Livestock Breeding Management Method, Device, Equipment and Storage Medium Based on RPA and AI

Abstract

The present application proposes a livestock breeding management method, device, equipment and storage medium based on RPA and AI. The method applied to the server includes: obtaining the original livestock information detected by at least one sensor; analyzing the original livestock information by using computer vision algorithms and / or machine learning algorithms to obtain the livestock entry information required by the livestock information management system; and sending the livestock entry information to a robotic process automation (RPA) robot so that the RPA robot enters the livestock entry information into the livestock information management system. The present application can not only automatically obtain the livestock entry information by using AI technology, but also automatically enter the livestock entry information into the livestock information management system by using RPA technology, without hiring manpower to regularly detect and analyze the livestock information in the farm to obtain the livestock entry information, and even more without manually entering these livestock entry information into the livestock information management system, thereby improving the efficiency of obtaining and entering the livestock entry information.

Description

Technical Field

[0001] This application relates to the technical field of process automation, and particularly to a livestock breeding management method, device, equipment, and storage medium based on RPA and AI. Background Art

[0002] Robotic Process Automation (RPA) is to simulate human operations on a computer through specific "robot software" and automatically execute process tasks according to rules.

[0003] Artificial Intelligence (AI) is a technical science that studies, develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence.

[0004] Livestock breeding has grown in scale from household breeding to individual business breeding and then to enterprise breeding, with increasingly fierce competition. The livestock breeding process is relatively complex, including seed selection, breeding, parturition, observing whether the behavior is normal, whether the environment is comfortable, and evaluating the quality of livestock meat, etc. In related technologies, in order to improve breeding quality, breeding efficiency, and breeding income statistics efficiency, etc., for large-scale individual business breeding and enterprise breeding, it is often necessary to hire a large number of people to regularly detect and analyze the behavior of each livestock in the farm, the environment where the livestock is located, and identify the quality of meat products, etc., and then enter the analyzed information into the livestock information management system, so as to manage and statistically analyze the livestock breeding situation through the livestock information management system. Therefore, although the livestock information management system that can be electronically managed and statistically analyzed has been used in the above livestock breeding management system, a large amount of manpower and time are required to obtain and enter the information required by the livestock information management system. Summary of the Invention

[0005] Embodiments of this application provide a livestock breeding management method, device, equipment, and storage medium based on RPA and AI to solve the problem in related technologies that a large amount of manpower and time are required to obtain and enter the information required by the livestock information management system. The technical solutions are as follows:

[0006] In a first aspect, embodiments of this application provide a livestock breeding management method based on RPA and AI. The method is applied to a server and includes:

[0007] S1. Obtain the original livestock information detected by at least one sensor;

[0008] S2. Analyze the original livestock information by using computer vision CV algorithms and / or machine learning ML algorithms to obtain the livestock entry information required by the livestock information management system;

[0009] S3. Send the livestock entry information to the Robotic Process Automation (RPA) robot so that the RPA robot can enter the livestock entry information into the livestock information management system.

[0010] In one embodiment, the livestock entry information includes any one or a combination of the following:

[0011] Individual behavior information of livestock, aggressive behavior information between livestock, livestock birthing information, livestock temperature perception information, livestock meat quality information.

[0012] In one embodiment, when the livestock entry information includes the individual behavior information of livestock and the original livestock information includes livestock videos, S2 includes:

[0013] S21. Use the Computer Vision (CV) algorithm to identify the livestock contained in each frame of the livestock image in the livestock video, track and identify the individual behavior of the livestock, and obtain the individual behavior information of each livestock. Among them, the individual behavior information includes normal individual behavior and / or abnormal individual behavior. The normal individual behavior includes at least one of eating, drinking, lying down, walking, and excreting. The abnormal individual behavior includes at least one of abnormal body surface color, abnormal mental state, and diarrhea.

[0014] In one embodiment, when the identified individual behavior information includes abnormal individual behavior, the method further includes:

[0015] S4. Send the warning prompt information containing abnormal individual behavior to the RPA robot so that the RPA robot can enter the warning prompt information into the livestock abnormal warning system.

[0016] In one embodiment, when the livestock entry information includes the aggressive behavior information between livestock and the original livestock information includes livestock videos, S2 includes:

[0017] S22. Use the Computer Vision (CV) algorithm to identify the livestock contained in each frame of the livestock image in the livestock video, and divide the area where the identified livestock is located into multiple livestock limb areas;

[0018] S23. For two livestock to be identified, if the first livestock limb area in the multiple livestock limb areas of one livestock overlaps with the second livestock limb area in the multiple livestock limb areas of the other livestock, it is determined that there is aggressive behavior between the two livestock to be identified.

[0019] In one embodiment, when the livestock entry information includes livestock birthing information and the original livestock information includes livestock videos, S2 includes:

[0020] S24. Use the Computer Vision (CV) algorithm to identify the livestock in the livestock image containing the maternal livestock confinement pen in the livestock video;

[0021] S25. Determine whether the radius of the preset area in the area where the identified livestock is located is greater than the preset radius threshold. If it is greater than the preset radius threshold, determine that the identified livestock is a female livestock, where the preset area is an area capable of distinguishing the gender of livestock;

[0022] S26. If the area included in a female livestock confinement pen contains female livestock and other livestock smaller in size than the female livestock, determine that the female livestock has given birth.

[0023] In one implementation, when the livestock input information includes livestock temperature perception information and the original livestock information includes livestock videos, S2 includes:

[0024] S27. Use the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and use the lying posture recognition model to identify the lying postures of the livestock in the lying state, where the lying postures include lateral lying and sternal lying, and the lying posture recognition model is a neural network model trained according to multiple livestock images with livestock lying posture marks added;

[0025] S28. Count the proportions of livestock in lateral lying and sternal lying respectively;

[0026] S29. If the proportion of livestock in lateral lying is greater than the preset proportion threshold, determine that the livestock temperature perception information is that the livestock body perception temperature is high;

[0027] S210. If the proportion of livestock in sternal lying is greater than the preset proportion threshold, determine that the livestock temperature perception information is that the livestock body perception temperature is low;

[0028] S211. If the proportion of livestock in lateral lying is less than or equal to the preset proportion threshold and the proportion of livestock in sternal lying is less than or equal to the preset proportion threshold, determine that the livestock temperature perception information is that the livestock body perception temperature is appropriate.

[0029] In one implementation, when the livestock input information includes livestock meat quality information and the original livestock information includes livestock meat block images, S2 includes:

[0030] S214. By inputting the livestock meat block image into the meat quality recognition model, identify the livestock meat quality level of the livestock meat block included in the livestock meat block image, where the livestock meat quality information includes the livestock meat quality level, and the meat quality recognition model is a neural network model trained according to multiple livestock meat block images with livestock meat quality level marks added.

[0031] In one implementation, the livestock meat quality level includes the livestock meat tissue texture level, and the meat quality recognition model includes the meat tissue texture recognition model;

[0032] And / or, the livestock meat quality level includes the livestock meat color level, and the meat quality recognition model includes the meat color recognition model.

[0033] In a second aspect, an embodiment of the present application provides a livestock breeding management method based on RPA and AI. The method is applied to a robotic process automation (RPA) robot, and the method includes:

[0034] S4. Receive livestock entry information sent by a server, where the livestock entry information is obtained by the server analyzing the original livestock information detected by at least one sensor using computer vision (CV) algorithms and / or machine learning (ML) algorithms;

[0035] S5. Enter the livestock entry information into a livestock information management system.

