Intelligent Early Warning System and Method for Animal Tracking Based on Big Data

Abstract

This invention relates to the technical field of alarm systems and discloses an intelligent early warning system and method for animal movement based on big data. It performs dynamic early warning analysis of potential dangers in known breeding areas based on the characteristics of farmed animals, their real-time movement locations, and historical location event information. This enables the scientific establishment of an early warning system for potential dangers in known breeding areas based on historical data. Furthermore, it enables intelligent early warning of potential dangers in unknown breeding areas based on real-time environmental image information of farmed animals, combined with AI intelligent algorithms and image information of potential dangers in unknown breeding areas stored in big data. This enables the scientific establishment of an early warning system for potential dangers in unknown breeding areas based on real-time environmental characteristics, and the scientific establishment of an early warning system for animal movement locations based on known and unknown breeding scenarios.

Description

Technical Field

[0001] This invention relates to the technical field of alarm systems, specifically to an intelligent early warning system and method for animal movement based on big data. Background Technology

[0002] Animal husbandry utilizes the physiological functions of domesticated animals such as livestock and poultry, or wild animals such as deer, musk deer, foxes, minks, otters, and quails, through artificial breeding and raising to convert plant energy from pasture and feed into animal energy, producing livestock products such as meat, eggs, milk, wool, cashmere, hides, silk, and medicinal materials. Currently, to obtain green and natural products, outdoor grazing is necessary, along with artificial or remote monitoring to provide early warnings about the safety of farmed animals. However, current early warning systems for farmed animals cannot classify and categorize them based on known and unknown grazing areas, thus reducing the safety of outdoor grazing.

[0003] Chinese invention patent CN105303786B, published on April 13, 2018, discloses a method, device, and system for early warning of escape from transgenic animals. The method involves acquiring breeding information of transgenic animals and the status information of the monitoring equipment, collected by the monitoring equipment; determining the escape warning level of the transgenic animals based on the breeding information, the status information of the monitoring equipment, and the current breeding status of the transgenic animals; and determining the corresponding escape warning measures based on the escape warning level and a preset correspondence between escape warning levels and escape warning measures. However, this technical solution cannot provide scientific and intelligent early warning for environmental hazards during the breeding process of transgenic animals. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the current issue that safety warnings for farmed animals cannot be based on different zoning and monitoring methods for known and unknown farming areas, thus reducing the safety of free-range animals, this paper aims to achieve the following: real-time collection of farmed animal location information; autonomous identification of farmed animal location area types; scientific identification of farmed animal location hazards by zoning and classification; intelligent and rapid early warning of farmed animal location hazards; improved accuracy and quality of farmed animal behavior hazard warnings; and reduced injuries from free-range farmed animals.

[0006] (II) Technical Solution

[0007] This invention is achieved through the following technical solution: an intelligent early warning method for animal movement patterns based on big data, the method comprising the following steps:

[0008] S1. Collect real-time location information of farmed animals, identify the types of areas covered by the tracks of farmed animals in real time during the farming process, and obtain data on the identification of the types of areas covered by the tracks of farmed animals.

[0009] S2. When the breeding area is known, collect the characteristic information of the farmed animals, search for historical danger events of the farmed animals in the known breeding area, and obtain the historical location event information of the farmed animals in real time; identify the potential danger features of the farmed animals in the known breeding area, and obtain the danger information of the farmed animals in real time in the known breeding area. If there is no danger, directly end the current animal movement danger warning operation; if there is danger, directly execute step S4.

[0010] S3. When the breeding area is unknown, collect real-time environmental image information of the location of the breeding animals; identify potential danger features of the location of the breeding animals in the unknown breeding area, and obtain real-time danger information of the location of the breeding animals in the unknown breeding area; when there is no danger, directly end the current animal track danger warning operation; when there is danger, directly execute step S4.

[0011] S4. Based on the real-time location danger features of farmed animals in known or unknown farming areas, match the danger warning voice information of farmed animals in the farming coverage area, obtain the real-time danger warning voice information of farmed animals, and execute the animal movement danger warning operation.

[0012] Preferably, the steps for collecting real-time location information of farmed animals and identifying the types of areas covered by the farmed animals' tracks during the farming process to obtain farmed animal track area type identification data are as follows:

[0013] S11. The spatial location information of the target farmed animals in the breeding area is collected online through the location sensor, and the real-time location information of the farmed animals is obtained. The target farmed animals include cattle, sheep, deer, musk deer and mink. The real-time location information of the farmed animals includes the longitude, latitude and altitude of the real-time location of the farmed animals.

[0014] S12. Based on the real-time location information of the farmed animals and the location information matrix of the farmed coverage area, perform real-time coverage area type identification processing of the farmed animals' tracks during the farming process to obtain farmed animal track area type identification data, which includes known farmed areas and unknown farmed areas.

[0015] Preferably, the real-time coverage area type identification processing of the farmed animals' tracks is performed based on the real-time track location information and the farmed coverage area location information matrix to obtain farmed animal track area type identification data. The operation steps for obtaining farmed animal track area type identification data, which includes known farmed areas and unknown farmed areas, are as follows:

[0016] S121. Establish a location information matrix for aquaculture coverage areas. ,in Represents the first known aquaculture area Location information of known aquaculture coverage areas The first one indicates the unknown breeding area The location information of unknown aquaculture coverage areas; the location information of known aquaculture coverage areas refers to the longitude, latitude, and altitude of familiar aquaculture areas where aquaculture animals are grazed at fixed times; the location information of unknown aquaculture coverage areas refers to the longitude, latitude, and altitude of unfamiliar aquaculture areas where aquaculture animals are prohibited from grazing.

[0017] S122. Combine the real-time movement location information of the farmed animals with the location information matrix of the farmed coverage area. The known aquaculture coverage area location information described in the text and the location information of the unknown aquaculture coverage area The location coordinates of the tracks of farmed animals are matched separately, and the data for identifying the type of farmed animal track area is generated based on the matching results.

[0018] When the real-time location of farmed animals and If the location coordinates of the track of the farmed animal are successfully matched, it means that the current location of the target farmed animal is in a known farming area. Then, the farmed animal track area type identification data is output as a known farming area.

[0019] When the real-time location of farmed animals and If the location coordinates of the animal tracks are successfully matched, it means that the current location of the target animal tracks is in an unknown breeding area. In this case, the animal track area type identification data is output as an unknown breeding area.

[0020] Preferably, when the breeding area is known, the following steps are taken: Collect the characteristic information of the farmed animals; search for historical hazard events related to the movement of farmed animals within the known breeding area to obtain real-time historical location event information of the farmed animals; identify potential hazard features of the movement of farmed animals within the known breeding area to obtain real-time hazard information of the movement of farmed animals within the known breeding area; if there is no hazard, directly end the current animal movement hazard warning operation; if there is a hazard, directly execute step S4 as follows:

[0021] S21. When the animal track area type identification data is a known breeding area, the breed, weight, sex and age characteristics of the target animal are collected online through the breeding supervision platform to obtain the animal characteristic information.

