A dairy cow breeding health data management system based on intelligent ecological deployment

Abstract

The application discloses a kind of dairy cow breeding health data management systems based on intelligent ecological deployment, including dairy cow identity feature and binding module, multi-dimensional sign state data acquisition association module, sign state multi-interval trend analysis module, dairy cow normal activity behavior quantitative analysis module, expert database health state evaluation module and intelligent data monitoring platform based on, dairy cow physical identification and biological mottling feature are fused in advance, double binding and checking are carried out to dairy cow characteristics, health examination corridor is set, complete body type sign data of dairy cow are automatically synchronized and obtained, and daily activity monitoring is combined, the time series sign data collected is analyzed in combination with different growth stages of dairy cow Period, by establishing individualized dynamic behavior baseline for each dairy cow, the relative deviation degree of its daily activity is evaluated, expert knowledge base is established, individual sign data obtained by real-time monitoring is compared and evaluated quantitatively with standard data of corresponding stage.

Description

Technical Field

[0001] This invention relates to the field of dairy farming, specifically a dairy farming health data management system based on intelligent ecological deployment. Background Technology

[0002] With the development of large-scale, intensive dairy farming, precise and efficient management of individual dairy cow health has become a core requirement for improving farming efficiency and animal welfare. Traditional dairy cow health management relies heavily on the daily observation and experience of farmers, resulting in low efficiency, strong subjectivity, and difficulty in achieving early warning and individualized intervention. Furthermore, body size data is crucial for assessing the growth and development of heifers, the body structure of adult cattle, and nutritional levels. Currently, this mainly relies on regular manual measurements, which are time-consuming, labor-intensive, and produce discrete data, failing to generate continuous dynamic growth curves. It also fails to deeply analyze continuous time-series data of key indicators such as weight and body size, making it impossible to calculate rates of change, identify growth plateaus or abnormal decline patterns, leading to missed opportunities for optimal intervention. Typically, population averages or fixed thresholds are used as benchmarks, ignoring behavioral differences between individuals. The inability to establish individualized dynamic behavioral baselines results in insufficient sensitivity to subtle behavioral changes, hindering early disease warnings. This application aims to integrate physical identifiers and biological markings of dairy cows in advance, doubly binding and verifying cow characteristics to ensure that the health data of each cow can be uniquely, continuously, accurately associated, and traced. It also establishes a health check corridor to automatically and synchronously acquire complete body shape and vital sign data of dairy cows, and combines this with daily activity monitoring to construct a complete individual physiological and behavioral profile covering both static body shape and dynamic behavior. The collected time-series vital sign data is analyzed in conjunction with different growth stages of the dairy cows to analyze their rate of change, trends, and abnormal patterns. By establishing an individualized dynamic behavior baseline for each dairy cow, the relative deviation of its daily activity level is assessed. Simultaneously, an expert knowledge base containing standard vital sign data for different growth stages is established to quantitatively compare and evaluate the individual vital sign data obtained from real-time monitoring with the standard data for the corresponding stage, achieving automated and intelligent monitoring and early warning of the dairy cow population's health. Summary of the Invention

[0003] The purpose of this invention is to provide a dairy cow breeding health data management system based on intelligent ecological deployment to solve the problems in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A dairy cow health data management system based on intelligent ecological deployment includes a dairy cow identity feature and binding module, a multi-dimensional vital sign data collection and association module, a vital sign multi-interval trend analysis module, a dairy cow routine activity behavior quantitative analysis module, an expert database-based health status assessment module, and an intelligent data monitoring platform. The cow identification and binding module assigns a corresponding smart collar to each cow. The smart collar has a unique electronic code built in and collects the cow's markings. The cow's markings, electronic code and basic cow information are combined to generate a cow identity ID. The multi-dimensional vital signs data acquisition and association module uses several monitoring sensors to intelligently select key three-dimensional coordinate points of different dairy cows' torsos in a specific monitoring corridor, and collects the vital signs data of dairy cows in real time. The vital signs data include the torso size, weight and daily activity level of dairy cows. The multi-interval trend analysis module for vital signs will collect and preprocess the body size of dairy cows, set multiple time windows to identify the changing trends of body size data of different dairy cows, associate it with the growth stage of dairy cows, and perform annotation analysis and early warning on body size data of different dairy cows at multiple stages. The quantitative analysis module for routine dairy cow activity behavior collects daily activity data from dairy cows across multiple time windows, analyzes the dynamic activity baseline of routine dairy cow activity behavior, and analyzes the deviation between the daily activity level of different dairy cows and the dynamic activity baseline under continuous time windows to provide early warning of abnormal dairy cow behavior. The health status assessment module based on the expert database compares the real-time physical signs of dairy cows with the standard physical signs data of the corresponding growth stages in the expert database, calculates the conformity of the standard physical signs of dairy cows, and assesses the health status of dairy cows.

