Poultry behavior abnormality detection method based on intelligent monitoring system

Abstract

The present invention relates to the field of image recognition technology, and in particular to a method for detecting abnormal poultry activity behaviors based on an intelligent monitoring system. The method determines the energy deficiency of chickens based on the aggregation and travel distance of chickens in the chicken house monitoring video; determines the morphological abnormality of chickens based on the abnormal morphological changes of the chicken outlines; determines the number of times chickens are pecked on the anus based on the head-tail contact moments and movement speeds of different chickens; determines the disease risk of chickens based on the abnormal morphological changes and the number of times chickens are pecked on the anus; determines the gregarious tendency of chickens based on the stability of the gathering group in which the chickens are gathered; determines the hidden danger manifestation degree of chickens based on the energy deficiency, disease risk and gregarious tendency of chickens, and provides abnormal warnings for the behavior of chickens. The present invention combines the manifestation of chicken diseases with the characteristics of natural gregariousness for analysis to obtain more accurate abnormal behavior detection results of chickens.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a method for detecting abnormal behavior of poultry activities based on an intelligent monitoring system. Background Art

[0002] When farmed chickens become sick, they often experience severe symptoms such as dehydration and anemia. These symptoms can cause an imbalance in body temperature, which in turn causes sick chickens to instinctively move closer to high heat sources due to a drop in body temperature, forming a clustering phenomenon. Therefore, traditional intelligent monitoring systems analyze the clustering behavior of farmed chickens to determine the warning level of their disease manifestations.

[0003] In actual chicken farming scenarios, since chickens themselves have certain social and gregarious behaviors, they live in the form of multiple small groups according to the social structure. This natural gregarious aggregation characteristic in the chicken flock is easily misjudged as a symptom when traditionally reflecting the hidden dangers of chicken disease based on the aggregation behavior of farmed chickens, thereby reducing the detection accuracy of abnormal activities and behaviors of farmed chickens. Summary of the Invention

[0004] In order to solve the technical problem that the traditional method of reflecting chicken disease risks based on the aggregation behavior of farmed chickens is easily misjudged as a symptom, the purpose of the present invention is to provide a method for detecting abnormal poultry activity behavior based on an intelligent monitoring system. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for detecting abnormal poultry activity behavior based on an intelligent monitoring system, the method comprising:

[0006] Obtain chicken house surveillance video and corresponding surveillance images;

[0007] Determine the energy deficiency of each chicken based on the chicken flock gathering situation and the chicken's movement distance in the chicken house monitoring video;

[0008] The abnormal morphological changes of the chickens in the monitoring images at different times are used to determine the abnormality of the chickens' morphology. The number of times the chickens were pecked at the anus is determined based on the head-tail contact time and the movement speed of the chickens. The disease risk of the chickens is determined by combining the abnormal morphological changes of the chickens and the number of times they were pecked at the anus.

[0009] Determine the flocking tendency of the chickens based on the stability of the flocking group where the chickens gather;

[0010] The chicken's hidden danger manifestation degree is determined based on the chicken's energy deficiency degree, disease risk degree, and flocking tendency degree; and an abnormal warning is issued for the chicken's behavioral activities based on the chicken's hidden danger manifestation degree.

[0011] Furthermore, the energy deficiency of each chicken is determined based on the chicken flocking situation and the chicken travel distance in the chicken house monitoring video, including:

[0012] Determine the aggregation tendency of each chicken based on the aggregation of chickens in the chicken house monitoring video;

[0013] The energy deficiency of each chicken is determined by combining the travel distance and aggregation tendency of each chicken in the chicken house monitoring video.

[0014] Furthermore, determining the degree of abnormality of the chicken's morphology based on abnormal changes in the morphology of the chicken's outline in the monitoring images of the chicken at different times includes:

[0015] Extract the head vector and tail vector of each chicken in the chicken silhouette;

[0016] The degree of morphological abnormality of the chicken when it has a hunched head and arched back is determined based on the angle formed by the head vector and the tail vector.

[0017] Furthermore, the method of determining the number of times a chicken is pecked in the anus based on the head-tail contact moments of different chickens and the movement speeds of the chickens includes:

[0018] Taking any chicken as the target chicken, determine the number of contact interactions when the tail vector of the target chicken comes into contact with the head vector of other chickens;

[0019] Obtain the angle corresponding to the head vector of the target chicken in two adjacent frames of monitoring images after contact occurs, as well as the moving speed of the target chicken after contact occurs;

[0020] Determining the chicken's anal pecking performance based on the angle and the movement rate;

[0021] The number of times the anal pecking performance level in the contact interaction number is greater than the preset normal range is regarded as the number of times the chicken has been anally pecked.