[0022] S22. Based on the real-time location information of the farmed animals and the set of historical dangerous events of farmed animals in the known farming area, search for the historical dangerous events of farmed animals in the known farming area to obtain the real-time location and historical location event information of farmed animals.

[0023] S23. Based on the characteristics of the farmed animals and the historical location event information of the farmed animals' real-time tracks, identify the potential danger features of the farmed animals' tracks within the known farming area to obtain the danger information of the farmed animals' real-time tracks within the known farming area; if there is no danger, directly end the current animal track danger warning operation; if there is danger, directly execute step S4.

[0024] Preferably, the steps for searching for historical danger event information of farmed animals within the known farming area based on the real-time location information of the farmed animals and the set of historical danger event information of farmed animals in the known farming area are as follows:

[0025] S221. Establish a set of historical hazardous events information for farmed animals in known farming areas. ,in This indicates the location information of the known aquaculture coverage area. The information includes historical hazardous events of farmed animals in the known farming areas at the corresponding geographical locations. This information represents the location information of hazardous events that occurred within the known farming areas and the characteristics of the injured farmed animals. Hazardous events include farmed animals becoming trapped in swamps, falling from steep slopes, falling off cliffs, being hit by falling rocks, falling into water pits, falling into ditches, ingesting poisonous plants, and being injured by wild animals.

[0026] S222. The Aho-Corasick text algorithm is used to combine the real-time location information of the farmed animals with the historical hazardous event information of farmed animals in the known farming area. Information on historical hazardous events involving farmed animals in known farming areas as described in the document. Perform a search for the coordinates of the tracks of farmed animals to retrieve historical hazardous events related to farmed animals in the known farming areas, corresponding to the real-time track locations of the farmed animals. The data is then used to generate real-time location and historical location event information of farmed animals. This information represents information on dangerous events that have occurred and characteristic information of the injured farmed animals at the real-time grazing locations of the target farmed animals in a known farming area.

[0027] Preferably, based on the farmed animal characteristic information and the farmed animal real-time tracking location historical location event information, potential danger features of farmed animal tracking locations are identified within the known farming area to obtain real-time tracking location danger information of farmed animals in the known farming area; when there is no danger, the current animal tracking danger warning operation is directly terminated; when there is danger, the operation steps of step S4 are directly executed as follows:

[0028] S231. Match the farmed animal feature information with the farmed animal real-time track location and historical location event information, and generate farmed animal real-time track location danger information in the known farming area based on the farmed animal feature matching result.

[0029] When the farmed animal feature information is successfully matched with the farmed animal real-time track location historical location event information, it means that the target farmed animal at the current track location in the known farmed area is associated with the features of farmed animals that have been injured in the past. Then, the danger information of the farmed animal real-time track location in the known farmed area is output as there is danger. At this time, the farmed animal real-time track location historical location event information is output and step S4 is executed directly.

[0030] If the animal characteristics information fails to match the real-time track location and historical location event information of the farmed animals, it means that the target farmed animal at the current track location in the known farming area is not associated with the characteristics of the farmed animals that were previously injured. In this case, the real-time track location danger information of the farmed animals in the known farming area is output as "no danger exists", and the current animal track danger warning operation is terminated directly.

[0031] Preferably, when the breeding area is unknown, real-time environmental image information of the animal tracks is collected; potential danger features of the animal tracks are identified within the unknown breeding area to obtain real-time danger information of the animal tracks in the unknown breeding area; when there is no danger, the current animal track danger warning operation is terminated directly; when there is danger, step S4 is executed directly as follows:

[0032] S31. When the animal track area type identification data is an unknown breeding area, the surrounding environment image information of the real-time grazing track location of the target animal in the unknown breeding area is collected online using a portable cloud camera, and a set of real-time track location environment image information of the animal is obtained. ,in Indicates the number of collections Real-time tracking and location information of farmed animals in their environment;

[0033] S32. Based on the set of environmental image information of the real-time movement location of farmed animals and the image information matrix of dangerous features of farmed animals in unknown farming areas, identify the potential dangerous features of the movement location of farmed animals in unknown farming areas to obtain the real-time movement location danger information of farmed animals in unknown farming areas; when there is no danger, directly end the current animal movement danger warning operation; when there is danger, directly execute step S4.

[0034] Preferably, based on the set of environmental image information of the real-time movement location of farmed animals and the image information matrix of dangerous features of farmed animal movement in unknown farming areas, potential dangerous features of farmed animal movement locations in unknown farming areas are identified to obtain real-time dangerous information of farmed animal movement locations in unknown farming areas; when there is no danger, the current animal movement danger warning operation is directly terminated; when there is danger, the operation steps of step S4 are directly executed as follows:

[0035] S321. Establish a matrix of image information on the dangerous characteristics of animal tracks in unknown breeding areas. ,in Indicates the first Dangerous animal behavior image information in unknown breeding areas corresponding to dangerous events involving farmed animals; the dangerous animal behavior image information in unknown breeding areas represents the dangerous characteristic standard image information of the free-range behavior of farmed animals in unknown breeding areas set for dangerous events involving different types of farmed animals;

[0036] S322, Set the real-time movement location and environmental image information of the farmed animals. The real-time tracking location and environmental image information of farmed animals described in the article Image information matrix of dangerous features of farmed animals in the unknown farming area The image information on the dangerous characteristics of animal tracks in the unknown breeding area described in the text Image feature analysis was performed to search for environmental image information related to the real-time movement and location of the farmed animals. Matching image information of dangerous features of farmed animals in the unknown farming area The corresponding information on dangerous events involving farmed animals is used to generate real-time location and danger information for farmed animals in unknown farming areas through data identification. The specific steps for generating this real-time location and danger information for farmed animals in unknown farming areas are as follows:

[0037] S3221. Initialize the search space: Determine the matrix of dangerous search roles' animal movement hazard feature image information in the unknown breeding area. Search boundaries in the search space; a matrix of dangerous search roles and the dangerous features of animal tracks in the unknown breeding area. The formula for the search boundary in the search space is as follows: ,in Indicates a dangerous search role exist The matrix of dangerous features of animal tracks in the unknown breeding area. The location in the search space, and These represent a matrix of image information representing the dangerous features of animals being raised in an unknown breeding area, indicating the presence of a dangerous searcher. The upper and lower bounds of the search space; Indicates the value Random numbers within the interval;

[0038] S3222. Initialize the location of the danger search role: [Information matrix of danger feature image data of animal tracks in the unknown breeding area] Randomly select an image from the search space that corresponds to the real-time location and environment of the farmed animal. Matching initial image information of dangerous features of farmed animals in the unknown farming area Location;

