A method and system for identifying estrus events of dairy cows based on behavior time series change point detection and extracting estrus expression traits

Abstract

The present application relates to a kind of based on behavior time series change point detection cow estrus event identification and estrus expression character extraction method, system, method includes: S11, obtain cow behavior time series data;S12, determine the time range to be analyzed and construct behavior time series analysis window;S13, change point detection is carried out to behavior time series, obtain candidate change point;S14, after processing to candidate change point, determine effective estrus event interval;S15, determine estrus event start time and estrus event end time;S16, calculate the basic statistics required for estrus expression character extraction;S17, extract estrus expression character;S18, output estrus event identification state information and estrus expression character extraction result.The present application can reduce the dependence on commercial system output result and preset threshold rule, realize the standardization of estrus event boundary demarcation, and improve the explainability and batch application consistency of estrus expression character extraction.

Description

Technical Field

[0001] This invention relates to the field of dairy farming technology, particularly dairy breeding technology, and specifically to a method and system for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection. Background Technology

[0002] Reproductive efficiency in dairy cows is a crucial factor affecting the economic benefits of large-scale dairy farming. Timely and accurate identification of estrus and appropriate scheduling of insemination are essential for improving conception rates, shortening the non-pregnant period, and reducing reproductive management costs.

[0003] Existing technologies for identifying estrus and recording estrus-related traits in dairy cows mainly include the following categories:

[0004] The first category is the method of manual observation and recording. This method involves feeding and management personnel regularly observing whether dairy cows exhibit signs such as mounting, standing to accept mounting, increased activity, changes in appetite, vulvar redness and swelling, or mucus secretion, and combining this with insemination records to determine estrus. Based on this, records can be generated indicating whether estrus has occurred, the timing of estrus, and estrus performance scores.

[0005] The second category consists of detection methods based on physiological or reproductive indicators. These methods assess ovarian activity, ovulation status, and estrus by detecting hormone levels such as progesterone, estradiol, and luteinizing hormone in blood or breast milk, or by observing follicle development, dominant follicle size, presence of the corpus luteum, and ovulation through rectal examination or ultrasound. These methods generate physiological records such as progesterone concentration, follicle diameter, corpus luteum status, whether ovulation occurred, ovulation time, or estrous cycle stage. These records are highly accurate and can serve as verification of estrus or ovulation status, but typically require sampling, manual manipulation, or specialized equipment for periodic monitoring.

[0006] The third category involves direct application of estrus index and alarm results from commercial automated behavior monitoring systems. Existing commercial automated behavior monitoring systems typically process sensor-collected behavioral data using built-in algorithms and output records such as estrus alarms, estrus indexes, alarm intensity, and system-determined activity peaks via accompanying software. Based on this, these records can be directly extracted, or simple statistical processing such as difference and ratio analysis can be performed to generate estrus-related indicators. This type of method reduces the burden of manual observation and has already been used in production practice to assist in insemination decisions.

[0007] The fourth category is rule-based determination and indicator calculation methods based on basic behavioral records. This method utilizes basic behavioral data such as activity levels, rumination time, and feeding behavior obtained through manual observation, video recording, pedometer recording, or derived from behavioral sensors. It determines estrus-related time periods based on rules such as increased individual activity levels, decreased rumination time, or changes relative to baseline, and calculates estrus-related indicators such as estrus duration, peak activity, and behavioral variation. In existing research and applications, this type of method typically relies on preset thresholds, fixed observation windows, or manually set rules to determine estrus intervals or estrus-related time periods.

[0008] Existing related technologies still have shortcomings, mainly in the following aspects:

[0009] First, while manual observation can directly record the external signs of estrus in dairy cows, this method is greatly affected by factors such as observation frequency, personnel experience, herd size, and observation period, making it prone to omissions or errors. Records of estrus time, presence of estrus, or estrus performance scores generated through manual observation are usually highly subjective. In particular, estrus performance scores are often categorical variables, making it difficult to continuously and precisely reflect the magnitude and duration of changes in estrus behavior.

[0010] Second, while physiological indicator-based detection methods can accurately reflect ovarian activity, hormonal changes, or ovulation status, these methods typically require multiple consecutive samplings, laboratory testing, or specialized equipment, resulting in high manpower costs and making low-cost monitoring difficult to implement in large-scale dairy herds. Furthermore, records of progesterone concentration, follicle diameter, and corpus luteum status primarily reflect physiological processes and cannot directly assess the intensity, duration, and dynamic changes of estrus behavior.

[0011] Third, when relying on commercial automated behavior monitoring systems to extract estrus characteristics, the internal algorithms and index construction methods of different systems are often not fully disclosed. Therefore, it is difficult to clearly define the basis for the formation of relevant indicators or thresholds and their applicability to different herds, pastures, and management conditions. As herd genetic levels, reproductive management methods, and behavioral patterns change, fixed thresholds or system-specific thresholds may need to be reassessed and updated in a timely manner. Furthermore, estrus behavior itself is a dynamic process, including stages such as gradual intensification, reaching a peak, short-term fluctuations, interruptions, and recovery. Existing methods based on preset thresholds are generally more suitable for capturing the core stages where behavioral changes are more pronounced, but may be insufficient for describing special cases such as gradual onset and short-term interruptions.

[0012] Fourth, methods for estrus identification and extraction of estrus-related indicators based on behavioral data have high practical application value. However, existing methods often rely on preset thresholds to determine the boundaries of estrus events, that is, to determine the start and end of estrus based on whether the estrus index, activity level, or behavioral change exceeds a predetermined value. The definition of related indicators is usually affected by the data source, system algorithm, threshold setting, or manual rules.

[0013] Therefore, there is still room for improvement in the standardization of estrus event boundaries and the consistent extraction of estrus expression traits using existing technologies. Summary of the Invention

[0014] To address the aforementioned issues, this invention provides a method and system for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection. This reduces reliance on the output results of commercial systems and preset threshold rules, standardizes the delineation of estrus event boundaries, and improves the interpretability and consistency of estrus expression trait extraction in batch applications.

