmental detection of the silkworm environmental area, determine the abnormal area of the silkworm rearing room according to the silkworm rearing environment and the silkworm activity trajectory, and mark the abnormal images of the silkworms;

[0008] Determine the abnormal skin features and posture features of the silkworm based on the abnormal image of the silkworm, and determine the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory;

[0009] The type of silkworm disease is predicted based on the abnormal skin characteristics, abnormal posture characteristics and excrement morphology of the silkworm, and the environmental control logic of the silkworm rearing room is triggered according to the type of silkworm disease and the silkworm rearing environment.

[0010] An embodiment of the present invention provides a system for predicting silkworm diseases based on a sericulture environment. The system for predicting silkworm diseases based on a sericulture environment is applied to the above-mentioned method for predicting silkworm diseases based on a sericulture environment. The system for predicting silkworm diseases based on a sericulture environment includes:

[0011] The silkworm breeding distribution map module is used to monitor the silkworm breeding room in real time and collect the silkworm breeding distribution map of the silkworm breeding room;

[0012] The silkworm identification module is used to determine the silkworm activity trajectory and silkworm environment area based on the identification of the silkworm distribution map;

[0013] An anomaly detection module is used to determine the silkworm rearing environment based on the environmental detection of the silkworm environmental area, determine the abnormal area of the silkworm rearing room according to the silkworm rearing environment and the silkworm activity trajectory, and mark the abnormal images of the silkworm;

[0014] A morphology detection module is used to determine abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and to determine the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory;

[0015] The prediction module is used to predict the type of silkworm disease based on the abnormal skin characteristics, abnormal posture characteristics and excrement morphology of the silkworm, and trigger the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] In an embodiment of the present invention, through the method in the embodiment of the present invention, the sericulture environment of the silkworms is determined based on the environmental detection of the sericulture environment area, and the abnormal area of the sericulture room is determined according to the sericulture environment of the silkworms and the activity trajectory of the silkworms, and the abnormal image of the sericulture room is marked, which is compatible with the overall consideration of the sericulture environment of the silkworms and the activity trajectory of the silkworms, ensures the accurate detection of the abnormal area of the sericulture room, and ensures the accuracy of the abnormal image of the silkworms.

[0018] Therefore, the abnormal skin features and abnormal posture features of the silkworm are determined based on the abnormal images of the silkworm, and the morphology of the silkworm's excrement is determined along the detection of the silkworm's activity trajectory; the type of silkworm disease of the silkworm is predicted based on the abnormal skin features, abnormal posture features and excrement morphology of the silkworm, and the environmental control logic of the silkworm rearing room is triggered according to the type of silkworm disease of the silkworm and the silkworm rearing environment of the silkworm, which is compatible with the overall consideration of the abnormal skin features, abnormal posture features and excrement morphology of the silkworm, realizes the accurate prediction of the type of silkworm disease of the silkworm, triggers the environmental control logic of the silkworm rearing room, and timely optimizes the silkworm rearing environment of the silkworm in the silkworm rearing room. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 1 is a flow chart of a method for predicting silkworm diseases based on a silkworm rearing environment in an embodiment of the present invention;

[0020] Figure 2 1 is a flow chart of step S11 in the method for predicting silkworm diseases based on the silkworm rearing environment in an embodiment of the present invention;

[0021] Figure 3 1 is a flow chart of step S12 in the method for predicting silkworm diseases based on the silkworm rearing environment in an embodiment of the present invention;

[0022] Figure 41 is a flow chart of step S13 in the method for predicting silkworm diseases based on the silkworm rearing environment in an embodiment of the present invention;

[0023] Figure 5 1 is a flow chart of step S14 in the method for predicting silkworm diseases based on the silkworm rearing environment in an embodiment of the present invention;

[0024] Figure 6 1 is a flow chart of step S15 in the method for predicting silkworm diseases based on the silkworm rearing environment in an embodiment of the present invention;

[0025] Figure 7 It is a schematic diagram of the structure of the silkworm disease prediction system based on the silkworm breeding environment in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0027] See also Figures 1 to 7 A method for predicting silkworm diseases based on a silkworm rearing environment is applied to a scenario of predicting silkworm diseases based on a silkworm rearing environment; the method for predicting silkworm diseases based on a silkworm rearing environment comprises:

[0028] Step S11: real-time monitoring of the silkworm rearing room and collection of the silkworm rearing distribution map of the silkworm rearing room;

[0029] Step S12: determining the silkworm activity trajectory and silkworm environment area based on the identification of the silkworm distribution map;

[0030] Step S13: determining the silkworm rearing environment based on the environmental detection of the silkworm environmental area, determining abnormal areas in the silkworm rearing room according to the silkworm rearing environment and the silkworm activity trajectory, and marking abnormal images of the silkworms;

[0031] Step S14: determining abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and determining the form of the silkworm's excrement by detecting along the silkworm's movement trajectory;

[0032] Step S15: predicting the type of silkworm disease based on the abnormal skin features, abnormal posture features, and excrement morphology of the silkworm, and triggering the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment;

[0033] refer to Figure 2 ,In step S11, the silkworm rearing room is monitored in real time, and a silkworm rearing distribution map of the silkworm rearing room is collected;

[0034] In the specific implementation process of the present invention, the specific steps are:

[0035] S111: The position of the silkworm rearing room is collected, and the corresponding visual detection component is triggered to respond according to the position of the silkworm rearing room. At this time, the visual detection component is converted from the standby state to the visual detection state during the response process, and the spatial position of each visual detection component relative to the silkworm rearing room is collected;

[0036] S112: Determine a real-time monitoring area of the silkworm room based on the spatial position of each visual detection component relative to the silkworm room and the visual detection range of the visual detection component. At this time, the real-time monitoring area of the silkworm room covers the activity trajectory of the silkworms and the surrounding environment of the silkworms;

[0037] S113: Each visual detection component moves in a circular motion along the internal space of the silkworm rearing room to collect multiple internal images of the silkworm rearing room. At the same time, the internal distribution map of the silkworm rearing room is determined based on the position of the silkworm rearing room and the database of the silkworm rearing room. The silkworm rearing distribution map of the silkworm rearing room is determined based on the synthesis of the internal distribution map of the silkworm rearing room and multiple internal images.

[0038] In the embodiment of the present application, the position of the silkworm room is collected, and the response of the corresponding visual detection element is triggered according to the position of the silkworm room. At this time, the visual detection element is converted from the standby state to the visual detection state during the response process, and the spatial position of each visual detection element relative to the silkworm room is collected;

[0039] At this point, determine the precise location of the silkworm room in physical space, which is usually done through GPS positioning (if the silkworm room is outdoors or in a large facility and GPS signal is reachable), indoor positioning technology (such as Wi-Fi positioning, Bluetooth beacons, ultrasonic positioning, etc.) or manual input of coordinates.

[0040] Once the location of the silkworm room is determined, the system will trigger the response of the associated visual detection components (such as cameras) according to the preset rules or algorithms. These visual detection components will switch from standby state to visual detection state and prepare to start collecting image data. At this time, high-definition cameras are installed in every corner of the silkworm room, and these cameras are connected to a central control system. When the system determines the location of the silkworm room through indoor positioning technology, it will send instructions to the nearest camera (or a group of cameras) to switch them from standby state to active state and start capturing images inside the silkworm room.

[0041] The specific position of each visual inspection part in the silkworm rearing room is recorded, which is usually achieved through pre-installed position sensors (such as accelerometers, gyroscopes, magnetometers, etc.) or through calibration at spatial reference points set in the silkworm rearing room. At the same time, a position sensor is installed on each camera (for example, an IMU module integrating accelerometers, gyroscopes and magnetometers). These sensors can record the position and orientation of the camera in real time. In addition, some known spatial reference points (such as marks on the wall) are also set in the silkworm rearing room. These reference points are used to calibrate the position and orientation of the camera to ensure that the acquired image matches the actual layout of the silkworm rearing room.

[0042] Furthermore, the real-time monitoring area of the silkworm room is determined according to the spatial position of each visual detection component relative to the silkworm room and the visual detection range of the visual detection component. At this time, the real-time monitoring area of the silkworm room covers the activity trajectory of the silkworms and the surrounding environment of the silkworms, and is compatible with the overall consideration of the spatial position of each visual detection component relative to the silkworm room and the visual detection range of the visual detection component, thereby ensuring the accuracy of the real-time monitoring area of the silkworm room.

[0043] At this time, the precise position of each visual detection component (such as a camera) in the silkworm rearing room is obtained, which is usually achieved through calibration with pre-installed position sensors (such as GPS, Wi-Fi locators, Bluetooth beacons combined with mobile devices, inertial measurement units (IMUs), etc.) or spatial reference points set in the silkworm rearing room; the accuracy of the position information is crucial for the subsequent determination of the monitoring area; at this time, the position sensors include accelerometers, gyroscopes, and magnetometers, which can measure the three-dimensional position and orientation of the camera; the spatial reference points are marks on the wall, fixed points on the ground, or pre-installed beacons.