[0039] S3223. Calculate fitness: Calculate the matrix of hazard search role's hazard feature image information of animal tracks in the unknown breeding area according to the requirements of the problem. Search the search space to find the dangerous features of animal tracks in the unknown breeding area at the current location. Real-time tracking location and environmental image information of the farmed animals The fitness value; where the fitness value The calculation formula is as follows: ,in Image information representing the dangerous characteristics of animal tracks in the unknown breeding area. and the real-time tracking location and environmental image information of the farmed animals The Middle Image feature points, ,in Image information representing the dangerous characteristics of animal tracks in the unknown breeding area. and the real-time tracking location and environmental image information of the farmed animals The total number of feature points in the image; Represents an indicator function that takes the value 0 or 1, and satisfies The value is 1 when it is used, where Image information representing the dangerous characteristics of animal tracks in the unknown breeding area. and the real-time tracking location and environmental image information of the farmed animals The Middle The coordinate difference of each image feature point Image information representing the dangerous characteristics of animal tracks in the unknown breeding area. and the real-time tracking location and environmental image information of the farmed animals Threshold for the coordinate difference of feature points in the image;

[0040] S3224. Moving the Hazard Search Role: Based on the problem-solving strategy, the hazard search role is moved and the hazard feature image information matrix of farmed animals' tracks in the unknown farming area is updated. The formula for calculating the updated position of a dangerous search role within the search space is:

[0041] ,in Indicates the range of values ​​as Random numbers in an interval , Indicates the maximum number of iterations; and All are values Random numbers within a range; It is a random number that follows a standard normal distribution; and Represents the hunger weighting factor. Indicates the value Random numbers within a range; This matrix represents the image information matrix indicating the dangerous features of farmed animals in the unknown breeding area, indicating the presence of dangerous search roles. Search the search space to find environmental image information related to the real-time movement and location of the farmed animals. The most matching image information of dangerous features of farmed animals in the unknown farming area The globally optimal position; Indicates the first In the next iteration, the danger search role is identified by the image matrix of dangerous features of animal tracks in the unknown breeding area. The location in the search space, Indicates the first In the next iteration, the danger search role is identified by the image matrix of dangerous features of animal tracks in the unknown breeding area. The position in the search space;

[0042] S3225. Update fitness: Calculate the matrix of hazard search roles' hazard feature image information on animal tracks in the unknown breeding area. Image information on the danger characteristics of farmed animals in the unknown breeding area at a new location in the search space. Real-time tracking location and environmental image information of the farmed animals fitness value;

[0043] S3226. Update the optimal solution: If the danger search role calculated in step S3225 is in the unknown breeding area, the danger feature image information matrix of the breeding animal tracks is updated. The search for new locations in the unknown breeding area yielded dangerous feature images of farmed animals' tracks. Real-time tracking location and environmental image information of the farmed animals The fitness value is better than the current location of the unknown breeding area's animal behavior danger feature image information. Real-time tracking location and environmental image information of the farmed animals The fitness value then represents the danger feature image information of the traces of farmed animals in the unknown farming area at the new location. Updated to the optimal solution;

[0044] S3227. When the maximum number of iterations is met, output the environmental image information of the real-time movement location of the farmed animal. The most matching image information of dangerous features of farmed animals in the unknown farming area Otherwise, return to step S3223 until the maximum number of iterations is satisfied;

[0045] S3228. Obtain the danger feature image information of farmed animals in the unknown farming area output in step S3227. The corresponding information on dangerous events involving farmed animals is used to construct real-time location and danger information of farmed animals in unknown farming areas through data identification.

[0046] when and Image feature matching was successful, indicating that the target farmed animal's environment at the current location of its tracks exists in the unknown farming area. If a dangerous event occurs in the breeding or raising of animals, the real-time location information of the animals in the unknown breeding area is output as a danger, and step S4 is executed directly.

[0047] when and If no matching of image features is found, it indicates that there are no dangerous events related to farmed animals in the environment surrounding the target farmed animal at the current location in the unknown farming area. In this case, the real-time location danger information of the farmed animal in the unknown farming area is output as "no danger exists", and the current animal movement danger warning operation is terminated directly.

[0048] Preferably, the operation steps for obtaining real-time animal hazard warning voice information by matching the hazard characteristics of the real-time movement location of farmed animals in known or unknown farming areas, and executing animal hazard warning operations are as follows:

[0049] S41. Construct a matrix of voice information prompts indicating dangers related to the movement of farmed animals in the farmed area. ,in Indicates the first The dangerous activity of farmed animals in the corresponding breeding coverage area is indicated by the dangerous animal movement warning voice information; the dangerous animal movement warning voice information in the breeding coverage area refers to the standard animal grazing danger warning voice information set for different types of dangerous activities.

[0050] S42. Using a bidirectional search algorithm, the generated real-time location and historical location event information of farmed animals or the real-time location and danger information of farmed animals in unknown farming areas are compared with the dangerous voice information matrix of farmed animal location danger warnings in the farming coverage area. The above-mentioned voice message provides warnings about the dangers posed by the tracks of farmed animals in the aquaculture coverage area. Perform keyword matching for farmed animal danger events to search for real-time location and historical location event information of farmed animals or real-time location danger information of farmed animals in unknown farming areas, and corresponding to the farmed animal danger warning voice information in the farming coverage area. And through data identification, real-time danger alert voice information about the movement of farmed animals is generated;

[0051] S43. Play the corresponding voice message of the real-time danger warning message of the farmed animals through a portable loudspeaker to drive the target farmed animals away from the danger at their current location and perform the animal movement danger warning operation.

[0052] A big data-based intelligent early warning system for animal tracks is used to implement the big data-based intelligent early warning method for animal tracks. The system includes a livestock track area identification module, a livestock track hazard analysis module, and a livestock track hazard early warning module.

[0053] The livestock track area identification module includes a livestock track location acquisition unit, a livestock coverage area location storage unit, and a livestock track area type identification unit.

[0054] The animal track location acquisition unit collects real-time track location information of the farmed animals through a location sensor; the farmed area location storage unit stores the farmed area location information; the animal track area type identification unit performs real-time track area type identification processing on the farmed animals' tracks during the farming process based on the real-time track location information and the farmed area location information to obtain farmed animal track area type identification data.

[0055] The farmed animal movement hazard analysis module includes a farmed animal characteristic information acquisition unit, a known farmed animal historical hazard event storage unit, a farmed animal movement location historical hazard event information search unit, a farmed animal movement hazard identification unit in known farmed areas, a farmed animal movement location environmental image acquisition unit, a farmed animal movement hazard characteristic image storage unit in unknown farmed areas, and a farmed animal movement hazard identification unit in unknown farmed areas.

[0056] The farmed animal characteristic information collection unit collects farmed animal characteristic information through the farmed animal supervision platform; the known farmed animal historical danger event storage unit stores known farmed animal historical danger event information in the known farmed area; the farmed animal track location historical danger event information search unit searches for farmed animal historical danger event information in the known farmed area based on the farmed animal real-time track location information and the known farmed animal historical danger event information, to obtain farmed animal real-time track location historical location event information; the known farmed animal track danger identification unit identifies known farmed animal danger based on the farmed animal characteristic information and the farmed animal real-time track location historical location event information. The system identifies potential hazards in the track locations of farmed animals within a given area, obtaining real-time hazard information on the track locations of farmed animals in known farming areas. The farmed animal track location environmental image acquisition unit acquires real-time environmental image information on the track locations of farmed animals using a portable cloud camera. The unknown farmed animal track hazard feature image storage unit stores image information on the track hazard features of farmed animals in unknown farming areas. The unknown farmed animal track hazard identification unit identifies potential hazards in the track locations of farmed animals within the unknown farming area based on the real-time environmental image information and the image information on the hazard features of farmed animals in unknown farming areas, obtaining real-time hazard information on the track locations of farmed animals in unknown farming areas.

