Control method, device, and storage medium of a farming management system

Abstract

The application discloses a kind of control method, equipment and storage medium of breeding management system, it is related to animal breeding technical field, including: based on the image data collected by camera device, the current body size parameter of target cow, current estimated weight and identity mark are determined;The historical weight sequence and historical body size sequence of the target cow are extracted from the database by the identity mark;Based on historical body size sequence and current body size parameter, the standard growth model is fine-tuned, and the personalized weight growth trajectory corresponding to the target cow is generated;According to current estimated weight, historical weight sequence and personalized weight growth trajectory, determine growth condition deviation;According to growth condition deviation and current estimated weight, determine the feeding scheme of target cow, solve the problem that no individual growth data can not be formulated precise adaptive feeding scheme in large-scale breeding, adapt the fine management needs of large-scale breeding cattle.

Description

Technical Field

[0001] This application relates to the field of animal husbandry technology, and in particular to a control method, equipment and storage medium for a husbandry management system. Background Technology

[0002] In large-scale cattle farming, farmers need to adjust feeding plans based on changes in cattle weight to ensure proper growth and improve efficiency. However, current methods for obtaining cattle weight rely heavily on manual, irregular weighing. Weighing large, live cattle is complex, time-consuming, and labor-intensive, easily causing stress to the animals. Furthermore, it's difficult to conduct these weighings frequently, resulting in sparse and outdated weight data that fails to reflect the cattle's true growth status in real time. Therefore, farmers often rely on limited historical weighing records or group average growth curves for extensive feeding decisions, unable to develop personalized feeding plans tailored to each cattle's growth characteristics, and struggling to adapt to the individual growth differences in cattle raised in large-scale operations.

[0003] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a control method, equipment and storage medium for a livestock management system, which aims to solve the technical problem of cumbersome manual weighing in large-scale cattle farming, which makes it impossible to accurately formulate feeding programs suitable for individual cattle growth.

[0005] To achieve the above objectives, this application proposes a control method for an aquaculture management system, the control method comprising: Based on the image data collected by the camera device, the current body size parameters, current estimated weight, and identification of the target cow are determined; The historical weight sequence and historical body size sequence of the target cow are extracted from the database using the identification identifier; Based on the historical body size sequence and the current body size parameters, the standard growth model corresponding to the target cow is fine-tuned to generate a personalized weight growth trajectory for the target cow. Based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory, the deviation of the target cow's growth status is determined; Based on the deviation from the growth status and the current estimated weight, a feeding plan for the target cattle is determined, the feeding plan including feed formulation and feeding amount.

[0006] In one embodiment, the step of determining the current body size parameters and estimated weight of the target cow based on image data acquired by the camera device includes: The image data is preprocessed, and the preprocessed image data is then segmented to obtain valid image data containing only the target cow body. Feature point detection and extraction are performed on the effective image data to obtain the image pixel feature parameters corresponding to the body length, body height, chest circumference, abdominal circumference, hip width, and head circumference of the target cow. Based on the pixel size calibration algorithm, the image pixel feature parameters are converted into physical size parameters to obtain the current body size parameters of the target cow; The current body size parameters are input into a pre-trained cattle weight prediction model. The cattle weight prediction model calculates the current estimated weight based on the current body size parameters. The cattle weight prediction model is a random forest regression model trained based on measured weight data and corresponding measured body size parameters of the same breed of cattle as the target cattle.

[0007] In one embodiment, before the step of determining the target cow's current body size parameters, current estimated weight, and identification based on image data acquired by the camera device, the following steps are included: The target cow is monitored using the camera device to obtain monitoring images of the target cow; The target cow in the monitoring image is subjected to posture recognition to obtain the posture recognition result corresponding to the monitoring image. The posture recognition result is used to indicate whether the posture category of the target cow is a preset posture category. If the pose recognition result corresponding to the monitored image is yes, the monitored image is identified as the image data.

[0008] In one embodiment, the step of fine-tuning the standard growth model corresponding to the target cow based on the historical body size sequence and the current body size parameters to generate a personalized weight growth trajectory for the target cow includes: Obtain the breed of the target cattle, and extract the first association rule between body size and weight, the second association rule between time and body size growth rate, and the standard body size growth trajectory generated based on the second association rule from the standard growth model corresponding to the breed. Based on the historical body size sequence and the current body size parameters, the second association rule is used to predict the theoretical body size of the target cow at each time point, thereby forming a personalized body size growth trajectory for the target cow. The ratio of the personalized body size growth trajectory to the standard body size growth trajectory at each time point is calculated to obtain the body size scale factor of the target cattle. The body size scale factor is used to characterize the scaling ratio of the body size of the target cattle relative to the average body size of the same breed population. Based on the body size scale factor, the first association rule is scaled and corrected to generate a third association rule between the body size and weight of the target cow. Based on the personalized body size growth trajectory, the third association rule is used to predict the theoretical weight of the target cow at each time point, thus forming the personalized weight growth trajectory.

[0009] In one embodiment, the step of determining the deviation of the target cow's growth status based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory includes: Extract the current theoretical weight corresponding to the current time point from the personalized weight growth trajectory; Calculate the difference between the current estimated weight and the current theoretical weight, and divide the difference by the current theoretical weight to obtain the current deviation at the current time point; The historical theoretical weight within the target historical range is extracted from the personalized weight growth trajectory, and the historical estimated weight within the target historical range is extracted from the historical weight sequence; Calculate the historical difference between the historical estimated weight and the historical theoretical weight at the corresponding historical time point, and divide the historical difference by the historical theoretical weight to obtain the historical deviation at the corresponding historical time point; Determine the first weighted value corresponding to the current deviation and the second weighted value corresponding to the historical deviation; Based on the first weighted value and the second weighted value, the current deviation and the historical deviation are weighted and fused to generate the growth status deviation.

[0010] In one embodiment, the step of determining the feeding plan for the target cattle based on the deviation from the growth status and the current estimated weight includes: Based on the deviation of the growth status, determine the feeding strategy corresponding to the target cattle; Select the candidate feed formula corresponding to the feeding strategy from a set of preset candidate feed formulas as the feed formula; Calculate the feeding amount corresponding to the target cow based on the current estimated weight; The feeding amount and the feed formula are determined as the feeding plan.

[0011] In one embodiment, the step of determining the feeding plan for the target cattle based on the deviation from the growth status and the current estimated weight further includes: The feed composition database and the pre-constructed feeding rule library are retrieved. The feed composition database stores the nutritional parameters, energy values ​​and prices of various feeds. The feeding rule library includes feed matching rules and constraint prohibition rules for the same breed of cattle under different growth status labels. The deviation from the growth status is mapped using rules to determine the growth status label corresponding to the target cow. The growth status deviation, the current estimated weight, the growth status label, and the feed composition database are input into the large language model. The large language model then filters and combines various feeds in the feed composition database based on the feeding rule base to generate a list of candidate feeding programs. Based on the prices of various types of feed, the cost of each candidate feeding scheme in the candidate feeding scheme list is calculated, and the candidate feeding scheme with a cost lower than a preset cost is selected as the feeding scheme.

