Intelligent fence monitoring and management system for animal husbandry

Abstract

The invention provides an intelligent fence monitoring and management system for animal husbandry, and particularly relates to the technical field of electric digital data processing, and the system comprises a video obtaining module which is used for obtaining a monitoring video corresponding to at least one monitoring area; the area analysis module is used for analyzing each monitoring area based on the monitoring video to obtain a monitoring object appearing in the early warning area; the monitoring data acquisition module is used for acquiring motion information and environment information of a monitored object; the first calculation module is used for calculating a motion trend degree and an environment trend degree according to the motion information and the environment information respectively; the second calculation module is used for calculating the fence crossing possibility based on the motion trend degree and the environment trend degree; and the early warning sending module is used for judging whether to send early warning information about the monitored object to the terminal or not based on the fence crossing possibility. According to the invention, objects in the early warning area are monitored in real time in combination with the environment and monitoring object behaviors, the animal husbandry monitoring efficiency is effectively improved, and the management efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric digital data processing, and in particular to an intelligent fence monitoring and management system for animal husbandry. Background Art

[0002] In the existing animal husbandry management, in order to prevent livestock from climbing over fences, real-time status monitoring can be used to analyze the movement status of livestock using computer vision technology to determine whether there is a risk of livestock climbing over fences. However, traditional animal husbandry fences are generally made of metal materials such as wire mesh and steel pipes. Such fences are exposed to harsh environments such as wind and rain, and livestock biting for a long time, and are prone to rust, fracture, structural deformation, and other problems. Therefore, if only the movement status of livestock in the image is analyzed without considering the environmental status of the fence, there will still be a serious risk of livestock escape. Summary of the invention

[0003] In order to solve the above technical problems, the purpose of the present invention is to provide an intelligent fence monitoring and management system for animal husbandry. In order to achieve the above purpose, the technical solution adopted by the present invention is as follows: The present application provides an intelligent fence monitoring and management system for animal husbandry, comprising: a video acquisition module, used to acquire monitoring videos corresponding to at least one monitoring area, each of which includes a fence and an early warning area adjacent to the fence; an area analysis module, used to analyze each of the monitoring areas based on the monitoring video to obtain monitoring objects appearing in the early warning area; a monitoring data acquisition module, used to acquire motion information and environmental information of the monitoring object, the motion information including speed and motion direction within a preset time period, the end time of the preset time period being the current time, and the environmental information being fence parameters within the monitoring area where the monitoring object is located; a first calculation module, used to calculate the motion trend degree and environmental trend degree according to the motion information and environmental information respectively; The second calculation module is used to calculate the possibility of fence crossing based on the movement trend degree and the environmental trend degree; the warning sending module is used to determine whether to send warning information about the monitored object to the terminal based on the possibility of fence crossing.

[0004] In one possible implementation, the method for calculating the motion trend degree includes: calculating the moving angle based on the fence parameters and the motion direction at the current moment, the moving angle being the angle between the motion direction and the fence normal vector, and the fence normal vector being the normal vector at the fence point closest to the monitored object; taking the fence extension direction as a reference and combining the moving angle, projecting and decomposing the speeds corresponding to the current moment and the previous moment respectively to obtain the fence speed projected in the fence extension direction, the previous moment being the starting moment of a preset time period; and calculating the motion trend degree based on the moving angle, the fence speed components corresponding to the current moment and the adjacent moments.

[0005] In a possible implementation, the method for calculating the movement trend degree also includes: obtaining historical data of the monitored object, the historical data including the number of historical times of climbing over a fence and the historical height of the fence climbed over; and correcting the movement trend degree based on the historical data to obtain a corrected movement trend degree.