[0015] The technical solution of the present invention is as follows:

[0016] The first aspect of this invention provides a method for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, comprising:

[0017] S11, Obtain time-series data on dairy cow behavior, including cow identification, collection time, activity level indicators, and rumination time;

[0018] S12, Determine the time range to be analyzed and construct the behavioral time series analysis window;

[0019] S13, Based on the constructed behavioral time series analysis window, change point detection is performed on the behavioral time series to obtain candidate change points;

[0020] S14, perform stratification and post-processing on candidate change points to determine the effective estrus event interval;

[0021] S15, Based on the determined effective estrus event interval, determine the start time and end time of the estrus event;

[0022] S16, Based on the determined start and end times of the estrus event, calculate the basic statistics required for extracting estrus expression traits;

[0023] S17, Based on the calculated basic statistics, extract estrus expression traits;

[0024] S18 outputs the estrus event recognition status information and the estrus expression trait extraction results.

[0025] Preferably, in S12, determining the time range to be analyzed and constructing the behavioral time series analysis window includes:

[0026] S121. Based on the reproductive management stage to be evaluated, determine the time range to be analyzed. The time range is the stage in which dairy cows may experience changes in estrus behavior, including the postpartum voluntary waiting period, the time range before and after mating, the time range before and after estrus synchronization treatment, or other time ranges in which changes in estrus behavior of dairy cows need to be evaluated.

[0027] S122, Within the selected time range, summarize the activity level indicators on a daily basis to obtain the daily activity level sum, and define the day with the highest daily activity level sum as... Subsequently A behavioral time series analysis window was constructed around the center, which could cover the behavioral changes before and after the increase in activity, and retain the non-estrus baseline behavioral data required for subsequent calculation of the basic statistics of estrus expression traits.

[0028] Preferably, in S122, with Constructing a behavioral time series analysis window centered on the subject further includes:

[0029] judge Are there any missing activity levels or rumination time records for the day before and after? Does the data have fewer than 3 complete data days on either side within a 4-day period?

[0030] If so, the corresponding analysis window will be marked as not meeting the analysis requirements, and subsequent estrus event interval delineation and estrus expression trait extraction will be stopped for that analysis window;

[0031] Otherwise, select separately. Qianhe Back distance The most recent 3 complete data days, and with The data from that day together constitute a 7-day behavioral time series analysis window to ensure data integrity.

[0032] Preferably, in S13, the step of performing change point detection on the behavioral time series to obtain candidate change points includes:

[0033] Based on the constructed behavioral time series analysis window, the Pruned ExactLinear Time change point detection method based on nonparametric cost function is adopted to detect change points for activity level indicators and rumination time respectively, and the positions where the two behavioral indicators change in the time series are identified as candidate change points.

[0034] Preferably, in S14, the step of performing stratified processing on candidate change points and determining the effective estrus event interval includes:

[0035] S141, Regularly screen and correct the change points detected in the time series of activity indicators to remove unreasonable results caused by boundary effects, short-term fluctuations or missing change points;

[0036] S142, the remaining activity variable points are arranged in chronological order, and candidate estrus event intervals are formed by grouping two adjacent variable points together;

[0037] S143, determine whether only one valid estrus event is identified. If so, identify it as a valid estrus event within the behavioral time series analysis window. Otherwise, determine that the reliability of the identification result of the window is low and terminate the extraction of estrus expression traits in the window.

[0038] Preferably, in S15, determining the start time and end time of the estrus event includes:

[0039] Based on the first and second boundary points of the determined effective estrus event interval, the first recording period after the first boundary point is defined as the start time of the estrus event, and the first recording period after the second boundary point is defined as the end time of the estrus event.

[0040] Preferably, in S16, the basic statistics required for calculating the extraction of estrus expression traits include:

[0041] S161, Based on the determined start and end times of the estrus event, the estrus period and non-estrus baseline period are divided within the 7-day behavioral time series analysis window; wherein, the estrus period is the time interval between the start and end times of the estrus event, and the non-estrus baseline period is the remaining time interval within the 7-day behavioral time series analysis window excluding the estrus period.

[0042] S162, calculate the basic statistics for the estrus period and the baseline non-estrus period, including the maximum activity level during the estrus period, the average activity level during the baseline non-estrus period, the standard deviation of the activity level during the baseline non-estrus period, the average rumination time during the baseline non-estrus period, and the standard deviation of the rumination time during the baseline non-estrus period.

[0043] Preferably, in S17, the estrus expression traits include: estrus duration, peak activity level, estrus intensity, and rumination variation; wherein, estrus duration is the time interval between the end time of the estrus event and the start time of the estrus event; peak activity level represents the standardized deviation of the maximum activity level index during estrus relative to the baseline of non-estrus activity level; estrus intensity represents the standardized cumulative increase of the activity level index during estrus relative to the baseline of non-estrus activity level; and rumination variation represents the standardized cumulative deviation of rumination time during estrus relative to the baseline of non-estrus rumination.

[0044] A second aspect of the present invention provides a system for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, comprising:

[0045] The data acquisition module is configured to acquire time-series data on dairy cow behavior, including cow identification, collection time, activity level indicators, and rumination time.

[0046] The analysis window construction module is configured to determine the time range to be analyzed and construct a behavioral time series analysis window.

[0047] The change point detection module is configured to perform change point detection on the behavioral time series based on the constructed behavioral time series analysis window, and obtain candidate change points;

[0048] The stratified post-processing module is configured to perform stratified post-processing on candidate variable points and determine the effective estrus event interval;

[0049] The estrus boundary determination module is configured to determine the start time and end time of the estrus event based on the determined valid estrus event interval.

[0050] The basic statistics calculation module is configured to calculate the basic statistics required for extracting estrus expression traits based on the determined start and end times of the estrus event.

[0051] The estrus expression trait extraction module is configured to extract estrus expression traits based on the calculated basic statistics;

[0052] The results output module is configured to output estrus event recognition status information and estrus expression trait extraction results.

[0053] A third aspect of the present invention provides a computing device, comprising:

[0054] Memory, used to store one or more programs;

[0055] One or more processors are configured to execute the one or more programs to implement the method.

[0056] A fourth aspect of the present invention provides a computer-readable storage medium storing a program that, when executed by a processor, implements the method described herein.