[0022] Furthermore, the determination of the disease risk of the chickens based on abnormal morphological changes of the chickens and the number of anal pecking behaviors includes:

[0023] The disease risk of the chickens is determined based on abnormal morphological changes, the number of anal pecking incidents, and the severity of anal pecking.

[0024] Furthermore, the determination of the chickens' gregariousness tendency based on the stability of the aggregation group in which the chickens gather comprises:

[0025] Cluster the surveillance images of the chicken house surveillance video to determine the cluster where each chicken is located;

[0026] Take any chicken as the target chicken and obtain the number of times other chickens are in the same cluster with the target chicken as the number of co-existences;

[0027] The other birds that have been in the same household with the target bird for more than the preset stability threshold are regarded as stable birds of the target bird;

[0028] The mean value of the ratio of stable chickens to unstable chickens in the cluster where the target chicken is located in different monitoring images is obtained as the gregarious tendency degree of the target chicken.

[0029] Furthermore, the chicken's hidden danger manifestation degree is determined by combining the chicken's energy deficiency degree, disease risk degree, and flocking tendency degree, including:

[0030] Determining the disease compliance of the chicken based on the energy deficiency degree and the disease risk degree;

[0031] The hidden danger manifestation degree of the chickens is determined by combining the symptom compliance and the flocking tendency of the chickens.

[0032] Furthermore, the abnormal early warning of the chicken's behavior activities based on the chicken's hidden danger manifestation degree includes:

[0033] Calculate the normalized value of the difference between the chicken's hidden danger performance level and the preset hidden danger threshold as the activity warning level; when the activity warning level is within the preset normal range, the chicken is judged to be normal;

[0034] When the activity warning level is within the preset abnormal range, it is determined that the chicken has abnormal activity.

[0035] Furthermore, the method of determining the aggregation tendency of each chicken based on the aggregation of the chickens in the chicken house monitoring video includes:

[0036] Cluster the chickens in each frame of the monitoring image in the chicken house monitoring video to obtain multiple clusters;

[0037] Taking any chicken as the target chicken, the number of chickens in the cluster where the target chicken is located in each frame of the monitoring image is used as the first tendency of the target chicken in each frame of the monitoring image;

[0038] Calculating the negative correlation mapping value of the distance between the target chicken and the cluster center in each frame of the monitoring image as the second tendency of the target chicken in each frame of the monitoring image;

[0039] Calculate the product of the first tendency and the second tendency of the target chicken in each frame of the monitoring image as the single-frame tendency;

[0040] The average of the single-frame tendencies of the target chickens in all monitoring images of the chicken house monitoring video is used as the aggregation tendency of the target chickens.

[0041] Furthermore, the step of determining the energy deficiency of each chicken by combining the travel distance and aggregation tendency of each chicken in the chicken house monitoring video includes:

[0042] The ratio of each chicken's aggregation tendency and its travel distance is used as the energy deficiency of each chicken.

[0043] In a second aspect, a poultry activity behavior abnormality detection system based on an intelligent monitoring system is provided, the system comprising the following modules:

[0044] Monitoring data acquisition module, used to obtain chicken house monitoring video and corresponding monitoring images;

[0045] The energy analysis module is used to determine the energy deficiency of each chicken based on the chicken flocking situation and the chicken's movement distance in the chicken house monitoring video;

[0046] The morphological analysis module is used to determine the degree of abnormal morphology of the chicken based on abnormal morphological changes of the chicken's outline in the monitoring images of the chicken at different times; determine the number of times the chicken was pecked at the anus based on the time of head-tail contact and the movement speed of the chicken; and determine the disease risk of the chicken based on the abnormal morphological changes of the chicken and the number of times the chicken was pecked at the anus;

[0047] The flocking analysis module is used to determine the flocking tendency of chickens based on the stability of the flocking groups in which the chickens gather;

[0048] The abnormality detection module is used to determine the hidden danger manifestation degree of the chickens based on the energy deficiency degree, disease risk degree and flocking tendency degree of the chickens; and to issue abnormal warnings for the chickens' behavioral activities based on the hidden danger manifestation degree of the chickens.