[0036] In an implementation manner, the livestock entry information includes any one or a combination of the following:

[0037] Individual behavior information of livestock, attack behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

[0038] In a third aspect, an embodiment of the present application provides a livestock breeding management device based on RPA and AI. The device is applied to a server, and the device includes:

[0039] An acquisition unit for acquiring the original livestock information detected by at least one sensor;

[0040] An analysis unit for analyzing the original livestock information using computer vision (CV) algorithms and / or machine learning (ML) algorithms to obtain the livestock entry information required by the livestock information management system;

[0041] A sending unit for sending the livestock entry information to a robotic process automation (RPA) robot so that the RPA robot enters the livestock entry information into the livestock information management system.

[0042] In an implementation manner, the livestock entry information includes any one or a combination of the following:

[0043] Individual behavior information of livestock, attack behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

[0044] In an implementation manner, the analysis unit includes:

[0045] An individual behavior analysis module, which is used to, when the livestock input information includes the individual behavior information of livestock and the original livestock information includes livestock videos, use computer vision CV algorithms to identify the livestock included in each frame of livestock images in the livestock videos, track the livestock, and identify the individual behaviors of the livestock, so as to obtain the individual behavior information of each livestock. The individual behavior information includes normal individual behaviors and / or abnormal individual behaviors. The normal individual behaviors include at least one of eating, drinking, lying down, walking, and excreting. The abnormal individual behaviors include at least one of abnormal body surface color, abnormal mental state, and diarrhea.

[0046] In one implementation, the sending unit is further configured to, when the identified individual behavior information includes abnormal individual behaviors, send the warning prompt information including the abnormal individual behaviors to the RPA robot, so that the RPA robot enters the warning prompt information into the livestock abnormal warning system.

[0047] In one implementation, the analysis unit includes:

[0048] An attack behavior analysis module, which is used to, when the livestock input information includes the attack behavior information between livestock and the original livestock information includes livestock videos, use computer vision CV algorithms to identify the livestock included in each frame of livestock images in the livestock videos, and divide the area where the identified livestock is located into multiple livestock limb areas; for two livestock to be identified, if there is an overlap between the first livestock limb area in the multiple livestock limb areas of one livestock and the second livestock limb area in the multiple livestock limb areas of the other livestock, it is determined that there is an attack behavior between the two livestock to be identified.

[0049] In one implementation, the analysis unit includes:

[0050] A parturition analysis module, which is used to, when the livestock input information includes livestock parturition information and the original livestock information includes livestock videos, use computer vision CV algorithms to identify the livestock in the livestock images including the maternal livestock confinement pen in the livestock videos; determine whether the radius of a preset area in the area where the identified livestock is located is greater than a preset radius threshold. If it is greater than the preset radius threshold, it is determined that the identified livestock is a maternal livestock, where the preset area is an area capable of distinguishing the gender of livestock; if the area included in a maternal livestock confinement pen contains a maternal livestock and other livestock smaller in size than the maternal livestock, it is determined that the maternal livestock has given birth.

[0051] In one implementation, the analysis unit includes:

[0052] A temperature perception analysis module is used to, when the livestock input information includes livestock temperature perception information and the original livestock information includes livestock videos, use a computer vision (CV) algorithm to identify the livestock contained in each frame of the livestock images in the livestock videos, and use a lying posture recognition model to identify the lying postures of the livestock in the lying state, where the lying postures include lateral lying and sternal lying, and the lying posture recognition model is a neural network model trained based on multiple livestock images with livestock lying posture marks added; respectively count the proportions of livestock in lateral lying and sternal lying; if the proportion of livestock in lateral lying is greater than a preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock's perceived temperature is high; if the proportion of livestock in sternal lying is greater than a preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock's perceived temperature is low; if the proportion of livestock in lateral lying is less than or equal to the preset proportion threshold and the proportion of livestock in sternal lying is less than or equal to the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock's perceived temperature is appropriate.

[0053] In one implementation, the analysis unit includes:

[0054] A meat quality analysis module is used to, when the livestock input information includes livestock meat quality information and the original livestock information includes livestock meat block images, input the livestock meat block images into a meat quality recognition model to identify the livestock meat quality level of the livestock meat blocks contained in the livestock meat block images, where the livestock meat quality information includes the livestock meat quality level, and the meat quality recognition model is a neural network model trained based on multiple livestock meat block images with livestock meat quality level marks added.

[0055] In one implementation, the livestock meat quality level includes the livestock meat tissue texture level, and the meat quality recognition model includes a meat tissue texture recognition model; and / or, the livestock meat quality level includes the livestock meat color level, and the meat quality recognition model includes a meat color recognition model.

[0056] Fourthly, an embodiment of the present application provides a livestock breeding management device based on RPA and AI. The device is applied to a robotic process automation (RPA) robot, and the device includes:

[0057] A receiving unit is used to receive the livestock input information sent by the server, where the livestock input information is obtained by the server through analyzing the original livestock information detected by at least one sensor using a computer vision (CV) algorithm and a machine learning (ML) algorithm;

[0058] An input unit is used to input the livestock input information into the livestock information management system.

[0059] In one implementation, the livestock input information includes any one or a combination of the following:

[0060] The individual behavior information of livestock, the aggressive behavior information between livestock, the livestock birth information, the livestock temperature perception information, the livestock meat quality information.

[0061] In one embodiment, the receiving unit is further configured to receive, when the individual behavior information includes abnormal individual behavior, a warning prompt message including the abnormal individual behavior sent by the server;

[0062] The input unit is further configured to input the warning prompt message into the livestock abnormal warning system.

[0063] In a fifth aspect, an embodiment of the present application provides a server, including: a processor and a memory, where instructions are stored in the memory, and the instructions are loaded and executed by the processor to implement the method in any one of the embodiments of the first aspect above.

[0064] In a sixth aspect, an embodiment of the present application provides a terminal, including: a processor and a memory, where instructions are stored in the memory, and the instructions are loaded and executed by the processor to implement the method in any one of the embodiments of the second aspect above.

[0065] In a seventh aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the method in any one of the embodiments of the above aspects is implemented.

[0066] The advantages or beneficial effects in the above technical solutions at least include:

[0067] 1. The server can first obtain the detected original livestock information from at least one sensor, then use computer vision (CV) algorithms and / or machine learning (ML) algorithms to analyze the original livestock information to obtain the livestock input information required by the livestock information management system, and finally send the livestock input information to a robotic process automation (RPA) robot, and the RPA robot inputs the livestock input information into the livestock information management system. It can be seen that the embodiment of the present application can not only automatically obtain the livestock input information required by the livestock information management system by using artificial intelligence (AI) technology, but also automatically input the livestock input information into the livestock information management system by using RPA technology, without hiring manpower to regularly detect and analyze the livestock information in the farm to obtain the livestock input information, and even more without manually inputting the livestock input information into the livestock information management system. Moreover, the working efficiency of the automated livestock breeding management method composed of sensors, servers, and RPA robots in the present application is much higher than the manual efficiency.

[0068] 2. The embodiments of the present application can use sensors (such as video sensors) to capture livestock videos, and use computer vision CV algorithms to identify the livestock included in each frame of livestock images in the livestock videos, and identify the individual behavior information of each livestock, whether there is an attack behavior between livestock, whether the maternal livestock gives birth, etc., without the need for manual recording of this information one by one in the farm, thus not only saving manpower, but also improving the efficiency of obtaining this livestock entry information.