[0012] In one embodiment, after the step of determining the feeding plan for the target cattle based on the deviation in growth status and the current estimated weight, the method further includes: Obtain a first warning threshold and a second warning threshold, wherein the second warning threshold is greater than the first warning threshold; When the absolute value of the deviation in growth status is greater than or equal to the first warning threshold and less than the second warning threshold, the system displays the first warning information, which includes the deviation in growth status, the image data of the target cattle, and the feeding plan, and sends the first warning information to all managers via email. When the absolute value of the deviation from the growth condition is greater than or equal to the second warning threshold, the system controls the display of the second warning information and the control terminal to issue a warning sound to indicate the second warning information, and also notifies all the management personnel by telephone.

[0013] Furthermore, to achieve the above objectives, this application also proposes a control device for an aquaculture management system, the control device comprising: The weighing module is used to determine the current body size parameters, current estimated weight, and identification of the target cow based on the image data collected by the camera device. The data extraction module is used to extract the historical weight sequence and historical body size sequence of the target cow from the database using the identity identifier; The standard weight prediction module is used to fine-tune the standard growth model corresponding to the target cow based on the historical body size sequence and the current body size parameters, and generate a personalized weight growth trajectory corresponding to the target cow. The deviation calculation module is used to determine the deviation of the target cow's growth status based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory. The plan generation module is used to determine the feeding plan for the target cattle based on the deviation of the growth status and the current estimated weight. The feeding plan includes feed formulation and feeding amount.

[0014] In addition, to achieve the above objectives, this application also proposes a control device for an aquaculture management system, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the control method for the aquaculture management system as described above.

[0015] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the control method of the aquaculture management system as described above.

[0016] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the control method for the aquaculture management system described above.

[0017] The one or more technical solutions proposed in this application have at least the following technical effects: Based on image data collected by a camera device, the current body size parameters, current estimated weight, and identification of the target cattle are determined, avoiding physical weighing and achieving high-frequency, stress-free, and low-cost data collection. The historical weight sequence and historical body size sequence of the target cattle are extracted from the database using the target cattle's identification. Based on the historical body size sequence and current body size parameters, the standard growth model corresponding to the target cattle is fine-tuned to generate a personalized weight growth trajectory for the target cattle, solving the problem that the general standard model cannot match the individual growth differences of the target cattle, and providing a unique reference benchmark that fits the target cattle's own skeleton / body type for subsequent growth assessment. Based on the current estimated weight, historical weight sequence, and personalized weight growth trajectory, the deviation of the target cattle's growth status is determined, providing a data-driven and quantitative decision-making basis for adjusting personalized feeding plans. Based on the deviation of growth status and current estimated weight, a customized feeding plan is determined for the target cattle. The feeding plan includes feed formula and feeding amount, which solves the problem of not being able to formulate a precise feeding plan in large-scale farming due to the lack of individual growth data. Through personalized feeding plan, each target cattle can receive nutritional supply that matches its current growth needs, which meets the needs of refined and intelligent management in large-scale cattle farming. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an embodiment of the control method for the aquaculture management system of this application. Figure 2 A schematic diagram of the framework of a breeding management system provided in this application; Figure 3 This is a schematic diagram of the module structure of the control device of the aquaculture management system according to an embodiment of this application; Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the control method of the aquaculture management system in this application embodiment. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution of this application embodiment is as follows: Based on image data collected by a camera device, determine the current body size parameters, current estimated weight, and identification of the target cow; extract the historical weight sequence and historical body size sequence of the target cow from the database using the identification; fine-tune the standard growth model corresponding to the target cow based on the historical body size sequence and current body size parameters to generate a personalized weight growth trajectory for the target cow; determine the deviation of the target cow's growth status based on the current estimated weight, historical weight sequence, and personalized weight growth trajectory; and determine the feeding plan for the target cow based on the deviation of the growth status and the current estimated weight, the feeding plan including feed formulation and feeding amount.

[0024] In this embodiment, for ease of description, the following description uses the aquaculture management system as the executing entity.

[0025] In large-scale cattle farming, current technology requires farmers to adjust feeding plans based on cattle weight changes to ensure proper growth and improve efficiency. However, current methods rely heavily on manual, irregular weighing. Weighing large, live cattle is complex, time-consuming, and stressful, and cannot be performed frequently, resulting in sparse and outdated weight data that fails to reflect the cattle's true growth status in real time. Consequently, farmers often rely on limited historical weighing records or group average growth curves for extensive feeding decisions, making it impossible to develop personalized feeding plans for each cattle to suit their individual growth characteristics and adapt to the individual growth differences in large-scale farming.

[0026] This application provides a solution that, based on image data collected by a camera device, determines the current body size parameters, current estimated weight, and identification of a target cattle. This eliminates the tedious manual weighing process, achieving automated and efficient collection of target cattle weight data, while providing accurate individual baseline data for subsequent analysis. Next, relying on the identification, the historical weight and body size sequences of the target cattle are retrieved. By comparing the individual's historical body size sequence with the current body size parameters, the standard growth model corresponding to the target cattle is fine-tuned, generating a unique personalized weight growth trajectory. This solves the problem that general growth models cannot match individual growth differences, establishing a growth judgment benchmark that fits the target cattle themselves. Subsequently, by combining the target cattle's current estimated weight, historical weight sequence, and personalized weight growth trajectory, the deviation from the target cattle's growth status is quantitatively determined, transforming growth status judgment from experience-based to data-driven, forming a scientific basis for adjusting feeding programs. Finally, based on the growth deviation and the current estimated weight, the feed formula and feeding amount are determined in a targeted manner, and a precise feeding plan containing the feed formula and feeding amount is dynamically generated. This solves the technical problem that the data is missing due to the cumbersome manual weighing, which makes it impossible to accurately formulate a feeding plan that is suitable for the individual growth of cattle, and realizes precise, timely and individualized feeding control.

[0027] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or a control device for an aquaculture management system capable of the above functions. The following description uses an aquaculture management system as an example to illustrate this embodiment and the subsequent embodiments.

[0028] Based on this, the embodiments of this application provide a control method for an aquaculture management system, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the control method for the aquaculture management system of this application.

[0029] In this embodiment, the control method of the aquaculture management system is applied to the data processing terminal of the aquaculture management system. The aquaculture management system includes a data processing terminal and a camera device that is communicatively connected to the data processing terminal. The connection relationship of the aquaculture management system is: camera device - local gateway - data processing terminal.