[0049] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores executable code, and when the processor executes the executable code, it implements the various possible implementations of the first aspect.

[0050] In a fourth aspect, an embodiment of the present invention provides a computer program product, which includes: computer program code, which, when running on a computer, enables the computer to execute the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0051] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed in a computer, the computer is caused to execute various possible implementations of the first aspect.

[0052] The embodiments of the present invention have at least the following beneficial effects:

[0053] The present invention first evaluates the aggregation of chickens according to the flocking situation of chickens, and then characterizes the behavioral activity of chickens in combination with the chickens' behavior paths, determines the energy deficiency of chickens through the aggregation situation and the chickens' behavior paths, and then obtains the disease risk according to the impact of symptoms on the morphological posture and anal pecking behavior of chickens. At the same time, the stability relationship of the members of the flock gathered in the process of the chickens' aggregation tendency is analyzed to obtain the flocking tendency degree that characterizes the natural gregarious tendency degree of the chickens, and obtains the hidden danger manifestation degree of the chickens through the flocking tendency degree, energy deficiency degree and disease risk degree, and then realizes abnormal early warning of the behavioral activities of each chicken based on the hidden danger manifestation degree of multiple chickens. Compared with the traditional method of directly analyzing the aggregation of chicken flocks, the present invention can further combine the distinguishing characteristics of chicken disease manifestations and natural gregarious manifestations for analysis, and obtain more accurate abnormal behavior detection results of chicken flocks based on the intelligent monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 This is a flow chart of a method for detecting abnormal poultry activity behavior based on an intelligent monitoring system provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0056] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail the specific implementation method, structure, characteristics and effects of the abnormal poultry activity behavior detection method based on the intelligent monitoring system proposed by the present invention.

[0057] In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0058] In the description of the embodiments of the present invention, unless otherwise specified, " / " means or, for example, A / B can mean A or B: "and / or" in the text is only a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present invention, "multiple" refers to two or more than two.

[0059] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to imply or suggest relative importance or implicitly indicate the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0060] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0061] The embodiments of the present invention are described below with reference to the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided by the embodiments of the present invention are also applicable to similar technical problems.

[0062] The embodiment of the present invention provides a specific implementation method of a method for detecting abnormal poultry activity behaviors based on an intelligent monitoring system, which is applicable to chicken farming scenarios. When farmed chickens become sick, they will experience dehydration and anemia, which will in turn cause an imbalance in body temperature. At this time, the sick chickens will instinctively move closer to the flocks of chickens with higher aggregation to obtain the missing heat in their bodies. The intelligent monitoring system in the chicken house is used to judge the symptoms of each farmed chicken. When the chicken house intelligent monitoring system analyzes the abnormal symptom behavior of chickens, the traditional method is to judge the symptom warning level of the chickens based on the chickens' aggregation tendency. However, due to the natural gregariousness of the chickens themselves, analysis errors will be caused in the judgment of the chickens' symptoms. Therefore, the present invention combines the detailed distinguishing features of the natural gregariousness and symptom aggregation of chickens in actual scenarios for analysis to obtain abnormal chicken activity behavior detection results with higher detection accuracy.

[0063] The specific scheme of the method for detecting abnormal poultry activity behaviors based on the intelligent monitoring system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0064] See also Figure 1 , which shows a flowchart of a method for detecting abnormal poultry activity behavior based on an intelligent monitoring system according to an embodiment of the present invention, the method comprises the following steps:

[0065] Step S100: Acquire chicken house surveillance video and corresponding surveillance images.

[0066] First, the chickens are divided into different chicken houses according to their birth batches. Each chicken house is equipped with intelligent monitoring equipment to collect monitoring videos of the chicken houses.

[0067] A "label" with a different value is hung on the neck of each chicken, and the label value is different for different chickens.

[0068] The chicken house monitoring video of the chicken flock is collected through intelligent monitoring equipment, and the chicken house monitoring video should be able to cover all areas in the chicken house that may be involved by the chicken flock.

[0069] Extract the chicken house monitoring video of the chicken flock within a day and perform denoising preprocessing on the chicken house monitoring video.

[0070] Finally, the preprocessed chicken house monitoring video is uploaded to the data acquisition system for subsequent time domain analysis. The chicken house monitoring video is frame-processed to obtain the monitoring image corresponding to the chicken house monitoring video. Each monitoring image corresponds to a different acquisition time.