[0057] The animal movement danger warning module includes an animal movement danger warning voice storage unit for the animal movement danger area, an animal movement danger warning voice matching unit, and an animal movement danger warning unit.

[0058] The animal movement hazard warning voice storage unit in the aquaculture coverage area is used to store animal movement hazard warning voice information in the aquaculture coverage area; the animal movement hazard warning voice matching unit performs matching processing on the animal movement hazard warning voice information in the aquaculture coverage area based on the real-time movement location historical location event information of the aquaculture animals or the real-time movement location hazard information of the aquaculture animals in unknown aquaculture areas and the animal movement hazard warning voice information in the aquaculture coverage area to obtain real-time animal movement hazard warning voice information; the animal movement hazard warning unit performs animal movement hazard warning operation based on the real-time animal movement hazard warning voice information and in conjunction with a portable loudspeaker.

[0059] (III) Beneficial Effects

[0060] This invention provides an intelligent early warning system and method for animal movement patterns based on big data. It has the following beneficial effects:

[0061] I. Real-time location information of farmed animals is collected through location sensors, providing real data support for the zoned monitoring of farmed animal location. Based on the real-time location information of farmed animals, combined with numerical analysis and farmed area location information based on big data storage, the farmed animal tracks are scientifically zoned and identified according to known and unknown farmed areas, improving the accuracy and precision of farmed animal track hazard warning detection.

[0062] II. Obtaining characteristic information of farmed animals through the aquaculture supervision platform provides reliable data support for scientific early warning of dangers posed by farmed animals in known aquaculture areas; conducting dynamic early warning analysis of potential dangers posed by farmed animals in known aquaculture areas based on characteristic information, real-time location, and historical location event information, enabling the scientific establishment of an early warning system for potential dangers posed by farmed animals in known aquaculture areas based on historical data, thereby improving the reliability of early warnings; using portable cloud cameras to dynamically and accurately monitor real-time location and environmental images of farmed animals in unknown aquaculture areas, providing accurate data for monitoring potential dangers posed by farmed animals in unknown aquaculture areas; conducting intelligent early warning of potential dangers posed by farmed animals in unknown aquaculture areas based on real-time location and environmental image information, combined with AI intelligent algorithms and big data-stored characteristic image information of potential dangers posed by farmed animals in unknown aquaculture areas, enabling the scientific establishment of an early warning system for potential dangers posed by farmed animals in unknown aquaculture areas based on real-time environmental characteristics, and enabling the scientific establishment of an early warning system for farmed animals in known aquaculture areas according to known and unknown aquaculture scenarios, thereby improving the applicability and quality of early warning detection of potential dangers posed by farmed animals.

[0063] Third, by combining real-time location and historical location event information of farmed animals or real-time location and danger information of farmed animals in unknown farming areas with intelligent search algorithms and scientifically established voice warning information for dangers of farmed animals in the farming coverage area, the system can efficiently match the voice warning information for dangers of farmed animals in the farming coverage area. This enables the independent creation of danger warning voice messages based on the danger type of farmed animals, improving the response effect of danger warnings for farmed animals. Furthermore, by combining real-time danger warning voice information for farmed animals with portable loudspeakers, the system can dynamically customize and execute animal danger warning operations, achieving efficient and accurate warning prompts for dangers of farmed animals, and improving the accuracy and speed of response to danger warnings for farmed animals. Attached Figure Description

[0064] Figure 1 A schematic diagram of the modules of the intelligent early warning system for animal movement based on big data provided by the present invention;

[0065] Figure 2 A flowchart of the intelligent early warning method for animal movement based on big data provided by the present invention. Detailed Implementation

[0066] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] An embodiment of the intelligent early warning system and method for animal movement based on big data is as follows:

[0068] Example 1:

[0069] Please see Figures 1-2 A smart early warning method for animal movement based on big data, the method includes the following steps:

[0070] S1. Collect real-time location information of farmed animals, identify the types of areas covered by the tracks of farmed animals in real time during the farming process, and obtain data on the identification of the types of areas covered by the tracks of farmed animals.

[0071] S2. When the breeding area is known, collect the characteristic information of the farmed animals, search for historical danger events of the farmed animals in the known breeding area, and obtain the historical location event information of the farmed animals in real time; identify the potential danger features of the farmed animals in the known breeding area, and obtain the danger information of the farmed animals in real time in the known breeding area. If there is no danger, directly end the current animal movement danger warning operation; if there is danger, directly execute step S4.

[0072] S3. When the breeding area is unknown, collect real-time environmental image information of the location of the breeding animals; identify potential danger features of the location of the breeding animals in the unknown breeding area, and obtain real-time danger information of the location of the breeding animals in the unknown breeding area; when there is no danger, directly end the current animal track danger warning operation; when there is danger, directly execute step S4.

[0073] S4. Based on the real-time location danger features of farmed animals in known or unknown farming areas, match the danger warning voice information of farmed animals in the farming coverage area, obtain the real-time danger warning voice information of farmed animals, and execute the animal movement danger warning operation.

[0074] For further details, please refer to Figures 1-2 The steps for collecting real-time location information of farmed animals and identifying the types of areas covered by their tracks during the farming process to obtain data on the types of farmed animal track areas are as follows:

[0075] S11. The spatial location information of the target farmed animals in the breeding area is collected online through the location sensor, and the real-time location information of the farmed animals is obtained. The target farmed animals include cattle, sheep, deer, musk deer and mink. The real-time location information of the farmed animals includes the longitude, latitude and altitude of the farmed animals' real-time location.

[0076] S12. Based on the real-time location information of farmed animals and the location information matrix of the farmed coverage area, the real-time coverage area type identification of farmed animals during the farming process is performed to obtain farmed animal track area type identification data, which includes known farmed areas and unknown farmed areas.

[0077] Based on the real-time location information of farmed animals and the location information matrix of the farmed area, the real-time coverage area type of farmed animal tracks is identified during the farming process. This results in farmed animal track area type identification data, which includes both known and unknown farmed areas. The operational steps are as follows:

[0078] S121. Establish a location information matrix for aquaculture coverage areas. ,in Represents the first known aquaculture area Location information of known aquaculture coverage areas The first one indicates the unknown breeding area Location information for unknown aquaculture areas; location information for known aquaculture areas refers to the longitude, latitude, and altitude of familiar aquaculture areas where animals are grazed at fixed times; location information for unknown aquaculture areas refers to the longitude, latitude, and altitude of unfamiliar aquaculture areas where grazing is prohibited.