[0069] 3. The embodiments of the present application can use sensors (such as video sensors) to capture livestock videos, and use computer vision CV algorithms to identify the livestock included in each frame of livestock images in the livestock videos, use a lying posture recognition model to identify the lying postures of the livestock in the lying state, and determine the livestock temperature perception information by counting the proportion of each lying posture, without the need for manual recording of the livestock temperature perception information one by one in the farm, thus not only saving manpower, but also improving the efficiency of obtaining the livestock temperature perception information.

[0070] 4. The embodiments of the present application can use sensors (such as image sensors) to capture livestock meat block images, and use a meat quality recognition model (including a meat tissue texture recognition model and / or a meat color recognition model) to identify the livestock meat quality level, without the need for manual identification of each piece of meat one by one through experience, thus not only saving manpower, but also improving the efficiency of obtaining the livestock meat quality information.

[0071] 5. When the livestock individual behavior information identified by the computer vision CV algorithm includes abnormal behaviors, the warning prompt information including the abnormal individual behavior can be sent to the RPA robot, so that the RPA robot can enter the warning prompt information into the livestock abnormal warning system, so that the administrator can timely discover and handle the abnormal behaviors of the livestock through the livestock abnormal warning system, and avoid causing irreparable losses.

[0072] The above summary is only for the purpose of the specification and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the present application will be readily apparent by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In the drawings, unless otherwise specified, the same reference numerals throughout the several views denote the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed according to the present application and should not be regarded as limiting the scope of the present application.

[0074] Figure 1 It is a schematic flow chart of a livestock breeding management method based on RPA and AI provided by an embodiment of the present application;

[0075] Figure 2 Schematic flowchart of another livestock breeding management method provided by an embodiment of the present application based on RPA and AI;

[0076] Figure 3 Schematic interface diagram of entering livestock delivery information into a livestock information management system by an RPA robot provided by an embodiment of the present application;

[0077] Figure 4 Architecture diagram of a livestock breeding management system provided by an embodiment of the present application based on RPA and AI;

[0078] Figure 5 Block diagram of the composition of a livestock breeding management device provided by an embodiment of the present application based on RPA and AI;

[0079] Figure 6 Block diagram of the composition of another livestock breeding management device provided by an embodiment of the present application based on RPA and AI;

[0080] Figure 7 Architecture diagram of a server provided by an embodiment of the present application;

[0081] Figure 8 Architecture diagram of another server provided by an embodiment of the present application;

[0082] Figure 9 Architecture diagram of a terminal provided by an embodiment of the present application;

[0083] Figure 10 Architecture diagram of another terminal provided by an embodiment of the present application. Detailed implementation manners

[0084] The following details each embodiment of the present application. The examples of each embodiment are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present application and should not be construed as a limitation to the present application.

[0085] In the description of the embodiments of the present application, the term "sensor" is a device for detecting livestock itself and the livestock environment, including video sensors, image sensors, etc. The original data obtained by sensor detection can be called "original livestock information". The term "server" is a background device that processes the original livestock information detected by the sensor and obtains livestock entry information, which can be a single server or a server cluster. The term "terminal" refers to a user terminal, including desktop computers, laptops, mobile phones, etc. An RPA robot and a livestock information management system can be installed on the terminal.

[0086] In the description of the embodiments of the present application, the term "livestock" refers to livestock that are raised and bred by humans for utilization and are beneficial to agricultural production, including pigs, cows, sheep, chickens, ducks, geese, etc. "Livestock entry information" refers to the information that needs to be entered into the livestock information management system for administrators to view and further perform statistical operations, including but not limited to any one or more combinations of the following: individual behavior information of livestock, attack behavior information between livestock, livestock birth information, livestock temperature perception information, and livestock meat quality information. Among them, the term "individual behavior information" refers to the behavior information generated during the individual activities of livestock, the "attack behavior information between livestock" refers to the mutual attack behavior generated during the activities of at least two livestock, the "livestock birth information" refers to the relevant information of the birth of female livestock, the "livestock temperature perception information" refers to the information on whether the perceived temperature of livestock is appropriate, and the "livestock meat quality information" refers to the information on the meat quality level of livestock.

[0087] In the description of the embodiments of the present application, the term "livestock information management system" refers to a system used to store and manage livestock entry information, and the term "livestock abnormal warning system" refers to a system used to store and output warning prompt information. The livestock abnormal warning system can be a subsystem of the livestock information management system or other systems independent of the livestock information management system.

[0088] In the description of the embodiments of the present application, the term "computer vision CV" is a science that studies how to enable machines to "see". Further, it refers to using cameras and computers to replace the human eye to perform machine vision such as target recognition, tracking, and measurement on targets, and further performing graphic processing to make the computer process information that is more suitable for human eye observation or transmission to instrument detection.

[0089] In the description of the embodiments of the present application, the term "Machine Learning ML" is an interdisciplinary field that involves multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how a computer simulates or implements human learning behaviors to acquire new knowledge or skills and reorganizes the existing knowledge structure to continuously improve its own performance. The terms "lying posture recognition model", "meat quality recognition model", "meat tissue texture recognition model", and "meat color recognition model" are neural network models for classification trained using the Machine Learning ML algorithm. Among them, the lying posture recognition model is a neural network model trained based on multiple livestock images with livestock lying posture marks, and is used to recognize the lying posture of livestock; the meat quality recognition model is a neural network model trained based on multiple livestock meat block images with livestock meat quality level marks, and is used to recognize the livestock meat quality level, and the meat quality recognition model includes a meat tissue texture recognition model and / or a meat color recognition model; the meat tissue texture recognition model is a neural network model trained based on multiple livestock meat block images including livestock meat tissue texture level marks, and is used to recognize the livestock meat tissue texture level; the meat color recognition model is a neural network model trained based on multiple livestock meat block images including livestock meat color level marks, and is used to recognize the livestock meat color level.

[0090] In the description of the embodiments of the present application, the term "OCR" refers to Optical Character Recognition, specifically the process in which an electronic device examines the characters printed on paper, determines their shapes by detecting dark and bright patterns, and then translates the shapes into computer text using character recognition methods; that is, for printed characters, an optical method is used to convert the text in a paper document into an image file of black and white dot matrices, and an image recognition software is used to convert the text in the image into a text format for further editing and processing by a word processing software.

[0091] In the process of livestock breeding, it is often necessary to first manually collect and analyze the livestock entry information, and then manually enter the livestock entry information into the livestock information management system. This work is not only highly repetitive and low in difficulty, but also very labor-consuming and time-consuming. The RPA technology can, through the user interface, intelligently understand the existing applications on the electronic device, and automate repetitive, rule-based, and large-volume routine operations, such as automatically repeatedly reading emails, reading Office components, operating databases and web pages, client software, etc., collecting data and performing cumbersome calculations, and batch generating the required files and reports. Thus, the RPA technology can greatly reduce the investment in labor costs and effectively improve office efficiency. The AI technology can break through fixed rules and simulate human thinking and consciousness to automate the processing of some more complex application scenarios. RPA has unique advantages: low code and non-invasive. Low code means that RPA does not require a high level of IT to operate, and business personnel who do not understand programming can also develop processes; non-invasive means that RPA can simulate human operations without opening the software system interface. However, traditional RPA has certain limitations: it can only be based on fixed rules and the application scenarios are limited. With the continuous development of the AI technology, the deep integration of RPA and AI overcomes the limitations of traditional RPA. RPA + AI = Hand work + Head work, which is greatly changing the value of the labor force. Based on this, the embodiments of the present application provide a method for automating livestock breeding management by combining the two technologies of RPA and AI, which can not only save labor, but also improve the efficiency of obtaining and entering livestock entry information.