[0071] Step S200: Determine the energy deficiency of each chicken based on the chicken flocking situation and the chicken travel distance in the chicken house monitoring video.

[0072] Considering that the disease can cause dehydration and anemia in chickens, causing the chickens' body temperature to be unbalanced, the sick chickens will need to move closer to the chicken group to obtain heat. Therefore, the tendency of chickens to gather in the chicken group can reflect the hidden dangers of the chickens' illness to a certain extent. Therefore, this step first analyzes the aggregation of chickens in the chicken house and determines the aggregation tendency of each chicken based on the aggregation of chickens in the chicken house monitoring video.

[0073] First, by performing YOLO target recognition on the label numbers of the chickens in the chicken house surveillance video, the chickens corresponding to each label are identified, and then the position coordinates of the chickens corresponding to each label within a day are obtained through YOLO target tracking technology.

[0074] In one day, an analysis time is determined every 5 minutes, and the total number of extracted The analysis time is 5 minutes, of which 5 minutes is an empirical value and can be adjusted according to the actual scenario. This analysis time is used to analyze the clustering of clusters in the chicken house surveillance video.

[0075] The chickens in each frame of the chicken house surveillance video are clustered to obtain multiple clusters. More specifically, for a single analysis moment, a two-dimensional chicken flock scatter plot is constructed. The position coordinates of each chicken at that moment are converted to the two-dimensional chicken flock scatter plot, and the position of each chicken sample point is then mapped to the two-dimensional chicken flock scatter plot.

[0076] Taking the two-dimensional chicken flock scatter plot of the chicken house corresponding to the i-th analysis time as an example, DBSCAN density clustering is performed on each chicken sample point in the two-dimensional chicken flock scatter plot to obtain multiple density clusters. In the subsequent steps, density clusters are referred to as clusters. The neighborhood radius for clustering is selected as 2m, and the minimum number of sample points is selected as 8. The parameters can be adjusted by the implementer according to the specific scenario.

[0077] Arrange the chickens from small to large according to the label value, and record the chicken ranked in the kth position as the chicken with label k, that is, the kth chicken is the chicken with label k.

[0078] For the chicken with label k in the two-dimensional chicken scatter plot at the i-th analysis moment, obtain the distance between the chicken and the center of the cluster to which it belongs , and at the same time obtain the number of all chicken sample points in the density cluster to which the chicken belongs .

[0079] Taking the chicken with label k as an example, if the number of all chicken sample points in the density cluster to which the chicken belongs is The more, the stronger the group performance of the chicken group that the chicken tends to be, and if the distance between the chicken and the center of the density cluster is The closer it is, the greater the flock tendency of the chicken with label k at the i-th analysis moment, and then combined with all The aggregation tendency of the chickens with label k is obtained at the analysis moment .

[0080] Taking any chicken as the target chicken, the number of chickens in the cluster where the target chicken is located in each frame of the monitoring image is used as the first tendency of the target chicken in each frame of the monitoring image; the negative correlation mapping value of the distance between the target chicken and the cluster center in each frame of the monitoring image is calculated as the second tendency of the target chicken in each frame of the monitoring image; the product of the first tendency and the second tendency of the target chicken in each frame of the monitoring image is calculated as the single-frame tendency; the average of the single-frame tendencies of the target chicken in all monitoring images of the chicken house monitoring video is used as the aggregation tendency of the target chicken.

[0081] In some embodiments, the aggregation tendency of the chickens with label k is The calculation formula of the aggregation tendency is: ; Where T is the number of surveillance images corresponding to the analysis time.

[0082] Finally, the aggregation tendency of all the chickens corresponding to the tags in the chicken house is calculated and recorded, that is, the aggregation tendency of each chicken is obtained.

[0083] Furthermore, the distinguishing characteristics between natural gregariousness and disease-induced aggregation in chickens are analyzed to determine the disease risk manifestation degree of each bird. Considering that chickens naturally gregarious, directly using the tendency of chickens to gregariously reflect disease risk can introduce certain judgment errors. Therefore, this step is followed by analyzing the detailed differences between the energy, posture, and behavior of the diseased chickens and normal chickens to determine the disease conformity. Simultaneously, the stability of the members of the flocks in which the chickens tend to gregariously gather is used to determine the natural gregariousness tendency of each bird. The natural gregariousness tendency is then combined with the disease manifestation conformity analysis to determine the disease risk manifestation degree reflecting the disease risk of each bird.