[0005] Further configuration: The cow identity feature and binding module includes a cow effective feature detection and acquisition submodule and a multi-feature fusion identity ID generation submodule. The cow effective feature detection and acquisition submodule includes a smart collar allocation unit and a stripe feature acquisition unit. The smart collar allocation unit obtains the total number of captive cows and manually assigns smart collars to the captive cows. Each smart collar has a unique electronic code, and the electronic code data of each cow is acquired. The stripe feature acquisition unit includes several cameras. While each cow is manually fitted with a smart collar, the cameras capture images of the left, right, and rear sides of the cow's torso. A set of images of cow torsos was collected. The collected images were preprocessed to remove images with resolution below a set threshold. The preprocessed images of each cow's torso were associated with the corresponding cow's electronic coding data. The cow stripe contours within each torso image were located and marked. Cow torso images with marked stripe contours were defined as valid feature images of the cow. The valid feature image sets of each cow were summarized. The average ratio of the marked stripe contour pixels to the total pixels within each cow's valid feature image set was analyzed and defined as the comprehensive feature value of each cow's stripe pattern. The multi-feature fusion identity ID generation submodule obtains the basic information of each cow, which is manually entered. The basic information includes the cow's date of birth, breed, and sex. The electronic code and stripe feature values ​​of each cow are combined and input into a hash function to generate a fixed hash value. The hash value of each cow is defined as its identity ID. The basic information of each cow is bound to its identity ID, electronic code, and set of valid feature images. The bound information is then summarized and uploaded.

[0006] Further configuration: The multi-dimensional vital sign data acquisition and association module includes an integrated detection corridor deployment submodule, a dairy cow key body shape parameter intelligent selection submodule, and a dairy cow activity data synchronous acquisition submodule. The integrated detection corridor deployment submodule includes a health corridor positioning unit and a sensor integrated deployment unit. The health corridor positioning unit obtains the location of the manually set health check corridor and shares the location with several administrators within the intelligent data monitoring platform. The sensor integrated deployment unit includes several visual sensors and weighing sensors, which are deployed at different locations in the manually set health check corridor.

[0007] Further settings: The intelligent selection submodule for key body shape parameters of dairy cows uses several visual sensors to collect three-dimensional data of dairy cows standing in the health check corridor. The three-dimensional data includes multi-coordinate data of the back contour and multi-coordinate data of the side contour. The module analyzes the different physical characteristics of the dairy cows, including height, width, length and weight. Extract any coordinate point on the plane of the cow's feet and any coordinate point on the top of its back. Set any coordinate point on the plane of the cow's feet as... Any coordinate point at the top of the back is ,make sure and The projections on the horizontal plane lie on the same vertical line, that is, aligned on the same horizontal line, and the calculation is performed. and The vertical distance between them is used as the current real-time height of the cow. ; Take any coordinate point at the very front of the cow's head and any coordinate point at the base of its tail. Let the coordinate point at the very front of the cow's head be... any coordinate point of the tail root ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time length of the current cow. ; Choose any center coordinates of the outermost edges of the left and right sides of the cow's back outline, with the selected left and right coordinates located at the center of the cow's back outline. Set the center coordinates of the outermost edge of the left side of the cow. The center coordinates of the outermost right edge of the cow ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time width of the current cow. ; The weight of each cow is monitored by weighing sensors as it passes through each health check corridor, and the weight data for each cow is obtained. ; The dairy cow activity data synchronization and collection submodule includes a step counting sensor, which is set inside the smart collar. The step counting sensor detects the daily activity of each dairy cow and obtains the height, width, length, weight and daily activity of each dairy cow. This data is then linked to the cow's identity ID and detection timestamp and uploaded and stored as a health data file for each dairy cow.