[0030] The camera device can be a camera deployed in the cattle shed passageway or identification area to continuously collect video or image streams of cattle. The camera device connects to the local gateway via wired (e.g., Ethernet) or wireless (Wi-Fi / 4G / 5G) methods. One end of the local gateway connects to the camera device, and the other end connects to the data processing terminal via a local area network (LAN) or the internet. The data processing terminal can be deployed as a server in a local data center or cloud platform, running algorithms for image analysis, weight prediction, growth modeling, and feeding decisions. Alternatively, the data processing terminal can be integrated into a display terminal, directly running processing and display functions on a high-performance terminal (e.g., a workstation). The display terminal can show weight data, curves, and monitoring data. Users can select a target cow in the monitoring video by clicking or other operations, and then jump to display the relevant data of that target cow. The data processing terminal receives data from the local gateway, processes it, and pushes the results to the user terminal, such as a computer, tablet, or mobile phone. Figure 2 , Figure 2 This is a schematic diagram of the framework of a breeding management system provided in this application.

[0031] In this embodiment, the control method of the aquaculture management system includes steps 101-105: Step 101: Based on the image data collected by the camera device, determine the current body size parameters, current estimated weight, and identification of the target cow.

[0032] Specifically, the camera device refers to image acquisition equipment deployed in cattle sheds, passageways, or feeding areas to acquire optical images of the target cattle from the front, side, etc. The image data consists of two-dimensional or three-dimensional digital images of the target cattle captured by the camera device, including appearance information such as body shape, outline, and body size. Current body size parameters are the current body size data of the target cattle measured from the images, including body length, height, chest circumference, abdominal circumference, hip width, and head circumference. The identification identifier is a unique ID for each target cattle, which can be derived from radio frequency identification (RFID) ear tags, facial / stripe biometric matching, or image recognition binding, used to accurately locate individual files in the database.

[0033] In some embodiments, optical image data of target cattle is collected by camera devices deployed in the breeding area. Based on image recognition and body shape analysis technology, non-contact weight prediction is performed on the target cattle. First, the current body size parameters such as body length, height, and chest circumference of the target cattle are extracted from the images. Then, the corresponding estimated weight of the target cattle is calculated using the correspondence between body size and weight. Individual growth data of the target cattle can be obtained without manual weighing. Simultaneously, image recognition algorithms (such as deep learning-based target detection models) are used to analyze the images and match facial features or RFID visual tags on the cattle's body surface, thereby quickly identifying the unique identifier of the target cattle. Using camera devices to collect images of target cattle non-contactly, and obtaining body size parameters and estimated weight through image analysis, replaces traditional manual weighing and simultaneously achieves cattle identification. This enables automated, efficient, and accurate collection of individual growth data in large-scale farming, providing accurate and continuous basic individual data for subsequent retrieval of historical data, adaptation to standard growth models, determination of growth deviation, and generation of personalized feeding plans.

[0034] Optionally, before determining the target cow's current body size parameters, current estimated weight, and identification based on image data acquired by the camera device, the following steps are included: The target cattle are monitored using a camera device to obtain monitoring images of the target cattle; The pose recognition of the target cow in the monitoring image is performed to obtain the pose recognition result corresponding to the monitoring image. The pose recognition result is used to indicate whether the pose category of the target cow is a preset pose category. If the pose recognition result corresponding to the monitored image is yes, the monitored image is identified as image data.

[0035] Specifically, the monitoring images are raw image frames containing the target cow captured in real time by the camera device. These images may contain various postures (such as head down, turning around, occlusion, etc.) and vary in quality. Posture recognition uses a computer vision model to determine the target cow's body orientation and standing position in the monitoring images, such as whether it is facing the camera directly and whether its limbs are naturally upright. Preset posture categories refer to ideal postures suitable for body size measurement, such as the target cow's frontal or lateral position, head raised, limbs not crossed, and torso unobstructed. Body size parameters of the target cow are most accurately extracted under preset posture categories. Image data (valid images) are high-quality images that meet the measurement requirements and are retained after posture screening, serving as input for subsequent body size extraction and weight prediction.

[0036] For example, a camera device deployed in the target cattle pen passage or identification area continuously monitors the target cattle, acquiring multiple frames of monitoring images containing the target cattle's body in real time. Each frame is then input into a pre-trained target cattle posture recognition model, such as a deep learning-based convolutional neural network or keypoint detection model. The target cattle posture recognition model can automatically analyze the target cattle's body orientation, head position, and limb posture in the image, and output corresponding posture classification results, such as standard lateral position, forward position, head down, turning, or occlusion posture categories. The standard lateral or forward position can be defined as a preset posture category, i.e., an ideal posture suitable for body size parameter extraction, where the cattle's outline is complete, the torso has no overlapping occlusion, the limbs are naturally standing, and the spine is basically horizontal. The system determines whether the posture classification result of the current monitoring image belongs to the preset posture category, obtaining a yes or no posture recognition result. If it is yes, the monitoring image is marked as valid image data and sent to the subsequent body size parameter extraction and weight prediction steps; if it is no, the monitoring image is discarded or temporarily stored for recapture. By introducing a posture screening mechanism, the distortion of body size measurement caused by non-ideal postures of the target cattle (such as head down eating, turning around to walk, or overlapping with other target cattle) can be effectively avoided, thereby improving the stability and reliability of the weight estimation results and laying a high-quality data foundation for generating accurate personalized growth trajectories and feeding programs.

[0037] Step 102: Extract the historical weight sequence and historical body size sequence of the target cow from the database using the target cow's identification.

[0038] Optionally, the breed of the target cattle can be obtained, and a standard growth model corresponding to the breed can be extracted as the standard growth model for the target cattle.

[0039] Specifically, the database is a structured data system storing the entire lifecycle breeding data of all target cattle, which may include identification ID, breed, date of birth, historical body measurements, estimated / measured weight, feeding records, etc. The historical weight sequence is a chronologically ordered list of all past weight data for the target cattle (which may include image-based weight estimates or occasional measured values), reflecting the actual growth trajectory of the target cattle. The historical body measurement sequence is a chronologically ordered list of past body measurement parameters for the target cattle, such as chest circumference, body length, and height, used to analyze body shape development trends. The breed refers to the genetic type of the target cattle, such as Simmental, Angus, Holstein, etc. The standard growth model is a set of body measurement-weight association rules, time-body measurement growth rate rules, and standard body measurement growth trajectories constructed based on growth data of the same breed population.

[0040] In some embodiments, after obtaining the identity of the current target cattle, the identity is used as a database query key to retrieve the historical records of the target cattle's entire life cycle from the livestock management system's database. This includes weight data sorted by timestamp (i.e., historical weight sequences) and corresponding body size parameters such as body length, height, chest circumference, abdominal circumference, hip width, and head circumference (i.e., historical body size sequences). Simultaneously, based on the breed information associated with the identity (e.g., Simmental, Angus, etc.), the system loads a standard growth model matching the breed from a pre-set model library. This model describes the typical evolution of the breed's weight or body size over time under ideal feeding conditions. The historical weight and body size sequences reflect the actual growth trajectory of the target cattle, while the standard growth model provides a priori baseline for the population. The historical weight and body size sequences, along with the standard growth model, can all serve as input for subsequent personalized modeling. By acquiring historical weight sequences, historical body size sequences, and standard growth models, it is helpful to organically integrate current observations with individual history and breed standards. This ensures that the subsequent growth assessment (personalized weight growth trajectory) takes into account individual developmental differences and conforms to biological laws, thus providing a reliable and structured data foundation for accurately judging nutritional status and formulating feeding programs.