[0092] With reference to the following description and drawings, these and other aspects of the embodiments of the present application will become clear. In these descriptions and drawings, some specific embodiments of the embodiments of the present application are specifically disclosed to represent some ways of implementing the principles of the embodiments of the present application. However, it should be understood that the scope of the embodiments of the present application is not limited thereto. On the contrary, the embodiments of the present application include all changes, modifications, and equivalents that fall within the spirit and connotation of the appended claims.

[0093] Figure 1 is a flowchart of a livestock breeding management method based on RPA and AI provided by an embodiment of the present application. This method is applied to a server, such as Figure 1 shown, and this method may include the following steps:

[0094] Step S101: The server obtains the original livestock information detected by at least one sensor.

[0095] In order to monitor the livestock in a farm, at least one sensor can be installed in the farm to measure the original livestock information, so that at least one sensor regularly uploads the measured original livestock information to the server, or after receiving the original livestock information acquisition instruction sent by the server, uploads the original livestock information within the time period indicated by the original livestock information acquisition instruction to the server. The at least one sensor can include a video sensor (or camera) for recording the living process of the livestock and an image sensor for collecting images of livestock meat, and the number and installation positions of the video sensor and the image sensor can be set according to the size of the farm. When the at least one sensor includes a video sensor, the original livestock information includes livestock videos; when the at least one sensor includes an image sensor, the original livestock information includes images of livestock meat.

[0096] Step S102: The server analyzes the original livestock information by using computer vision CV algorithms and / or machine learning ML algorithms to obtain the livestock entry information required by the livestock information management system.

[0097] Among them, the livestock entry information includes, but is not limited to, any one or a combination of the following: individual behavior information of livestock, attack behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

[0098] The following elaborates on each method for obtaining livestock entry information respectively:

[0099] (1) When the livestock entry information includes the individual behavior information of livestock and the original livestock information includes livestock videos, the specific process for obtaining the individual behavior information of livestock in the embodiments of the present application includes: using computer vision CV algorithms to identify the livestock included in each frame of livestock image in the livestock video, tracking and identifying the individual behavior of the livestock, and obtaining the individual behavior information of each livestock.

[0100] Among them, the individual behavior information includes normal individual behavior and / or abnormal individual behavior. The normal individual behavior includes at least one of eating, drinking, lying down, walking, and excreting. The abnormal individual behavior includes at least one of abnormal body surface color, abnormal mental state, and diarrhea. When the individual behavior information of each livestock within a preset time period is obtained in the embodiments of the present application, the individual behavior information can also include the behavior duration of a single behavior in the normal individual behavior and / or the behavior duration of a single behavior in the abnormal individual behavior. For example, the behavior duration of a single behavior in the normal individual behavior includes the eating duration, drinking duration, lying down duration, walking duration, or excreting duration of the livestock within the preset time period. In addition, the food intake and water intake can also be counted. The behavior duration of a single behavior in the abnormal individual behavior includes the abnormal body surface color duration, abnormal mental state duration, or diarrhea duration of the livestock within the preset time period.

[0101] In the process of using computer vision CV algorithms to identify livestock in each frame of livestock images in livestock videos, the background difference method can be used to distinguish the background area and livestock area in the livestock images. The farm can set up a feeding area, a drinking area, and an activity area. After identifying the livestock in the livestock image, the livestock can be tracked. When a certain livestock enters the feeding area and performs a feeding action, the individual behavior information of the livestock can be determined to include feeding. Similarly, when a certain livestock enters the drinking area and performs a drinking action, the individual behavior information of the livestock can be determined to include drinking. When a certain livestock enters the activity area and has walking, lying, or excretion actions, the individual behavior information of the livestock can be determined to include walking, lying, or excretion. And when the image features of the excrement match the preset diarrhea image features, the individual behavior information of the livestock can be determined to include diarrhea. For abnormal body surface color, it can be determined whether the skin color value of the livestock (such as purple skin) is within the preset abnormal color value range. If it is within the preset abnormal color value range, it is determined that the body surface color is abnormal. For abnormal mental state, it can be determined whether the livestock has abnormal mental state by comprehensively judging the eye closure degree, skin color, and limb movements of the livestock.

[0102] In summary, the embodiments of the present application can use sensors to capture livestock videos and use computer vision CV algorithms to identify livestock in each frame of livestock images in the livestock videos, realizing the tracking and individual behavior recognition of livestock, so that the individual behavior information of each livestock such as feeding, drinking, lying, walking, and excretion can be obtained, without the need for manual recording of the individual behavior of each livestock in the farm one by one. Furthermore, not only manpower is saved, but also the efficiency of obtaining individual behavior information is improved.

[0103] In one implementation, when the identified individual behavior information includes abnormal individual behavior, in order to enable the administrator to timely learn and handle the abnormal behavior of livestock and avoid losses (such as the death of individual livestock, the death of a large number of livestock due to the rapid spread of plague, etc.), the server can also send the warning prompt information containing the abnormal individual behavior to the RPA robot, so that the RPA robot can enter the warning prompt information into the livestock abnormal warning system. Among them, the livestock abnormal warning system can be a subsystem within the livestock information management system or a system independent of the livestock information management system. The RPA robot can directly enter the warning prompt information into the livestock abnormal warning system, and the livestock abnormal warning system outputs warning prompt information in formats such as voice and pop-up windows, or automatically triggers a text message of the warning prompt information to a designated administrator. The RPA robot can also directly send the warning prompt information to the designated administrator in the form of text messages or emails through the livestock abnormal warning system. In addition, the livestock abnormal warning system can also include an online medical function, sending the abnormal behavior of livestock to a remote veterinarian for remote diagnosis in order to cure the abnormal livestock as soon as possible.

[0104] (2) When the livestock input information includes the attack behavior information between livestock and the original livestock information includes livestock videos, the specific process for the embodiments of the present application to obtain the attack behavior information between livestock is as follows: using the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and dividing the area where the identified livestock is located into multiple livestock limb areas; for two livestock to be identified, if there is an overlap between the first livestock limb area in the multiple livestock limb areas of one livestock and the second livestock limb area in the multiple livestock limb areas of the other livestock, it is determined that there is an attack behavior between the two livestock to be identified.

[0105] Among them, the multiple livestock limb areas include a head area, a body area, limb areas (one leg corresponds to one area), and a tail area. The first livestock limb area and the second livestock limb area can be the same type of area or different areas. For example, when the head area of one livestock overlaps with the body area of another livestock, it can be determined that one livestock is attacking the body of the other livestock with its head, and there is an attack behavior between them. In addition, when the first livestock limb area and the second livestock limb area are livestock limb areas for breeding, the attack behavior is a breeding behavior.

[0106] In summary, the embodiments of the present application can use a sensor to capture livestock videos, use the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, divide the area where the identified livestock is located into multiple livestock limb areas, and determine whether there is an attack behavior between livestock by whether there is an overlap between the livestock limb areas of different livestock, without the need for manual recording in the farm whether there is an attack behavior between every two livestock, thereby not only saving manpower but also improving the efficiency of obtaining the attack behavior information between livestock.