[0044] Understand each camera's field of view, focal length, and viewing angle, as these parameters determine the area of the silkworm room that the camera can clearly capture. The visual detection range is usually determined by the camera's specifications or actual on-site testing. In this case, the camera's visual detection range is affected by factors such as lens type (such as wide-angle lens, standard lens, telephoto lens), focal length adjustment, image sensor size, and resolution.

[0045] Combined with the spatial position and visual detection range of the visual inspection parts, determine which areas in the silkworm rearing room are covered by the camera. These areas constitute the real-time monitoring area and should cover the main activity tracks and surrounding environment of the silkworms to ensure that the status of the silkworms can be fully monitored. At the same time, it is necessary to use geographic information system (GIS) or computer-aided design (CAD) software to visualize the layout of the silkworm rearing room and the camera, as well as the relative position relationship between them. Through simulation or actual testing, optimize the position and focal length of the camera to ensure maximum coverage and minimize overlap of the monitoring area.

[0046] Specifically, assume that the silkworm room is a rectangular room with a length of 10 meters, a width of 8 meters, and a height of 3 meters. To monitor the activities of the silkworms, a high-definition camera (Camera A, B, C, D) is installed in each of the four corners of the room, and these cameras are connected to a central control system. Camera A is located in the upper left corner of the room with coordinates (0,0,3); Camera B is located in the upper right corner of the room with coordinates (10,0,3); Camera C is located in the lower left corner of the room with coordinates (0,8,3); and Camera D is located in the lower right corner of the room with coordinates (10,8,3). These coordinates are calibrated using spatial reference points pre-set in the room (such as markers on the wall).

[0047] Each camera is equipped with a wide-angle lens with a 120-degree viewing angle and a fixed focal length. Through field testing, the visual detection range of each camera was determined, that is, the area of the silkworm room that it can clearly capture. Combining the spatial position and visual detection range of the camera, GIS software was used to generate a layout map of the silkworm room and cameras. By analyzing the layout map, the real-time monitoring areas in the silkworm room were determined. These areas cover the main activity trajectories of the silkworms (such as the feed delivery area and the silkworm rack area) and the surrounding environment (such as walls, doors and windows). To ensure the continuity and no blind spots of monitoring, the position and focal length of the camera were fine-tuned to ensure that the overlapping area between adjacent cameras was minimized while maximizing the coverage of key areas in the silkworm room. Through these steps, the system can efficiently determine the real-time monitoring area of the silkworm room and provide comprehensive image data support for subsequent silkworm disease prediction.

[0048] Therefore, each visual detection component moves in a circle along the internal space of the silkworm room to collect multiple internal images of the silkworm room. At the same time, the internal distribution map of the silkworm room is determined based on the position of the silkworm room and the database of the silkworm room. The silkworm distribution map of the silkworm room is determined based on the synthesis of the internal distribution map of the silkworm room and multiple internal images, which is compatible with the overall consideration of the position of the silkworm room and the database of the silkworm room, and ensures the accuracy of the internal distribution map of the silkworm room.

[0049] At this time, the visual detection device (such as a camera) is moved in a circular motion along the interior space of the silkworm room to collect internal images of the silkworm room from multiple angles and positions. This is achieved by installing a track system, using robots or drones and other technologies. The purpose of the circular motion is to ensure that every corner of the silkworm room can be captured, thereby generating a comprehensive image data set. At this time, the circular motion requires a precise control algorithm to ensure that the camera moves along the predetermined path and speed. In addition, in order to avoid image blur or overlap, the camera's shooting frequency and exposure time need to be controlled.

[0050] Based on the location of the silkworm room and the known silkworm room database information, the internal distribution map of the silkworm room is determined, which usually includes the physical layout of the silkworm room (such as silkworm racks, feed delivery area, ventilation equipment, etc.) as well as obstacles or special areas; at this time, the database of the silkworm room contains information such as architectural drawings, previous monitoring records, equipment layout diagrams, etc.; by combining this information with the actual location of the silkworm room, a detailed internal distribution map is generated.

[0051] The multiple internal images collected are synthesized to generate a silkworm distribution map of the silkworm room, which usually involves image stitching, deduplication, enhancement and other processing steps; the synthesized image should be able to clearly show the distribution and density of silkworms in the silkworm room; at the same time, image synthesis requires the use of image processing software or algorithms, such as image registration, image fusion, image segmentation, etc.; in addition, in order to improve the accuracy and efficiency of the synthesis, high-performance computing resources or parallel processing technology are required.

[0052] Specifically, assume that the silkworm breeding room is a rectangular room with a length of 10 meters, a width of 8 meters, and a height of 3 meters, with multiple silkworm racks and feed delivery areas inside; in order to generate a silkworm distribution map, it is decided to use a camera with automatic movement function (such as a camera installed on a track) for image acquisition.

[0053] The camera is mounted on a track that loops around the interior of the silkworm rearing room. The track's height and angle are designed to capture top and side views of the silkworm racks and feed placement area. The camera moves along the track at a constant speed and stops at preset positions to capture images. To ensure image quality, the camera's shooting frequency is set to one frame per second, and the exposure time is automatically adjusted based on indoor lighting conditions. In addition, to avoid image overlap, the distance between adjacent shooting points is precisely calculated and set.

[0054] Using the architectural drawings of the silkworm room and previous surveillance records as a reference, combined with the image information collected by the camera, the physical layout of the silkworm room was determined, including the location and number of silkworm racks, the location and size of the feed placement area, the layout of the ventilation equipment, etc.; by combining this information with the actual location of the silkworm room, a detailed internal distribution map was generated, which shows all the key areas and equipment in the silkworm room.

[0055] The multiple internal images collected are imported into image processing software and aligned using an image registration algorithm. Then, an image fusion algorithm is used to smoothly transition the overlapping parts of adjacent images to generate a seamless panoramic view of the silkworm room. Next, an image segmentation algorithm is used to segment the individual silkworms and groups in the panoramic view, and a silkworm distribution map is generated based on their position and density. The map shows the distribution of silkworms in the silkworm room, as well as dense or sparse areas. Through these steps, the system can efficiently generate a silkworm distribution map of the silkworm room, providing strong data support for subsequent silkworm disease prediction and environmental regulation.

[0056] refer to Figure 3 , in step S12, the silkworm activity trajectory and the silkworm environment area are determined based on the identification of the silkworm distribution map;

[0057] In the specific implementation process of the present invention, the specific steps are:

[0058] S121: Marking the current positions of multiple silkworms based on the detection of the silkworm distribution map, where each silkworm has a corresponding identity tag; determining multiple position nodes that each silkworm has passed through based on the current position of each silkworm and the image corresponding to each silkworm; and determining the corresponding silkworm activity trajectory based on the synthesis of the multiple position nodes that each silkworm has passed through;

[0059] S122: determining a silkworm activity area based on the silkworm activity trajectory and the current position of each silkworm, and determining a remaining area based on a comparison between the silkworm activity area and the silkworm breeding distribution map;

[0060] S123: Determine the silkworm environment area based on the division of the remaining area. At this time, the silkworm environment area is located around the silkworm activity area and covers the silkworm activity area.

[0061] In an embodiment of the present application, the current positions of multiple silkworms are marked based on the detection of the silkworm breeding distribution map. At this time, each silkworm has a corresponding identity mark; multiple position nodes passed by each silkworm are determined according to the current position of each silkworm and the image corresponding to each silkworm, and the corresponding silkworm activity trajectory is determined based on the synthesis of the multiple position nodes passed by each silkworm, which is compatible with the overall consideration of the synthesis of the multiple position nodes passed by each silkworm, thereby ensuring the accuracy of the corresponding silkworm activity trajectory.

[0062] At this time, a silkworm distribution map (generated from images of the silkworm room taken by a camera using image processing technology) is used to detect the current location of the silkworms. The silkworm distribution map is usually a two-dimensional plan that marks key elements in the silkworm room, such as silkworm racks, feed placement areas, etc., as well as the locations of the silkworms detected by image processing technology. Detecting the location of silkworms involves image processing technologies such as target detection, image segmentation, or machine learning algorithms, which can identify individual silkworms in the image and give their location coordinates. Each detected silkworm is assigned a unique identity tag in the form of numbers, letters, or a QR code for subsequent tracking and analysis.

[0063] Optionally, assume that there is a high-definition camera in the silkworm rearing room, which regularly takes images of the silkworm rearing room and transmits these images to a computer for processing; the image processing software on the computer uses a target detection algorithm to identify the position of the silkworms from each image and mark them on the silkworm distribution map; for example, in one image, the software identifies three silkworms, located at coordinates (x1, y1), (x2, y2) and (x3, y3), respectively, and these three silkworms are marked as silkworm A, silkworm B and silkworm C, respectively.

[0064] The silkworm distribution map at consecutive time points (or a sequence of continuously taken images) and the identity mark of each silkworm are used to determine the movement path of the silkworms. By comparing images at different time points, the position changes of each silkworm from the previous time point to the current time point are tracked. These position change points (i.e., position nodes) constitute the key points of the silkworm's movement path. Each position node records the position coordinates of the silkworm at a certain point in time.