[0041] Step 103: Based on the historical body size sequence and the current body size parameters, fine-tune the standard growth model corresponding to the target cow to generate a personalized weight growth trajectory for the target cow.

[0042] Specifically, fine-tuning refers to calibrating or adjusting the standard growth model based on the actual body size and development of the target cow to match its innate skeletal size or developmental phase, rather than directly applying the population curve. The personalized weight growth trajectory is a reference curve generated after adaptation, unique to the target cow, showing its weight changes over time. It represents "the weight that should be achieved at the standard growth rate under the actual skeletal structure," and is used to assess whether the actual weight has met the target.

[0043] In some embodiments, historical body size sequences and current body size parameters are used to reflect the past and present body size development of the target cattle. While maintaining the time-body size growth rate in the standard growth model, the body size-weight correlation rules within the model are individually adapted, transforming the general population standard into an individual standard tailored to the target cattle's body type. This ultimately yields a personalized weight growth trajectory reflecting the target cattle's expected standard weight at each time point under the standard growth rate. By converting general population growth patterns into an individualized weight reference benchmark that aligns with the actual skeletal development of the target cattle, a precise and specific reference benchmark is provided for subsequent calculations of growth deviation, avoiding judgment errors caused by using population standards and thus improving the accuracy of feeding program development.

[0044] Step 104: Determine the deviation of the target cow's growth status based on the current estimated weight, historical weight sequence, and personalized weight growth trajectory.

[0045] Specifically, the deviation of growth status refers to a numerical indicator calculated using specific mathematical methods to quantify the degree to which the "current estimated weight" and "historical weight sequence" deviate from the "personalized weight growth trajectory".

[0046] In some embodiments, the current specific standard weight (current theoretical weight) corresponding to the current time point and the historical specific standard weight (historical theoretical weight) matched with each time point of the historical weight sequence are first extracted from the personalized weight growth trajectory. The static deviation rate between the current estimated weight and the current specific standard weight is calculated, and then the dynamic average deviation rate between the historical weight sequence and the corresponding historical specific standard weight is calculated. Preset weights are assigned to the two types of deviation rates, and the growth status deviation is obtained by weighted summation. This transforms the growth status of the target cattle from a qualitative description to a quantitative judgment, accurately reflecting the difference between actual growth and its own standard, avoiding misjudgments caused by group standards, and providing a direct decision-making basis for the subsequent generation of personalized feeding plans.

[0047] Step 105: Determine the feeding plan for the target cattle based on the deviation in growth status and the current estimated weight. The feeding plan includes feed formulation and feeding amount.

[0048] Specifically, the current estimated weight is the latest weight data of the target cattle, which determines the baseline nutritional requirements for adjusting the feeding program. The feeding program is a set of structured operational parameters, mainly including feed formulation and feed volume. The feed formulation refers to the proportion of various feed ingredients (such as corn, soybean meal, alfalfa, and premix) that make up the daily ration. The feed volume refers to the total weight of the above-mentioned feed formulation that should be provided to the target cattle daily.

[0049] In some embodiments, based on the current estimated weight and nutritional requirement model, the basal energy, crude protein, and other nutritional indicators required for the target cattle to maintain and gain weight are determined. Subsequently, the nutritional requirements are adjusted according to the direction and magnitude of the deviation from the growth status. For example, if the deviation is negative (underweight), the energy and protein supply levels are increased; if the deviation is positive (overweight), the energy density is reduced and the proportion of roughage is increased. Next, using a feed composition database and the adjusted nutritional requirements as constraints, a nutritionally sound and cost-effective feed formula is generated according to preset rules. The feeding amount is then calculated based on the energy concentration and total nutritional requirements of the feed formula. By transforming the growth assessment results (growth status deviation) into executable, precise feeding instructions, dynamic matching of nutritional supply and individual growth goals is achieved, avoiding the one-size-fits-all feeding methods of traditional farming and improving feeding efficiency and growth outcomes.

[0050] Based on the control method of the livestock management system provided in this application, the current body size parameters, current estimated weight, and identification of the target cattle are determined based on image data collected by a camera device, avoiding physical weighing and achieving high-frequency, stress-free, and low-cost data collection. The historical weight and body size sequences of the target cattle are extracted from the database using their identification. Based on the historical body size sequences and current body size parameters, the standard growth model corresponding to the target cattle is fine-tuned to generate a personalized weight growth trajectory for each target cattle. This solves the problem that the general standard model cannot match the individual growth differences of the target cattle, providing a unique reference benchmark that fits the target cattle's own skeletal structure / body type for subsequent growth assessment. Based on the current estimated weight, historical weight sequence, and personalized weight growth trajectory, the deviation of the target cattle's growth status is determined, providing a data-driven and quantitative decision-making basis for adjusting personalized feeding plans. Based on the deviation of growth status and the current estimated weight, a unique feeding plan is determined for the target cattle. The feeding plan includes feed formulation and feeding amount, which solves the problem of not being able to formulate a precise feeding plan in large-scale farming due to the lack of individual growth data. Through personalized feeding plans, each cattle can receive nutritional supply that matches its current growth needs, which meets the needs of refined and intelligent management in large-scale cattle farming.

[0051] In some embodiments, the steps of determining the current body size parameters and current estimated weight of the target cow based on image data acquired by the camera device include: The image data is preprocessed, and the background of the preprocessed image data is segmented to obtain effective image data containing only the target cow body; Feature point detection and extraction are performed on the effective image data to obtain the image pixel feature parameters corresponding to the target cow's body length, body height, chest circumference, abdominal circumference, hip width, and head circumference; Based on the pixel size calibration algorithm, the image pixel feature parameters are converted into physical size parameters to obtain the current body size parameters of the target cow; The current body size parameters are input into a pre-trained cattle weight prediction model. The cattle weight prediction model calculates the current estimated weight based on the current body size parameters. The cattle weight prediction model is a random forest regression model trained based on measured weight data and corresponding measured body size parameters of the same breed of cattle.