[0107] (3) In order to facilitate the delivery management of female livestock, a confinement pen is generally specially set for female livestock, and one female livestock is raised in each confinement pen. Therefore, when the livestock input information includes livestock delivery information and the original livestock information includes livestock videos, the specific process for the embodiments of the present application to obtain the livestock delivery information is as follows: using the computer vision CV algorithm to identify the livestock in the livestock image including the confinement pen for female livestock in the livestock video; determining whether the radius of a preset area in the area where the identified livestock is located is greater than a preset radius threshold, and if it is greater than the preset radius threshold, determining that the identified livestock is a female livestock; if the area included in a confinement pen for female livestock includes a female livestock and other livestock smaller in size than the female livestock, it is determined that the female livestock has given birth.

[0108] Among them, the preset area is the area capable of distinguishing the genders of livestock. For example, the preset area is the breast area. The preset radius threshold can be measured based on the radii of the preset areas of a large number of female livestock, and the average value of multiple radii can be calculated. When it is first recognized that a female livestock has given birth, the acquisition date of the image of the female livestock that has given birth is determined as the birth date of the female livestock. It is also possible to count the number of other livestock smaller in size than the female livestock to determine the number of livestock offspring.

[0109] In summary, the embodiments of the present application can use a sensor to capture livestock videos and use computer vision CV algorithms to identify female livestock in the livestock videos and birth information such as whether the female livestock has given birth, without the need for manual regular observation in the farm to determine whether the female livestock has given birth. This not only saves manpower but also improves the efficiency of obtaining livestock birth information.

[0110] (4) When the livestock input information includes livestock temperature perception information and the original livestock information includes livestock videos, the specific process for the embodiments of the present application to obtain livestock temperature perception information includes: using computer vision CV algorithms to identify the livestock included in each frame of the livestock image in the livestock video, and using a lying posture recognition model to identify the lying postures of the livestock in the lying state; respectively counting the proportions of livestock in the side-lying and sternum-lying postures; if the proportion of side-lying livestock is greater than the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels hot; if the proportion of sternum-lying livestock is greater than the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels cold; if the proportion of side-lying livestock is less than or equal to the preset proportion threshold and the proportion of sternum-lying livestock is less than or equal to the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels comfortable.

[0111] Among them, the lying postures include side-lying and sternum-lying. When a livestock lies on its side, the contact area between the livestock body and the outside is large, which can accelerate heat dissipation. Therefore, when the proportion of side-lying livestock is large, it indicates that the livestock body feels hot; when a livestock lies with its sternum down, the contact area between the livestock body and the outside is small, which can prevent heat dissipation as much as possible. Therefore, when the proportion of sternum-lying livestock is large, it indicates that the livestock body feels cold; when the number of livestock in the two lying postures is about the same, it indicates that it is just the lying posture habit of individual livestock, and the livestock body feels comfortable. The lying posture recognition model is a neural network model trained based on multiple livestock images with lying posture marks added. That is, the lying posture types of a large number of collected livestock images can be manually marked first, and then machine learning ML can be performed using the neural network model to obtain a lying posture recognition model capable of identifying livestock lying postures. The preset proportion threshold can be determined according to actual experience. For example, it can be 80%.

[0112] In summary, the embodiments of the present application can use sensors (such as image sensors) to capture images of livestock meat, and use a meat quality recognition model (including a meat tissue texture recognition model and / or a meat color recognition model) to identify the livestock meat quality level, without the need for manual identification of each piece of meat through experience, thus not only saving manpower but also improving the efficiency of obtaining livestock meat quality information.

[0113] In addition, in order to facilitate the administrator to better adjust the environmental temperature and / or humidity, at least one temperature sensor and / or at least one humidity sensor can be installed in the farm to measure the environmental temperature information and / or environmental humidity information, and upload the environmental temperature information and / or environmental humidity information to the server. The server sends the environmental temperature information and / or environmental humidity information to the RPA robot, and the RPA robot enters the environmental temperature information and / or environmental humidity information into the livestock information management system, so that the administrator can determine how to adjust the environmental temperature and / or humidity by comparing the environmental information (including environmental temperature information and / or environmental humidity information) measured by the sensor with the livestock temperature perception information.

[0114] (5) When the livestock entry information includes livestock meat quality information and the original livestock information includes an image of a livestock meat block, the livestock meat quality level of the livestock meat block included in the livestock meat block image is identified by inputting the livestock meat block image into a meat quality recognition model.

[0115] Among them, the livestock meat quality information includes the livestock meat quality level, and the meat quality recognition model is a neural network model trained according to multiple livestock meat block images with livestock meat quality level labels added. The livestock meat quality level includes the livestock meat tissue texture level, and the meat quality recognition model includes a meat tissue texture recognition model; and / or, the livestock meat quality level includes the livestock meat color level, and the meat quality recognition model includes a meat color recognition model. The meat tissue texture recognition model can be trained according to multiple livestock meat block images including livestock meat tissue texture level labels, and the meat tissue texture recognition model can be a k-Nearest Neighbor (kNN) model or other classification models. The meat color recognition model can be trained according to multiple livestock meat block images including livestock meat color level labels, and the meat color recognition model can be a Back Propagation (BP) model, a Support Vector Machine (SVM) model, or other classification models.

[0116] In summary, the embodiments of the present application can use sensors (such as image sensors) to capture images of livestock meat, and use a meat quality recognition model (including a meat tissue texture recognition model and / or a meat color recognition model) to identify the livestock meat quality level, without the need for manual identification of each piece of meat through experience, thus not only saving manpower but also improving the efficiency of obtaining livestock meat quality information.

[0117] Step S103: The server sends the livestock entry information to the RPA robot so that the RPA robot can enter the livestock entry information into the livestock information management system.

[0118] Among them, the number of livestock information management systems can be one or multiple. When there is one livestock information management system, it can be a comprehensive management system; when there are multiple livestock information management systems, different livestock information management systems can have different management focuses. For example, some livestock information management systems focus on managing the individual behavior information of livestock, and some livestock information management systems focus on managing the delivery information of livestock, etc.

[0119] The livestock breeding management method based on RPA and AI provided by the embodiments of the present application enables the server to first obtain the detected original livestock information from at least one sensor, then use computer vision CV algorithms and / or machine learning ML algorithms to analyze these original livestock information to obtain the livestock entry information required by the livestock information management system, and finally send the livestock entry information to the RPA robot so that the RPA robot can enter the livestock entry information into the livestock information management system. It can be seen from this that the embodiments of the present application can not only automatically obtain the livestock entry information required by the livestock information management system using AI technology, but also automatically enter the livestock entry information into the livestock information management system using RPA technology, without hiring manpower to regularly detect and analyze the livestock information in the farm to obtain the livestock entry information, and even more without manually entering these livestock entry information into the livestock information management system. Moreover, the working efficiency of the automated livestock breeding management method composed of sensors, servers, and RPA robots in this application is much higher than the manual efficiency.

[0120] Figure 2 It is a flowchart of the livestock breeding management method based on RPA and AI provided by another embodiment of the present application. This method is applied to the RPA robot, as Figure 2 shown, and this method may include the following steps:

[0121] Step S201: The RPA robot receives the livestock entry information sent by the server.