[0084] First, the YOLO target tracking technology is used to obtain the travel distance data of the chickens corresponding to each label in one day. The travel distance of the chicken with label k in the history of one day is recorded as .

[0085] Because illness can make chickens weak and lack energy, which in turn leads to a decrease in the chickens' behavioral activity, which is reflected in a reduction in the distance they walk. Therefore, the lower the distance a sick individual chicken walks compared to the average level of the flock, and the higher the chickens' tendency to gather, the more energy-deficient the chickens are. Combined with the travel distance and gathering tendency of each chicken in the chicken house monitoring video, the energy deficiency of each chicken can be determined.

[0086] The energy deficiency performance of the chicken labeled k : .

[0087] Calculate the level of energy starvation displayed by all birds in the house.

[0088] Step S300: Determine the degree of abnormal morphology of the chicken based on abnormal morphological changes of the chicken's outline in the monitoring images of the chicken at different times; determine the number of times the chicken was pecked on the anus based on the different head-tail contact moments and the chicken's movement speed; and determine the chicken's disease risk based on the abnormal morphological changes of the chicken and the number of times the chicken was pecked on the anus.

[0089] The body shape and appearance of sick chickens can reflect certain symptoms of their disease. For example, sick chickens are weaker than normal chickens, so they tend to show hunched backs and hunched heads. Based on the abnormal changes in the outline of the chickens in the monitoring images at different times, the degree of abnormality of the chickens' morphology is determined:

[0090] First, for a single chicken, 10 side view image frames of the chicken are randomly selected from the monitoring video.

[0091] For a single side view image frame of a single chicken, the trained YOLO target recognition model is used to identify the outline area of ​​the chicken in the side view image frame, recorded as the chicken outline, and the head area and tail area of ​​the chicken are identified.

[0092] The closed area centroid extraction technology is used to obtain the overall centroid of the chicken in the corresponding contour area, the centroid of the head area, and the centroid of the tail area.

[0093] Taking the contour area corresponding to the j-th side view image frame of the chicken with label k as an example, the vector from the overall center of mass of the chicken contour area to the center of mass of the head is recorded as the head vector, and the vector from the overall center of mass of the chicken contour area to the center of mass of the tail is recorded as the tail vector.

[0094] The degree of abnormality of the chicken's morphology when it has a hunched head and arched back is determined based on the angle formed by the head vector and the tail vector.

[0095] Get the angle between the chicken's head vector and tail vector When the chicken's morphology tends to be more hunched and its back is more arched, the chicken's head and back are both offset toward the center of the body, and the angle between the head vector and the tail vector is The smaller it is, and the more the chicken shows the characteristics of shrinking head and arching back, the greater the abnormal morphology of the chicken is. Calculate the abnormal morphology of the chicken with label k in the jth side image frame. : .

[0096] Calculate and record morphological abnormalities for all birds.

[0097] As chickens gather together, their excrement is distributed in one area. The fermentation of feces leads to an increase in ammonia concentration. Ammonia stimulates chickens to peck their anus, thereby increasing the risk of cross-infection among sick chickens. Therefore, it is necessary to analyze the anus pecking behavior of chickens based on the degree of morphological abnormality, and then obtain the disease risk of each chicken. :

[0098] First, for a single chicken, obtain the head vector and tail vector of the chicken at each historical moment in a day.

[0099] The time frame image corresponding to the moment when the end of the chicken's tail vector contacts and interacts with the end of the head vector of another chicken is extracted, and this part of the time frame image is defined as the interaction time frame image of the chicken.

[0100] When two chickens engage in anal pecking, not only does there exist an interaction between the head vector and the tail vector, but the front chicken will also quickly move forward due to the pain after being pecked in the anus to create distance from the rear chicken. At this time, the front chicken will quickly adjust the tilt angle of its head to enter a fast running state.

[0101] Taking any chicken as the target chicken, the number of contact interactions in which the tail vector of the target chicken comes into contact with the head vector of another chicken is determined; the angle corresponding to the head vector of the target chicken in two adjacent frames of monitoring images after the contact occurs and the movement rate of the target chicken after the contact occur are obtained; the chicken's anal pecking performance degree is determined by combining the angle and the movement rate; the number of times the anal pecking performance degree in the number of contact interactions is greater than a preset normal range is regarded as the number of times the chicken has been pecked on the anus.