[0052] Specifically, preprocessing includes denoising, illumination equalization, and image enhancement to improve the robustness of subsequent segmentation and detection. Background segmentation uses semantic segmentation models (such as U-Net) to separate the foreground of the target cow from the complex background of the target cowshed, obtaining a masked image containing only the main body of the target cow (effective image data) and avoiding background interference. Feature point detection and extraction uses keypoint detection algorithms, such as High-Resolution Network (HRNet) and Deep Labelling of Cut-out Body Parts (DeepLabCut), to locate anatomical landmarks of the target cow (such as the scapula, hip tuberosity, and bottom of the chest) for calculating the geometric distance of the body size. Pixel size calibration algorithms are used to establish the mapping relationship between image pixels and actual physical dimensions, converting pixel values ​​into true physical lengths. As an example, the raw image data acquired by the camera device is preprocessed, including Gaussian filtering for noise reduction, illumination equalization, and contrast enhancement, to improve image quality. Subsequently, a pre-trained semantic segmentation model is used to perform background segmentation on the preprocessed image, generating a binary mask containing only the target cattle, thus obtaining effective image data unaffected by fences, the ground, or other target cattle. Next, a target cattle body keypoint detection network, such as a customized model based on the HRNet architecture, is run on this effective image to detect and extract key feature points of the target cattle. This automatically locates anatomical landmarks such as the scapula, hip tuberosity, withers, midpoint of the chest floor, and tail root, and calculates the corresponding pixel distances or contour parameters for body length (from nose tip to tail root), body height (from withers to the ground), chest circumference (perimeter fitted by an ellipse of the chest cross-section), abdominal circumference, hip width, and head circumference. The corresponding image pixel feature parameters for each dimension of body size are then calculated. Next, a pixel size calibration algorithm is employed. Using a pre-calibrated mapping relationship between physical dimensions and image pixels, the aforementioned pixel feature parameters are converted into actual physical size parameters, thus obtaining the accurate current body size parameters of the target cattle. Finally, these body size parameters are input into a pre-trained cattle weight prediction model. This model is a random forest regression model, and its training data comes from historical measured records of the same breed of cattle, including multiple sets of simultaneously acquired measured weight data (actual weight) and their corresponding measured body size parameters. During training, the random forest regression model learns the complex nonlinear relationship between body size and weight (such as the product effect of the square of chest circumference and body length), enabling it to robustly predict the current estimated weight. The pre-trained cattle weight prediction model performs inference calculations based on the input current body size parameters and outputs the corresponding current estimated weight.

[0053] By using image analysis and machine learning models, non-contact measurement of body size and weight prediction of target cattle is achieved, realizing automated, stress-free, and non-contact body size and weight collection, replacing traditional manual weighing, improving data collection efficiency and accuracy, and providing reliable basic data for subsequent personalized growth model adaptation, growth deviation judgment, and feeding program generation.

[0054] In some embodiments, the step of fine-tuning the standard growth model corresponding to the target cow based on historical body size sequences and current body size parameters to generate a personalized weight growth trajectory for the target cow includes: Obtain the breed of the target cattle, and extract the first association rule between body size and weight, the second association rule between time and body size growth rate, and the standard body size growth trajectory generated based on the second association rule from the standard growth model corresponding to the breed. Based on historical body size sequences and current body size parameters, the second association rule is used to predict the theoretical body size of the target cow at each time point, thus forming a personalized body size growth trajectory for the target cow. The ratio of the personalized body size growth trajectory to the standard body size growth trajectory is calculated at each time point to obtain the body size scale factor of the target cattle. The body size scale factor is used to characterize the scaling ratio of the target cattle's body size relative to the average body size of the same breed population. The first association rule is scaled and corrected based on the body size factor to generate a third association rule between the body size and weight of the target cow. Based on personalized body size growth trajectories, the third association rule is used to predict the theoretical weight of the target cattle at each time point, thus forming a personalized weight growth trajectory.

[0055] Specifically, the first association rule is the mapping relationship between body size and weight of the target cattle breed group; the second association rule is the relationship between time and body size growth rate; the standard body size growth trajectory is the body size change curve over time under the average level of the group; the historical body size sequence and the current body size parameters are the actual body size data of the target cattle individual; and the personalized body size growth trajectory is the body size change trajectory fitted based on individual data.

[0056] As an example, from the standard growth model of the target cattle breed, the first correlation rule between body size and weight, the second correlation rule between time and body size growth rate, and the standard body size growth trajectory of the target cattle's average body size over time, generated based on the growth rate rule, are precisely extracted to clarify the growth pattern benchmark at the group level. Then, based on the target cattle's historical body size sequence and combined with current body size parameters, the second correlation rule from the standard growth model is used to maintain the group's standard growth rate. The scaling relationship between actual body size data and the standard body size curve is fitted to deduce the theoretical body size value of the target cattle at each time point, forming a personalized body size growth trajectory that fits the individual's growth rhythm and reflects its own developmental characteristics. Finally, the personalized body size growth trajectory is compared with the standard body size growth trajectory at each corresponding time point to obtain a body size scale factor that characterizes the scaling ratio of the target cattle's body size relative to the average body size of the same breed group, quantifying the body size differences between individuals and the group. Based on this body size factor, the first association rule between body size and weight extracted from the standard growth model is scaled and corrected. The general body size-weight mapping relationship is adjusted to a specific association rule adapted to the target cow's body shape characteristics, namely the third association rule, ensuring that the correspondence between body size and weight matches the individual's actual situation. Finally, using the personalized body size growth trajectory as input, the theoretical weight of the target cow at each time point is predicted point by point through the corrected specific body size-weight association rule (i.e., the third association rule), ultimately forming a complete and accurate personalized weight growth trajectory. Through rule extraction, trajectory fitting, and parameter correction, a personalized weight growth trajectory that fits the target cow's own physique development characteristics is generated, providing an accurate reference for subsequent growth status assessment. This ensures that the subsequent assessment of growth deviation is more consistent with the actual growth situation of the target cow, avoiding misjudgments caused by group standards and individual differences.

[0057] In some embodiments, the step of determining the deviation of the target cow's growth status based on the current estimated body weight, historical body weight sequence, and personalized body weight growth trajectory includes: Extract the current theoretical weight corresponding to the current time point from the personalized weight growth trajectory; Calculate the difference between the current estimated weight and the current theoretical weight, and divide the difference by the current theoretical weight to obtain the current deviation at the current time point; Extract multiple historical theoretical weights within the target historical range from personalized weight growth trajectories, and extract multiple historical predicted weights within the target historical range from historical weight sequences. Calculate the historical difference between each historical estimated weight and the corresponding historical theoretical weight at each historical time point, and divide the historical difference by the historical theoretical weight to obtain the historical deviation for each historical time point. Determine the first weighted value corresponding to the current deviation and the second weighted value corresponding to each historical deviation. The smaller the time difference between the historical time point corresponding to the historical deviation and the current time point, the larger the value of the second weighted value. Based on the first weighted value and the second weighted value, the current deviation and the historical deviation are weighted and fused to generate the growth status deviation.

[0058] Specifically, a preset time window (such as the last 15 days or the last 5 records) is used to limit the range of historical data to be fused, ensuring relevance. The first weighting value is the weight assigned to the current deviation. (e.g., 0.5), reflecting the importance of the current state; the second weighting value is the weight assigned to each historical deviation. Its value decreases exponentially as the time difference between the historical time point and the current time point increases, meaning that the more recent the historical data, the higher the weight.