[0006] In a possible implementation, the environmental information also includes the shortest distance from the monitored object to the fence at the current moment, and the height of the fence in the warning area where the monitored object is located at the current moment. The method for calculating the environmental trend degree includes: obtaining an adjacent area based on the motion information analysis, and the adjacent area is the next warning area entered by the monitored object according to the motion direction at the current moment; obtaining the fence height of the adjacent area to obtain the adjacent fence height; analyzing the target frame to obtain the fence damage area, and the target frame is the corresponding frame of the monitoring video at the target moment, and the target moment is the moment when the monitored object first appears; and calculating the environmental trend degree based on the environmental information, the adjacent fence height, and the fence damage area.

[0007] In one possible implementation, the second calculation module includes: a data update module, which is used to continuously monitor the monitored object in the warning area, record the monitoring duration and update the motion trend degree at each moment; and a correction module, which is used to perform a weighted summation of the motion trend degree and the environmental trend degree according to the monitoring duration to obtain the possibility of fence crossing.

[0008] In a possible implementation, the data update module includes: a data secondary acquisition module, used to reacquire the motion information of the monitored object and calculate the second-order motion trend degree; a judgment module, used to judge whether the second-order motion trend degree is the same as the motion trend degree, and if not, update the motion trend degree to the second-order motion trend degree.

[0009] In a possible implementation, the warning sending module includes: a logic module, which is used to generate a warning message when the movement trend is greater than a preset judgment threshold; and a response module, which is used to respond to the generation result of the warning message and send the warning message to the terminal.

[0010] In a possible implementation, the preset judgment threshold is 0.8.

[0011] In a possible implementation, the speed is acquired by an acceleration sensor worn on the monitored object.

[0012] In a possible implementation, the movement direction is acquired by a gyroscope worn on the monitored object.

[0013] The present invention has the following beneficial effects: In the present invention, by combining the environmental dimension and the behavioral dimension of the monitored object itself, the monitored object located in the early warning area is monitored in real time in terms of behavior and position, which can effectively improve the monitoring efficiency of animal husbandry. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 A schematic diagram of the structure of an intelligent fence monitoring and management system for animal husbandry provided by Embodiment 1 of the present invention; Figure 2 Schematic diagram of a flow chart of a method for calculating a motion trend degree in Embodiment 1 of the present invention; Figure 3 This is a flow chart of a method for calculating an environmental trend degree in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the structure of the warning sending module in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the structure of the second computing module in Example 2 of the present invention. DETAILED DESCRIPTION

[0016] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation method, structure, features and effects of an intelligent fence monitoring and management system for animal husbandry proposed by the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0017] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0018] Embodiment 1: The specific scheme of an intelligent fence monitoring and management system for animal husbandry provided by the present invention is described in detail below with reference to the accompanying drawings.

[0019] See also Figure 1 , which shows a schematic diagram of the structure of an intelligent fence monitoring and management system for animal husbandry provided by an embodiment of the present invention. In this embodiment, the intelligent fence monitoring and management system for animal husbandry includes: The video acquisition module is used to acquire a surveillance video corresponding to at least one surveillance area, each of which includes a fence and an early warning area adjacent to the fence.

[0020] It should be noted that in this embodiment, the surveillance video is collected by multiple cameras using 120-degree wide-angle lenses installed in the ranch. These cameras are all waterproof and dustproof to maintain long-term steady-state operation, and the installation height is recommended to be 2.5-3 meters at the top of the fence, which can not only avoid animal collisions, but also obtain the best monitoring angle. At the same time, it can be understood that in this embodiment, each camera corresponds to the collection of surveillance video of a monitoring area, and the coverage of adjacent cameras has an overlapping area of ​​5% to 10%, ensuring that there is no blind spot in the monitoring. After the camera is installed at a position higher than the fence, its 120-degree wide-angle lens can completely cover the monitoring area with a depth of 8-10 meters, so that the shooting area of ​​each camera not only contains a section of the fence, but also can synchronously monitor the range of 3-5 meters inside and outside the fence. Through multi-lens video stitching algorithm and geographic information system (GIS) coordinate mapping, the digital stitching of multiple shooting areas can achieve continuous video coverage of all fence sections.