[0122] Among them, the livestock entry information is obtained by the server using computer vision CV algorithms and / or machine learning ML algorithms to analyze the original livestock information detected by at least one sensor. The livestock entry information includes any one or a combination of the following: the individual behavior information of livestock, the attack behavior information between livestock, the livestock delivery information, the livestock temperature perception information, and the livestock meat quality information.

[0123] Step S202: The RPA robot enters the livestock entry information into the livestock information management system.

[0124] When the RPA robot logs in to the livestock information management system, the livestock information management system can pop up a login interface containing a verification code image. In this case, the RPA robot can perform OCR recognition on the verification code image, obtain the verification code content in the verification code image, and input the verification code content into the corresponding edit box, so as to successfully log in to the corresponding system. Among them, the livestock information management system can be an application software or a website, and the form of the livestock information management system is not limited in the embodiments of the present application.

[0125] Exemplarily, as Figure 3 shown, when the livestock being raised are pigs and the livestock entry information includes livestock birth information, after the RPA robot logs in to the livestock information management system and enters the pig birth record interface, it can enter information such as the identification of the sows that have given birth (such as the pre-set sow numbers), the birth date, and the number of piglets into this interface. Among them, the breeding-related information in the interface is pre-entered information based on the attack behavior information between livestock.

[0126] The livestock breeding management method based on RPA and AI provided by the embodiments of the present application enables the RPA robot to receive the livestock entry information obtained by the server through analyzing the original livestock information detected by at least one sensor using computer vision CV algorithms and / or machine learning ML algorithms, and automatically enter the livestock entry information into the livestock information management system. It can be seen from this that the embodiments of the present application can not only automatically obtain the livestock entry information required by the livestock information management system using AI technology, but also automatically enter the livestock entry information into the livestock information management system using RPA technology, without hiring manpower to regularly detect and analyze the information of each livestock in the farm to obtain the livestock entry information, and even more without manually entering this livestock entry information into the livestock information management system. Moreover, the working efficiency of the automated livestock breeding management method composed of sensors, servers, and RPA robots in the present application is much higher than that of manual work.

[0127] Figure 4 Fig. shows a livestock breeding management system based on RPA and AI provided by another embodiment of the present application. The system includes a server 310, a terminal 320, and at least one sensor 330. The terminal includes an RPA robot 321 and a livestock information management system 322;

[0128] The sensor 330 is configured to detect the original livestock information and send the original livestock information to the server 310;

[0129] A server 310, configured to receive the original livestock information sent by at least one sensor 330, analyze the original livestock information by using a computer vision (CV) algorithm and / or a machine learning (ML) algorithm to obtain the livestock entry information required by the livestock information management system 322, and send the livestock entry information to the RPA robot 321;

[0130] An RPA robot 321, configured to enter the livestock entry information into the livestock information management system 322.

[0131] For the specific functions of the server, RPA robot, livestock information management system, and at least one sensor in the system provided by the embodiments of the present application, reference may be made to the corresponding descriptions in the above method, which will not be elaborated here.

[0132] Figure 5 Fig. shows a livestock breeding management device based on RPA and AI provided by another embodiment of the present application. The device is applied to a server and includes:

[0133] An acquisition unit 410, configured to acquire the original livestock information detected by at least one sensor;

[0134] An analysis unit 420, configured to analyze the original livestock information by using a computer vision (CV) algorithm and / or a machine learning (ML) algorithm to obtain the livestock entry information required by the livestock information management system;

[0135] A sending unit 430, configured to send the livestock entry information to a robotic process automation (RPA) robot so that the RPA robot can enter the livestock entry information into the livestock information management system.

[0136] In one implementation, the livestock entry information includes any one or a combination of the following:

[0137] The individual behavior information of livestock, the aggression behavior information between livestock, the livestock birth information, the livestock temperature perception information, the livestock meat quality information.

[0138] In one implementation, the analysis unit 420 includes:

[0139] An individual behavior analysis module, configured to, when the livestock entry information includes the individual behavior information of livestock and the original livestock information includes a livestock video, use a computer vision (CV) algorithm to identify the livestock included in each frame of the livestock image in the livestock video, track and identify the individual behavior of the livestock, and obtain the individual behavior information of each livestock. The individual behavior information includes normal individual behavior and / or abnormal individual behavior. The normal individual behavior includes at least one of eating, drinking, lying, walking, and excreting, and the abnormal individual behavior includes at least one of abnormal body surface color, abnormal mental state, and diarrhea.

[0140] The sending unit 430 is further configured to, when the recognized individual behavior information includes abnormal individual behavior, send a warning prompt information including the abnormal individual behavior to the RPA robot, so that the RPA robot enters the warning prompt information into the livestock abnormal warning system.

[0141] In one embodiment, the analysis unit 420 includes:

[0142] The attack behavior analysis module is configured to, when the livestock input information includes the attack behavior information between livestock and the original livestock information includes the livestock video, use the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and divide the area where the recognized livestock is located into multiple livestock limb areas; for two livestock to be recognized, if there is an overlap between the first livestock limb area in the multiple livestock limb areas of one livestock and the second livestock limb area in the multiple livestock limb areas of the other livestock, it is determined that there is an attack behavior between the two livestock to be recognized.

[0143] In one embodiment, the analysis unit 420 includes:

[0144] The parturition analysis module is configured to, when the livestock input information includes the livestock parturition information and the original livestock information includes the livestock video, use the computer vision CV algorithm to identify the livestock in the livestock image including the maternal livestock confinement pen in the livestock video; determine whether the radius of the preset area in the area where the recognized livestock is located is greater than the preset radius threshold, and if it is greater than the preset radius threshold, determine that the recognized livestock is a maternal livestock, where the preset area is an area capable of distinguishing the gender of the livestock; if the area included in a maternal livestock confinement pen contains a maternal livestock and other livestock smaller than the maternal livestock, it is determined that the maternal livestock has given birth.

[0145] In one embodiment, the analysis unit 420 includes:

[0146] The temperature perception analysis module is configured to, when the livestock input information includes the livestock temperature perception information and the original livestock information includes the livestock video, use the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and use the lying posture recognition model to identify the lying postures of the livestock in the lying state, where the lying postures include lying on the side and lying on the sternum, and the lying posture recognition model is a neural network model trained according to multiple livestock images with livestock lying posture marks added; respectively count the proportions of livestock lying on the side and lying on the sternum; if the proportion of livestock lying on the side is greater than the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels hot; if the proportion of livestock lying on the sternum is greater than the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels cold; if the proportion of livestock lying on the side is less than or equal to the preset proportion threshold and the proportion of livestock lying on the sternum is less than or equal to the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels comfortable.

[0147] In one embodiment, the analysis unit 420 includes:

[0148] A meat quality analysis module, which is used to, when the livestock input information includes livestock meat quality information and the original livestock information includes livestock meat block images, identify the livestock meat quality level of the livestock meat blocks included in the livestock meat block images by inputting the livestock meat block images into a meat quality recognition model. Among them, the livestock meat quality information includes the livestock meat quality level, and the meat quality recognition model is a neural network model trained according to multiple livestock meat block images with livestock meat quality level labels added.

[0149] In one embodiment, the livestock meat quality level includes the livestock meat tissue texture level, and the meat quality recognition model includes a meat tissue texture recognition model; and / or, the livestock meat quality level includes the livestock meat color level, and the meat quality recognition model includes a meat color recognition model.