[0057] The advantages of this invention compared to existing technologies are as follows: This invention proposes a method and system for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection. This reduces reliance on the output results of commercial systems and preset threshold rules, achieves standardized delineation of estrus event boundaries, and improves the interpretability and consistency of estrus expression trait extraction in batch applications. Specifically, this is reflected in:

[0058] (I) A method for identifying the boundary of estrus events based on structural changes in behavioral time series is proposed. This invention starts with time series data of individual dairy cow behavior and introduces a change point detection strategy. By identifying structural changes in activity levels and rumination time series, the time positions at which dairy cows enter and leave estrus-related behavioral states are marked. This method transforms estrus event identification from fixed threshold judgment to boundary delineation based on changes in individual behavioral states, making the determination of the onset and end times of estrus more consistent with the behavioral change process itself.

[0059] (II) A complete data processing workflow and system implementation method, from behavioral data processing to estrus expression trait extraction, have been established. This invention not only proposes specific processing methods for estrus event boundary identification, but also integrates steps such as data acquisition, analysis window construction, change point detection, stratified post-processing, estrus boundary determination, basic statistical calculation, estrus expression trait extraction, and result output into a complete workflow executable by a software system. Through this workflow, automatic behavioral monitoring data is no longer only used to obtain estrus judgment results, but can output continuous estrus expression traits such as estrus duration, peak activity level, estrus intensity, and rumination changes under unified rules, thereby providing a more complete data foundation for the quantification of dairy cow estrus behavior, reproductive management, and subsequent genetic analysis.

[0060] (III) In the change point detection stage, the location of behavioral state transitions can be objectively identified from the behavioral time series. This invention performs change point detection on activity level indicators and rumination time series respectively, transforming changes in behavioral levels before and after estrus into candidate change points that can be used for boundary delineation. Compared with judgments based solely on fixed thresholds or single peaks, this approach focuses more on the state transition process in the individual dairy cow's behavioral time series, which helps improve the objectivity and interpretability of the candidate estrus boundary acquisition process.

[0061] (iv) In determining the effective estrus event interval, the statistical change point results and the characteristics of dairy cow estrus behavior are combined for rule-based screening and correction. This invention removes change points near the window boundary, limits the duration of candidate intervals, requires candidate intervals to include the maximum activity level and reach a certain range above the individualized activity baseline, and uses rumination time change points to correct the boundary when necessary. This eliminates candidate results caused by boundary effects, short-term fluctuations, missing change points, or unreasonable interval durations. The effective estrus event intervals obtained in this way not only come from time series changes but also conform to the physiological and production realities of dairy cow estrus behavior, which helps to improve the stability and biological rationality of the estrus event boundary delineation results.

[0062] (v) In the process of extracting estrus expression traits, the estrus event can be expanded from a simple "whether it occurs" judgment to a quantifiable and comparable continuous estrus expression phenotype. The estrus expression traits such as estrus duration, peak activity level, estrus intensity, and rumination changes calculated and obtained by this invention can be used to describe the strength of estrus expression and behavioral changes in individual dairy cows, providing a rich data foundation for subsequent reproductive management and genetic analysis.

[0063] It should be understood that the implementation of any embodiment of the present invention does not mean that it will simultaneously possess or achieve multiple or all of the above-mentioned beneficial effects. Attached Figure Description

[0064] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0065] The structures, proportions, sizes, etc., shown in the accompanying drawings of this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0066] Figure 1 A schematic flowchart of a method for identifying estrus events and extracting estrus expression traits in dairy cows, provided as an embodiment of the present invention;

[0067] Figure 2 A schematic flowchart illustrating the specific execution of the method for identifying estrus events and extracting estrus expression traits in dairy cows provided in the embodiments of the present invention;

[0068] Figure 3 A schematic diagram illustrating the delineation of the boundary of estrus events and the definition of estrus expression traits in dairy cows, provided for embodiments of the present invention;

[0069] Figure 4 This is a schematic block diagram of the system for identifying estrus events and extracting estrus expression traits in dairy cows provided in an embodiment of the present invention;

[0070] Figure 5 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention; and

[0071] Figure 6 This is a schematic block diagram of a computing device provided in an embodiment of the present invention. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments and descriptions of the present invention are used to explain the present invention, but are not intended to limit the present invention.

[0073] It should be understood that the terms "comprising / including," "consisting of," or any other variations are intended to cover non-exclusive inclusion, such that a product, apparatus, process, or method that comprises a list of elements includes not only those elements but may also include, where necessary, other elements not expressly listed, or elements inherent to such a product, apparatus, process, or method. Without further limitation, an element defined by the phrases "comprising / including," "consisting of," does not exclude the presence of additional identical elements in the product, apparatus, process, or method that includes said element.

[0074] As mentioned earlier, existing technologies still have shortcomings in the standardized delineation of estrus event boundaries and the consistent extraction of estrus expression traits. The intensity, duration, and behavioral changes of estrus are related to the subsequent reproductive performance of dairy cows. In applications such as farm reproductive management, production management, and genetic improvement, it is necessary to obtain comparable estrus expression traits stably in large herds of dairy cows. Automated behavior monitoring systems can continuously collect behavioral data such as dairy cow activity levels and rumination time, providing a data foundation for the batch extraction of these traits. However, the estrus index and its derived indicators in existing automated behavior monitoring systems are mainly geared towards estrus detection and insemination prompts in daily production management. Different systems often use their own algorithms, thresholds, and indicator definitions. Although such outputs can be used to assist in daily reproductive management, it is still difficult to extract unified and interpretable definitions of the start and end boundaries of estrus events, the magnitude of behavioral changes, and the duration of the process in cross-system, cross-farm, or large-scale data analysis scenarios. There is a lack of a universal technical process that delineates estrus events and extracts estrus expression features in batches based on original behavioral time series and according to unified rules.

[0075] Therefore, this invention provides a method, system, and storage medium for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time-series change point detection. The method takes dairy cow behavioral time-series data output by an automatic behavior monitoring system as input, identifies estrus event boundaries using behavioral state changes, and extracts estrus expression traits based on the defined estrus interval and non-estrus baseline interval. This method does not rely on estrus alarm results or estrus index thresholds output by commercial systems as the primary basis for defining estrus event boundaries, and can be used for offline, batch, and standardized analysis of dairy cow estrus behavior processes. The system, used to execute the above method, may include a data acquisition module, an analysis window construction module, a change point detection module, a hierarchical post-processing module, an estrus boundary determination module, a basic statistics calculation module, an estrus expression trait extraction module, and a result output module. The storage medium stores a computer program for executing the above method; when this computer program is executed by a processor, the process of identifying dairy cow estrus events and extracting estrus expression traits is implemented.