[0079] S122. Combine the real-time location information of farmed animals with the location information matrix of the farmed coverage area. Known location information of aquaculture coverage area Location information of unknown aquaculture coverage areas The location coordinates of the tracks of farmed animals are matched separately, and the data for identifying the type of farmed animal track area is generated based on the matching results.

[0080] When the real-time location of farmed animals and If the location coordinates of the track of the farmed animal are successfully matched, it means that the current location of the target farmed animal is in a known farming area. In this case, the output data for the farmed animal track area type identification is "known farming area".

[0081] When the real-time location of farmed animals and If the location coordinates of the animal tracks are successfully matched, it means that the current location of the target animal track is in an unknown breeding area. In this case, the output data for the animal track area type identification is "unknown breeding area".

[0082] The animal tracking location acquisition unit uses location sensors to collect real-time tracking location information of farmed animals, providing real data support for zoned monitoring of real-time tracking location of farmed animals. The farmed area location storage unit and the farmed animal tracking area type identification unit work together to scientifically identify the tracking of farmed animals in the farmed process according to known and unknown farmed areas based on real-time tracking location information of farmed animals, combined with numerical analysis and farmed area location information based on big data storage, thereby improving the accuracy and precision of dangerous early warning detection of farmed animal tracking.

[0083] For further details, please refer to Figures 1-2 When the breeding area is known, collect the characteristic information of the farmed animals, search for historical hazard events related to the movement of farmed animals within the known breeding area, and obtain the historical location event information of the farmed animals' real-time movement locations; identify potential hazard features of the farmed animals' movement locations within the known breeding area, and obtain the hazard information of the farmed animals' real-time movement locations within the known breeding area. If there is no hazard, directly end the current animal movement hazard warning operation; if there is a hazard, directly execute step S4 as follows:

[0084] S21. When the data for identifying the type of animal track area is a known breeding area, the breed, weight, sex and age characteristics of the target breeding animal are collected online through the breeding supervision platform, and the breeding animal characteristic information is obtained.

[0085] S22. Based on the real-time location information of farmed animals and the set of historical dangerous events of farmed animals in known farming areas, search for the historical dangerous events of farmed animals in known farming areas to obtain the real-time location and historical location event information of farmed animals.

[0086] S23. Based on the characteristics of farmed animals and the real-time location and historical location event information of farmed animals, identify the potential danger characteristics of farmed animal locations within the known farming area to obtain the real-time location danger information of farmed animals in the known farming area; if there is no danger, directly end the current animal movement danger warning operation; if there is danger, directly execute step S4.

[0087] The steps for obtaining the real-time location and historical location event information of farmed animals by searching for their locations within a known farming area based on real-time location information and historical event information of farmed animals in a known farming area are as follows:

[0088] S221. Establish a set of historical hazardous events information for farmed animals in known farming areas. ,in Indicates the location information of the known aquaculture coverage area. Information on historical hazardous events involving farmed animals in the known farming area at the corresponding geographical location; this information represents the location of hazardous events involving farmed animals that occurred within the known farming area and the characteristics of the injured farmed animals; among these hazardous events are: farmed animals getting stuck in swamps, farmed animals falling from steep slopes, farmed animals falling off cliffs, farmed animals being hit by falling rocks, farmed animals falling into water pits, farmed animals falling into ditches, farmed animals ingesting poisonous plants, and farmed animals being injured by wild animals;

[0089] S222. The Aho-Corasick text algorithm is used to combine the real-time location information of farmed animals with the historical hazardous event information of farmed animals in known farming areas. Information on historical hazardous events involving farmed animals in known farming areas The search is conducted to locate the coordinates of farmed animals' tracks, and to retrieve historical hazardous events related to farmed animals in known farming areas corresponding to the real-time locations of these animals. The data is then used to generate real-time location and historical location event information for farmed animals. This information represents information on dangerous events that have occurred and characteristics of the injured farmed animals at their real-time grazing locations within a known farming area.

[0090] Based on the characteristic information of farmed animals and the historical location and event information of farmed animals' real-time tracks, potential danger features of farmed animal tracks within the known farming area are identified, and the real-time track location danger information of farmed animals in the known farming area is obtained. When there is no danger, the current animal track danger warning operation is directly terminated; when there is danger, the operation steps of S4 are directly executed as follows:

[0091] S231. Match the characteristic information of farmed animals with the real-time location and historical location event information of farmed animals, and generate danger information of the real-time location of farmed animals in known farming areas based on the matching results.

[0092] When the characteristic information of farmed animals is successfully matched with the real-time location and historical location event information of farmed animals, it means that the target farmed animal at the current location in the known farming area is associated with the characteristics of farmed animals that have been injured in the past. Then, the danger information of the real-time location of farmed animals in the known farming area is output as "danger exists". At this time, the real-time location and historical location event information of farmed animals is output and step S4 is executed directly.

[0093] If the animal characteristics information fails to match the real-time location and historical location event information of the farmed animals, it means that the target farmed animal at the current location in the known farming area is not related to the characteristics of the farmed animals that were previously injured. In this case, the real-time location danger information of the farmed animals in the known farming area is output as "no danger exists", and the current animal movement danger warning operation is terminated directly.

[0094] When the breeding area is unknown, collect real-time environmental image information on the location of the farmed animals; identify potential danger features of the farmed animals' locations within the unknown breeding area to obtain real-time danger information on the location of the farmed animals' locations in the unknown breeding area; if there is no danger, directly end the current animal movement danger warning operation; if there is danger, directly execute the operation steps of S4 as follows:

[0095] S31. When the data for identifying the type of animal track area is an unknown breeding area, use a portable cloud camera to collect images of the surrounding environment of the target animal's real-time grazing location in the unknown breeding area, and obtain a set of environmental images of the animal's real-time track location. ,in Indicates the number of collections Real-time tracking and location information of farmed animals in their environment;

[0096] S32. Based on the set of environmental image information of real-time track location of farmed animals and the image information matrix of dangerous features of track location of farmed animals in unknown farming areas, identify the potential dangerous features of track location of farmed animals in unknown farming areas, and obtain the dangerous information of real-time track location of farmed animals in unknown farming areas; when there is no danger, directly end the current animal track danger warning operation; when there is danger, directly execute step S4.

[0097] Based on the set of environmental image information of real-time track locations of farmed animals and the image information matrix of hazard features of farmed animal track locations in unknown farming areas, potential hazard features of farmed animal track locations in unknown farming areas are identified, and real-time hazard information of farmed animal track locations in unknown farming areas is obtained. When there is no hazard, the current animal track hazard warning operation is directly terminated; when there is hazard, step S4 is directly executed. The operation steps are as follows:

[0098] S321. Establish a matrix of image information on the dangerous characteristics of animal tracks in unknown breeding areas. ,in Indicates the first Dangerous animal activity images in unknown breeding areas corresponding to dangerous events involving farmed animals; dangerous animal activity images in unknown breeding areas represent standard images of dangerous animal activity patterns in unknown breeding areas set for different types of dangerous events involving farmed animals.