[0150] Figure 6 Fig. shows a livestock breeding management device based on RPA and AI provided by another embodiment of the present application. This device is applied to an RPA robot and includes:

[0151] A receiving unit 510, which is used to receive the livestock input information sent by the server. Among them, the livestock input information is obtained by the server analyzing the original livestock information detected by at least one sensor using computer vision CV algorithms and machine learning ML algorithms;

[0152] An input unit 520, which is used to input the livestock input information into the livestock information management system.

[0153] For the functions of the modules in each device of the embodiments of the present application and the beneficial effects of the device, reference can be made to the corresponding descriptions in the above methods, which will not be elaborated here.

[0154] Figure 7 Fig. shows a structural block diagram of a server provided by another embodiment of the present application. The server includes: a processor 610 and a memory 620. Instructions are stored in the memory 620, and these instructions are loaded and executed by the processor 610 to implement the methods in any of the above method embodiments. The number of the memory 620 and the processor 610 can be one or more.

[0155] As Figure 8 shown, the server further includes:

[0156] A communication interface 630, which is used to communicate with external devices and perform data interaction and transmission.

[0157] If the memory 620, the processor 610, and the communication interface 630 are implemented independently, the memory 620, the processor 610, and the communication interface 630 can be interconnected through a bus and communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only a thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0158] In a specific implementation, if the memory 620, the processor 610, and the communication interface 630 are integrated on a single chip, the memory 620, the processor 610, and the communication interface 630 can communicate with each other through an internal interface.

[0159] Figure 9 A block diagram of a terminal provided in another embodiment of the present application is shown. The terminal includes: a processor 710 and a memory 720. Instructions are stored in the memory 720 and loaded and executed by the processor 710 to implement the method in any of the above method embodiments. The number of the memory 720 and the processor 710 can be one or more.

[0160] As Figure 10 shown, the terminal further includes:

[0161] A communication interface 730, configured to communicate with external devices and perform data interaction and transmission;

[0162] A display 740, configured to output and display each interface in the livestock information management system and / or the livestock anomaly warning system.

[0163] If the memory 720, the processor 710, the communication interface 730, and the display 740 are implemented independently, the memory 720, the processor 710, the communication interface 730, and the display 740 can be interconnected via a bus to complete communication with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0164] In a specific implementation, if the memory 720, the processor 710, and the communication interface 730 are integrated on a single chip, the memory 720, the processor 710, and the communication interface 730 can complete communication with each other through an internal interface.

[0165] The embodiments of the present application provide a computer-readable storage medium. A computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, the method provided in any one of the above method embodiments is implemented.

[0166] The embodiments of the present application further provide a chip. The chip includes a processor for calling and running an instruction stored in a memory, so that a communication device installed with the chip executes the method provided in the embodiments of the present application.

[0167] The embodiments of the present application further provide a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, the output interface, the processor, and the memory are connected through an internal connection path. The processor is configured to execute code in the memory. When the code is executed, the processor is configured to execute the method provided in the embodiments of the application.

[0168] It should be understood that the above-mentioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc. It is worth noting that the processor can be a processor that supports the advanced RISC machines (ARM) architecture.

[0169] The above-mentioned memory can include read-only memory and random access memory, and can also include non-volatile random access memory. The memory can be volatile memory or non-volatile memory, or can include both volatile and non-volatile memory. Among them, the non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), or flash memory. The volatile memory can include random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchlink DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0170] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0171] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0172] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of the features. In the description of the present application, "a plurality" means two or more unless otherwise specifically defined.

[0173] Any process or method description represented in the flowchart or described in other ways herein can be understood as representing a module, segment, or part of the code including one or more executable instructions for implementing a specific logical function or process. And the scope of the preferred embodiments of the present application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the involved functions, rather than in the order shown or discussed.

[0174] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a sequenced list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in connection with these instruction execution systems, apparatus, or devices.

[0175] It should be understood that each part of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above-described example methods can be completed by a program instructing relevant hardware. The program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0176] In addition, in each embodiment of the present application, the functional units can be integrated into one processing module, or each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disc, etc.

[0177] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various changes or substitutions, and these should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A livestock breeding management method based on RPA and AI, the method being applied to a server, characterized in that, The method includes: S1. Obtain the original livestock information detected by at least one sensor; S2. Analyze the original livestock information by using a computer vision (CV) algorithm and / or a machine learning (ML) algorithm to obtain the livestock entry information required by the livestock information management system; S3. Send the livestock entry information to a robotic process automation (RPA) robot so that the RPA robot enters the livestock entry information into the livestock information management system; When the livestock entry information includes livestock temperature perception information and the original livestock information includes a livestock video, S2 includes: S27. Use the computer vision (CV) algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and use a lying posture recognition model to identify the lying postures of the livestock in the lying state, where the lying postures include lateral lying and sternum lying, and the lying posture recognition model is a neural network model trained based on multiple livestock images with livestock lying posture marks added; S28. Count the proportions of livestock in lateral lying and sternum lying respectively; S29. If the proportion of livestock in lateral lying is greater than a preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is high; S210. If the proportion of livestock in sternum lying is greater than the preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is low; S211. If the proportion of livestock in lateral lying is less than or equal to the preset proportion threshold and the proportion of livestock in sternum lying is less than or equal to the preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is appropriate; In the case where at least one temperature sensor and / or at least one humidity sensor is installed in the farm, receive the environmental temperature information measured by at least one temperature sensor and / or, receive the environmental humidity information measured by at least one humidity sensor, and send the environmental temperature information and / or the environmental humidity information to the RPA robot so that the RPA robot enters the environmental temperature information and / or the environmental humidity information into the livestock information management system, so that the administrator can determine how to adjust the environmental temperature and / or humidity by comparing the environmental information measured by the sensor with the livestock temperature perception information, where the environmental information includes the environmental temperature information and / or the environmental humidity information.

2. The method according to claim 1, wherein The livestock entry information includes any one or a combination of the following: Individual behavior information of livestock, aggressive behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

3. The method according to claim 2, wherein When the livestock entry information includes the individual behavior information of the livestock and the original livestock information includes a livestock video, S2 includes: S21. Use the computer vision (CV) algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and perform tracking and individual behavior recognition on the livestock to obtain the individual behavior information of each livestock, where the individual behavior information includes normal individual behavior and / or abnormal individual behavior, and the normal individual behavior includes at least one of eating, drinking, lying, walking, and excreting, and the abnormal individual behavior includes at least one of abnormal body surface color, abnormal mental state, and diarrhea.

4. The method according to claim 3, characterized in that, When the identified individual behavior information includes abnormal individual behavior, the method further includes: S4. Sending a warning prompt message including the abnormal individual behavior to the RPA robot, so that the RPA robot enters the warning prompt message into the livestock abnormal warning system.

5. The method according to claim 2, characterized in that, When the livestock entry information includes the attack behavior information between livestock and the original livestock information includes a livestock video, S2 includes: S22. Using the computer vision CV algorithm to identify the livestock included in each frame of the livestock image in the livestock video, and dividing the area where the identified livestock is located into multiple livestock limb areas; S23. For two livestock to be identified, if there is an overlap between a first livestock limb area in the multiple livestock limb areas of one livestock and a second livestock limb area in the multiple livestock limb areas of the other livestock, it is determined that there is an attack behavior between the two livestock to be identified.