[0076] The implementation of the present invention will be described in detail below with reference to preferred embodiments.

[0077] This invention provides a method 10 for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, specifically as follows: Figure 1 The process shown, method 10 includes: S11, acquiring time series data of dairy cow behavior; S12, determining the time range to be analyzed and constructing a behavior time series analysis window; S13, based on the constructed behavior time series analysis window, performing change point detection on the behavior time series to obtain candidate change points; S14, performing stratified post-processing on the candidate change points and determining the effective estrus event interval; S15, based on the determined effective estrus event interval, determining the estrus event start time and estrus event end time; S16, based on the determined estrus event start time and estrus event end time, calculating the basic statistics required for estrus expression trait extraction; S17, based on the calculated basic statistics, extracting estrus expression traits; S18, outputting estrus event identification status information and estrus expression trait extraction results.

[0078] The following will continue to combine Figure 2 The specific execution of each method and step is explained in detail.

[0079] S11, Obtain time series data on cow behavior.

[0080] This study receives or imports time-series data on dairy cow behavior generated by automated behavior monitoring systems, video acquisition devices, pedometers, or other devices capable of acquiring behavioral information, or data that has been processed. The data includes cow identification, collection time, activity level indicators, and rumination time. Activity level indicators reflect changes in dairy cow activity levels, and rumination time reflects changes in rumination behavior. This method is suitable for behavioral time-series data recorded or summarized at fixed time intervals; in this study, the data used were activity levels and rumination times summarized over 2 hours.

[0081] It should be noted that the behavioral time series data required here only uses cattle identification, collection time, activity level indicators, and rumination time. In actual use, if there is a personalized analysis time range, such as extracting data within a certain number of days after calving, then the corresponding calving date of the cattle needs to be used; similarly, if the data is extracted within a certain number of days before and after mating, then the mating date of the cattle needs to be provided.

[0082] The purpose of this step is to obtain the basic behavioral data required for subsequent estrus event boundary delineation and estrus expression trait extraction.

[0083] S12, Determine the time range to be analyzed and construct the behavioral time series analysis window. This may further include:

[0084] S121. Determine the time frame to be analyzed based on the reproductive management stage to be evaluated. This time frame should encompass the period during which changes in estrus behavior are likely to occur in dairy cows, including the postpartum voluntary waiting period, the time frame before and after mating, the time frame before and after estrus synchronization treatment, or other time frames where changes in estrus behavior need to be evaluated. This method is not suitable for excessively long and undefined time frames, such as the entire lactation period, nor is it suitable for time frames where estrus behavior is clearly unlikely to occur.

[0085] S122, Within the selected time range, summarize the activity level indicators on a daily basis to obtain the daily activity level sum, and define the day with the highest daily activity level sum as... ; then with A behavioral time-series analysis window is constructed around the central point, enabling it to cover the behavioral changes before and after the activity spike, while retaining the non-estrus baseline behavioral data needed for subsequent calculations of baseline statistics for estrus expression traits. In practice, this method employs... Add 3 days before and after The data from that day were used to create a 7-day behavioral time series analysis window. The activity levels and rumination times within this window retained the original recording intervals; in this study, data was actually used at 2-hour intervals.

[0086] To ensure data integrity, if The day before or The following day, there was a lack of activity level or rumination time records, or Within the first 4 days If any side within the subsequent 4-day period has fewer than 3 complete data days, the analysis window is marked as not meeting the analysis requirements, and subsequent estrus interval delineation and estrus expression trait extraction for that analysis window are terminated. After quality control, each is retained... Qianhe Back distance The most recent 3 complete data days, and with The data from that day together constitute a 7-day behavioral time series analysis window used for analysis.

[0087] S13, perform change point detection on behavioral time series.

[0088] Based on the behavioral time series analysis window constructed in the previous step, the PrunedExact Linear Time change point detection method, based on a nonparametric cost function, is used to detect change points for both activity level and rumination time, identifying the locations where the two behavioral indicators undergo state changes in the time series as candidate change points. The detected candidate change points are used for subsequent estrus event interval construction, boundary screening, and boundary correction.

[0089] S14, perform stratified post-processing on candidate change points to determine the effective estrus event interval. This may further include:

[0090] S141, Regularly screen and correct the change points detected in the time series of activity indicators to remove unreasonable results caused by boundary effects, short-term fluctuations or missing change points;

[0091] Here, during the screening process, specific points within a 6-hour range at the beginning or end of the analysis window are removed to reduce the impact of boundary effects on the delineation of estrus intervals.

[0092] It should be noted that candidate estrus intervals must meet the following conditions:

[0093] First, the candidate interval should include the maximum activity level indicator within a 7-day analysis window;

[0094] Second, the average activity level within the candidate interval should be at least 1.5 standard deviations higher than the individualized activity baseline of the cow. Here, the individualized activity baseline and standard deviation refer to the average activity level and standard deviation of activity level calculated from records other than the candidate estrus interval within the current 7-day analysis window.

[0095] Third, the length of the candidate interval should be between 4 and 32 hours.

[0096] The above conditions are used to exclude unreasonable candidate intervals caused by window boundary effects, short-term abnormal fluctuations, or durations that do not conform to the physiological process of estrus in dairy cows.

[0097] S142, the remaining activity level change points are arranged in chronological order, and candidate estrus intervals are formed by grouping two adjacent change points together;

[0098] S143, determine whether only one valid estrus event is identified. If so, identify it as a valid estrus event within the behavioral time series analysis window. Otherwise, determine that the identification result of the window is unreliable and terminate the extraction of estrus expression traits in the window.

[0099] If the candidate intervals formed based on activity level change points do not meet the duration requirements, the boundaries of the candidate intervals are corrected using change points detected by rumination time series analysis to help determine the estrus start or end boundaries. If no valid change points are detected in the activity level time series, or only one change point is detected, but the activity level shows a significant single-peak increase with the peak value at least 5 standard deviations above the individualized activity baseline, then the two time points with the highest activity level are selected as candidate boundary points. When the interval between these two candidate boundary points does not exceed 6 hours, a candidate estrus interval can be formed based on these two time points.

[0100] After the above stratification and post-processing, if only one candidate estrus interval meets the above conditions, it is identified as a valid estrus event within the analysis window. If no candidate estrus interval meets the conditions, or if multiple candidate estrus intervals meet the conditions simultaneously, it is recorded as no unique valid estrus event was identified within the analysis window, and the calculation of estrus expression traits for that window is terminated.