[0059] As an example, the current theoretical weight, which perfectly corresponds to the current time point, is accurately extracted from the personalized weight growth trajectory. This current theoretical weight is a unique standard weight generated based on the target cow's own body shape characteristics. Then, a formula is used... = (Current estimated weight - Current theoretical weight) / Current theoretical weight. This calculates the current deviation, visually reflecting the relative difference between the current actual weight and the specified standard. This represents the current deviation. Next, a target historical range is set, and multiple historical theoretical weights within this range are extracted from the personalized weight growth trajectory. Simultaneously, multiple historical estimated weights at corresponding time points are extracted from the historical weight sequence, ensuring a one-to-one correspondence between the data's time dimension; this is achieved through the formula... = (Historical deviation at the i-th historical time point - Historical estimated weight at the i-th historical time point) / Historical estimated weight at the i-th historical time point To determine the historical deviation at the i-th historical time point, calculate the historical deviation for each historical time point to capture the historical growth deviation pattern. Then, set the first weighting value for the current deviation. (e.g., 0.5), and assign a second weighting value to each historical deviation according to the rule that "the smaller the time difference between the historical time point and the current time point, the greater the weight". If exponentially decaying weights are used... , Let be the time difference between the i-th historical time point and the current time point. Let be the attenuation coefficient, and satisfy... Finally, through the formula The current deviation is weighted and fused with each historical deviation to generate the final growth status deviation. .

[0060] To avoid the random errors of data from a single point in time, a dynamic weighted fusion algorithm is used to comprehensively consider the current growth status and historical growth trends, accurately quantifying the degree of deviation of the actual growth of the target cattle from the individual-specific standard. This enables a comprehensive and objective assessment of the deviation, thereby providing a scientific basis for subsequent adjustments to the feeding program.

[0061] In some embodiments, the step of determining the feeding program for the target cattle based on the deviation from growth status and the current estimated weight includes: Determine the feeding strategy for the target cattle based on the degree of deviation in growth status; Select the candidate feed formula corresponding to the feeding strategy from a set of preset candidate feed formulas as the feed formula; Calculate the feed amount corresponding to the target cattle based on the current estimated weight; The feeding amount and feed formula are determined as the feeding plan.

[0062] Specifically, the feeding strategy is a nutritional adjustment direction determined based on growth deviation, such as significantly promoting growth, moderately promoting growth, moderately controlling weight, and maintaining the status quo. The candidate feed formulations are pre-constructed multiple feed ratio schemes adapted to different feeding strategies. Each scheme includes the proportion of raw materials (e.g., 60% corn, 20% soybean meal, 18% alfalfa, and 2% premix) and the corresponding nutrient concentrations (ME, CP, etc.), which correspond one-to-one with the feeding strategy.

[0063] As an example, based on the value and sign of the deviation from the growth status, the corresponding feeding strategy category is determined from preset rules. For instance, if the deviation from the growth status is negative and significantly lower than the normal range (e.g., lower than...), the feeding strategy category is determined. If the growth rate is 5%, it is considered a significant growth retardation, and the "significant growth promotion" strategy can be selected; if the growth status deviation is negative and slightly below the normal range (e.g., below 5%), it is considered a significant growth retardation. 3%) is considered a slight growth retardation, and a "moderate growth promotion" strategy can be chosen; if it is within the normal range (e.g. If the growth deviation is between 5% and +3%, the "Maintain Status quo" strategy is selected; if the deviation is positive and higher than the normal range (e.g., higher than +3%), the "Moderate Weight Control" strategy is selected. Subsequently, the system retrieves a standard feed formula linked to this feeding strategy from a pre-configured library of candidate feed formulas. This formula includes fixed proportions and nutrient concentrations of ingredients such as corn, soybean meal, alfalfa, and premixes. Next, based on the current estimated weight and a nutritional requirement model (e.g., NRC standards), the system calculates the target cow's daily net energy requirement and, based on the energy concentration of the selected feed formula, deduces the daily feed intake to meet nutritional needs. Finally, the selected feed formula and the calculated feed intake are combined to form a structured feeding plan. By transforming abstract growth deviation data into specific, operable feeding plans, this system achieves refined feeding for each cow, precisely meeting the individual growth needs of the target cattle, thereby helping to improve breeding efficiency and growth quality.

[0064] In some embodiments, the step of determining a feeding program for a target cattle based on the deviation from growth status and the current estimated weight further includes: The feed composition database and the pre-built feeding rule library are accessed. The feed composition database stores the nutritional parameters, energy values ​​and prices of various feeds. The feeding rule library includes feed matching rules and restrictions / prohibition rules for the same breed of cattle under different growth status labels. The deviation from growth status is mapped by rules to determine the growth status label corresponding to the target cow; The deviation of growth status, current estimated weight, growth status label and feed composition database are input into the big language model. The big language model filters and combines various feeds in the feed composition database based on the feeding rule base to generate a list of candidate feeding programs. Based on the prices of various feeds, the cost of each candidate feeding scheme in the candidate feeding scheme list is calculated, and the candidate feeding scheme with a cost lower than the preset cost is selected as the feeding scheme.

[0065] Specifically, the feed composition database is a structured database storing nutritional parameters (crude protein, crude fiber, etc.), energy values, and market prices of various feeds (such as corn, soybean meal, alfalfa, etc.). The feeding rule base refers to a set of rules containing feed adaptation rules (such as the required proportion of high-protein feed during weight gain stages) and restrictive rules (such as prohibited feed types at specific stages) for cattle of the target breed at different growth stages. The Large Language Model (LLM) is a language model fine-tuned for the livestock industry, such as Llama-3 or Qwen, capable of understanding feed nutrition, rule logic, and optimization objectives, used to generate reasonable formula combinations under rule constraints. Growth status labels are the results of semantic classification of deviations in growth status, such as severely underweight, slightly overweight, and normal, facilitating rule matching and LLM understanding.

[0066] As an example, a pre-built feed composition database and feeding rule base are retrieved. The feed composition database stores the nutrient parameters, energy values, and unit prices of various feeds such as corn, soybean meal, and alfalfa. The feeding rule base contains adaptation rules for the same breed of cattle at different growth stages and conditions (e.g., crude protein ≥14% for lean fattening cattle) and restrictive rules. Subsequently, the deviation of continuous numerical growth status (e.g., ...) is used to... The 6.2% deviation rate is transformed into semantic growth status labels (e.g., "severely underweight") through rule mapping. Next, the deviation rate of growth status, current estimated weight, growth status labels, and feed composition database are used as contextual inputs to a large language model fine-tuned for the livestock field. Constraints from the feeding rule base are explicitly embedded in the prompt. Based on its understanding of nutritional logic and rules, the large language model filters compliant raw materials from the feed composition database and generates multiple candidate feeding plans that meet nutritional requirements and safety restrictions. Each candidate feeding plan includes feed type, ratio, estimated nutritional level, and daily feed amount. Finally, the system calculates the daily feeding cost of each candidate feeding plan based on the price of various feeds and determines the plan with a cost lower than the preset cost (e.g., 8 yuan / day) as the final feeding plan. If multiple plans meet the criteria, the plan with the highest nutritional suitability is selected first. While ensuring nutritional scientific accuracy and breed suitability, a large language model is used to perform semantic understanding and constraint reasoning on massive feed combinations to generate multiple compliant, feasible and low-cost candidate feeding schemes. The optimal solution of the feeding scheme is then selected based on economic factors, achieving precise, economical and executable intelligent feeding.