[0102] Taking the r-th interactive time frame image of a chicken with label k as an example, the angle formed by the head vector of the chicken in the r-th interactive time frame image and the head vector in the next time frame image adjacent to the r-th interactive time frame is obtained respectively. The larger the angle, the greater the performance of the chicken in adjusting the head tilt angle to enter the fast running state after the interaction.

[0103] Furthermore, for the chickens after the interaction, the YOLO target tracking technology is used to obtain the distance traveled by the chickens within 2 seconds after the interaction. The distance traveled is compared with the time period of 2 seconds to obtain the movement rate of the chickens after the interaction. The movement rate of the chicken with the label k in the rth interaction time frame after the interaction is recorded as The faster the movement speed, the greater the tendency of the front chicken and the back chicken to distance themselves from each other after the interaction.

[0104] Therefore The bigger, The faster it is, the higher the degree of anal pecking by other chickens occurs in the rth interaction time frame for the chicken with label k. So let’s calculate the anal pecking degree of the chicken with label k in the rth interaction time frame. : ;norm is the normalization function. Get the mean of the pissing performance of the chicken with label k in all interaction time frames as the pissing performance of the chicken .

[0105] The present invention selects a preset normal range of [0, 0.8]. If the corresponding anal pecking behavior degree of a chicken in a certain interaction time frame is higher than 0.8, it is considered that the chicken is anally pecked at the interaction moment.

[0106] Then extract the number of times the chicken with label k was pecked in the past day .

[0107] The disease risk of the chickens is determined based on abnormal morphological changes, the number of anal pecking incidents, and the severity of anal pecking.

[0108] If a chicken has a higher morphological abnormality and there is more anal pecking between this chicken and other chickens with higher morphological abnormalities, the disease risk of this chicken is higher. So let's calculate the disease risk of the chicken with label k. : . Calculate and record the disease risk of each chicken.

[0109] Step S400: determining the flocking tendency of the chickens based on the stability of the flocking group in which the chickens gather.

[0110] When chickens gather in groups naturally, they form a stable hierarchy by foraging and roosting together to maintain group relationships. When chickens gather due to illness, it is more likely that the temperature discomfort in the body prompts sick chickens to temporarily gather in the group. Therefore, the stability of the chickens' tendency to flock together is analyzed to obtain the degree of flocking tendency of each chicken. The gregariousness degree reflects the natural gregariousness tendency of the chickens.

[0111] First, for a single chicken, a total of Density clustering results at the analysis time.

[0112] All chickens with label k Taking the density clustering results of k as an example, we extract the label types of all chickens in all density clustering results, and then extract the frequency of occurrence of each label type in all density clustering results. This can also be understood as taking the chicken with label k as the target chicken and obtaining the number of times other chickens are in the same cluster as the target chicken at different analysis times as the number of co-occurrences.

[0113] Other birds that have been cohabited with the target bird more than the preset stability threshold are regarded as stable birds of the target bird.

[0114] In the embodiment of the present invention, the preset stability threshold is 30, that is, if the frequency of occurrence of a chicken with a certain label in all clustering results corresponding to the chickens with label k is higher than 30, the chicken with the label type is determined to be a stable member of the chickens with label k and recorded as a stable chicken. Then, all the chickens appearing in the clustering results are divided into stable chickens and unstable chickens of the chickens with label k. The number of stable chickens is recorded as , the number of unstable chickens is .

[0115] The normalized value of the mean of the proportion of stable chickens to unstable chickens in the cluster where the target chicken is located in different monitoring images is obtained as the gregariousness tendency of the target chicken.

[0116] Among all the clustering results corresponding to the chickens labeled k, if there are more stable chickens and fewer unstable chickens, it means that the chickens labeled k tend to have smaller changes in group membership, which means that the chickens are more likely to be selected due to natural gregarious relationships.

[0117] Calculate and record the social tendency of all birds.

[0118] Step S500 , determining the chicken's hidden danger manifestation degree based on the chicken's energy deficiency degree, disease risk degree, and flocking tendency degree; and issuing an abnormal warning for the chicken's behavior activities based on the chicken's hidden danger manifestation degree.

[0119] Furthermore, if a chicken's energy deficiency score is higher and its disease risk score is higher, then the chicken's disease compliance score is higher. Therefore, the disease compliance score of the chicken is determined by combining the energy deficiency score and the disease risk score. In some embodiments, the normalized value of the product of the energy deficiency score and the disease risk score is used as the disease compliance score of the chicken.