[0101] The purpose of this step is to perform regular screening and correction on the candidate boundaries obtained from change point detection, remove unreasonable results caused by boundary effects, short-term fluctuations or missing change points, and thus obtain stable, interpretable and biologically reasonable estrus event intervals.

[0102] S15, determine the start time and end time of the estrus event.

[0103] The identified valid estrus event intervals have a first boundary point and a second boundary point. The first recorded period after the first boundary point is defined as the start time of the estrus event, and the first recorded period after the second boundary point is defined as the end time of the estrus event. The data actually used in this study is 2-hour summary data, therefore corresponding to the first 2-hour recorded period after the boundary point.

[0104] S16, Calculate the basic statistics required for extracting estrus expression traits. This may further include:

[0105] S161, based on the determined start and end times of the estrus event, the estrus period and non-estrus baseline period are divided within the 7-day behavioral time series analysis window; wherein, the estrus period is the time interval between the start and end times of the estrus event, and the non-estrus baseline period is the remaining time interval within the 7-day behavioral time series analysis window excluding the estrus period; as shown in Table 1, Table 1 shows the original records of a dairy cow in the 7-day behavioral time series analysis window and its estrus period division results in an embodiment of the present invention, demonstrating the data types processed by the present invention, the recording time interval, and the results of dividing specific behavioral records into the estrus period or non-estrus baseline period according to the start and end times of the estrus event.

[0106] Table 1. Window record of 7-day behavioral time series analysis and estrus period segmentation results of an individual dairy cow.

[0107] ID492703 2021 / 8 / 28 0:00 42 7 Non-estrus baseline period ID492703 2021 / 8 / 28 2:00 27 39 Non-estrus baseline period ID492703 2021 / 8 / 28 4:00 28 80 Non-estrus baseline period ID492703 2021 / 8 / 28 6:00 42 1 Non-estrus baseline period ID492703 2021 / 8 / 28 8:00 29 36 Non-estrus baseline period ID492703 2021 / 8 / 28 10:00 42 39 Non-estrus baseline period ID492703 2021 / 8 / 28 12:00 41 4 Non-estrus baseline period ID492703 2021 / 8 / 28 14:00 40 45 Non-estrus baseline period ID492703 2021 / 8 / 28 16:00 41 32 Non-estrus baseline period ID492703 2021 / 8 / 28 18:00 46 35 Non-estrus baseline period ID492703 2021 / 8 / 28 20:00 24 71 Non-estrus baseline period ID492703 2021 / 8 / 28 22:00 34 51 Non-estrus baseline period ID492703 2021 / 8 / 29 0:00 35 18 Non-estrus baseline period ID492703 2021 / 8 / 29 2:00 29 65 Non-estrus baseline period ID492703 2021 / 8 / 29 4:00 31 83 Non-estrus baseline period ID492703 2021 / 8 / 29 6:00 35 9 Non-estrus baseline period ID492703 2021 / 8 / 29 8:00 39 77 Non-estrus baseline period ID492703 2021 / 8 / 29 10:00 36 26 Non-estrus baseline period ID492703 2021 / 8 / 29 12:00 41 33 Non-estrus baseline period ID492703 2021 / 8 / 29 14:00 30 75 Non-estrus baseline period ID492703 2021 / 8 / 29 16:00 48 20 Non-estrus baseline period ID492703 2021 / 8 / 29 18:00 40 42 Non-estrus baseline period ID492703 2021 / 8 / 29 20:00 29 70 Non-estrus baseline period ID492703 2021 / 8 / 29 22:00 39 53 Non-estrus baseline period ID492703 2021 / 8 / 30 0:00 43 15 Non-estrus baseline period ID492703 2021 / 8 / 30 2:00 28 70 Non-estrus baseline period ID492703 2021 / 8 / 30 4:00 26 83 Non-estrus baseline period ID492703 2021 / 8 / 30 6:00 41 9 Non-estrus baseline period ID492703 2021 / 8 / 30 8:00 33 23 Non-estrus baseline period ID492703 2021 / 8 / 30 10:00 40 13 Non-estrus baseline period ID492703 2021 / 8 / 30 12:00 37 59 Non-estrus baseline period ID492703 2021 / 8 / 30 14:00 33 57 Non-estrus baseline period ID492703 2021 / 8 / 30 16:00 31 36 Non-estrus baseline period ID492703 2021 / 8 / 30 18:00 56 1 Non-estrus baseline period ID492703 2021 / 8 / 30 20:00 29 74 Non-estrus baseline period ID492703 2021 / 8 / 30 22:00 31 35 Non-estrus baseline period ID492703 2021 / 8 / 31 0:00 43 35 Non-estrus baseline period ID492703 2021 / 8 / 31 2:00 33 68 Non-estrus baseline period ID492703 2021 / 8 / 31 4:00 73 34 estrus ID492703 2021 / 8 / 31 6:00 51 21 estrus ID492703 2021 / 8 / 31 8:00 125 8 estrus ID492703 2021 / 8 / 31 10:00 153 0 estrus ID492703 2021 / 8 / 31 12:00 143 0 estrus ID492703 2021 / 8 / 31 14:00 119 10 estrus ID492703 2021 / 8 / 31 16:00 110 0 estrus ID492703 2021 / 8 / 31 18:00 66 50 estrus ID492703 2021 / 8 / 31 20:00 34 73 estrus ID492703 2021 / 8 / 31 22:00 32 74 Non-estrus baseline period ID492703 2021 / 9 / 1 0:00 44 27 Non-estrus baseline period ID492703 2021 / 9 / 1 2:00 28 68 Non-estrus baseline period ID492703 2021 / 9 / 1 4:00 32 63 Non-estrus baseline period ID492703 2021 / 9 / 1 6:00 41 18 Non-estrus baseline period ID492703 2021 / 9 / 1 8:00 26 28 Non-estrus baseline period ID492703 2021 / 9 / 1 10:00 36 12 Non-estrus baseline period ID492703 2021 / 9 / 1 12:00 42 44 Non-estrus baseline period ID492703 2021 / 9 / 1 14:00 29 73 Non-estrus baseline period ID492703 2021 / 9 / 1 16:00 39 0 Non-estrus baseline period ID492703 2021 / 9 / 1 18:00 40 45 Non-estrus baseline period ID492703 2021 / 9 / 1 20:00 26 67 Non-estrus baseline period ID492703 2021 / 9 / 1 22:00 32 32 Non-estrus baseline period ID492703 2021 / 9 / 2 0:00 37 31 Non-estrus baseline period ID492703 2021 / 9 / 2 2:00 28 62 Non-estrus baseline period ID492703 2021 / 9 / 2 4:00 26 72 Non-estrus baseline period ID492703 2021 / 9 / 2 6:00 41 27 Non-estrus baseline period ID492703 2021 / 9 / 2 8:00 27 47 Non-estrus baseline period ID492703 2021 / 9 / 2 10:00 39 1 Non-estrus baseline period ID492703 2021 / 9 / 2 12:00 37 47 Non-estrus baseline period ID492703 2021 / 9 / 2 14:00 31 34 Non-estrus baseline period ID492703 2021 / 9 / 2 16:00 37 26 Non-estrus baseline period ID492703 2021 / 9 / 2 18:00 38 17 Non-estrus baseline period ID492703 2021 / 9 / 2 20:00 28 71 Non-estrus baseline period ID492703 2021 / 9 / 2 22:00 34 75 Non-estrus baseline period ID492703 2021 / 9 / 3 0:00 33 34 Non-estrus baseline period ID492703 2021 / 9 / 3 2:00 30 45 Non-estrus baseline period ID492703 2021 / 9 / 3 4:00 33 78 Non-estrus baseline period ID492703 2021 / 9 / 3 6:00 43 9 Non-estrus baseline period ID492703 2021 / 9 / 3 8:00 29 12 Non-estrus baseline period ID492703 2021 / 9 / 3 10:00 35 4 Non-estrus baseline period ID492703 2021 / 9 / 3 12:00 42 25 Non-estrus baseline period ID492703 2021 / 9 / 3 14:00 35 57 Non-estrus baseline period ID492703 2021 / 9 / 3 16:00 44 28 Non-estrus baseline period ID492703 2021 / 9 / 3 18:00 43 30 Non-estrus baseline period ID492703 2021 / 9 / 3 20:00 31 45 Non-estrus baseline period ID492703 2021 / 9 / 3 22:00 33 38 Non-estrus baseline period