[0065] By connecting the position node sequences of each silkworm in chronological order, the corresponding silkworm activity trajectory is synthesized; the activity trajectory is usually represented as a series of continuous position points, which are connected by lines to form a curve representing the silkworm's movement path; the activity trajectory is used to analyze the silkworm's behavioral patterns, activity range, and interaction with other silkworms or environmental factors.

[0066] Furthermore, the silkworm activity area is determined based on the silkworm activity trajectory and the current position of each silkworm, and the remaining area is determined based on the comparison between the silkworm activity area and the silkworm breeding distribution map, which is compatible with the overall consideration of the silkworm activity trajectory and the current position of each silkworm, ensuring the accuracy of the silkworm activity area.

[0067] At this point, the silkworm activity trajectories obtained in the previous steps and the current location information of each silkworm are used to determine the main activity area of the silkworms in the silkworm rearing room. The activity area is usually one or more two-dimensional areas, which cover the locations where silkworms frequently appear and move. Methods for determining the activity area include cluster analysis, spatial statistical analysis or density-based grid division, etc. These methods can identify the areas with the most intensive silkworm activity in the silkworm rearing room. The size and shape of the activity area vary depending on the type and number of silkworms, the environmental conditions of the silkworm rearing room, and the management strategy.

[0068] Optionally, assume that multiple silkworm activity trajectories have been obtained through the previous steps, and the current location of each silkworm is known; by analyzing these trajectories and location information, it is found that the silkworms are mainly concentrated in a corner of the silkworm room, which has multiple silkworm racks and is close to the feed delivery area; using the cluster analysis method, this corner is divided into a silkworm activity area; the area is an ellipse or irregular shape, depending on the actual distribution of the silkworms.

[0069] After determining the silkworm activity area, it is necessary to compare it with the silkworm distribution map to determine the remaining part of the silkworm room that is not covered by the silkworm activity area; the remaining area is the blank area, edge area or other functional area (such as ventilation holes, doorways, etc.) in the silkworm room, where silkworms have less activity or no activity at all; methods for determining the remaining area include spatial subtraction, image mask operation or rule-based segmentation, etc. These methods can remove the active area from the silkworm distribution map to obtain the remaining area; understanding the remaining area is of great significance for optimizing the silkworm environment, preventing the spread of diseases and improving silkworm breeding efficiency.

[0070] Optionally, the silkworm activity area has been determined to be an elliptical area located in the corner of the silkworm breeding room; now, this activity area is compared with the silkworm breeding distribution map, and the activity area is removed through a spatial subtraction operation to obtain the remaining area; the remaining area includes other corners of the silkworm breeding room, areas near the door and ventilation holes, and gaps between silkworm racks, etc. These areas are represented by different colors or marks on the silkworm breeding distribution map to facilitate subsequent analysis and management.

[0071] Therefore, the silkworm environment area is determined based on the division of the remaining area. At this time, the silkworm environment area is located on the periphery of the silkworm activity area and covers the silkworm activity area, thereby introducing the covered silkworm activity area.

[0072] At this time, the environmental areas of the silkworms are further divided according to the remaining areas determined in the previous steps; the environmental areas of the silkworms refer to those areas where the silkworms do not move directly but have an important impact on their growth, development and health. These areas are usually located around the silkworm activity areas, including vents, doorways, near light sources, temperature and humidity control equipment, etc.

[0073] The method of determining environmental zones involves spatial analysis, environmental factor assessment or judgment based on sericulture management experience; by analyzing the spatial layout of the remaining areas, environmental factors and the biological characteristics of silkworms, those environmental areas that have an important impact on silkworms are identified; the size and shape of environmental zones vary depending on the specific conditions and management strategies of the silkworm room; it is important to ensure that these areas can effectively cover the silkworm activity area and provide them with a suitable growth environment.

[0074] Optionally, it is assumed that the remaining areas in the silkworm rearing room have been determined through the previous steps. These areas include the edges of the silkworm rearing room, near the door, around the vents, and the gaps between certain silkworm racks; when analyzing these remaining areas, it is found that the areas near the door and around the vents are the areas where the airflow and temperature and humidity changes in the silkworm rearing room are most obvious; at the same time, these areas are also close to the silkworm activity area, and therefore have an important impact on the growth environment of the silkworms; based on these analyses, the areas near the door and around the vents are designated as environmental areas for silkworms, and these areas are represented by specific marks or colors in the silkworm rearing room to facilitate subsequent environmental monitoring and management.

[0075] When determining the environmental areas for silkworms, it is necessary to ensure that these areas are located on the periphery of the silkworm activity area and can effectively cover the activity area, which means that the environmental area should be large enough to include all important factors around the activity area and ensure that the silkworms in the activity area can obtain a suitable growth environment; covering the activity area does not mean that the environmental area must completely surround the activity area, but that the environmental area should extend to the edge of the activity area and slightly exceed it when necessary to ensure that no important environmental factors are missed; through reasonable planning and layout, ensure that the spatial relationship between the silkworm environmental area and the activity area is optimized, thereby providing the best growth conditions for the silkworms.

[0076] Optionally, when determining the silkworm environmental areas, ensure that these areas are located on the periphery of the silkworm activity area and extend to the edge of the activity area; for example, in the environmental area near the door, it is delineated as an area extending a certain distance outward from the door to ensure that all silkworms affected by the airflow at the door are covered.

[0077] Similarly, the environmental area around the ventilation holes is also demarcated as an area extending outward from the ventilation holes to a certain extent to ensure that all silkworms affected by the airflow and temperature and humidity changes at the ventilation holes are covered; through such a layout, it is ensured that the silkworm environmental area can effectively cover the silkworm activity area and provide a suitable growth environment for the silkworms.

[0078] In some embodiments of the present application, a region matching table is collected, and the region matching table is shown in Table 1:

[0079] Table 1. Regional matching table

[0080]

[0081] refer to Figure 4 In step S13, the silkworm rearing environment is determined based on the environmental detection of the silkworm environmental area, and the abnormal area of the silkworm rearing room is determined according to the silkworm rearing environment and the silkworm activity trajectory, and the abnormal image of the silkworm is marked;

[0082] In the specific implementation process of the present invention, the specific steps are:

[0083] S131: Collecting a mulberry silkworm environmental area, determining a plurality of environmental detection nodes based on the mulberry silkworm environmental area and the silkworm activity trajectory, determining corresponding environmental parameters based on environmental detections of the plurality of environmental detection nodes, and determining a sericulture environment for the mulberry silkworms based on the locations of the plurality of environmental detection nodes, the corresponding environmental parameters, and the regional morphology of the mulberry silkworm environmental area;

[0084] S132: collecting silkworm activity trajectories, and determining a plurality of silkworm activity nodes according to the division of the silkworm activity trajectories, and determining corresponding silkworm activity events according to the tracing back of the plurality of silkworm activity nodes, wherein the silkworm activity events include the silkworm's activity posture, resting posture, and excretion action;

[0085] S133: Determine a first abnormal parameter based on multiple silkworm activity events and the silkworm rearing environment, determine a second abnormal parameter based on multiple silkworm activity events and the type of silkworms, determine the abnormal area of the silkworm rearing room based on the mapping relationship between the first abnormal parameter, the second abnormal parameter and the abnormal area, and collect abnormal images of the silkworms based on real-time monitoring of the abnormal area in the silkworm rearing room.

[0086] In an embodiment of the present application, the mulberry silkworm environmental area is collected, a plurality of environmental detection nodes are determined based on the mulberry silkworm environmental area and the silkworm activity trajectory, corresponding environmental parameters are determined based on the environmental detection of the plurality of environmental detection nodes, and the sericulture environment of the mulberry silkworms is determined based on the locations of the plurality of environmental detection nodes and the corresponding environmental parameters and the regional morphology of the mulberry silkworm environmental area. This combines the overall consideration of the locations of the plurality of environmental detection nodes and the corresponding environmental parameters and the regional morphology of the mulberry silkworm environmental area to ensure the accuracy of the sericulture environment of the mulberry silkworms.

[0087] At this time, the silkworm environmental area is collected. Based on the information of the silkworm environmental area and the activity trajectory of the silkworms, a series of key environmental monitoring nodes need to be determined. These nodes are usually areas where the silkworms are active frequently and are easily affected by the environment. When determining the nodes, the biological characteristics of the silkworms (such as sensitivity to temperature, humidity, and light), the structural characteristics of the silkworm rearing room, and the spatial distribution of environmental factors must be considered.

[0088] Optionally, in the silkworm rearing room, based on the activity trajectory of the silkworms, locations such as near the silkworm racks, under the ventilation equipment, near the door, and in the corners of the silkworm rearing room are determined as environmental detection nodes. These locations can reflect the environmental conditions of the silkworm activity area and are easy to deploy sensors.

[0089] After identifying the environmental monitoring nodes, corresponding sensors need to be deployed at each node to monitor key environmental parameters in real time, including temperature, humidity, light intensity, and gas concentrations (such as carbon dioxide and oxygen). The selection and deployment of sensors should consider their measurement range, accuracy, stability, and compatibility with the environmental monitoring nodes. For this purpose, temperature and humidity sensors are deployed at environmental monitoring nodes near the silkworm racks to monitor temperature and humidity changes in the silkworm activity area. Gas concentration sensors are deployed at nodes below ventilation equipment to monitor the gas environment within the silkworm breeding room.