[0008] Further settings: The multi-interval trend analysis module for bovine vital signs includes a sub-module for monitoring and analyzing changes in bovine vital signs and a sub-module for labeling and associating multi-stage vital signs data. The bovine vital signs monitoring and analysis sub-module acquires the health data file of each bovine, sorts the bovine test data in chronological order, preprocesses the test data, compares the test data pairwise in sequence, compares the deviation between two consecutive test values, and if the deviation is greater than a set threshold, an anomaly is marked on the data tested in the two time windows and sent to a human for data review. The system acquires the height, width, length, and weight data of each cow within a continuous time window, continuously monitors whether the height, width, length, and weight data of the cows increase within the continuous time window, identifies situations where the height, width, length, and weight data of the cows do not increase or decrease within a continuous set number of time windows, marks the time window with such a situation as an early warning, and extracts the cow's vital signs data within that time window and sends it to a human for data review. The multi-stage vital sign data annotation and association submodule extracts height, width, length, and weight data of different dairy cows under different time windows, screens the current growth cycle of the dairy cow corresponding to each time window, associates and annotates the growth cycle stage of the dairy cow for the time window of each dairy cow, and summarizes the vital sign data of each dairy cow in continuous time windows under different growth cycles.

[0009] Further configuration: The quantitative analysis module for dairy cow routine activity behavior includes a multi-time-window dynamic update sub-module for dairy cow routine activity and an activity level classification report and early warning sub-module. The multi-time-window dynamic update sub-module for dairy cow routine activity acquires the daily activity of each dairy cow. At the same time, based on each defined dairy cow growth cycle stage, it sets short-term and long-term time windows for each growth cycle. The time intervals of the short-term and long-term time windows are set manually. It calculates the clipped average of dairy cow daily activity within the short-term and long-term time windows respectively. It extracts the dairy cow daily activity within different short-term and long-term time windows within the same growth cycle. It calculates the clipped average of dairy cow daily activity within several short-term time windows and the clipped average of dairy cow daily activity within several long-term time windows respectively.

[0010] Further settings: The activity level grading report early warning submodule sets the cut-off average daily activity level of a dairy cow within different short-term time windows during the same growth cycle. The average daily activity level of a dairy cow within different long-term time windows during the same growth cycle is: The average activity levels of the trimming process were compared and analyzed within several short-term and long-term time windows, based on the formula: in The maximum difference threshold between the average activity levels within several artificially set short-term and long-term time windows is used. When the average activity levels within these short-term and long-term time windows satisfy the above formula, it is determined that the activity levels of dairy cows within these short-term and long-term time windows are highly stable within the same growth cycle stage. The normal baseline for daily activity levels of dairy cows during this growth cycle stage is then calculated. According to the formula: The normal baseline for daily activity levels at the current growth stage of dairy cows is calculated. The current daily activity level of dairy cows is then compared with the normal baseline, and the daily activity level of dairy cows is set as follows: The daily activity level of dairy cows on a certain date To determine if a cow's behavior is highly active on a given day, the daily activity level of the cows on a given day is... To determine if a cow's behavior is low activity on a given day, the daily activity level of the cows on a given day is... To determine whether the cows' behavior on that day was at a normal level of activity, the daily behavioral activity of cows at different growth stages was analyzed. If the average activity level over several short-term and long-term time windows does not meet the above formula, it is determined that the daily activity level of dairy cows in the current growth cycle is disordered. The current dairy cow's identity ID is marked, and the dairy cow's identity ID and the daily activity level of dairy cows in the current growth cycle are sent to a human for early warning review.

[0011] Further configuration: The health status assessment module based on the expert database includes a multi-growth stage real-time vital sign status comparison and assessment submodule and a dairy cow health status marking submodule. The multi-growth stage real-time vital sign status comparison and assessment submodule pre-builds an expert database that stores standard vital sign data of dairy cows at different growth stages. It compares the real-time detected vital sign data of each dairy cow at different growth stages with the standard vital sign data of dairy cows at different growth stages in the expert database. It compares the deviation values ​​between the real-time detected vital sign data and the standard vital sign data of dairy cows at the same growth cycle stage. The data of dairy cows with deviation values ​​greater than a set threshold are sent to the dairy cow health status marking submodule. The dairy cow health status marking submodule marks the dairy cow with abnormal health status for the current comparison, extracts the real-time detection vital sign data value of the dairy cow in the current growth stage cycle, and sends the marked dairy cow's identity ID and the extracted real-time detection vital sign data value to the intelligent data monitoring platform for manual review.