6. The method according to claim 2, wherein When the livestock entry information includes the livestock birth information and the original livestock information includes a livestock video, S2 includes: S24. Using the computer vision CV algorithm to identify the livestock in the livestock image including the maternal livestock confinement pen in the livestock video; S25. Judging whether the radius of a preset area in the area where the identified livestock is located is greater than a preset radius threshold. If it is greater than the preset radius threshold, it is determined that the identified livestock is a maternal livestock, where the preset area is an area capable of distinguishing the gender of livestock; S26. If the area included in a maternal livestock confinement pen contains a maternal livestock and other livestock smaller than the maternal livestock, it is determined that the maternal livestock has given birth.

7. The method according to any one of claims 2-6, characterized in that, When the livestock entry information includes the livestock meat quality information and the original livestock information includes a livestock meat block image, S2 includes: S212. By inputting the livestock meat block image into a meat quality recognition model, recognizing the livestock meat quality level of the livestock meat block included in the livestock meat block image, where the livestock meat quality information includes the livestock meat quality level, and the meat quality recognition model is a neural network model trained according to multiple livestock meat block images with livestock meat quality level marks added.

8. The method according to claim 7, characterized in that The livestock meat quality level includes the livestock meat tissue texture level, and the meat quality recognition model includes a meat tissue texture recognition model; And / or, the livestock meat quality level includes the livestock meat color level, and the meat quality recognition model includes a meat color recognition model.

9. A livestock breeding management method based on RPA and AI, the method is applied to a robotic process automation (RPA) robot, characterized in that, The method includes: S5. Receiving the livestock entry information sent by the server, where the livestock entry information is obtained by the server analyzing the original livestock information detected by at least one sensor using the computer vision CV algorithm and / or the machine learning ML algorithm; S6. Entering the livestock entry information into the livestock information management system; When the livestock entry information includes the livestock temperature perception information and the original livestock information includes a livestock video, analyzing the original livestock information detected by at least one sensor using the computer vision CV algorithm and / or the machine learning ML algorithm to obtain the livestock entry information, including: Identify the livestock included in each frame of livestock images in the livestock video using the computer vision CV algorithm, and identify the lying postures of the livestock in the lying state using a lying posture recognition model, where the lying postures include lateral lying and sternal lying, and the lying posture recognition model is a neural network model trained based on multiple livestock images with livestock lying posture marks added; Statistically calculate the proportions of livestock in lateral lying and sternal lying respectively; If the proportion of livestock in lateral lying is greater than a preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is high; If the proportion of livestock in sternal lying is greater than the preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is low; If the proportion of livestock in lateral lying is less than or equal to the preset proportion threshold, and the proportion of livestock in sternal lying is less than or equal to the preset proportion threshold, determine that the livestock temperature perception information is that the livestock's perceived temperature is appropriate; When at least one temperature sensor and / or at least one humidity sensor are installed in the farm, the method further includes: Receive the ambient temperature information measured by at least one temperature sensor and / or the ambient humidity information measured by at least one humidity sensor sent by the server, and enter the ambient temperature information and / or the ambient humidity information into the livestock information management system, so that the administrator can determine how to adjust the ambient temperature and / or humidity by comparing the ambient information measured by the sensor with the livestock temperature perception information, where the ambient information includes the ambient temperature information and / or the ambient humidity information.

10. The method according to claim 9, wherein The livestock entry information includes any one or a combination of more than one of the following: Individual behavior information of livestock, aggressive behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

11. A livestock breeding management device based on RPA and AI, the device is applied to a server, and is characterized in that, The device includes: An acquisition unit for acquiring the original livestock information detected by at least one sensor; An analysis unit for analyzing the original livestock information using the computer vision CV algorithm and / or the machine learning ML algorithm to obtain the livestock entry information required by the livestock information management system; A sending unit for sending the livestock entry information to a robotic process automation RPA robot, so that the RPA robot enters the livestock entry information into the livestock information management system; The analysis unit includes: A temperature perception analysis module, which is used to, when the livestock input information includes livestock temperature perception information and the original livestock information includes livestock videos, use computer vision CV algorithms to identify the livestock included in each frame of livestock images in the livestock videos, and use a lying posture recognition model to identify the lying postures of the livestock in the lying state. Among them, the lying postures include side lying and sternum lying. The lying posture recognition model is a neural network model trained according to multiple livestock images with livestock lying posture marks added; respectively count the proportions of livestock in side lying and sternum lying; if the proportion of livestock in side lying is greater than a preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels hot; if the proportion of livestock in sternum lying is greater than a preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels cold; if the proportion of livestock in side lying is less than or equal to the preset proportion threshold, and the proportion of livestock in sternum lying is less than or equal to the preset proportion threshold, it is determined that the livestock temperature perception information is that the livestock body feels comfortable; The device is further used for: when at least one temperature sensor and / or at least one humidity sensor are installed in the farm, receiving the ambient temperature information measured by at least one temperature sensor, and / or, receiving the ambient humidity information measured by at least one humidity sensor, and sending the ambient temperature information and / or the ambient humidity information to the RPA robot, so that the RPA robot enters the ambient temperature information and / or the ambient humidity information into the livestock information management system, so that the administrator can determine how to adjust the ambient temperature and / or humidity by comparing the ambient information measured by the sensor with the livestock temperature perception information, where the ambient information includes the ambient temperature information and / or the ambient humidity information.

12. The device according to claim 11, characterized in that, The livestock input information includes any one or a combination of the following: Individual behavior information of livestock, aggressive behavior information between livestock, livestock birth information, livestock temperature perception information, livestock meat quality information.

13. A livestock breeding management device based on RPA and AI, the device is applied to a robotic process automation (RPA) robot, and is characterized in that, The device includes: A receiving unit for receiving livestock entry information sent by a server, where the livestock entry information is obtained by the server analyzing the original livestock information detected by at least one sensor using computer vision CV algorithms and machine learning ML algorithms; wherein, when the livestock entry information includes livestock temperature perception information and the original livestock information includes livestock videos, analyzing the original livestock information detected by at least one sensor using computer vision CV algorithms and / or machine learning ML algorithms to obtain the livestock entry information includes: using the computer vision CV algorithm to identify the livestock included in each frame of livestock image in the livestock video, and using a lying posture recognition model to identify the lying postures of the livestock in the lying state, where the lying postures include lateral lying and sternal lying, and the lying posture recognition model is a neural network model trained according to multiple livestock images with livestock lying posture marks added; respectively counting the proportions of livestock in lateral lying and sternal lying; if the proportion of livestock in lateral lying is greater than a preset proportion threshold, determining that the livestock temperature perception information is that the livestock body temperature perception is high; if the proportion of livestock in sternal lying is greater than the preset proportion threshold, determining that the livestock temperature perception information is that the livestock body temperature perception is low; if the proportion of livestock in lateral lying is less than or equal to the preset proportion threshold and the proportion of livestock in sternal lying is less than or equal to the preset proportion threshold, determining that the livestock temperature perception information is that the livestock body temperature perception is appropriate; An entry unit for entering the livestock entry information into a livestock information management system.

14. A server, characterized in that, Including: A processor and a memory, where instructions are stored in the memory and are loaded and executed by the processor to implement the method according to any one of claims 1 to 8.

15. A terminal, characterized in that, Including: A processor and a memory, where instructions are stored in the memory and are loaded and executed by the processor to implement the method according to claim 9 or 10.

16. A computer-readable storage medium storing a computer program therein, characterized in that, The computer program, when executed by a processor, implements the method according to any one of claims 1 - 10.

Images