[0090] Based on the locations of multiple environmental monitoring nodes, their corresponding environmental parameters, and the morphology of the silkworm environmental area, a comprehensive assessment and determination of the silkworm rearing environment is conducted. This typically involves performing spatial, time series, or statistical analysis on the collected environmental data to identify hotspots, cold spots, or abnormal areas within the silkworm rearing room. Based on the analysis results, adjustments are made to the indoor environmental conditions, such as adjusting temperature and humidity and improving ventilation, to ensure an optimal growth environment for the silkworms. Alternatively, analysis of the temperature and humidity data from various environmental monitoring nodes within the silkworm rearing room may reveal significant fluctuations in temperature and humidity near a particular silkworm rack, potentially detrimental to silkworm growth. Adjustments are then made to the temperature and humidity control devices in that area to reduce these fluctuations and maintain suitable temperature and humidity conditions.

[0091] Specifically, it is assumed that there are three rows of silkworm racks in the silkworm room, with appropriate spacing between each row of racks; based on the activity trajectory of the silkworms and the structural characteristics of the silkworm room, six environmental monitoring nodes were identified: three are located near the silkworm racks (denoted as A, B, and C respectively), two are located in the corners of the silkworm room (denoted as D and E), and one is located below the ventilation equipment (denoted as F); at each node, temperature and humidity sensors and gas concentration sensors are deployed; through real-time monitoring of the data from these sensors, it was found that the temperature and humidity near node A fluctuated greatly, while the carbon dioxide concentration below node F was higher; based on these data, the temperature and humidity control device near node A was adjusted to reduce fluctuations; at the same time, the ventilation volume in the silkworm room was increased to reduce the carbon dioxide concentration below node F; through this series of operations, the environmental conditions in the silkworm room were successfully optimized, providing a more suitable growth environment for the silkworms.

[0092] Furthermore, the silkworm activity trajectories are collected, and multiple silkworm activity nodes are determined based on the division of the silkworm activity trajectories. The corresponding silkworm activity events are determined based on the tracing of multiple silkworm activity nodes. The silkworm activity events cover the silkworm's activity posture, resting posture and excretion movements, and are compatible with the overall consideration of tracing multiple silkworm activity nodes to ensure the accuracy of the corresponding silkworm activity events.

[0093] At this point, technologies such as image recognition, video tracking, or RFID are used to capture the silkworms' movements within the sericulture room in real time. The collected data should include the silkworms' location information (e.g., x, y coordinates) and time information (e.g., timestamps) to facilitate subsequent analysis of their movement patterns and frequency. The data collection process must ensure accuracy and continuity to avoid missing data due to missed or inaccurate data. Alternatively, a camera can be used to capture images of the silkworms and use computer vision algorithms to identify their position and posture. The captured video can be analyzed frame by frame to track the silkworms' movements. RFID tags can be attached to the silkworms and the tag information can be read using an RFID reader to obtain the silkworms' location data.

[0094] Based on the collected silkworm activity trajectories, the trajectories need to be divided into multiple key activity nodes; activity nodes are usually areas where silkworms stay for a long time, are active frequently, or have specific behavior patterns; when dividing activity nodes, factors such as the silkworm's activity density, activity speed, and direction changes are considered; at this time, according to the density distribution of silkworms in the activity area, high-density areas are divided into activity nodes; by analyzing the behavior patterns of silkworms (such as foraging, resting, excretion, etc.), areas with similar behaviors are divided into the same activity node.

[0095] After determining the silkworm activity nodes, each node needs to be traced back to determine the specific activity events of the silkworms at these nodes; the activity events should cover various behaviors of the silkworms, including active postures (such as crawling, foraging), resting postures (such as standing still) and excretion movements, etc.; the tracing process requires the use of technical means such as image recognition and behavioral analysis to carefully classify and identify the behavior of the silkworms; at this time, the collected images are classified using machine learning algorithms to identify the different postures and behaviors of the silkworms; by analyzing the behavioral sequences of the silkworms at the activity nodes, activity events with specific significance are identified.

[0096] Specifically, suppose that the activity trajectories of silkworms are collected in a silkworm rearing room using image recognition technology. By analyzing the trajectory data, it is found that the silkworms are mainly concentrated in three areas: near the silkworm racks, near the feed area, and near the rest area. A camera is used to capture the activity images of silkworms in the silkworm rearing room and record the location and time information of each silkworm. Through the density clustering method, the areas near the silkworm racks, near the feed area, and near the rest area are divided into three key activity nodes.

[0097] At the activity nodes near the silkworm racks, the foraging postures and activity events of the silkworms were identified through image classification and behavior sequence analysis; at the activity nodes near the feeding area, the foraging postures of the silkworms were also identified, but also included behavioral events of competing for feed; at the activity nodes near the rest area, the resting events such as the silkworms' stationary postures and excretion movements were identified; through this series of steps, the activity trajectories of the silkworms were successfully collected, the key activity nodes were determined, and the specific activity events at each node were traced. This information is of great significance for understanding the behavioral patterns of silkworms, optimizing the silkworm breeding environment, and improving breeding efficiency.

[0098] Therefore, the first abnormal parameter is determined according to multiple mulberry silkworm activity events and the sericulture environment of the mulberry silkworms, the second abnormal parameter is determined according to multiple mulberry silkworm activity events and the types of mulberry silkworms, the abnormal area of the silkworm rearing room is determined based on the mapping relationship between the first abnormal parameter, the second abnormal parameter and the abnormal area, and the abnormal images of the silkworms are collected based on the real-time monitoring of the abnormal area in the silkworm rearing room. This is compatible with the overall consideration of the first abnormal parameter, the second abnormal parameter and the mapping relationship between the abnormal area, ensuring the accuracy of the abnormal area of the silkworm rearing room, and is compatible with the overall consideration of the silkworm rearing environment and the activity trajectory of the silkworms, ensuring the accurate detection of the abnormal area of the silkworm rearing room, and further ensuring the accuracy of the abnormal images of the silkworms.

[0099] At this time, based on multiple mulberry silkworm activity events and the mulberry silkworm breeding environment, the first abnormal parameter that is inconsistent with the normal state is determined; the first abnormal parameter includes the mulberry silkworm activity frequency, activity range, activity intensity, etc., as well as temperature, humidity, light, gas concentration, etc. related to the mulberry silkworm breeding environment; when determining the first abnormal parameter, it is necessary to establish a standard range or threshold of the normal state for comparison with the current state.

[0100] Optionally, statistical analysis is performed on multiple silkworm activity events, and statistical quantities such as the average value and standard deviation of the activity frequency and activity range are calculated and compared with the standard range of normal conditions; data such as temperature, humidity, light, and gas concentration of the silkworm rearing environment are monitored and analyzed in real time and compared with the threshold values of normal conditions.

[0101] Based on multiple silkworm activity events and silkworm species, a second abnormal parameter that is inconsistent with the normal behavior pattern is determined; the second abnormal parameter includes abnormal behaviors such as the silkworm's activity posture, resting posture, and excretion movement; when determining the abnormal parameter, the biological characteristics and behavior pattern of the silkworm species need to be considered.

[0102] At this time, image recognition, machine learning and other technical means are used to identify and analyze the silkworms' activity postures, resting postures, excretion movements and other behaviors, and compare them with normal behavior patterns; the silkworms' activity sequences are analyzed to identify abnormal behaviors or changes in behavior patterns.

[0103] Based on the mapping relationship between the first abnormal parameter, the second abnormal parameter and the abnormal area, the abnormal area in the silkworm breeding room is determined; the abnormal area mapping relationship refers to associating the abnormal parameters with the specific area in the silkworm breeding room to quickly locate the abnormal area; at the same time, using the geographic information system (GIS) or spatial data analysis technology, the abnormal parameters are associated with the spatial position in the silkworm breeding room to determine the abnormal area; according to the intensity or frequency of the abnormal parameters, a heat map of the silkworm breeding room is generated to intuitively display the abnormal area.

[0104] Based on real-time monitoring of abnormal areas in the silkworm rearing room, abnormal images of silkworms are collected; the abnormal images should be able to clearly show the abnormal behavior or abnormal state of the silkworms for subsequent analysis and processing; at the same time, the abnormal areas are monitored in real time using cameras or image acquisition equipment; when an abnormality occurs, the abnormal images are captured in time and saved for subsequent analysis.

[0105] Specifically, assume that in a silkworm rearing room, image recognition technology and environmental monitoring equipment are used to conduct real-time monitoring and analysis of silkworm activity events and the silkworm rearing environment. At this time, the first abnormal parameter is determined: through statistical analysis of the data, it is found that the activity frequency of silkworms in a certain area is significantly reduced, and the temperature and humidity data in the area also deviate from the normal threshold. Therefore, the reduced activity frequency and abnormal temperature and humidity are determined as the first abnormal parameters.

[0106] Determine the second abnormal parameter: Through behavioral pattern recognition and behavioral sequence analysis, it was found that the silkworms in this area showed abnormal resting postures and excretion movements, such as standing still for a long time and abnormal excretion; therefore, abnormal resting postures and abnormal excretion movements were determined as the second abnormal parameter.