[0067] In some embodiments, after determining the feeding program for the target cattle based on the deviation from growth status and the current estimated weight, the method further includes: Obtain a first warning threshold and a second warning threshold, wherein the second warning threshold is greater than the first warning threshold; When the absolute value of the deviation in growth status is greater than or equal to the first warning threshold and less than the second warning threshold, the control displays the first warning information, which includes the deviation in growth status, image data of the target cattle and the feeding plan, and sends the first warning information to all managers via email. When the absolute value of the deviation in growth status is greater than or equal to the second warning threshold, the control system displays the second warning information and issues an alarm sound to the user terminal to indicate the second warning information. It also notifies all managers by telephone. The second warning information includes the deviation in growth status, image data of the target cattle, and feeding plan.

[0068] Specifically, the first warning threshold refers to the critical value of deviation in growth status that triggers a primary warning (e.g., 0.1). The second warning threshold refers to the critical value that triggers a high-level warning (e.g., 0.2), and the value must be greater than the first warning threshold. The user end refers to the terminal equipment used by management personnel, such as aquaculture monitoring platforms, mobile apps, etc.

[0069] As an example, the system obtains a preset first warning threshold and a second warning threshold (e.g., the first warning threshold is set to 0.1 and the second warning threshold is set to 0.2). Then, it determines the relationship between the absolute value of the deviation in growth status and the two thresholds: if the absolute value of the deviation in growth status is ≥ the first warning threshold and < the second warning threshold (e.g., -0.15, 0.18), the first warning is triggered. On the one hand, the system controls the breeding monitoring platform or display terminal to display the first warning information, which includes the specific value of the deviation in growth status, the original image data of the target cattle, and the optimized feeding plan. On the other hand, the system automatically sends the first warning information to the email addresses of all managers through preset email addresses to ensure that everyone is aware of it. If the absolute value of the deviation in growth status is less than the first warning threshold, the deviation, image data of the target cattle, and feeding plan are simply sent to the email addresses of all managers to notify them. If the absolute value of the deviation is greater than or equal to the second warning threshold (e.g., -0.25, 0.3), a second warning is triggered. This not only controls the display terminal to show the second warning information, including the deviation in growth status, image data of the target cattle, and feeding plan, but also controls the managers' user terminals (e.g., mobile APP, monitoring terminal) to emit an alarm sound (e.g., continuous beeping). Simultaneously, a pre-set telephone notification system automatically dials all managers' phones to inform them of the second warning information and key details via voice broadcast. By establishing a tiered response mechanism, differentiated warning methods are adopted for different degrees of growth abnormalities. Primary abnormalities are addressed through email and display warnings to ensure information delivery, while severe abnormalities are addressed through audible and visual warnings and telephone notifications to force reminders. This ensures that managers can quickly grasp the abnormal situation and intervene in a timely manner according to the attached feeding plan, effectively reducing breeding risks and improving the efficiency of abnormality handling.

[0070] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the control method of the aquaculture management system of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0071] This application also provides a control device for an aquaculture management system; please refer to [reference needed]. Figure 3 The control devices of the aquaculture management system include: The weighing module 301 is used to determine the current body size parameters, current estimated weight, and identification of the target cow based on the image data collected by the camera device. Data extraction module 302 is used to extract the historical weight sequence and historical body size sequence of the target cow from the database through the identity identifier; The standard weight prediction module 303 is used to fine-tune the standard growth model corresponding to the target cow based on the historical body size sequence and the current body size parameters, and generate a personalized weight growth trajectory for the target cow. The deviation calculation module 304 is used to determine the deviation of the target cow's growth status based on the current estimated weight, historical weight sequence, and personalized weight growth trajectory. The plan generation module 305 is used to determine the feeding plan for the target cattle based on the deviation of growth status and the current estimated weight. The feeding plan includes feed formulation and feeding amount.

[0072] The control device for the livestock management system provided in this application, employing the control method of the livestock management system in the above embodiments, can solve the technical problem in large-scale cattle farming where manual weighing is cumbersome and makes it impossible to accurately formulate feeding plans suitable for individual cattle growth. Compared with the prior art, the beneficial effects of the control device for the livestock management system provided in this application are the same as those of the control method for the livestock management system provided in the above embodiments, and other technical features in the control device for the livestock management system are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0073] This application provides a control device for an aquaculture management system. The control device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the control method of the aquaculture management system in the above embodiment 1.

[0074] The following is for reference. Figure 4The diagram illustrates a structural schematic of a control device suitable for implementing the aquaculture management system of the embodiments of this application. The control device for the aquaculture management system in the embodiments of this application may include, but is not limited to, mobile terminals such as laptops, tablets (Portable Application Description, PADs), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 4 The control device of the aquaculture management system shown is merely an example and should not impose any limitation on the function and scope of use of the embodiments of this application.

[0075] like Figure 4 As shown, the control device of the aquaculture management system may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1002 or a program loaded from storage device 1003 into random access memory (RAM) 1004. The random access memory 1004 also stores various programs and data required for the operation of the control device of the aquaculture management system. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the control equipment of the aquaculture management system to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows the control equipment of an aquaculture management system with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented or possessed alternatively.

[0076] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0077] The control device for the livestock management system provided in this application, employing the control method of the livestock management system in the above embodiments, can solve the technical problem in large-scale cattle farming where manual weighing is cumbersome and makes it impossible to accurately formulate feeding plans suitable for individual cattle growth. Compared with the prior art, the beneficial effects of the control device for the livestock management system provided in this application are the same as those of the control method for the livestock management system provided in the above embodiments, and other technical features of the control device for this livestock management system are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0078] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0079] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0080] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the control method of the aquaculture management system in the above embodiments.

[0081] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0082] The aforementioned computer-readable storage medium may be included in the control equipment of the aquaculture management system; or it may exist independently and not be installed in the control equipment of the aquaculture management system.

[0083] The aforementioned computer-readable storage medium carries one or more programs that, when executed by the control equipment of the livestock management system, cause the control equipment of the livestock management system to: determine the current body size parameters, current estimated weight, and identification of the target cattle based on image data collected by the camera device; extract the historical weight sequence and historical body size sequence of the target cattle from the database using the identification; fine-tune the standard growth model corresponding to the target cattle based on the historical body size sequence and the current body size parameters to generate a personalized weight growth trajectory corresponding to the target cattle; determine the growth status deviation of the target cattle based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory; and determine the feeding plan for the target cattle based on the growth status deviation and the current estimated weight, wherein the feeding plan includes feed formulation and feeding amount.

[0084] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0085] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation that may be implemented in systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0086] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0087] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., computer programs) for executing the control method of the above-described livestock management system. This solves the technical problem in large-scale cattle farming where manual weighing is cumbersome and makes it impossible to accurately formulate feeding plans suitable for individual cattle growth. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the control method of the livestock management system provided in the above embodiments, and will not be repeated here.