[0021] Specifically, in this embodiment, the warning area is based on the fence as the boundary body, and the range of the strip area is expanded 0.5 meters into the pasture. Those skilled in the art should understand that in the implementation process, other values ​​can be set according to factors such as the breed difference of the monitored object (such as sheep and cattle), the material of the fence (electrical fence / wooden fence), etc. In this embodiment, it is recommended that 0.3-0.7 meters is more appropriate. Therefore, based on the above division results, in this embodiment, the possibility of the monitored object crossing the fence can be confirmed according to the area where the monitored object is located.

[0022] The area analysis module is used to analyze each monitoring area based on the monitoring video to obtain the monitoring objects appearing in the warning area.

[0023] It should be noted that the monitoring objects mentioned in this embodiment can be animals raised in pastures such as sheep, beef cattle, and horses. In this module, the system identifies the monitoring object in real time from the video through the target recognition algorithm, and its technical implementation relies on the video analysis architecture deployed in the edge computing unit or the video analysis architecture of the central processor located at the administrator's residence. Regarding the target recognition algorithm, those skilled in the art can use the YOLOv5 (YouOnly Look Once version5) or Faster R-CNN (Region-based Convolutional NeuralNetworks) model based on the convolutional neural network to implement it, among which the YOLO series algorithm is more suitable for real-time monitoring scenarios due to its single-stage detection characteristics, while Faster R-CNN has advantages in target positioning accuracy. It should be supplemented that, regarding the model training data, this embodiment recommends that animal image samples under different lighting conditions (morning mist, noon strong light, dusk backlight) should be included, and data enhancement processing (random rotation, noise injection, contrast adjustment) should be performed to improve the robustness of the model. Among them, the method for using YOLO or Faster R-CNN is a prior art and will not be repeated in this embodiment.

[0024] The monitoring data acquisition module is used to obtain the motion information and environmental information of the monitored object, the motion information includes the speed and motion direction within a preset time period, the end time of the preset time period is the current time, and the environmental information is the fence parameters within the monitoring area where the monitored object is located.

[0025] First of all, the speed and direction of movement mentioned in this embodiment can be collected by the three-axis speed sensor and the gyroscope respectively. Specifically, the sensor combination is integrated inside the waterproof electronic collar. The sensor combination is integrated inside the waterproof electronic collar, and the electronic collar is worn by the monitored object and fixed by an anti-disassembly lock. Regarding the way in which the motion information is transmitted to the central processor located at the administrator's residence, a layered transmission architecture is adopted: through the low-power Internet of Things communication module built into the electronic collar, it is transmitted to the receiving module set on the fence post through Bluetooth 5.0, zeebig communication protocol or LoRa communication protocol in 128-bit AES encryption, and the receiving module is then transmitted to the central processor located at the administrator's residence through a fiber optic network or wireless communication for multi-threaded parallel processing.

[0026] Secondly, the duration of the preset time period mentioned in this embodiment is set to be synchronized with the timed wake-up cycle of the IoT communication module, and can be specifically configured as a communication interval of 30 seconds / time. This setting enables the data acquisition timestamp and the transmission cycle to form a time alignment window, which can not only reduce the power consumption of the wireless module, but also ensure the consistency of the timing of motion trajectory analysis calculation.

[0027] In addition, regarding the fence parameters in the environmental information mentioned in the embodiment, the system uses the unique identification code corresponding to each camera to query the fence parameters monitored by the camera in the preset database for acquisition. The preset database adopts a time series database architecture and contains an associated data set of unique identification codes and fence parameter tables (geographic coordinates, height). At the same time, in order to ensure the uniformity of subsequent spatial calculations, in this embodiment, the reference direction of the gyroscope is the same as the reference direction of the geographic coordinates describing the fence, thereby establishing a unified azimuth coordinate system.

[0028] The first calculation module is used to calculate the motion trend degree and the environmental trend degree according to the motion information and the environmental information respectively.