[0012] Compared with existing technologies, the beneficial effects of this invention are: it aims to integrate the physical identification and biological markings of dairy cows in advance, and to double-bind and verify the characteristics of dairy cows, ensuring that the health data of each dairy cow can be uniquely, continuously, accurately associated and traced. At the same time, a health check corridor is set up to automatically and synchronously acquire the complete body shape and vital signs data of dairy cows, and combine it with daily activity monitoring to construct a complete individual physiological and behavioral profile covering the static body shape and dynamic behavior of dairy cows. The collected time-series vital signs data are analyzed in conjunction with the different growth stages of dairy cows to analyze their rate of change, trends and abnormal patterns. By establishing an individualized dynamic behavior baseline for each dairy cow, the relative deviation of its daily activity level is assessed. At the same time, an expert knowledge base containing standard vital signs data of different growth stages is established, and the individual vital signs data obtained from real-time monitoring are quantitatively compared and evaluated with the standard data of the corresponding stage, so as to realize automated and intelligent monitoring and early warning of the health of the dairy cow herd. Attached Figure Description

[0013] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0014] Figure 1 This is a flowchart illustrating a dairy farming health data management system based on intelligent ecological deployment according to the present invention. Figure 2 This invention relates to a module of a dairy cow health data management system based on intelligent ecological deployment. Figure 1 ; Figure 3 This invention relates to a module of a dairy cow health data management system based on intelligent ecological deployment. Figure 2 ; Figure 4 This invention relates to a module of a dairy cow health data management system based on intelligent ecological deployment. Figure 3 ; Figure 5 This invention relates to a module of a dairy cow health data management system based on intelligent ecological deployment. Figure 4 ; Figure 6 This invention relates to a module of a dairy cow health data management system based on intelligent ecological deployment. Figure 5 . Detailed Implementation

[0015] 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.

[0016] Please see Figures 1-6 In this embodiment of the invention, a dairy cow health data management system based on intelligent ecological deployment is provided. The system includes a dairy cow identity feature and binding module, a multi-dimensional vital sign data collection and association module, a vital sign multi-interval trend analysis module, a dairy cow routine activity behavior quantitative analysis module, an expert database-based health status assessment module, and an intelligent data monitoring platform. Among them, the dairy cow breeding health data management system includes a dairy cow identity feature and binding module, a multi-dimensional vital sign data collection and association module, a vital sign multi-interval trend analysis module, a dairy cow routine activity behavior quantitative analysis module, and an expert database-based health status assessment module, all of which are connected to the intelligent data monitoring platform.

[0017] The cow identification and binding module assigns a corresponding smart collar to each cow. The smart collar has a unique electronic code built in and collects the cow's markings. The cow's markings, electronic code and basic cow information are combined to generate a cow identity ID. like Figure 2 It should also be specifically explained that the cow identification and binding module includes a cow effective feature detection and collection submodule and a multi-feature fusion identity ID generation submodule. The cow effective feature detection and collection submodule includes a smart collar allocation unit and a stripe feature collection unit. The smart collar allocation unit obtains the total number of captive cows and manually assigns smart collars to the captive cows. Each smart collar has a unique electronic code, and the electronic code data of each cow is obtained. The stripe feature collection unit includes several cameras. While manually attaching the smart collar to each cow, the cameras capture images of the left and right sides of the cow's torso. A set of images of cow torsos was collected from the rear. The collected images were preprocessed to remove images with a resolution below a set threshold. Each cow's preprocessed torso image set was associated with its corresponding electronic coding data. The cow stripe contours within each torso image were located and marked. Cow torso images with stripe contour location marks were defined as valid feature images of the cow. The valid feature image sets of each cow were summarized. The average ratio of the stripe contour pixels with location marks to the total pixels within each cow's valid feature image set was analyzed and defined as the comprehensive value of each cow's stripe feature. The multi-feature fusion identity ID generation submodule obtains the basic information of each cow, which is manually entered. The basic information includes the cow's date of birth, breed, and sex. The electronic code and stripe feature values ​​of each cow are combined and input into a hash function to generate a fixed hash value. The hash value of each cow is defined as its identity ID. The basic information of each cow is bound to its identity ID, electronic code, and set of valid feature images. The bound information is then summarized and uploaded.

[0018] The multi-dimensional vital signs data acquisition and association module uses several monitoring sensors to intelligently select key three-dimensional coordinate points of different dairy cows' torsos in a specific monitoring corridor, and collects the vital signs data of dairy cows in real time. The vital signs data include the torso size, weight and daily activity level of dairy cows. like Figure 3 It should also be noted that the multi-dimensional vital sign data acquisition and association module includes an integrated detection corridor deployment sub-module, a dairy cow key body shape parameter intelligent selection sub-module, and a dairy cow activity data synchronous acquisition sub-module. The integrated detection corridor deployment sub-module includes a health corridor positioning unit and a sensor integration deployment unit. The health corridor positioning unit obtains the location of the manually set health check corridor and shares the location with several administrators within the intelligent data monitoring platform. The sensor integration deployment unit includes several visual sensors and weighing sensors, which are deployed at different locations in the manually set health check corridor.