[0108] S162, calculate the basic statistics for the estrus period and the non-estrus baseline period respectively; wherein, the maximum value of the activity index among the recorded periods within the estrus period is taken as the maximum activity index of the estrus period; the average value of the activity index among the recorded periods within the non-estrus baseline period is taken as the average activity index of the non-estrus baseline period, and the standard deviation of the activity index of the non-estrus baseline period is calculated; the average rumination time among the recorded periods within the non-estrus baseline period is taken as the average rumination time of the non-estrus baseline period, and the standard deviation of the rumination time of the non-estrus baseline period is calculated. As shown in Table 2, Table 2 shows an example of the basic statistics further calculated based on the estrus period and non-estrus baseline period division results shown in Table 1. Accordingly, after determining the estrus period and the non-estrus baseline period, the behavioral records within the corresponding time intervals can be converted into individualized basic parameters required for subsequent calculation of estrus expression traits.

[0109] Table 2. Baseline statistics of estrus and non-estrus periods for an individual dairy cow.

[0110] cattle identification ID492703 Start time of estrus 2021 / 8 / 31 4:00 End time of estrus 2021 / 8 / 31 20:00 Maximum activity level during estrus 153.00 au / 2h Average activity level during the non-estrous baseline period 35.37 au / 2h Standard deviation of baseline activity levels during nonestrus 6.38 Average rumination time at baseline (non-estrus) 40.73 min / 2h Standard deviation of rumination time at baseline of nonestrus 24.29

[0111] S17, extract estrus expression traits.

[0112] Based on the basic statistics obtained in the previous step, calculate estrus expression traits such as estrus duration, peak activity level, estrus intensity, and rumination changes. Estrus duration is the time interval between the end and start of the estrus event, which can be expressed in hours. Peak activity level represents the standardized deviation of the maximum activity level during estrus from the individualized non-estrus activity baseline, and can be calculated based on the maximum activity level during estrus, the average activity level during the non-estrus baseline period, and the standard deviation of the activity level during the non-estrus baseline period. Estrus intensity represents the standardized cumulative increase in activity level during estrus relative to the non-estrus reference activity level; here, the non-estrus reference activity level refers to the expected activity level of the cow at the corresponding time point during estrus, assuming no estrus has occurred, and can be obtained by establishing a non-estrus state prediction model based on the non-estrus baseline activity level; among these, the non-estrus state prediction model includes the Prophet regression model and the linear time series model. In specific calculations, the optimal model suitable for the target individual cow can be determined based on preset model evaluation indicators to estimate the non-estrus reference activity level at the corresponding time point during estrus. As a simplified implementation, the average activity level at the baseline non-estrus period can be used to approximate the non-estrus reference activity level. Rumination variation represents the standardized cumulative deviation of rumination time during estrus relative to the non-estrus reference rumination level. Here, the non-estrus reference rumination level refers to the expected rumination level of the cow at the corresponding time point in the estrus period, assuming no estrus has occurred. This level can be obtained by establishing a non-estrus state prediction model based on the rumination time at the baseline non-estrus period. The non-estrus state prediction model includes Prophet regression models and linear time series models. In specific calculations, the optimal model suitable for the target cow can be determined based on preset model evaluation indicators to estimate the non-estrus reference rumination level at the corresponding time point in the estrus period. As a simplified implementation, the average rumination time at the baseline non-estrus period can also be used to approximate the non-estrus reference rumination level.

[0113] like Figure 3 As shown, Figure 3 This illustrates the origins and definitions of four estrus expression traits: estrus duration, peak activity level, estrus intensity, and rumination changes. From... Figure 3 As can be seen, taking the real input data within the 7-day behavioral time series analysis window of a certain dairy cow as an example, the changes in activity level indicators and rumination time before and after estrus are shown; the start time and end time of the estrus event determined based on the method of this invention are marked, and the 7-day behavioral time series analysis window is divided into the estrus period and the non-estrus baseline period accordingly; at the same time, it intuitively explains how to extract quantifiable estrus expression traits based on behavioral sequences such as activity level indicators and rumination time.