[0107] By combining the first and second abnormal parameters, as well as the mapping relationship between the abnormal areas, the specific area in the silkworm rearing room where the abnormality occurred was determined; through spatial analysis and thermal map analysis, the degree and scope of the abnormality in the area were intuitively seen; at the same time, after the abnormal area was determined, the area was monitored in real time using a camera; when the abnormality occurred again, the abnormal image was captured in time, and the relevant video data was saved for subsequent analysis.

[0108] In some embodiments of the present application, a first abnormal parameter matching table is collected, and the first abnormal parameter matching table is shown in Table 2:

[0109] Table 2. First abnormal parameter matching table

[0110]

[0111] At this time, the first abnormal parameters are determined to be: decreased activity frequency, expanded activity range, low temperature and humidity, and insufficient light.

[0112] Collect the second abnormal parameter matching table, the second abnormal parameter matching table is shown in Table 3:

[0113] Table 3. Second abnormal parameter matching table

[0114]

[0115] At this time, the second abnormal parameter was determined to be: slow crawling when foraging, body stretching when resting, and frequent defecation;

[0116] Collect the abnormal area mapping table, as shown in Table 4:

[0117] Table 4 Abnormal area mapping table

[0118]

[0119] The abnormal areas in the silkworm rearing room were identified as: Area A (reduced activity frequency, low temperature and humidity), Area B (expanded activity range, insufficient light), and Area C (slow crawling when foraging and frequent excretion).

[0120] refer to Figure 5 In step S14, the abnormal skin features and abnormal posture features of the silkworm are determined based on the abnormal image of the silkworm, and the morphology of the silkworm's excrement is determined along the detection of the silkworm's movement trajectory;

[0121] In the specific implementation process of the present invention, the specific steps are:

[0122] S141: collecting an abnormal image of the silkworm, and determining a skin region of the silkworm and a movement region of the silkworm according to the division of the abnormal image of the silkworm. At this time, the skin region of the silkworm and the movement region of the silkworm are in the same abnormal image;

[0123] S142: determining a plurality of skin color features based on the identification of the skin region of the silkworm, and determining an abnormal color feature based on matching the plurality of skin color features with a database of the silkworm breeding room, and determining the abnormal skin feature of the silkworm based on the position and shape of the abnormal color feature and the type of the silkworm;

[0124] S143: determining a plurality of posture features based on the identification of the movement area of the silkworm, and determining abnormal posture features based on matching the plurality of posture features with a database of the silkworm breeding room;

[0125] S144: Collect the silkworm's activity trajectory and perform detection along the silkworm's activity trajectory to collect images of the silkworm's excrement in the silkworm's activity trajectory, determine the silkworm's excrement morphology based on the recognition of the excrement image, and determine the abnormal feature combination based on the matching of the silkworm's excrement morphology, abnormal posture features and abnormal skin features, and autonomously update the features in the abnormal feature combination.

[0126] In an embodiment of the present application, an abnormal image of a silkworm is collected, and the skin area of the silkworm and the movement area of the silkworm are determined based on the division of the abnormal image of the silkworm. At this time, the skin area of the silkworm and the movement area of the silkworm are in the same abnormal image, and the skin area of the silkworm and the movement area of the silkworm are introduced to be in the same abnormal image.

[0127] At this time, abnormal images are collected and preprocessed, including denoising and contrast enhancement, to improve image quality. The preprocessed images should be clearer to facilitate the subsequent division of skin areas and action areas. At the same time, image processing techniques such as image segmentation and edge detection are used to identify the skin area of the silkworm. The skin area usually includes the body part of the silkworm, and its color, texture and other characteristics can be used for subsequent analysis.

[0128] Image processing technology is also used to identify the silkworm's motion area; the motion area usually includes the silkworm's limbs, head and other parts that can reflect its activity state; by analyzing the morphological changes of the motion area, it is determined whether the silkworm has abnormal behavior; at this time, the divided skin area and motion area are integrated into the same abnormal image for subsequent comprehensive analysis.

[0129] Specifically, suppose that in a silkworm-raising room, a high-definition camera is used to monitor silkworms in real time. One day, it is observed that the body color of a silkworm suddenly becomes abnormally dark, and its movements become sluggish. Therefore, this abnormal image is immediately captured.

[0130] During the image preprocessing stage, the images were denoised and contrast enhanced to make the appearance features of the silkworm more obvious; then, image processing technology was used to identify the silkworm's skin area (including its body part) and movement area (including its limbs and head); in the skin area, it was clearly seen that the silkworm's body color became dull, in sharp contrast to the bright color of normal silkworms; in the movement area, it was observed that the silkworm's limbs moved sluggishly and lacked vitality.

[0131] Finally, the divided skin area and movement area are integrated into the same abnormal image to facilitate subsequent comprehensive analysis of the silkworm's abnormal conditions. Through this abnormal image, it is intuitively seen that the silkworm's skin color and movement status are abnormal, thereby further judging its health status and taking corresponding treatment measures.

[0132] Furthermore, multiple skin color features are determined based on the identification of the skin area of the silkworm, and abnormal color features are determined based on the matching of multiple skin color features with the database of the silkworm breeding room. The abnormal skin features of the silkworm are determined based on the position, form and type of the abnormal color features, which is compatible with the overall consideration of the position, form and type of the abnormal color features, thereby ensuring the accuracy of the abnormal skin features of the silkworm.

[0133] At this time, within the divided silkworm skin area, color recognition technology (such as RGB color space analysis, HSV color space conversion, etc.) is used to extract multiple skin color features. These color features include the overall hue, brightness, saturation, etc. of the skin; ensuring that the extracted color features can accurately reflect the true color state of the silkworm skin.

[0134] The extracted skin color features are matched with the normal skin color features stored in the silkworm breeding room database; the normal skin color features in the database are derived based on a large number of observation data of healthy silkworms, representing the normal skin color range of silkworms at different growth stages and in different environments; at the same time, through comparison, color features that do not match the normal skin color features in the database, namely abnormal color features, are identified; abnormal color features are manifested as deviations in hue, abnormal brightness, and excessive or low saturation.

[0135] Carefully observe the location distribution and morphological characteristics of abnormal color features on the silkworm's skin; the location distribution involves the whole body or local, and the morphological characteristics include spots, stripes, fading, etc.; at the same time, according to the type of silkworm, its unique skin color characteristics and abnormal conditions are considered; different types of silkworms are more sensitive to certain color changes, or have specific skin color abnormality patterns; combined with the location, morphology and type of silkworm of the abnormal color features, a comprehensive judgment is made on the abnormal skin characteristics of the silkworm.

[0136] Specifically, suppose a silkworm is observed to have abnormal skin color in a silkworm rearing room. Within the demarcated skin area, the skin color features of the silkworm are extracted using color recognition technology, and it is found that its overall hue is yellowish, the brightness is low, and the saturation is also reduced. Then, these color features are matched with the normal skin color features stored in the silkworm rearing room database. Through comparison, it is found that the skin color features of this silkworm are significantly different from the normal features stored in the database, especially the yellowish hue is very obvious.

[0137] Therefore, it was determined that this silkworm had abnormal color characteristics; further observation revealed that the abnormal color characteristics were mainly distributed on the abdomen and back of the silkworm, and morphologically appeared as a uniform yellow coverage; finally, combined with the species of this silkworm (assuming it was a silkworm variety that was abnormally sensitive to yellow), it was judged that its abnormal skin characteristics were caused by a certain disease or malnutrition; therefore, corresponding treatment measures were taken, and the environment in the silkworm breeding room was adjusted in order to improve the health of the silkworm.

[0138] Furthermore, multiple posture features are determined based on the identification of the silkworm's movement area, and abnormal posture features are determined based on the matching of multiple posture features with the database of the silkworm breeding room. This is compatible with the overall consideration of the matching of multiple posture features with the database of the silkworm breeding room, ensuring the accuracy of the abnormal posture features.

[0139] At this time, within the already divided silkworm movement area, posture recognition technology (such as skeleton extraction, key point detection, movement classification, etc.) is used to extract multiple posture features. These posture features include the silkworm's limb angle, movement trajectory, movement speed, movement duration, etc.; ensuring that the extracted posture features can accurately reflect the silkworm's actual movement status.

[0140] The extracted posture features are matched with the normal posture features stored in the silkworm breeding room database; the normal posture features in the database are derived based on a large number of observation data of healthy silkworms in different situations, representing the normal posture range of silkworms in different growth stages and different activity states; at the same time, through comparison, posture features that do not match the normal posture features in the database, namely abnormal posture features, are identified; abnormal posture features are manifested as uncoordinated movements, abnormal speed, too long or too short duration, etc.

[0141] Specifically, suppose a silkworm is observed to be moving abnormally in a silkworm-raising room. Within the demarcated motion area, posture recognition technology is used to extract the silkworm's posture features. Through skeleton extraction and key point detection, it is found that the movements of the silkworm's limbs appear uncoordinated, especially the movement trajectories of the forelimbs and hind limbs, which show obvious deviations. In addition, it is also noted that the movement speed of this silkworm is significantly slower than that of other healthy silkworms.