[0029] In this embodiment, the position of the fence relative to the entire warning area is taken into account, and it is located at the boundary part of the area. By analyzing the moving speed and direction of the monitored object, the possibility of its tendency to cross the fence can be confirmed. Specifically, if the moving direction of the monitored object continues to point in the direction of the boundary, then the possibility of the monitored object's tendency to cross the fence will increase. In addition, when the moving direction of the monitored object is highly consistent with the boundary extension direction of the fence, it can be inferred that the monitored object may just be constantly observing the surrounding environment, and the possibility of its tendency to cross the fence is low. At the same time, the height and damage of the fence itself will also affect the monitored object, inducing it to have a tendency to cross the fence. Therefore, in this embodiment, the motion trend degree and environmental trend degree are calculated based on these factors to more accurately measure the behavior of the monitored object.

[0030] The calculation method of motion trend can be found in Figure 2 , the figure shows that the calculation method of the motion trend degree includes steps S1 to S3.

[0031] S1. Calculate a moving angle based on the fence parameters and the current direction of motion, where the moving angle is the angle between the direction of motion and the fence normal vector, and the fence normal vector is the normal vector at the fence point closest to the monitored object.

[0032] S2. Based on the extension direction of the fence and the moving angle, the speeds corresponding to the current moment and the previous moment are projected and decomposed respectively to obtain the fence speed projected in the extension direction of the fence, and the previous moment is the starting moment of the preset time period.

[0033] S3. Calculate the movement trend based on the moving angle, the fence speed components corresponding to the current moment and the adjacent moment.

[0034] Specifically, the calculation function of the movement trend is as follows: ; in, Indicates The movement trend of each monitored object; Represents the absolute value calculation function; Indicates The monitored object is The moving angle at the moment; Indicates The monitored object is The velocity at the moment is projected onto the fence velocity in the fence direction; Indicates The monitored object is The velocity at the moment is projected onto the fence velocity in the fence direction; Indicates The monitored object is The shortest distance to the fence at any moment; represents a non-zero constant. In this embodiment .

[0035] In the above calculation function, The moment represents the previous moment. The moment represents the current moment. Can be obtained by The target recognition algorithm is used to identify the target in the surveillance video at the moment, and the real distance is obtained by calculating the relevant distance of the image; the positioning module can also be set in the electronic collar, and the geographic coordinates of the monitored object can be obtained by the positioning module, and the geographic coordinates of the fence can be calculated. The specific implementation methods are all existing technologies and will not be repeated in this embodiment.

[0036] As described above, when the moving speed of the monitored object increases within a preset time period, and the moving direction of the monitored object is also toward the fence, it indicates that it is more likely to cross the fence. Specifically, the angle between the moving direction of the monitored object and the normal vector of the fence can reflect whether the animal is moving in the direction of the fence. The smaller the angle, the more likely the monitored object is to move directly toward the fence, and the greater the possibility of the monitored object crossing the fence. When the angle is close to 90 degrees (when the animal is walking almost in a straight line along the fence boundary), it may indicate that the monitored object is constantly observing the surrounding environment and is less likely to cross the fence.

[0037] At the same time, in this embodiment, it is also considered that historical behavior and data will affect the possibility of the monitored object crossing the fence. If the monitored object has tried to climb over the fence many times before or has successfully climbed over the fence in the historical records, the possibility of the monitored object crossing the fence will be greater. Therefore, in this embodiment, the movement trend degree is also corrected according to the historical behavior. See step S4 and step S5 for details.

[0038] S4. Obtain historical data of the monitored object, wherein the historical data includes the number of times the monitored object has climbed over a fence and the historical jumping height.

[0039] S5. Correct the movement trend degree based on the historical data to obtain a corrected movement trend degree.

[0040] Specifically, the correction function of the motion trend is as follows: ; in, Indicates The corrected motion trend of each monitored object; Indicates The number of times a monitored object has climbed over the fence in history; Indicates The highest historical jump height of each monitored object can be obtained by analyzing historical monitoring videos; Indicates the current moment The height of the fence within the monitoring area where the monitored object is located; Indicates The motion trend of a monitored object before correction.