[0019] To further explain, the intelligent selection submodule for key body shape parameters of dairy cows uses several visual sensors to collect three-dimensional data of dairy cows standing in the health check corridor. The three-dimensional data includes multi-coordinate data of the back contour and multi-coordinate data of the side contour. The module analyzes the different physical characteristics of the dairy cows, including height, width, length and weight. Extract any coordinate point on the plane of the cow's feet and any coordinate point on the top of its back. Set any coordinate point on the plane of the cow's feet as... Any coordinate point at the top of the back is ,make sure and The projections on the horizontal plane lie on the same vertical line, that is, aligned on the same horizontal line, and the calculation is performed. and The vertical distance between them is used as the current real-time height of the cow. ; Take any coordinate point at the very front of the cow's head and any coordinate point at the base of its tail. Let the coordinate point at the very front of the cow's head be... any coordinate point of the tail root ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time length of the current cow. ; Choose any center coordinates of the outermost edges of the left and right sides of the cow's back outline, with the selected left and right coordinates located at the center of the cow's back outline. Set the center coordinates of the outermost edge of the left side of the cow. The center coordinates of the outermost right edge of the cow ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time width of the current cow. ; The weight of each cow is monitored by weighing sensors as it passes through each health check corridor, and the weight data for each cow is obtained. ; The dairy cow activity data synchronization and collection submodule includes a step counting sensor, which is set inside the smart collar. The step counting sensor detects the daily activity of each dairy cow and obtains the height, width, length, weight and daily activity of each dairy cow. This data is then linked to the cow's identity ID and detection timestamp and uploaded and stored as a health data file for each dairy cow.

[0020] The multi-interval trend analysis module for vital signs will collect and preprocess the body size of dairy cows, set multiple time windows to identify the changing trends of body size data of different dairy cows, associate it with the growth stage of dairy cows, and perform annotation analysis and early warning on body size data of different dairy cows at multiple stages. like Figure 4 It should also be noted that the multi-interval trend analysis module for bovine vital signs includes a sub-module for monitoring and analyzing changes in bovine vital signs and a sub-module for labeling and associating multi-stage vital signs data. The bovine vital signs monitoring and analysis sub-module acquires the health data file of each bovine, sorts the bovine test data in chronological order, preprocesses the test data, compares the test data pairwise in sequence, compares the deviation between two consecutive test values, and if the deviation is greater than a set threshold, the data tested in the two time windows are marked as abnormal and sent to a human for data review. The system acquires the height, width, length, and weight data of each cow within a continuous time window, continuously monitors whether the height, width, length, and weight data of the cows increase within the continuous time window, identifies situations where the height, width, length, and weight data of the cows do not increase or decrease within a continuous set number of time windows, marks the time window with such a situation as an early warning, and extracts the cow's vital signs data within that time window and sends it to a human for data review. The multi-stage vital sign data annotation and association submodule extracts height, width, length, and weight data of different dairy cows under different time windows, screens the current growth cycle of the dairy cow corresponding to each time window, associates and annotates the growth cycle stage of the dairy cow for the time window of each dairy cow, and summarizes the vital sign data of each dairy cow in continuous time windows under different growth cycles.

[0021] The quantitative analysis module for routine dairy cow activity behavior collects daily activity data from dairy cows across multiple time windows, analyzes the dynamic activity baseline of routine dairy cow activity behavior, and analyzes the deviation between the daily activity level of different dairy cows and the dynamic activity baseline under continuous time windows to provide early warning of abnormal dairy cow behavior. like Figure 5 Furthermore, it should be noted that the quantitative analysis module for routine dairy cow activity includes a multi-time-window dynamic update sub-module for routine dairy cow activity and an activity level grading report and early warning sub-module. The multi-time-window dynamic update sub-module for routine dairy cow activity acquires the daily activity of each dairy cow. At the same time, based on each defined dairy cow growth cycle stage, it sets short-term and long-term time windows for each growth cycle. The time intervals of the short-term and long-term time windows are set manually. The module calculates the clipped average of the daily activity of dairy cows within the short-term and long-term time windows respectively. It also extracts the daily activity of dairy cows within different short-term and long-term time windows within the same growth cycle, and calculates the clipped average of the daily activity of dairy cows within several short-term time windows and several long-term time windows respectively.