[0142] Then, these posture features were matched with the normal posture features stored in the silkworm breeding room database; through comparison, it was found that the posture features of this silkworm were significantly different from the normal features stored in the database; in particular, the incoordination of the limbs and the slowness of the movement speed, both of which deviated significantly from the normal range; therefore, it was determined that this silkworm had abnormal posture features; combined with previous observations, it was speculated that this silkworm suffered from some disease that affected its motor ability, or was physically unwell due to environmental discomfort (such as too high or too low temperature, too high or too low humidity), which in turn affected its normal movements.

[0143] To verify this hypothesis, we further observed other behavioral characteristics of the silkworm and collected relevant physiological indicator data. At the same time, we also checked the environment in the silkworm breeding room to ensure the suitability of the environmental conditions. Finally, based on various information, we formulated corresponding treatment measures and made necessary adjustments to the environment in the silkworm breeding room.

[0144] Therefore, the silkworm's activity trajectory is collected, and detection is carried out along the silkworm's activity trajectory to collect images of excrement in the silkworm's activity trajectory. The silkworm's excrement morphology is determined based on the recognition of the excrement image, and the abnormal feature combination is determined based on the matching of the silkworm's excrement morphology, abnormal posture features and abnormal skin features. Autonomous update of features is performed in the abnormal feature combination, which is compatible with the overall consideration of the matching of the silkworm's excrement morphology, abnormal posture features and abnormal skin features, and ensures the accuracy of the abnormal feature combination.

[0145] At this time, use video tracking technology or infrared sensors and other equipment to continuously record the activity tracks of the silkworms in the silkworm breeding room; ensure the continuity and accuracy of the activity tracks for subsequent analysis of the silkworms' behavioral patterns and habits.

[0146] Based on the collected activity tracks, a detection path is planned, and the areas where the silkworms have passed are detected along the path; the focus of the detection is the areas where the silkworms excrete, such as mulberry leaf piles, silkworm feces basins, etc.; at the same time, during the detection process, a high-resolution camera or other image acquisition equipment is used to capture images of excrement on the silkworms' activity tracks; ensure that the image is clear and can accurately reflect the shape, color, texture and other characteristics of the excrement.

[0147] Image processing technology is used to identify and analyze the collected excrement images; the morphology of the excrement (such as shape, size, color, etc.) is identified and compared with the normal excrement morphology stored in the database; at this time, the identified excrement morphology is matched with the previously determined abnormal posture characteristics and abnormal skin characteristics; the abnormal characteristics of these three aspects are comprehensively considered to form an abnormal feature combination to comprehensively reflect the health status of the silkworm.

[0148] Furthermore, as time goes by and the health status of silkworms changes, the features in the abnormal feature combination are continuously updated; through continuous monitoring and data analysis, new abnormal features are discovered in a timely manner and incorporated into the abnormal feature combination.

[0149] Specifically, suppose the activity trajectory of a silkworm is continuously monitored in the silkworm-raising room, and it is found that its activity range gradually decreases and its movements become sluggish; the activity trajectory of the silkworm is recorded using video tracking technology, and it is found that it frequently moves near the mulberry leaf pile, but rarely moves to other areas; then, inspections are carried out along the activity trajectory, especially near the mulberry leaf pile and the silkworm feces basin; during the inspection process, a clear image of the excrement is captured, and it is found that the morphology of the excrement is abnormal, appearing dry and clumping, which is significantly different from the moist and loose morphology of normal excrement.

[0150] Then, the excrement images were identified and analyzed using image processing technology, confirming the abnormality of the excrement morphology. At the same time, the previously identified abnormal posture characteristics (such as slow movement and uncoordinated limbs) and abnormal skin characteristics (such as dull skin color and spots) were reviewed. Taking these three abnormal characteristics into consideration, an abnormal feature combination was formed, and it was believed that the silkworm was suffering from a certain disease or malnutrition. To verify this speculation, the silkworm's physiological indicator data was further collected, and sericulture experts were consulted. Finally, based on the new monitoring data and analysis results, the abnormal feature combination was independently updated. It was found that the health status of the silkworm had improved, but some abnormal characteristics still existed. Therefore, the treatment measures and environmental conditions were adjusted, and the silkworms continued to be monitored and analyzed in order to comprehensively and accurately reflect their health status.

[0151] In some embodiments of the present application, an abnormal feature category matching table is collected, and the abnormal feature category matching table is shown in Table 5:

[0152] Table 5 Abnormal feature category matching table

[0153]

[0154] Abnormal characteristics of the silkworms are evaluated against the weighted scoring table; for example, dry and caked excrement is scored as 4 points (assuming it is at the upper middle level of the scoring range), sluggish movement is scored as 5 points (severe), and dull skin color is scored as 2 points (mild); the scores of all abnormal characteristics are added together to obtain the total health score of the silkworm; in this example, the total score = 4 (excrement) + 5 (posture) + 2 (skin) = 11 points.

[0155] Based on the total score, the health status of the silkworms is divided into different levels (such as excellent, good, average, poor, and very poor). In this example, it is assumed that the total score of 11 points belongs to the "poor" level. At the same time, the silkworms are continuously monitored and the weights and score tables are adjusted according to the changes in their health status. The total health score and level of the silkworms are re-evaluated regularly. Through the above two methods, a more comprehensive and accurate understanding of the health status of the silkworms can be obtained, and corresponding treatment measures and environmental adjustment strategies can be taken.

[0156] refer to Figure 6 In step S15, the type of silkworm disease is predicted based on the abnormal skin features, abnormal posture features and the morphology of the silkworm's excrement, and the environmental control logic of the silkworm rearing room is triggered according to the type of silkworm disease and the silkworm rearing environment;

[0157] In the specific implementation process of the present invention, the specific steps are:

[0158] S151: collecting abnormal skin features, abnormal posture features, and excrement morphology of the silkworms, determining a first silkworm disease parameter based on the excrement morphology and skin abnormality features, and determining a second silkworm disease parameter based on the excrement morphology and abnormal posture features of the silkworms;

[0159] S152: Predicting the type of silkworm disease of the silkworm based on the mapping relationship between the first silkworm disease parameter, the second silkworm disease parameter, and the type of silkworm disease; determining the silkworm disease-affecting behavior based on the tracing of the silkworm disease type; and dynamically matching the silkworm disease-affecting behavior with the activity behavior corresponding to the silkworm activity trajectory to further review the silkworm disease type;

[0160] S153: After the review of the types of silkworm diseases of the mulberry silkworms is completed, the silkworm disease optimization measures are determined based on the types of silkworm diseases of the mulberry silkworms and the database of the silkworm rearing room, and a plurality of sub-optimization measures are formed according to the division of the silkworm disease optimization measures. The environmental control logic of the silkworm rearing room is determined according to the contents of the plurality of sub-optimization measures, the types of the mulberry silkworms and the silkworm rearing environment of the mulberry silkworms to trigger the environmental control event of the silkworm rearing room.

[0161] In an embodiment of the present application, abnormal skin characteristics, abnormal posture characteristics and excrement morphology of silkworms are collected, and the first silkworm disease parameter is determined based on the excrement morphology of the silkworms and the abnormal skin characteristics of the silkworms, and the second silkworm disease parameter is determined based on the excrement morphology of the silkworms and the abnormal posture characteristics of the silkworms. This takes into account the overall consideration of the excrement morphology of the silkworms and the abnormal posture characteristics of the silkworms, thereby ensuring the accuracy of the second silkworm disease parameter.

[0162] At this time, the abnormal skin characteristics of the silkworms are collected. At this time, high-definition cameras or microscopes and other equipment are used to carefully observe the skin of the silkworms; changes in skin color, texture, glossiness, etc. are recorded, and special attention is paid to whether there are abnormal phenomena such as spots, ulcers, swelling, etc. These abnormal characteristics indicate that the silkworms have a certain skin disease or are invaded by external parasites.

[0163] Collect abnormal posture characteristics of silkworms. At this time, record the postures of silkworms such as walking, eating, and resting through video tracking or manual observation; pay attention to whether the silkworms have abnormal postures such as slow movement, uncoordinated limbs, and twisted bodies. These abnormal characteristics indicate that the silkworms have nervous system problems, muscle diseases, or poisoning.

[0164] Collect the morphology of silkworm excrement. At this time, collect silkworm excrement samples and use microscopes or chemical analyzers to detect them. Observe the color, texture, shape and other characteristics of the excrement, paying special attention to the presence of abnormal substances (such as blood and pus). Changes in the morphology of excrement reflect the health of the silkworm's digestive system or whether it suffers from some infectious disease.

[0165] A comprehensive analysis is conducted based on the morphology of the silkworm's excrement and abnormal skin characteristics; based on the correspondence between known silkworm diseases and these characteristics, a preliminary judgment is made on the type of silkworm disease the silkworm suffers from; this judgment result is used as the first silkworm disease parameter for subsequent silkworm disease prediction and formulation of optimization measures; optionally, assuming that during the collection process, a silkworm is found to have dull skin color and irregular spots, and its excrement is dry and lumpy; based on these characteristics, a preliminary judgment is made that the silkworm suffers from malnutrition combined with skin disease, and this judgment result is the determined first silkworm disease parameter.