[0114] The purpose of this step is to transform the increased activity and rumination changes within the estrous zone into continuous estrus expression traits.

[0115] S18 outputs the estrus event recognition status information and the estrus expression trait extraction results.

[0116] The system outputs the identification status information and estrus expression trait extraction results for each estrus event of each dairy cow. The estrus event identification status information includes whether the corresponding analysis window meets the analysis requirements, whether a valid estrus event has been identified, and whether estrus expression trait extraction has been completed. The estrus expression trait extraction results include cow identification, estrus duration, peak activity level, estrus intensity, and rumination changes. As shown in Table 3, Table 3 illustrates the estrus event identification status and estrus expression trait extraction results for a specific dairy cow obtained further based on the basic statistics shown in Table 2, demonstrating the final output form of the method of this invention. It is evident that the method provided by this invention is effective and feasible.

[0117] Table 3. Results of estrus expression traits extracted from an individual dairy cow

[0118] ID492703 The only valid estrus event was identified. 16 h 18.45 87.16 -7.02

[0119] It is evident that the method provided by this invention has been innovatively improved, particularly in the following aspects:

[0120] (i) A method for identifying the boundary of estrus events based on changes in the structure of behavioral time series is proposed.

[0121] Existing estrus detection methods largely rely on estrus alarm results, estrus indices, or preset threshold rules output by commercial systems, typically focusing on determining whether a cow is in estrus. This invention starts with time-series data of individual cow behavior and introduces a change-point detection strategy. By identifying structural changes in activity levels and rumination time series, it marks the time points at which cows enter and leave estrus-related behavioral states. This method transforms estrus event identification from fixed threshold judgment to boundary delineation based on individual behavioral state transitions, making the determination of estrus onset and estrus event end times more consistent with the behavioral change process itself.

[0122] (ii) A complete data processing flow and system implementation method from behavioral data processing to estrus expression trait extraction were established.

[0123] While automated behavioral monitoring data is widely used in existing research and production applications for estrus monitoring and insemination alerts, its utilization often relies on estrus alarms or estrus indices output by commercial systems, or simply performs basic statistical calculations on behavioral indicators such as activity levels and rumination time. It lacks a complete technical process for defining estrus event boundaries and extracting estrus expression traits from a behavioral time series perspective. This invention establishes a complete software-executable process, including data acquisition, analysis window construction, change point detection, stratified post-processing, estrus boundary determination, basic statistical calculation, estrus expression trait extraction, and result output. This process not only identifies estrus events but also extracts continuous estrus expression traits such as estrus duration, peak activity level, estrus intensity, and rumination changes under unified rules, providing a more complete data foundation for the quantification of dairy cow estrus behavior, reproductive management, and subsequent genetic analysis.

[0124] (iii) Improve the interpretability and consistency of estrus expression traits extraction in batch applications.

[0125] The estrus event boundaries and estrus expression traits obtained by this invention have clear judgment logic and rule basis. The estrus event start time and estrus event end time are derived from structural changes in the behavioral time series. The retention, exclusion, and correction of effective estrus intervals are not based solely on mechanical screening of statistical results, but rather on combining the understanding of characteristics such as estrus duration, increased activity, and rumination changes in dairy cow estrus behavior and physiology research. Biologically reasonable limitations are imposed on candidate intervals, ensuring that the finally determined estrus event intervals reflect both the state transition in the behavioral time series and conform to the actual occurrence process of dairy cow estrus behavior. Simultaneously, this invention can batch process behavioral data from large-scale dairy cow groups, helping to reduce the impact of differences in human judgment and equipment default thresholds on the results, and improving the consistency of dairy cow estrus event identification and estrus expression trait extraction results in application.

[0126] This invention also provides a system 20 for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, used to execute the above method 10, such as... Figure 4 As shown, system 20 includes:

[0127] Data acquisition module 21 is configured to acquire time series data of dairy cow behavior, including cow identification, collection time, activity level indicators and rumination time;

[0128] Analysis window construction module 22 is configured to determine the time range to be analyzed and construct a behavioral time series analysis window;

[0129] The change point detection module 23 is configured to perform change point detection on the behavioral time series based on the constructed behavioral time series analysis window to obtain candidate change points;

[0130] The stratified post-processing module 24 is configured to perform stratified post-processing on candidate variable points and determine the effective estrus event interval;

[0131] The estrus boundary determination module 25 is configured to determine the start time and end time of the estrus event based on the determined valid estrus event interval;

[0132] The basic statistics calculation module 26 is configured to calculate the basic statistics required for extracting estrus expression traits based on the determined start time and end time of the estrus event.

[0133] The estrus expression trait extraction module 27 is configured to extract estrus expression traits based on the calculated basic statistics;

[0134] The result output module 28 is configured to output estrus event recognition status information and estrus expression trait extraction results.

[0135] It should be noted that system 20 is a system corresponding to method 10 described above. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0136] like Figure 5 As shown, embodiments of the present invention also provide an electronic device 30, including: a memory 31 for storing one or more computer programs; and one or more processors 32 for executing the one or more computer programs, wherein the computer programs, when run by the processors 32, perform the method 10 described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects. The electronic device 30 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown in this invention, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0137] like Figure 6As shown, electronic device 30 is a computing device or computer system, which may include CPU 301 (computing unit), which can perform various appropriate actions and processes according to a computer program stored in ROM 302 (read-only memory) or a computer program loaded from storage unit 308 into random access RAM 303 (memory). RAM 303 may also store various programs and data required for the operation of electronic device 30. CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. I / O interface 305 (input / output interface) is also connected to bus 304.

[0138] Multiple components in electronic device 30 are connected to I / O interface 305, including: input unit 306, such as keyboard, mouse, etc.; output unit 307, such as various types of displays, speakers, etc.; storage unit 308, such as disk, optical disk, etc.; and communication unit 309, such as network card, modem, wireless transceiver, etc. Communication unit 309 allows electronic device 30 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0139] CPU 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of CPU 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. CPU 301 performs the various methods and processes described above. For example, in some embodiments, method 10 may be implemented as a computer software program tangibly contained in a computer-readable storage medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 30 via ROM 302 and / or communication unit 309. When the computer program is loaded into RAM 303 and executed by CPU 301, one or more steps of method 10 may be performed. Alternatively, in other embodiments, CPU 301 may be configured to perform method 10 by any other suitable means (e.g., by means of firmware).