[0041] In the above correction function, Indicates The number of times a monitored object has climbed over the fence in history. If the number is greater, it means that the monitored object has a strong desire to climb over the fence, and the possibility of the monitored object climbing over the fence is greater; Indicates The maximum height that the monitored object can jump is greater than the fence height and the greater the difference, the higher the height of the monitored object. The more likely the monitored object is to successfully cross the fence. and For The movement trend of the monitored object crossing the fence is corrected to improve the subsequent The accuracy of calculating the fence crossing probability of each monitored object.

[0042] As mentioned above, in this embodiment, it is also considered that the current condition of the fence will affect the possibility of the monitored object crossing the fence. If there is a potential opportunity to climb over the fence in the current environment, such as a damaged fence, a high fence, etc., the monitored object is more likely to climb over the fence. Therefore, it is necessary to calculate the environmental trend degree based on the height and damage of the fence. For details, please refer to the calculation method of the environmental trend degree. Figure 3 , the figure shows that the calculation method of the environmental trend degree includes steps Step 1 to Step 3.

[0043] Step 1: Obtain an adjacent area based on the motion information analysis, where the adjacent area is the next warning area that the monitored object enters according to the motion direction at the current moment.

[0044] Step 2, obtaining the fence height of the adjacent area to obtain the adjacent fence height.

[0045] Step 3. Analyze the target frame to obtain the fence damage area, where the target frame is the corresponding frame of the monitoring video at the target time, and the target time is the time when the monitored object first appears.

[0046] Among them, the calculation method of the damaged area of ​​the fence can be based on the fence surface image analysis of the target frame, and after the effective area of ​​the fence is extracted by image segmentation technology, the integrity of the fence structure is identified by the edge detection algorithm. Specifically, it includes: performing HSV color space conversion on the target frame of the monitoring video, and segmenting the main area of ​​the fence using the preset fence material color threshold; using the Canny operator to detect the edge contour of the fence, and combining the Hough line transform to identify the fence grid structure; marking the abnormal area with fractures, depressions or missing materials as a damaged area, and finally obtaining the value of the damaged area of ​​the fence by calculating the ratio of the pixel area of ​​the damaged area to the total pixel area of ​​the fence, combined with the pre-calibrated unit pixel corresponding physical size parameters. At the same time, for those skilled in the art, a semantic segmentation model based on deep learning (such as U-Net architecture) can also be used to perform end-to-end damage identification on the fence surface. It is a prior art, and its process will not be repeated in this implementation.

[0047] Step 4. Calculate the environmental trend degree based on the environmental information, the height of the adjacent fence, and the damaged area of ​​the fence.

[0048] Specifically, the calculation function of environmental trend degree is as follows: ; in, Indicates Environmental trend of each monitored object; represents the maximum-minimum normalized function; Indicates The damaged area of ​​the fence in the monitoring area where the monitored object is located; Indicates the total area of ​​the fence in the monitoring area where the monitored object is located; No. The monitored object is The height of the fence in the monitoring area at any given moment; Indicates The monitored object is The height of the adjacent fence corresponding to the moment; Indicates The monitored object is The shortest distance to the fence at any moment; represents a non-zero constant. In this embodiment .

[0049] In the above calculation function, The greater the difference between the two, the The more likely a monitored object is to cross the fence; Indicates the damaged area ratio of the fence. The higher the ratio, the easier it is for the monitored object to cross the fence. The greater the possibility that a monitored object will cross the fence; Indicates The monitored object is The shortest distance from the current moment to the fence. The smaller the distance, the Therefore, the above calculation formula can better describe the impact of environmental factors on the behavior of monitoring objects crossing the fence.

[0050] The second calculation module is used to calculate the possibility of fence crossing based on the movement trend degree and the environmental trend degree.