[0166] Determine the second silkworm disease parameter. Similarly, conduct a comprehensive analysis based on the morphology of the silkworm's excrement and abnormal posture characteristics. Based on the correspondence between the silkworm disease and these characteristics, further determine the type of silkworm disease the silkworm suffers from. Use this judgment result as the second silkworm disease parameter and verify it with the first silkworm disease parameter to improve the accuracy of silkworm disease prediction.

[0167] Furthermore, based on the mapping relationship between the first silkworm disease parameter, the second silkworm disease parameter and the silkworm disease type, the silkworm disease type is predicted, the silkworm disease-affecting behavior is determined based on the tracing of the silkworm disease type, and the silkworm disease-affecting behavior is dynamically matched with the activity behavior corresponding to the silkworm activity trajectory to further review the silkworm disease type, which is compatible with the overall consideration of the tracing of the silkworm disease type and ensures the accuracy of the silkworm disease-affecting behavior.

[0168] At this time, a database of mapping relationships between silkworm disease types and the first and second silkworm disease parameters is established. This database should contain characteristic descriptions of various known silkworm diseases, corresponding silkworm disease parameters, and silkworm disease types. After the first and second silkworm disease parameters of the silkworm are collected, they are compared with the mapping relationships in the database to predict the type of silkworm disease the silkworm suffers from. The prediction result is one or more types of silkworm diseases, depending on the similarity of the silkworm disease parameters and the accuracy of the database.

[0169] Based on the predicted type of silkworm disease, the specific impact of the disease on the behavior of silkworms is traced and determined, including abnormal behavior of silkworms, changes in activity patterns, and behavioral differences from other healthy silkworms. These affected behaviors include decreased appetite, slow movement, and weakened response to external stimuli.

[0170] Using video tracking, sensor monitoring and other technical means, the activity trajectories and corresponding activity behaviors of mulberry silkworms are recorded; the recorded activity behaviors of mulberry silkworms are dynamically matched with the predicted behaviors affected by silkworm diseases, which includes comparing the similarity between the actual behavior patterns of mulberry silkworms and the predicted abnormal behavior patterns; the purpose of dynamic matching is to verify whether the predicted type of silkworm disease is accurate and whether the mulberry silkworms are indeed affected by the silkworm disease.

[0171] If the dynamic matching results show that the actual behavior of the silkworms is highly consistent with the predicted behavior affected by the silkworm disease, then the predicted type of silkworm disease is confirmed to be accurate; if the matching results are inconsistent or there are significant differences, then the predicted type of silkworm disease needs to be reviewed, which includes re-collecting and analyzing silkworm disease parameters, consulting silkworm experts or conducting more in-depth medical examinations; the purpose of the review is to ensure the accuracy of the silkworm disease type so as to formulate effective treatment and optimization measures.

[0172] In this example, suppose that in step S151, a silkworm is predicted to suffer from malnutrition combined with nervous system damage (based on the first and second silkworm disease parameters). In step S152, the affected behaviors, such as decreased appetite, slow movement, and decreased response to external stimuli, are first determined based on the predicted silkworm disease type. Next, the silkworm's movement trajectory and corresponding behaviors are recorded using video tracking technology. It is found that the silkworm indeed exhibits abnormal behaviors such as decreased appetite and slow movement, and these behaviors are highly consistent with the predicted behaviors affected by the silkworm disease.

[0173] Therefore, it is confirmed that the predicted type of silkworm disease is accurate, that is, the silkworm suffers from malnutrition and nervous system damage. This result provides an important basis for the subsequent formulation of effective treatment and optimization measures; however, if the actual behavior of the silkworm is found to be significantly different from the predicted disease-affected behavior during the dynamic matching process (for example, the silkworm shows extreme activity rather than sluggishness), then the predicted type of silkworm disease will be reviewed immediately to ensure accuracy and adjust subsequent treatment and optimization measures.

[0174] Therefore, after the review of the types of silkworm diseases of mulberry silkworms is completed, the silkworm disease optimization measures are determined based on the types of silkworm diseases of mulberry silkworms and the database of the silkworm rearing room, and multiple sub-optimization measures are formed according to the division of the silkworm disease optimization measures. The environmental control logic of the silkworm rearing room is determined according to the contents of the multiple sub-optimization measures, the types of mulberry silkworms and the silkworm rearing environment of the mulberry silkworms to trigger the environmental control events of the silkworm rearing room, which is compatible with the overall consideration of the contents of multiple sub-optimization measures, the types of mulberry silkworms and the silkworm rearing environment of the mulberry silkworms, ensures the accuracy of the environmental control logic of the silkworm rearing room, realizes the accurate prediction of the types of silkworm diseases of mulberry silkworms, triggers the environmental control logic of the silkworm rearing room, and timely optimizes the silkworm rearing environment of the mulberry silkworms in the silkworm rearing room.

[0175] At this time, after the review of the types of silkworm diseases, the database of the silkworm rearing room or the professional silkworm rearing guide is consulted according to the confirmed types of silkworm diseases to determine the optimization measures for the silkworm diseases. These optimization measures include adjusting the feed formula, increasing or decreasing the intake of specific nutrients, improving the sanitary conditions of the silkworm rearing environment, adjusting the temperature and humidity of the silkworm rearing room, etc. The goal of the optimization measures is to promote the healthy recovery of the silkworms, prevent further deterioration of the silkworm diseases, and improve the yield and quality of the cocoons.

[0176] The identified silkworm disease optimization measures are further subdivided into multiple specific sub-optimization measures, which should be more operational and easier to implement and monitor. For example, if the optimization measure is to adjust the feed formula, then the sub-optimization measures include increasing protein content, adding specific vitamins or minerals, etc. The division of sub-optimization measures helps to more accurately control the various variables in the silkworm rearing process, thereby more effectively treating silkworm diseases.

[0177] The environmental control logic of the silkworm room is determined based on the content of multiple sub-optimization measures, the type of silkworm (such as variety, growth stage, etc.), and the actual conditions of the silkworm environment (such as current temperature, humidity, and lighting conditions). The environmental control logic should clearly specify which environmental control events are triggered under what conditions to achieve the goals of the sub-optimization measures. For example, if the sub-optimization measure is to reduce the humidity in the silkworm room to reduce the growth of pathogens, then the environmental control logic will automatically start the dehumidification equipment when the humidity exceeds a certain threshold.

[0178] According to the determined environmental control logic, the corresponding environmental control events are triggered by automatic control systems or manual operations. These events include adjusting environmental factors such as temperature and humidity, light intensity, and ventilation rate to meet the requirements of sub-optimization measures. After the environmental control events are triggered, the reactions of the silkworms and changes in the silkworm rearing environment should be continuously monitored to ensure the effectiveness of the optimization measures and make adjustments as needed.

[0179] Specifically, assume that in step S152, a batch of silkworms is confirmed to suffer from bacterial gastrointestinal disease; in step S153, first, according to the characteristics of bacterial gastrointestinal disease, corresponding optimization measures are searched from the silkworm room database; it is determined that the sanitary quality of the feed needs to be improved, the antibacterial components of the feed need to be increased, and the disinfection work of the silkworm room needs to be strengthened; then, these optimization measures are subdivided into multiple sub-optimization measures; for example, improving the sanitary quality of the feed includes using fresh, uncontaminated mulberry leaves and adding an appropriate amount of antibacterial agents to the feed; strengthening disinfection work includes regularly cleaning and disinfecting the silkworm room thoroughly, and using ultraviolet lamps to kill pathogens in the air.

[0180] Then, based on the content of the sub-optimization measures and the growth environment of the silkworms, the environmental control logic of the silkworm room was determined; for example, it was decided to strictly clean and disinfect the mulberry leaves before feeding every day, and to install ultraviolet lamps in the silkworm room, which automatically turned on at night to kill pathogens; finally, the corresponding environmental control events were triggered; the disinfection equipment in the silkworm room was manually adjusted to ensure that it worked at the predetermined time intervals; at the same time, the feed formula was also adjusted, and an appropriate amount of antibacterial agents were added to improve the immunity of the silkworms; through the implementation of these optimization measures and environmental control events, the spread of bacterial gastrointestinal diseases was successfully controlled, the healthy recovery of the silkworms was promoted, and ultimately the yield and quality of the cocoons were improved.

[0181] See also Figure 7 , Figure 7 1 is a schematic diagram of the structure of a silkworm disease prediction system based on a silkworm rearing environment in an embodiment of the present invention; the silkworm disease prediction system based on a silkworm rearing environment includes:

[0182] The silkworm breeding distribution map module 21 is used to monitor the silkworm breeding room in real time and collect the silkworm breeding distribution map of the silkworm breeding room;

[0183] A silkworm identification module 22 is used to determine the silkworm activity trajectory and silkworm environment area based on the identification of the silkworm distribution map;

[0184] Anomaly detection module 23, for determining the silkworm rearing environment based on the environmental detection of the silkworm environmental area, determining abnormal areas in the silkworm rearing room according to the silkworm rearing environment and silkworm activity trajectory, and marking abnormal images of the silkworms;

[0185] a morphology detection module 24 for determining abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and determining the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory;

[0186] The prediction module 25 is used to predict the type of silkworm disease based on the abnormal skin features, abnormal posture features and the morphology of the silkworm's excrement, and trigger the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment.