[0022] To further clarify, the activity level grading report early warning submodule sets the cut-off average daily activity level of a dairy cow within different short-term time windows during the same growth cycle as [value missing]. The average daily activity level of a dairy cow within different long-term time windows during the same growth cycle is: The average activity levels of the trimming process were compared and analyzed within several short-term and long-term time windows, based on the formula: in The maximum difference threshold between the average activity levels within several artificially set short-term and long-term time windows is used. When the average activity levels within these short-term and long-term time windows satisfy the above formula, it is determined that the activity levels of dairy cows within these short-term and long-term time windows are highly stable within the same growth cycle stage. The normal baseline for daily activity levels of dairy cows during this growth cycle stage is then calculated. According to the formula: The normal baseline for daily activity levels at the current growth stage of dairy cows is calculated. The current daily activity level of dairy cows is then compared with the normal baseline, and the daily activity level of dairy cows is set as follows: The daily activity level of dairy cows on a certain date To determine if a cow's behavior is highly active on a given day, the daily activity level of the cows on a given day is... To determine if a cow's behavior is low activity on a given day, the daily activity level of the cows on a given day is... To determine whether the cows' behavior on that day was at a normal level of activity, the daily behavioral activity of cows at different growth stages was analyzed. If the average activity level over several short-term and long-term time windows does not meet the above formula, it is determined that the daily activity level of dairy cows in the current growth cycle is disordered. The current dairy cow's identity ID is marked, and the dairy cow's identity ID and the daily activity level of dairy cows in the current growth cycle are sent to a human for early warning review.

[0023] The health status assessment module based on the expert database compares the real-time physical signs of dairy cows with the standard physical signs data of the corresponding growth stages in the expert database, calculates the conformity of the standard physical signs of dairy cows, and assesses the health status of dairy cows.

[0024] like Figure 6 As shown, the health status assessment module based on the expert database includes a multi-growth stage real-time vital sign comparison and assessment submodule and a dairy cow health status marking submodule. The multi-growth stage real-time vital sign comparison and assessment submodule pre-builds an expert database that stores standard vital sign data of dairy cows at different growth stages. It compares the real-time vital sign data of each dairy cow at different growth stages with the standard vital sign data of dairy cows at different growth stages in the expert database. It compares the deviation values ​​between the real-time detected vital sign data and the standard vital sign data of dairy cows at the same growth cycle stage. The data of dairy cows with deviation values ​​greater than a set threshold are sent to the dairy cow health status marking submodule. The dairy cow health status marking submodule marks the dairy cow with abnormal health status for the current comparison, extracts the real-time detection vital sign data value of the dairy cow in the current growth stage cycle, and sends the marked dairy cow's identity ID and the extracted real-time detection vital sign data value to the intelligent data monitoring platform for manual review.

[0025] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A dairy cow farming health data management system based on intelligent ecological deployment, characterized in that: The system includes a dairy cow identity feature and binding module, a multi-dimensional vital sign data collection and association module, a vital sign multi-interval trend analysis module, a dairy cow routine activity behavior quantitative analysis module, an expert database-based health status assessment module, and an intelligent data monitoring platform. The cow identification feature and binding module assigns a corresponding smart collar to each cow. The smart collar has a unique electronic code built in. At the same time, it collects the cow's stripe features and integrates the cow's stripe features, electronic code and basic cow information to generate a cow identity ID. The multi-dimensional vital signs data acquisition and association module uses several monitoring sensors to intelligently select key three-dimensional coordinate points of different dairy cow trunks in a specific monitoring corridor, and collects the vital signs data of dairy cows in real time. The vital signs data include the cow's trunk size, weight and daily activity level. The multi-interval trend analysis module for vital signs will collect and preprocess the body size of dairy cows, set multiple time windows to identify the changing trends of body size data of different dairy cows, associate it with the growth stage of dairy cows, and perform annotation analysis and early warning on body size data of different dairy cows at multiple stages. The quantitative analysis module for routine dairy cow activity behavior collects daily activity data from dairy cows across multiple time windows, analyzes the dynamic activity baseline of routine dairy cow activity behavior, analyzes the deviation between the daily activity level of different dairy cows and the dynamic activity baseline under continuous time windows, and provides early warnings for abnormal dairy cow behavior. The health status assessment module based on the expert database compares the real-time physical signs of dairy cows with the standard physical signs data of the corresponding growth stages in the expert database, calculates the conformity of the standard physical signs of dairy cows, and assesses the health status of dairy cows.