[0051] Specifically, the calculation function of the fence crossing probability is as follows: ; in, Indicates Fence crossing possibility of each monitored object; represents the maximum-minimum normalized function; Indicates The corrected motion trend of each monitored object; Indicates Environmental trend of each monitored object; represents the first weight coefficient; represents the second weight coefficient. For those skilled in the art, other weight coefficients may also be selected, and this is not specifically limited in this embodiment.

[0052] In the above calculation function, by assigning weights to the possibility of crossing the fence and the environmental trend respectively, and performing weighted summation, we can obtain the evaluation of the possibility of crossing the fence in both the environmental dimension and the behavioral dimension of the monitored object itself, which can better describe whether the monitored object has the possibility of crossing the fence.

[0053] The warning sending module is used to determine whether to send warning information about the monitored object to the terminal based on the possibility of crossing the fence.

[0054] Specifically, see Figure 4 , the figure shows that this module also includes a logic module and a response module.

[0055] In this embodiment, the logic module is used to generate warning information when the movement trend degree is greater than a preset judgment threshold.

[0056] The preset judgment threshold is 0.8. For those skilled in the art, the preset judgment threshold may also be selected from other values ​​according to actual conditions, and this embodiment does not make any specific limitation thereto.

[0057] It should also be noted that in this embodiment, there is a preferred graded warning mechanism. When the motion trend degree is in the range of 0.8-0.9, a yellow warning is triggered, and a real-time monitoring video containing a timestamp, the geographic coordinates of the dangerous monitoring object, and the monitoring area where the dangerous monitoring object is located is generated. Among them, the dangerous monitoring object refers to the monitoring object when the motion trend degree is greater than the preset judgment threshold, and the geographic coordinates are collected by the positioning module set in the electronic collar. When the motion trend degree is in the range of 0.89-1.0, a red warning is triggered, and a real-time monitoring video containing a timestamp, the geographic coordinates of the dangerous monitoring object, and the monitoring area where the dangerous monitoring object is located is also generated. However, at the same time, the sound and light alarm device must be activated at the fence, and a three-level gradient alarm mode is formed by combining high-frequency buzzing and red light flashing to drive the monitored object away from the fence.

[0058] The response module is used to respond to the generation result of the warning information and send the warning information to the terminal.

[0059] In this embodiment, by combining the environmental dimension and the behavior dimension of the monitored object itself, the behavior and position of the monitored object are monitored in real time, which can effectively improve the monitoring efficiency of animal husbandry, reduce the frequency of manual inspections, reduce false alarms and human errors, and improve management efficiency.

[0060] Embodiment 2: This embodiment provides a smart fence monitoring and management system for animal husbandry. The difference between this embodiment and embodiment 1 is that the execution content of the second computing module is different. For details, see Figure 5 , the figure shows that the second computing module includes: The data updating module is used to continuously monitor the monitored object in the warning area, record the monitoring duration and update the movement trend at each moment.

[0061] The correction module is used to perform weighted summation of the movement trend degree and the environment trend degree according to the monitoring duration to obtain the possibility of fence crossing.

[0062] Specifically, the calculation function of the fence crossing probability is as follows: ; in, Indicates Fence crossing possibility of each monitored object; Represents the Sigmoid function; Indicates The corrected motion trend of each monitored object; Indicates Environmental trend of each monitored object; Indicates the monitoring duration in seconds.

[0063] In this embodiment, considering that if the current The longer the monitored object stays in the warning area, the smaller the impact of environmental factors is compared with the movement behavior, that is, the longer the stay time, the The greater the weight ratio of the movement trend of the monitored object, the smaller the weight ratio of the environmental trend; at the same time, if the stay time is too long, it should not increase infinitely, so the Sigmoid function is used to limit the positive infinite increase caused by the stay time. The shorter the time a monitored object stays in the warning area, the more serious the fence damage. The greater the probability that an animal will cross the fence, the greater the probability that an animal will cross the fence. Therefore, the above calculation function can better describe the above situation.