[0187] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for predicting silkworm diseases based on silkworm rearing environment, characterized in that: include: Real-time monitoring of the silkworm rearing room and collection of the silkworm rearing distribution map of the silkworm rearing room; Determine the silkworm activity trajectory and silkworm environment area based on the identification of silkworm distribution map; Determining the silkworm rearing environment based on environmental detection of the silkworm environmental area, determining abnormal areas of a silkworm rearing room based on the silkworm rearing environment and silkworm activity trajectories, and marking abnormal images of the silkworms, including: collecting the silkworm environmental area, determining multiple environmental detection nodes based on the silkworm environmental area and the silkworm activity trajectories, determining corresponding environmental parameters based on environmental detection of the multiple environmental detection nodes, and determining the silkworm rearing environment based on the locations of the multiple environmental detection nodes, the corresponding environmental parameters, and the regional morphology of the silkworm environmental area; collecting the silkworm activity trajectories, and determining multiple silkworm activity nodes based on the division of the silkworm activity trajectories, and determining corresponding silkworm activity events based on tracing back the multiple silkworm activity nodes, the silkworm activity events covering the activity posture, rest posture, and excretion of the silkworms; determining a first abnormal parameter based on the multiple silkworm activity events and the silkworm rearing environment, determining a second abnormal parameter based on the multiple silkworm activity events and the type of the silkworms, determining the abnormal area of the silkworm rearing room based on a mapping relationship between the first abnormal parameter, the second abnormal parameter, and the abnormal area, and collecting abnormal images of the silkworms based on real-time monitoring of the abnormal area in the silkworm rearing room; Determine the abnormal skin features and posture features of the silkworm based on the abnormal image of the silkworm, and determine the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory; The type of silkworm disease is predicted based on the abnormal skin characteristics, abnormal posture characteristics and excrement morphology of the silkworm, and the environmental control logic of the silkworm rearing room is triggered according to the type of silkworm disease and the silkworm rearing environment.

2. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 1, characterized in that: The real-time monitoring of the silkworm raising room and the collection of the silkworm raising distribution map of the silkworm raising room include: The position of the silkworm rearing room is collected, and the corresponding visual detection component is triggered to respond according to the position of the silkworm rearing room. At this time, the visual detection component is converted from the standby state to the visual detection state during the response process, and the spatial position of each visual detection component relative to the silkworm rearing room is collected; The real-time monitoring area of the silkworm breeding room is determined according to the spatial position of each visual detection component relative to the silkworm breeding room and the visual detection range of the visual detection component. At this time, the real-time monitoring area of the silkworm breeding room covers the activity trajectory of the silkworms and the surrounding environment of the silkworms; Each visual detection component moves in a circle along the internal space of the silkworm rearing room to collect multiple internal images of the silkworm rearing room. At the same time, the internal distribution map of the silkworm rearing room is determined based on the position of the silkworm rearing room and the database of the silkworm rearing room, and the silkworm rearing distribution map of the silkworm rearing room is determined based on the synthesis of the internal distribution map of the silkworm rearing room and multiple internal images.

3. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 1, characterized in that: The method of determining the silkworm activity trajectory and the silkworm environment area based on the identification of the silkworm breeding distribution map includes: Based on the detection of the silkworm distribution map, the current positions of multiple silkworms are marked. At this time, each silkworm has a corresponding identity mark; based on the current position of each silkworm and the image corresponding to each silkworm, multiple position nodes passed by each silkworm are determined, and the corresponding silkworm activity trajectory is determined based on the synthesis of the multiple position nodes passed by each silkworm; Determine the silkworm activity area based on the silkworm activity trajectory and the current location of each silkworm, and determine the remaining area based on the comparison between the silkworm activity area and the silkworm breeding distribution map; The silkworm environment area is determined based on the division of the remaining area. At this time, the silkworm environment area is located on the periphery of the silkworm activity area and covers the silkworm activity area.

4. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 1, characterized in that: The method of determining abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and determining the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory, includes: Collecting an abnormal image of the silkworm, and determining a skin area of the silkworm and a movement area of the silkworm according to the division of the abnormal image of the silkworm, wherein the skin area of the silkworm and the movement area of the silkworm are located in the same abnormal image; Determining multiple skin color features based on the identification of the silkworm's skin area, and determining abnormal color features based on matching the multiple skin color features with a database of the silkworm breeding room, and determining the abnormal skin features of the silkworm based on the location and morphology of the abnormal color features and the type of the silkworm; A plurality of posture features are determined based on the recognition of the movement area of the silkworm, and abnormal posture features are determined based on the matching of the plurality of posture features with the database of the silkworm breeding room.

5. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 4, characterized in that: The method of determining abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and determining the form of the silkworm's excrement by detecting the movement trajectory of the silkworm, further includes: The silkworm's activity trajectory is collected, and detection is performed along the silkworm's activity trajectory to collect images of the silkworm's excrement in the silkworm's activity trajectory. The silkworm's excrement morphology is determined based on the recognition of the excrement image, and the abnormal feature combination is determined based on the matching of the silkworm's excrement morphology, abnormal posture features and abnormal skin features, and the features in the abnormal feature combination are autonomously updated.

6. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 1, characterized in that: The method of predicting the type of silkworm disease based on the abnormal skin features, abnormal posture features and excrement morphology of the silkworm, and triggering the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment, includes: collecting abnormal skin features, abnormal posture features, and excrement morphology of the silkworms, determining a first silkworm disease parameter based on the excrement morphology and the abnormal skin features of the silkworms, and determining a second silkworm disease parameter based on the excrement morphology and the abnormal posture features of the silkworms; The type of silkworm disease of the silkworm is predicted based on the mapping relationship between the first silkworm disease parameter, the second silkworm disease parameter and the type of silkworm disease. The silkworm disease-affecting behavior is determined based on the tracing of the silkworm disease type. The silkworm disease-affecting behavior is dynamically matched with the activity behavior corresponding to the silkworm activity trajectory to further review the silkworm disease type.

7. The method for predicting silkworm diseases based on the silkworm rearing environment according to claim 6, characterized in that: The method of predicting the type of silkworm disease based on the abnormal skin features, abnormal posture features and excrement morphology of the silkworm, and triggering the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment, further includes: After the review of the types of silkworm diseases of the mulberry silkworms is completed, the silkworm disease optimization measures are determined based on the types of silkworm diseases and the database of the silkworm rearing room. According to the division of the silkworm disease optimization measures, multiple sub-optimization measures are formed. According to the contents of the multiple sub-optimization measures, the types of mulberry silkworms and the silkworm rearing environment, the environmental control logic of the silkworm rearing room is determined to trigger the environmental control events in the silkworm rearing room.

8. A silkworm disease prediction system based on silkworm rearing environment, characterized in that: The silkworm disease prediction system based on the silkworm rearing environment is applied to the silkworm disease prediction method based on the silkworm rearing environment as claimed in any one of claims 1 to 7, and the silkworm disease prediction system based on the silkworm rearing environment comprises: The silkworm breeding distribution map module is used to monitor the silkworm breeding room in real time and collect the silkworm breeding distribution map of the silkworm breeding room; The silkworm identification module is used to determine the silkworm activity trajectory and silkworm environment area based on the identification of the silkworm distribution map; An abnormality detection module is used to determine the silkworm rearing environment based on environmental detection of the silkworm environmental area, determine abnormal areas of the silkworm rearing room according to the silkworm rearing environment and the silkworm activity trajectory, and mark abnormal images of the silkworms, including: collecting the silkworm environmental area, determining multiple environmental detection nodes according to the silkworm environmental area and the silkworm activity trajectory, determining corresponding environmental parameters according to the environmental detection of the multiple environmental detection nodes, and determining the silkworm rearing environment according to the locations and corresponding environmental parameters of the multiple environmental detection nodes and the regional morphology of the silkworm environmental area; collecting the silkworm activity trajectory, and determining multiple silkworm activity nodes according to the division of the silkworm activity trajectory, and determining corresponding silkworm activity events according to the tracing of the multiple silkworm activity nodes, wherein the silkworm activity events cover the activity posture, rest posture and excretion of the silkworm; determining a first abnormality parameter according to the multiple silkworm activity events and the silkworm rearing environment, determining a second abnormality parameter according to the multiple silkworm activity events and the type of the silkworm, determining the abnormal area of the silkworm rearing room based on the mapping relationship between the first abnormality parameter, the second abnormality parameter and the abnormal area, and collecting abnormal images of the silkworm based on real-time monitoring of the abnormal area in the silkworm rearing room; A morphology detection module is used to determine abnormal skin features and abnormal posture features of the silkworm based on the abnormal image of the silkworm, and to determine the morphology of the silkworm's excrement by detecting the silkworm's movement trajectory; The prediction module is used to predict the type of silkworm disease based on the abnormal skin characteristics, abnormal posture characteristics and excrement morphology of the silkworm, and trigger the environmental control logic of the silkworm rearing room according to the type of silkworm disease and the silkworm rearing environment.

Images