2. The dairy cow breeding health data management system based on intelligent ecological deployment according to claim 1, characterized in that... The dairy cow identity feature and binding module includes a dairy cow effective feature detection and acquisition submodule and a multi-feature fusion identity ID generation submodule. The dairy cow effective feature detection and acquisition submodule includes a smart collar allocation unit and a stripe feature acquisition unit. The smart collar allocation unit obtains the total number of dairy cows in captivity and manually assigns smart collars to the captivity dairy cows. Each smart collar has a unique electronic code built in, and the electronic code data of each dairy cow is obtained. The stripe feature acquisition unit includes several cameras. While each dairy cow is manually wearing a smart collar, the cameras capture images of the left, right, and rear sides of the dairy cow's torso. A set of images of cow torsos was collected. The collected images were preprocessed, and images with resolution below a set threshold were removed. The preprocessed images of each cow torso were associated with the corresponding cow's electronic coding data. The cow stripe contours within each cow torso image were located and marked. Cow torso images with stripe contour location marks were defined as effective feature images of the cow. The effective feature image sets of each cow were summarized, and the average ratio of the marked stripe contour pixels to the total pixels within each cow's effective feature image set was analyzed and defined as the comprehensive value of each cow's stripe features. The multi-feature fusion identity ID generation submodule obtains the basic information of each cow, which is manually entered. The basic information includes the cow's date of birth, breed, and sex. The electronic code and stripe feature values ​​of each cow are combined and input into a hash function to generate a fixed hash value. The hash value of each cow is defined as its identity ID. The basic information of each cow is bound to its identity ID, electronic code, and set of valid feature images. The bound information is then summarized and uploaded.

3. The dairy cow health data management system based on intelligent ecological deployment according to claim 1, characterized in that... The multi-dimensional vital sign data acquisition and association module includes an integrated detection corridor deployment submodule, a dairy cow key body shape parameter intelligent selection submodule, and a dairy cow activity data synchronous acquisition submodule. The integrated detection corridor deployment submodule includes a health corridor positioning unit and a sensor integrated deployment unit. The health corridor positioning unit obtains the location of the manually set health check corridor and shares the location with several administrators within the intelligent data monitoring platform. The sensor integrated deployment unit includes several visual sensors and weighing sensors, which are deployed at different locations in the manually set health check corridor.

4. A dairy cow breeding health data management system based on intelligent ecological deployment according to claim 3, characterized in that... The intelligent selection submodule for key body shape parameters of dairy cows uses several visual sensors to collect three-dimensional data of dairy cows standing in the health check corridor. The three-dimensional data includes multi-coordinate data of the back contour and multi-coordinate data of the side contour. The module analyzes the different physical characteristics of the dairy cows, including the height, width, length and weight of the dairy cows. Extract any coordinate point on the plane of the cow's feet and any coordinate point on the top of its back. Set any coordinate point on the plane of the cow's feet as... Any coordinate point at the top of the back is ,make sure and The projections on the horizontal plane lie on the same vertical line, that is, aligned on the same horizontal line, and the calculation is performed. and The vertical distance between them is used as the current real-time height of the cow. ; Take any coordinate point at the very front of the cow's head and any coordinate point at the base of its tail. Let the coordinate point at the very front of the cow's head be... any coordinate point of the tail root ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time length of the current cow. ; Choose any center coordinates of the outermost edges of the left and right sides of the cow's back outline, with the selected left and right coordinates located at the center of the cow's back outline. Set the center coordinates of the outermost edge of the left side of the cow. The center coordinates of the outermost right edge of the cow ,make sure and The projections on the horizontal plane lie on the same horizontal line, that is, aligned on the same horizontal line, and the calculation is performed. and The straight-line distance between them is used as the real-time width of the current cow. ; The weight of each cow is monitored by weighing sensors as it passes through each health check corridor, and the weight data for each cow is obtained. ; The dairy cow activity data synchronization and collection submodule includes a step counting sensor, which is set inside the smart collar. The step counting sensor detects the daily activity of each dairy cow and obtains the height, width, length, weight and daily activity of each dairy cow. This data is then linked to the cow's identity ID and detection timestamp and uploaded and stored as a health data file for each dairy cow.