[0099] S322, Collect real-time tracking location and environmental image information of farmed animals Real-time tracking location and environmental image information of farmed animals Image information matrix of dangerous features of farmed animals in unknown breeding areas Image information on the dangerous characteristics of farmed animals in unknown breeding areas Image feature analysis was performed to search for environmental image information related to the real-time movement and location of farmed animals. Matching image information of dangerous features of farmed animals in unknown farming areas The corresponding information on dangerous events involving farmed animals is used to generate real-time location and hazard information for farmed animals in unknown farming areas through data identification. The specific steps for generating this information are as follows:

[0100] S3221. Initialize the search space: Determine the dangerous search role's dangerous feature image information matrix of animal tracks in unknown breeding areas. Search boundaries in the search space; dangerous search roles in unknown breeding areas; dangerous feature image information matrix of farmed animals. The formula for the search boundary in the search space is as follows: ,in Indicates a dangerous search role exist Image information matrix of dangerous features of farmed animals in unknown breeding areas The location in the search space, and These represent the image information matrix of dangerous features of farmed animals in unknown breeding areas, representing the search roles for dangerous animals. The upper and lower bounds of the search space; Indicates the value Random numbers within the interval;

[0101] S3222. Initialize the location of the danger search role: Matrix of danger feature image information of animal tracks in unknown breeding areas. Randomly select an image of the real-time location and environment of farmed animals within the search space. Matching image information of hazard features of farmed animals in the initial unknown farming area Location;

[0102] S3223. Calculate fitness: Calculate the matrix of hazard search role's hazard feature image information of animal tracks in unknown breeding areas, based on the requirements of the problem. Search the search space to find images of dangerous features and traces of farmed animals in unknown farming areas at the current location. Real-time tracking and location information of farmed animals The fitness value; where the fitness value The calculation formula is as follows: ,in Image information representing the dangerous characteristics of farmed animals in unknown farming areas. Real-time tracking and location information of farmed animals The Middle Image feature points, ,in Image information representing the dangerous characteristics of farmed animals in unknown farming areas. Real-time tracking and location information of farmed animals The total number of feature points in the image; Represents an indicator function that takes the value 0 or 1, and satisfies The value is 1 when it is used, where Image information representing the dangerous characteristics of farmed animals in unknown farming areas. Real-time tracking and location information of farmed animals The Middle The coordinate difference of each image feature point Image information representing the dangerous characteristics of farmed animals in unknown farming areas. Real-time tracking and location information of farmed animals Threshold for the coordinate difference of feature points in the image;

[0103] S3224. Moving the Hazard Search Role: Based on the problem-solving strategy, the hazard search role is moved and the hazard feature image information matrix of farmed animals in unknown farming areas is updated. The formula for calculating the updated position of a dangerous search role within the search space is:

[0104] ,in Indicates the range of values ​​as Random numbers in an interval , Indicates the maximum number of iterations; and All are values Random numbers within a range; It is a random number that follows a standard normal distribution; and Represents the hunger weighting factor. Indicates the value Random numbers within a range; This matrix represents the image information of dangerous features of farmed animals in unknown breeding areas, indicating the presence of dangerous search roles. Search the search space to find environmental image information related to the real-time movement and location of farmed animals. The most matching image information of dangerous features of farmed animals in unknown breeding areas. The globally optimal position; Indicates the first In the next iteration, the danger search role uses a matrix of image information of dangerous features of farmed animals in unknown farming areas. The location in the search space, Indicates the first In the next iteration, the danger search role uses a matrix of image information of dangerous features of farmed animals in unknown farming areas. The position in the search space;

[0105] S3225. Update fitness: Calculate the danger feature image information matrix of dangerous search roles in the traces of farmed animals in unknown farming areas. Image information on the danger characteristics of farmed animals in unknown breeding areas of new locations within the search space. Real-time tracking and location information of farmed animals fitness value;

[0106] S3226. Update the optimal solution: If the danger search role calculated in step S3225 has a danger feature image information matrix of animal tracks in the unknown breeding area, then... Searching for new locations in unknown breeding areas, including images of animal tracks and hazard characteristics. Real-time tracking and location information of farmed animals The fitness value is better than the current location of the unknown breeding area's animal track danger feature image information. Real-time tracking and location information of farmed animals The fitness value indicates the danger characteristics of animal tracks in the unknown breeding area at the new location. Updated to the optimal solution;

[0107] S3227. When the maximum number of iterations is met, output the real-time location and environmental image information of the farmed animals. The most matching image information of dangerous features of farmed animals in unknown breeding areas. Otherwise, return to step S3223 until the maximum number of iterations is satisfied;

[0108] S3228. Obtain the danger feature image information of farmed animals in the unknown farming area output in step S3227. The corresponding information on dangerous events involving farmed animals is used to construct real-time location and danger information of farmed animals in unknown farming areas through data identification.

[0109] when and Image feature matching was successful, indicating that the target farmed animal's environment at the current location of its tracks exists in the unknown farming area. If a dangerous event occurs in the breeding or raising of animals, the real-time location information of the breeding animals in the unknown breeding area will be output as a danger. At this time, the real-time location information of the breeding animals in the unknown breeding area will be output and step S4 will be executed directly.

[0110] when and If no matching of image features is found, it means that there are no dangerous events related to farmed animals in the environment surrounding the target farmed animal at the current location of the track in the unknown farming area. In this case, the real-time location danger information of farmed animals in the unknown farming area is output as "no danger". At this time, the current animal track danger warning operation is terminated directly.

[0111] The system utilizes a farmed animal characteristic information collection unit and a farmed animal monitoring platform to acquire farmed animal characteristic information, providing reliable data support for scientifically warning of potential dangers associated with farmed animal tracks in known farming areas. A farmed animal track location historical hazard event information search unit and a known farmed animal track hazard identification unit work together to perform dynamic early warning analysis of potential dangers associated with farmed animal tracks within known farming areas based on farmed animal characteristic information and real-time track location historical event information. This enables the scientific establishment of a known farmed animal track hazard early warning system based on historical data, improving the reliability of farmed animal track hazard early warning. Finally, a farmed animal track location environmental image acquisition unit uses a portable cloud camera to dynamically and accurately monitor unknown farming areas. Real-time environmental images of animal tracks provide accurate data for monitoring the dangers of animal tracks in unknown breeding areas. The image storage unit and the identification unit of animal tracks in unknown breeding areas work together to provide intelligent early warning of potential dangers based on real-time environmental images of animal tracks, combined with AI intelligent algorithms and big data-based images of animal tracks in unknown breeding areas. This enables the scientific establishment of an early warning system for animal tracks in unknown breeding areas based on real-time environmental characteristics, and to scientifically establish an early warning system for animal tracks based on known and unknown breeding scenarios, thereby improving the applicability and quality of animal track danger detection.