[0064] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0065] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An intelligent fence monitoring and management system for animal husbandry, characterized in that: include: A video acquisition module, used to acquire a surveillance video corresponding to at least one surveillance area, each of which includes a fence and an early warning area adjacent to the fence; A regional analysis module, used to analyze each of the monitoring areas based on the monitoring video to obtain monitoring objects appearing in the warning area; A monitoring data acquisition module, used to acquire motion information and environmental information of the monitored object, wherein the motion information includes speed and motion direction within a preset time period, the end time of the preset time period is the current time, and the environmental information is fence parameters within the monitoring area where the monitored object is located; A first calculation module, configured to calculate a moving angle according to the fence parameters and the moving direction at the current moment, and calculate a moving trend degree based on the moving angle and the fence speed components corresponding to the current moment and the adjacent moment; Analyzing the movement area trend based on the movement information, and obtaining the environmental trend degree in combination with the fence parameters; A second calculation module, configured to calculate the fence crossing possibility based on the movement trend degree and the environment trend degree; The warning sending module is used to determine whether to send warning information about the monitored object to the terminal based on the possibility of crossing the fence.

2. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The calculation method of motion trend degree includes: The movement angle is the angle between the movement direction and the fence normal vector, and the fence normal vector is the normal vector at the fence point closest to the monitored object; Based on the extension direction of the fence and the moving angle, the speeds corresponding to the current moment and the previous moment are projected and decomposed respectively to obtain the fence speed projected in the extension direction of the fence, and the previous moment is the starting moment of the preset time period; The motion trend degree is calculated based on the moving angle, the fence speed components corresponding to the current moment and the adjacent moment.

3. The intelligent fence monitoring and management system for animal husbandry according to claim 2 is characterized in that: The calculation method of the movement trend degree also includes: Acquire historical data of the monitored object, the historical data including the number of times the fence was climbed over and the height of the fence climbed over; The movement trend degree is corrected based on the historical data to obtain a corrected movement trend degree.

4. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The environmental information also includes the shortest distance from the monitored object to the fence at the current moment, and the height of the fence in the warning area where the monitored object is located at the current moment. The calculation method of the environmental trend degree includes: Obtaining an adjacent area based on the motion information analysis, where the adjacent area is the next warning area that the monitored object enters according to the motion direction at the current moment; Obtaining the fence height of the adjacent area to obtain the adjacent fence height; Analyze the target frame to obtain the fence damage area, wherein the target frame is the frame corresponding to the monitoring video at the target time, and the target time is the time when the monitoring object first appears; The environmental trend degree is calculated based on the environmental information, the height of the adjacent fence, and the damaged area of ​​the fence.

5. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The second calculation module includes: A data updating module, used to continuously monitor the monitored object in the warning area, record the monitoring duration and update the movement trend at each moment; The correction module is used to perform weighted summation of the movement trend degree and the environment trend degree according to the monitoring duration to obtain the possibility of fence crossing.

6. The intelligent fence monitoring and management system for animal husbandry according to claim 5, characterized in that: The data updating module comprises: A data secondary acquisition module is used to reacquire the motion information of the monitored object and calculate the second-order motion trend degree; The judgment module is used to judge whether the second-order motion trend degree is the same as the motion trend degree. If not, the motion trend degree is updated to the second-order motion trend degree.

7. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The early warning sending module includes: A logic module, used for generating warning information when the movement trend degree is greater than a preset judgment threshold; The response module is used to respond to the generation result of the warning information and send the warning information to the terminal.

8. The intelligent fence monitoring and management system for animal husbandry according to claim 6, characterized in that: The preset judgment threshold is 0.

8.

9. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The speed is acquired by an acceleration sensor worn on the monitored object.

10. The intelligent fence monitoring and management system for animal husbandry according to claim 1, characterized in that: The movement direction is acquired by a gyroscope worn on the monitored object.

Images