[0088] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the control method for the aquaculture management system described above.

[0089] The computer program product provided in this application can solve the technical problem of cumbersome manual weighing in large-scale cattle farming, which makes it impossible to accurately formulate feeding programs suitable for the individual growth of cattle. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the control method of the breeding management system provided in the above embodiments, and will not be repeated here.

[0090] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A control method for an aquaculture management system, characterized in that, The data processing terminal of the aquaculture management system is used in the control method of the aquaculture management system, which includes: Based on the image data collected by the camera device, the current body size parameters, current estimated weight, and identification of the target cow are determined; The historical weight sequence and historical body size sequence of the target cow are extracted from the database using the identification identifier; Based on the historical body size sequence and the current body size parameters, the standard growth model corresponding to the target cow is fine-tuned to generate a personalized weight growth trajectory for the target cow. Based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory, the deviation of the target cow's growth status is determined; Based on the deviation from the growth status and the current estimated weight, a feeding plan for the target cattle is determined, the feeding plan including feed formulation and feeding amount.

2. The control method of the aquaculture management system as described in claim 1, characterized in that, The steps of determining the current body size parameters and estimated weight of the target cow based on image data acquired by the camera device include: The image data is preprocessed, and the preprocessed image data is then segmented to obtain valid image data containing only the target cow body. Feature point detection and extraction are performed on the effective image data to obtain the image pixel feature parameters corresponding to the body length, body height, chest circumference, abdominal circumference, hip width, and head circumference of the target cow. Based on the pixel size calibration algorithm, the image pixel feature parameters are converted into physical size parameters to obtain the current body size parameters of the target cow; The current body size parameters are input into a pre-trained cattle weight prediction model. The cattle weight prediction model calculates the current estimated weight based on the current body size parameters. The cattle weight prediction model is a random forest regression model trained based on measured weight data and corresponding measured body size parameters of the same breed of cattle as the target cattle.

3. The control method of the aquaculture management system as described in claim 1, characterized in that, Before the step of determining the target cow's current body size parameters, current estimated weight, and identification based on image data collected by the camera device, the following steps are included: The target cow is monitored using the camera device to obtain monitoring images of the target cow; The target cow in the monitoring image is subjected to posture recognition to obtain the posture recognition result corresponding to the monitoring image. The posture recognition result is used to indicate whether the posture category of the target cow is a preset posture category. If the pose recognition result corresponding to the monitored image is yes, the monitored image is identified as the image data.

4. The control method of the aquaculture management system as described in claim 1, characterized in that, The step of fine-tuning the standard growth model corresponding to the target cow based on the historical body size sequence and the current body size parameters to generate a personalized weight growth trajectory for the target cow includes: Obtain the breed of the target cattle, and extract the first association rule between body size and weight, the second association rule between time and body size growth rate, and the standard body size growth trajectory generated based on the second association rule from the standard growth model corresponding to the breed. Based on the historical body size sequence and the current body size parameters, the second association rule is used to predict the theoretical body size of the target cow at each time point, thereby forming a personalized body size growth trajectory for the target cow. The ratio of the personalized body size growth trajectory to the standard body size growth trajectory at each time point is calculated to obtain the body size scale factor of the target cattle. The body size scale factor is used to characterize the scaling ratio of the body size of the target cattle relative to the average body size of the same breed population. Based on the body size scale factor, the first association rule is scaled and corrected to generate a third association rule between the body size and weight of the target cow. Based on the personalized body size growth trajectory, the third association rule is used to predict the theoretical weight of the target cow at each time point, thus forming the personalized weight growth trajectory.

5. The control method of the aquaculture management system as described in claim 1, characterized in that, The step of determining the deviation of the target cow's growth status based on the current estimated weight, the historical weight sequence, and the personalized weight growth trajectory includes: Extract the current theoretical weight corresponding to the current time point from the personalized weight growth trajectory; Calculate the difference between the current estimated weight and the current theoretical weight, and divide the difference by the current theoretical weight to obtain the current deviation at the current time point; The historical theoretical weight within the target historical range is extracted from the personalized weight growth trajectory, and the historical estimated weight within the target historical range is extracted from the historical weight sequence; Calculate the historical difference between the historical estimated weight and the historical theoretical weight at the corresponding historical time point, and divide the historical difference by the historical theoretical weight to obtain the historical deviation at the corresponding historical time point; Determine the first weighted value corresponding to the current deviation and the second weighted value corresponding to the historical deviation; Based on the first weighted value and the second weighted value, the current deviation and the historical deviation are weighted and fused to generate the growth status deviation.

6. The control method of the aquaculture management system as described in claim 1, characterized in that, The step of determining the feeding plan for the target cattle based on the deviation in growth status and the current estimated weight includes: Based on the deviation of the growth status, determine the feeding strategy corresponding to the target cattle; Select the candidate feed formula corresponding to the feeding strategy from a set of preset candidate feed formulas as the feed formula; Calculate the feeding amount corresponding to the target cow based on the current estimated weight; The feeding amount and the feed formula are determined as the feeding plan.

7. The control method of the aquaculture management system as described in claim 1, characterized in that, The step of determining the feeding plan for the target cattle based on the deviation in growth status and the current estimated weight further includes: The feed composition database and the pre-constructed feeding rule library are retrieved. The feed composition database stores the nutritional parameters, energy values ​​and prices of various feeds. The feeding rule library includes feed matching rules and constraint prohibition rules for the same breed of cattle under different growth status labels. The deviation from the growth status is mapped using rules to determine the growth status label corresponding to the target cow. The growth status deviation, the current estimated weight, the growth status label, and the feed composition database are input into the large language model. The large language model then filters and combines various feeds in the feed composition database based on the feeding rule base to generate a list of candidate feeding programs. Based on the prices of various types of feed, the cost of each candidate feeding scheme in the candidate feeding scheme list is calculated, and the candidate feeding scheme with a cost lower than a preset cost is selected as the feeding scheme.

8. The control method of the aquaculture management system as described in claim 1, characterized in that, After the step of determining the feeding plan for the target cattle based on the deviation of growth status and the current estimated weight, the method further includes: Obtain a first warning threshold and a second warning threshold, wherein the second warning threshold is greater than the first warning threshold; When the absolute value of the deviation in growth status is greater than or equal to the first warning threshold and less than the second warning threshold, the system displays the first warning information, which includes the deviation in growth status, the image data of the target cattle, and the feeding plan, and sends the first warning information to all managers via email. When the absolute value of the deviation from the growth condition is greater than or equal to the second warning threshold, the system controls the display of the second warning information and the control terminal to issue a warning sound to indicate the second warning information, and also notifies all the management personnel by telephone.

9. A control device for a livestock breeding management system, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the control method of the aquaculture management system as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the control method of the aquaculture management system as described in any one of claims 1 to 8.

Images