[0112] For further details, please refer to Figures 1-2 The operation steps for obtaining real-time animal hazard warning voice information and executing animal hazard warning operations based on the real-time location hazard features of farmed animals in known or unknown farming areas are as follows:

[0113] S41. Construct a matrix of voice information prompts indicating dangers related to the movement of farmed animals in the farmed area. ,in Indicates the first The audio alerts for dangerous animal behavior in the breeding coverage area corresponding to dangerous events involving farmed and farmed animals; the audio alerts for dangerous animal behavior in the breeding coverage area represent standard animal grazing warning audio messages set for different types of dangerous events involving farmed and farmed animals.

[0114] S42. Employ a bidirectional search algorithm to combine the generated real-time location and historical location event information of farmed animals or the real-time location and danger information of farmed animals in unknown farming areas with the voice information matrix of danger warnings for farmed animals in the farming coverage area. Danger warning voice messages regarding the tracks of farmed animals in the central aquaculture coverage area This involves keyword matching to search for information on the real-time location and historical location of farmed animals, or for real-time location and danger information of farmed animals in unknown farming areas, along with corresponding farmed animal hazard warning voice messages within the farmed area coverage region. And through data identification, real-time danger alert voice information about the movement of farmed animals is generated;

[0115] S43. Play real-time danger warning voice messages about the movement of farmed animals through a portable loudspeaker to drive the target farmed animals away from the danger at their current location and perform animal movement danger warning operations.

[0116] By employing a voice storage unit and a voice matching unit for hazard warnings of farmed animals within the farmed area, and combining real-time location and historical location event information of farmed animals or real-time location hazard information of farmed animals in unknown farmed areas with intelligent search algorithms, the system efficiently matches the voice information for hazard warnings of farmed animals within the farmed area with scientifically established hazard warning voice information. This enables the independent creation of hazard warning voices based on the hazard type of farmed animals, improving the response effect of hazard warnings. The hazard warning unit for farmed animals, based on real-time hazard warning voice information and combined with a portable loudspeaker, dynamically customizes the execution of hazard warning operations, achieving efficient and accurate hazard warning prompts and improving the accuracy and speed of hazard warning responses.

[0117] Example 2:

[0118] Please see Figures 1-2 The intelligent early warning system for animal tracks based on big data is used to realize the intelligent early warning method for animal tracks based on big data. The system includes a module for identifying the area of ​​farmed animal tracks, a module for analyzing the danger of farmed animal tracks, and a module for early warning of the danger of farmed animal tracks.

[0119] The livestock track area identification module includes a livestock track location acquisition unit, a livestock coverage area location storage unit, and a livestock track area type identification unit.

[0120] The system includes a farmed animal track location acquisition unit, which acquires real-time track location information of farmed animals through a location sensor; a farmed area location storage unit, which stores the farmed area location information; and a farmed animal track area type identification unit, which performs real-time track area type identification processing on the farmed animals' tracks based on the real-time track location information and the farmed area location information to obtain farmed animal track area type identification data.

[0121] The farmed animal movement hazard analysis module includes a farmed animal characteristic information collection unit, a known farmed area historical hazard event storage unit, a farmed animal movement location historical hazard event information search unit, a farmed animal movement hazard identification unit in known farmed areas, a farmed animal movement location environmental image collection unit, a farmed animal movement hazard feature image storage unit in unknown farmed areas, and a farmed animal movement hazard identification unit in unknown farmed areas.

[0122] The system comprises the following components: a farmed animal characteristic information collection unit, which collects farmed animal characteristic information through a farmed animal supervision platform; a known farmed animal historical hazard event storage unit, which stores known farmed animal historical hazard event information in known farmed areas; a farmed animal track location historical hazard event information search unit, which searches for farmed animal historical hazard event information within known farmed areas based on real-time farmed animal track location information and known farmed animal historical hazard event information, thus obtaining real-time farmed animal track location historical location event information; and a known farmed animal track hazard identification unit, which identifies hazard within known farmed areas based on farmed animal characteristic information and real-time farmed animal track location historical location event information. The system identifies potential hazards in the locations of farmed animals' tracks, obtaining real-time hazard information on the locations of farmed animals in known farming areas; a farmed animal track location environmental image acquisition unit acquires real-time environmental images of farmed animals' tracks using a portable cloud camera; a farmed animal track hazard feature image storage unit stores image information on the hazard features of farmed animals' tracks in unknown farming areas; and a farmed animal track hazard identification unit identifies potential hazards in the locations of farmed animals' tracks within unknown farming areas based on the real-time environmental images and the hazard feature images, obtaining real-time hazard information on the locations of farmed animals' tracks in unknown farming areas.

[0123] The early warning module for the movement of farmed animals includes a voice storage unit for warnings of the movement of farmed animals in the farmed area, a voice matching unit for warnings of the movement of farmed animals, and a warning unit for the movement of farmed animals.

[0124] A voice storage unit for warning of dangers related to the movement of farmed animals within the farmed area is used to store voice information warning of dangers related to the movement of farmed animals within the farmed area. A voice matching unit for warning of dangers related to the movement of farmed animals performs matching processing on the voice information warning of dangers related to the movement of farmed animals within the farmed area, based on historical location event information of the real-time movement of farmed animals or real-time location danger information of farmed animals in unknown farmed areas, to obtain real-time voice information warning of dangers related to the movement of farmed animals. A voice warning unit for warning of dangers related to the movement of farmed animals performs animal movement danger warning operations based on the real-time voice information warning of dangers related to the movement of farmed animals and in conjunction with a portable loudspeaker.