5. A dairy cow breeding health data management system based on intelligent ecological deployment according to claim 1, characterized in that... The multi-interval trend analysis module for bovine vital signs includes a bovine vital sign status change monitoring and analysis submodule and a multi-stage vital sign data labeling and association submodule. The bovine vital sign status change monitoring and analysis submodule acquires the health data file of each bovine, sorts the bovine test data in chronological order, preprocesses the test data, compares the test data pairwise in sequence, compares the deviation of two consecutive test values, and if the deviation is greater than a set threshold, the data tested in the two time windows are marked as abnormal and sent to a human for data review. The system acquires the height, width, length, and weight data of each cow within a continuous time window, continuously monitors whether the height, width, length, and weight data of the cows increase within the continuous time window, identifies situations where the height, width, length, and weight data of the cows do not increase or decrease within a continuous set number of time windows, marks the time window with such a situation as an early warning, and extracts the cow's vital signs data within that time window and sends it to a human for data review. The multi-stage vital sign data annotation and association submodule extracts height, width, length, and weight data of different dairy cows under different time windows, screens the current growth cycle of the dairy cow corresponding to each time window, associates and annotates the growth cycle stage of the dairy cow for the time window of each dairy cow, and summarizes the vital sign data of each dairy cow in continuous time windows under different growth cycles.

6. A dairy cow breeding health data management system based on intelligent ecological deployment according to claim 1, characterized in that... The quantitative analysis module for normal cow activity includes a multi-time-window dynamic update submodule for normal cow activity and an activity level classification report and early warning submodule. The multi-time-window dynamic update submodule for normal cow activity acquires the daily activity of each cow. At the same time, based on each defined cow growth cycle stage, it sets short-term and long-term time windows for each growth cycle. The time intervals of the short-term and long-term time windows are set manually. The module calculates the clipped average of the daily activity of cows within the short-term and long-term time windows. It also extracts the daily activity of cows within different short-term and long-term time windows within the same growth cycle, and calculates the clipped average of the daily activity of cows within several short-term time windows and several long-term time windows.

7. A dairy cow breeding health data management system based on intelligent ecological deployment according to claim 6, characterized in that... The activity level grading report early warning submodule sets the cut-off average daily activity level of a dairy cow within different short-term time windows during the same growth cycle as follows: The average daily activity level of a dairy cow within different long-term time windows during the same growth cycle is: The average activity levels of the trimming process were compared and analyzed within several short-term and long-term time windows, based on the formula: in The maximum difference threshold between the average activity levels within several artificially set short-term and long-term time windows is used. When the average activity levels within these short-term and long-term time windows satisfy the above formula, it is determined that the activity levels of dairy cows within these short-term and long-term time windows are highly stable within the same growth cycle stage. The normal baseline for daily activity levels of dairy cows during this growth cycle stage is then calculated. According to the formula: The normal baseline for daily activity levels at the current growth stage of dairy cows is calculated. The current daily activity level of dairy cows is then compared with the normal baseline, and the daily activity level of dairy cows is set as follows: The daily activity level of dairy cows on a certain date To determine if a cow's behavior is highly active on a given day, the daily activity level of the cows on a given day is... To determine if a cow's behavior is low activity on a given day, the daily activity level of the cows on a given day is... To determine whether the cows' behavior on that day was at a normal level of activity, the daily behavioral activity of cows at different growth stages was analyzed. If the average activity level over several short-term and long-term time windows does not meet the above formula, it is determined that the daily activity level of dairy cows in the current growth cycle is disordered. The current dairy cow's identity ID is marked, and the dairy cow's identity ID and the daily activity level of dairy cows in the current growth cycle are sent to a human for early warning review.

8. A dairy cow breeding health data management system based on intelligent ecological deployment according to claim 1, characterized in that... The expert database-based health status assessment module includes a multi-growth-stage real-time vital sign comparison and assessment submodule and a dairy cow health status marking submodule. The multi-growth-stage real-time vital sign comparison and assessment submodule pre-constructs an expert database storing standard vital sign data of dairy cows at different growth stages. It compares the real-time detected vital sign data of each dairy cow at different growth stages with the standard vital sign data of dairy cows at different growth stages in the expert database. It compares the deviation values ​​between the real-time detected vital sign data and the standard vital sign data of dairy cows at the same growth cycle stage, and sends the data of dairy cows with deviation values ​​greater than a set threshold to the dairy cow health status marking submodule. The dairy cow health status marking submodule marks the dairy cow with abnormal health status for the current comparison, extracts the real-time detection vital sign data value of the dairy cow in the current growth stage cycle, and sends the marked dairy cow's identity ID and the extracted real-time detection vital sign data value to the intelligent data monitoring platform for manual review.

Images