[0140] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform method 10 as described above, thereby realizing the process of identifying estrus events and extracting estrus expression traits in dairy cows. All implementations of method 10 described above are applicable to this embodiment and can achieve the same technical effects.

[0141] Those skilled in the art will recognize that the method steps described in conjunction with the embodiments disclosed in this invention can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0142] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and equipment described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0143] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0144] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0145] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code for implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps for performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0146] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, characterized in that, include: S11, Obtain time-series data on dairy cow behavior, including cow identification, collection time, activity level indicators, and rumination time; S12, Determine the time range to be analyzed and construct the behavioral time series analysis window; S13, Based on the constructed behavioral time series analysis window, change point detection is performed on the behavioral time series to obtain candidate change points; S14, perform stratification and post-processing on candidate change points to determine the effective estrus event interval; S15, Based on the determined effective estrus event interval, determine the start time and end time of the estrus event; S16, Based on the determined start and end times of the estrus event, calculate the basic statistics required for extracting estrus expression traits; S17, Based on the calculated basic statistics, extract estrus expression traits; S18 outputs the estrus event recognition status information and the estrus expression trait extraction results.

2. The method according to claim 1, characterized in that, In S12, determining the time range to be analyzed and constructing the behavioral time series analysis window includes: S121. Based on the reproductive management stage to be evaluated, determine the time range to be analyzed. The time range is the stage in which dairy cows may experience changes in estrus behavior, including the postpartum voluntary waiting period, the time range before and after mating, the time range before and after estrus synchronization treatment, or other time ranges in which changes in estrus behavior of dairy cows need to be evaluated. S122, Within the selected time range, summarize the activity level indicators on a daily basis to obtain the daily activity level sum, and define the day with the highest daily activity level sum as... Subsequently A behavioral time series analysis window was constructed around the center, which could cover the behavioral changes before and after the increase in activity, and retain the non-estrus baseline behavioral data required for subsequent calculation of the basic statistics of estrus expression traits.

3. The method according to claim 2, characterized in that, In S122, with Constructing a behavioral time series analysis window centered on the subject further includes: judge Did the activity level or rumination time records lack in the day before and after? Does the data have fewer than 3 complete data days on either side within a 4-day period? If so, the corresponding analysis window will be marked as not meeting the analysis requirements, and subsequent estrus event interval delineation and estrus expression trait extraction will be stopped for that analysis window; Otherwise, select separately. Qianhe Back distance The three most recent complete data days, and with The data from that day together constitute a 7-day behavioral time series analysis window to ensure data integrity.

4. The method according to claim 1, characterized in that, In S13, the step of performing change point detection on the behavioral time series to obtain candidate change points includes: Based on the constructed behavioral time series analysis window, the Pruned ExactLinear Time change point detection method based on nonparametric cost function is adopted to detect change points for activity level indicators and rumination time respectively, and the positions where the two behavioral indicators change in the time series are identified as candidate change points.

5. The method according to claim 1, characterized in that, In S14, the step of performing stratification and post-processing on candidate change points and determining the effective estrus event interval includes: S141, Regularly screen and correct the change points detected in the time series of activity indicators to remove unreasonable results caused by boundary effects, short-term fluctuations or missing change points; S142, the remaining activity variable points are arranged in chronological order, and candidate estrus event intervals are formed by grouping two adjacent variable points together; S143, determine whether only one valid estrus event is identified. If so, identify it as a valid estrus event within the behavioral time series analysis window. Otherwise, determine that the reliability of the identification result of the window is low and terminate the extraction of estrus expression traits in the window.

6. The method according to claim 1, characterized in that, In S15, determining the start time and end time of the estrus event includes: Based on the first and second boundary points of the determined effective estrus event interval, the first recording period after the first boundary point is defined as the start time of the estrus event, and the first recording period after the second boundary point is defined as the end time of the estrus event.

7. The method according to claim 1, characterized in that, In S16, the basic statistics required for calculating the extraction of estrus expression traits include: S161, Based on the determined start and end times of the estrus event, the estrus period and non-estrus baseline period are divided within the 7-day behavioral time series analysis window; wherein, the estrus period is the time interval between the start and end times of the estrus event, and the non-estrus baseline period is the remaining time interval within the 7-day behavioral time series analysis window excluding the estrus period. S162, calculate the basic statistics for the estrus period and the baseline non-estrus period, including the maximum activity level during the estrus period, the average activity level during the baseline non-estrus period, the standard deviation of the activity level during the baseline non-estrus period, the average rumination time during the baseline non-estrus period, and the standard deviation of the rumination time during the baseline non-estrus period.

8. The method according to claim 1, characterized in that, In S17, the estrus expression traits include: estrus duration, peak activity level, estrus intensity, and rumination variation; wherein, estrus duration is the time interval between the end time of the estrus event and the start time of the estrus event; peak activity level represents the standardized deviation of the maximum activity level index during estrus relative to the baseline of non-estrus activity level; estrus intensity represents the standardized cumulative increase of the activity level index during estrus relative to the baseline of non-estrus activity level; and rumination variation represents the standardized cumulative deviation of the rumination time during estrus relative to the baseline of non-estrus rumination.

9. A system for identifying estrus events and extracting estrus expression traits in dairy cows based on behavioral time series change point detection, characterized in that, include: The data acquisition module is configured to acquire time-series data on dairy cow behavior, including cow identification, collection time, activity level indicators, and rumination time. The analysis window construction module is configured to determine the time range to be analyzed and construct a behavioral time series analysis window. The change point detection module is configured to perform change point detection on the behavioral time series based on the constructed behavioral time series analysis window, and obtain candidate change points; The stratified post-processing module is configured to perform stratified post-processing on candidate variable points and determine the effective estrus event interval; The estrus boundary determination module is configured to determine the start time and end time of the estrus event based on the determined valid estrus event interval. The basic statistics calculation module is configured to calculate the basic statistics required for extracting estrus expression traits based on the determined start and end times of the estrus event. The estrus expression trait extraction module is configured to extract estrus expression traits based on the calculated basic statistics; The results output module is configured to output estrus event recognition status information and estrus expression trait extraction results.

10. A computing device, characterized in that, include: Memory, used to store one or more programs; One or more processors are configured to execute the one or more programs to implement the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 8.

Images