[0125] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent early warning of animal movement patterns based on big data, characterized in that, The method includes the following steps: S1. Collect real-time location information of farmed animals, identify the types of areas covered by the tracks of farmed animals in real time during the farming process, and obtain data on the identification of the types of areas covered by the tracks of farmed animals. The operation steps of S1 are as follows: S11. The spatial location information of the target farmed animals in the breeding area is collected online through the location sensor, and the real-time location information of the farmed animals is obtained. S12. Based on the real-time track location information of the farmed animals and the location information matrix of the farmed coverage area, perform real-time coverage area type identification processing of the track of the farmed animals during the farming process to obtain farmed animal track area type identification data, which includes known farmed areas and unknown farmed areas. The operation steps of S12 are as follows: S121. Establish a location information matrix for aquaculture coverage areas. The include and ;in Represents the first known aquaculture area Location information of known aquaculture coverage areas. The first one indicates the unknown breeding area Location information of an unknown aquaculture coverage area; S122, Combine the real-time location information of the farmed animals with the... The above and stated The location coordinates of the tracks of farmed animals are matched separately, and the data for identifying the type of farmed animal track area is generated based on the matching results. When the real-time location of farmed animals is If the location coordinates of the tracks of farmed animals are successfully matched, the farmed animal track area type identification data will be output as a known farmed area. When the real-time location of farmed animals is If the location coordinates of the farmed animal tracks are successfully matched, the farmed animal track area type identification data will be output as an unknown farmed area. S2. When the breeding area is known, collect the characteristic information of the farmed animals, search for historical danger events of the farmed animals in the known breeding area, and obtain the historical location event information of the farmed animals in real time; identify the potential danger features of the farmed animals in the known breeding area, and obtain the danger information of the farmed animals in real time in the known breeding area. If there is no danger, directly end the current animal movement danger warning operation; if there is danger, directly execute step S4. The operation steps of S2 are as follows: S21. When the animal track area type identification data is a known breeding area, the breed, weight, sex and age characteristics of the target animal are collected online through the breeding supervision platform to obtain the animal characteristic information. S22. Based on the real-time location information of the farmed animals and the set of historical dangerous events of farmed animals in the known farming area, search for the historical dangerous events of farmed animals in the known farming area to obtain the real-time location and historical location event information of farmed animals. The operation steps of S22 are as follows: S221. Establish a set of historical hazardous events information for farmed animals in known farming areas. The include ;in Indicates the Historical hazardous events involving farmed animals in the corresponding geographical location's known farming area; S222. The Aho-Corasick text algorithm is used to combine the real-time location information of the farmed animals with the... The above Perform a search for the coordinates of the tracks of farmed animals, and retrieve the coordinates of the real-time tracks of the farmed animals corresponding to the coordinates of the tracks. And through data identification, real-time location and historical location event information of farmed animals are generated; S23. Based on the characteristics of the farmed animals and the historical location event information of the farmed animals' real-time tracks, identify the potential danger features of the farmed animals' tracks within the known farming area to obtain the danger information of the farmed animals' real-time tracks within the known farming area; if there is no danger, directly end the current animal track danger warning operation; if there is danger, directly execute step S4. The operation steps of S23 are as follows: S231. Match the farmed animal feature information with the farmed animal real-time track location and historical location event information, and generate farmed animal real-time track location danger information in the known farming area based on the farmed animal feature matching result. When the animal feature information is successfully matched with the real-time track location and historical location event information of the animal, the danger information of the real-time track location of the known animal in the breeding area is output as danger exists. At this time, the real-time track location and historical location event information of the animal is output and step S4 is executed directly. If the animal characteristics information is not matched with the real-time location and historical location event information of the farmed animals, the real-time location danger information of the farmed animals in the known farming area is output as "no danger exists", and the current animal movement danger warning operation is terminated directly. S3. When the breeding area is unknown, collect real-time environmental image information of the location of the breeding animals; identify potential danger features of the location of the breeding animals in the unknown breeding area, and obtain real-time danger information of the location of the breeding animals in the unknown breeding area; when there is no danger, directly end the current animal track danger warning operation; when there is danger, directly execute step S4. The operation steps of S3 are as follows: S31. When the animal track area type identification data is an unknown breeding area, the surrounding environment image information of the real-time grazing track location of the target animal in the unknown breeding area is collected online using a portable cloud camera, and a set of real-time track location environment image information of the animal is obtained. The include ;in Indicates the number of collections Real-time tracking and location information of farmed animals in their environment; S32. Based on the set of environmental image information of the real-time movement location of farmed animals and the image information matrix of dangerous features of farmed animals in unknown farming areas, identify the potential dangerous features of the movement location of farmed animals in unknown farming areas to obtain the dangerous information of the real-time movement location of farmed animals in unknown farming areas; when there is no danger, directly end the current animal movement danger warning operation; when there is danger, directly execute step S4. S4. Based on the real-time location danger features of farmed animals in known or unknown farming areas, match the danger warning voice information of farmed animals in the farming coverage area, obtain the real-time danger warning voice information of farmed animals, and execute the animal movement danger warning operation.

2. The intelligent early warning method for animal movement based on big data according to claim 1, characterized in that: The operation steps of S32 are as follows: S321. Establish a matrix of image information on the dangerous characteristics of animal tracks in unknown breeding areas. The include ;in Indicates the first Image information on the dangerous characteristics of animal tracks in unknown breeding areas corresponding to dangerous events involving farmed and farmed animals; S322, the above The above With the The above Perform image feature analysis to search for images that match the... The matching The corresponding information on dangerous events involving farmed animals is used to generate real-time location and danger information of farmed animals in unknown farming areas through data identification. The specific steps for generating the real-time location and danger information of farmed animals in unknown farming areas are as follows: S3221. Initialize the search space: Determine the dangerous search roles in the [location / section]. The search boundary in the search space; S3222, Initialize the location of the danger search role: In the... Randomly select a match within the search space. Matching the initial statement Location; S3223. Calculate fitness: Calculate the hazard search role according to the requirements of the problem being solved. The search space is used to find the current location. With the fitness value; S3224. Move the hazard search role: Based on the problem-solving strategy, move the hazard search role and update the settings in the table. The position in the search space; S3225, Update Fitness: Calculate the danger search role in the... The new location in the search space With the fitness value; S3226. Update the optimal solution: If the danger search role calculated in step S3225 is in the... The new location searched in the middle With the The fitness value is better than the current position. With the The fitness value, then the fitness value at the new position Updated to the optimal solution; S3227. When the maximum number of iterations is satisfied, the output is the same as described above. The most matching Otherwise, return to step S3223 until the maximum number of iterations is satisfied; S3228, The output of step S3227 The corresponding information on dangerous events involving farmed animals is used to construct real-time location and danger information of farmed animals in unknown farming areas through data identification. when and If image feature matching is successful, the real-time location danger information of the farmed animals in the unknown breeding area is output as "danger exists". At this time, the real-time location danger information of the farmed animals in the unknown breeding area is output and step S4 is executed directly. when and If no matching of image features is found, the output of the real-time location danger information of the farmed animals in the unknown breeding area is "no danger," and the current animal movement danger warning operation is terminated directly.

3. The intelligent early warning method for animal movement based on big data according to claim 2, characterized in that: The operation steps for obtaining real-time animal hazard warning voice information and executing animal hazard warning operation S4 based on the real-time location hazard features of farmed animals in known or unknown farming areas are as follows: S41. Construct a matrix of voice information prompts indicating dangers related to the movement of farmed animals in the farmed area. The include ;in Indicates the first Voice alerts indicating potential dangers related to animal behavior within the designated breeding area, corresponding to dangerous incidents involving farmed or farmed animals. S42. Using a bidirectional search algorithm, the generated real-time location and historical location event information of the farmed animals or the real-time location and danger information of the farmed animals in the unknown farming area are compared with the... The above Perform keyword matching for farmed animal hazardous events to search for the real-time location and historical location event information of the farmed animals or the real-time location and hazardous information of farmed animals in the unknown farming area. And through data identification, real-time danger alert voice information about the movement of farmed animals is generated; S43. Play the corresponding voice message of the real-time danger warning message of the farmed animals through a portable loudspeaker to drive the target farmed animals away from the danger at their current location and perform the animal movement danger warning operation.

4. A big data-based intelligent animal movement early warning system, used to implement the big data-based intelligent animal movement early warning method as described in any one of claims 1-3, characterized in that: The system includes a module for identifying areas of farmed animal tracks, a module for analyzing the risks associated with farmed animal tracks, and a module for early warning of risks associated with farmed animal tracks.

Images