[0120] Obtain the symptom compliance of each chicken.

[0121] For an individual chicken, a higher symptom compliance score and a lower gregariousness tendency score, which represents a natural gregarious tendency, indicate a higher risk of symptom risk. Therefore, the symptom compliance score and gregariousness tendency score are combined to determine the hazard manifestation level of the chicken. In some embodiments, a normalized value of the ratio of the symptom compliance score to the gregariousness tendency score is calculated as the hazard manifestation level of the chicken.

[0122] Calculate and record the severity of symptoms for all birds.

[0123] Finally, based on the degree of hidden danger performance of the chickens, abnormal warnings are issued for the chickens' behavioral activities, which is also an abnormal warning for the chickens' health status.

[0124] Calculate the normalized value of the difference between the chicken's hidden danger performance level and the preset hidden danger threshold as the activity warning level; when the activity warning level is within the preset normal range, the chicken is judged to be normal;

[0125] When the activity warning level is within the preset abnormal range, it is determined that the chicken has abnormal activity.

[0126] In the embodiment of the present invention, the preset hidden danger threshold value is the average value of the hidden danger manifestation degrees of all chickens.

[0127] If the level of hidden danger manifestation of a single chicken's disease is higher than the average level of hidden danger manifestation of all chickens, the behavioral activities of the chicken should be given a higher level of warning.

[0128] In the embodiment of the present invention, the preset normal range is [0, 0.45], and the preset abnormal range is (0.45, 1].

[0129] As a preferred embodiment of the present invention, the warning level of the chicken's behavior and activity is divided into warning levels. If the warning level of the chicken's behavior and activity is between (0.88, 1], a level 1 health warning is issued for the chicken.

[0130] If the warning level of the chicken's behavior activity is between (0.45, 0.88], a level 2 warning will be issued for the chicken.

[0131] Finally, doctors set up different patient screening strategies based on the different warning levels of chickens, thereby completing the entire process of detecting abnormal poultry activity behaviors based on the intelligent monitoring system.

[0132] An embodiment of the present invention provides a poultry activity behavior abnormality detection system based on an intelligent monitoring system, the system comprising:

[0133] Monitoring data acquisition module, used to obtain chicken house monitoring video and corresponding monitoring images;

[0134] The energy analysis module is used to determine the energy deficiency of each chicken based on the chicken flocking situation and the chicken's movement distance in the chicken house monitoring video;

[0135] The morphological analysis module is used to determine the degree of abnormal morphology of the chicken based on abnormal morphological changes of the chicken's outline in the monitoring images of the chicken at different times; determine the number of times the chicken was pecked at the anus based on the time of head-tail contact and the movement speed of the chicken; and determine the disease risk of the chicken based on the abnormal morphological changes of the chicken and the number of times the chicken was pecked at the anus;

[0136] The flocking analysis module is used to determine the flocking tendency of chickens based on the stability of the flocking groups in which the chickens gather;

[0137] The abnormality detection module is used to determine the hidden danger manifestation degree of the chickens based on the energy deficiency degree, disease risk degree and flocking tendency degree of the chickens; and to issue abnormal warnings for the chickens' behavioral activities based on the hidden danger manifestation degree of the chickens.

[0138] Optionally, the transmission medium may be a wired link, such as but not limited to coaxial cable, optical fiber, and digital subscriber line, or a wireless link, such as but not limited to Wireless Fidelity (WIFI), Bluetooth, and mobile device network.

[0139] It should be noted that the device provided in the above embodiment is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above.

[0140] A schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Exemplarily, the computer device includes: a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the computer device can perform any of the aforementioned methods for detecting abnormal poultry behavior based on an intelligent monitoring system.

[0141] In addition, an embodiment of the present invention also protects a device, which may include a memory and a processor, wherein the memory stores executable program code, and the processor is used to call and execute the executable program code to perform the poultry activity behavior abnormality detection method based on the intelligent monitoring system provided by an embodiment of the present invention.

[0142] In embodiments of the present invention, the device may be divided into functional modules based on the above-described method examples. For example, these modules may correspond to individual functional modules, or two or more functions may be integrated into a single processing module. The integrated modules may be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and represents only a logical functional division. In actual implementation, other division methods may be employed.

[0143] In the case of dividing each module into modules corresponding to each function, the device may further include a signal uploading module, a determination module, an adjustment module, etc. It should be noted that all relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module and will not be repeated here.

[0144] It should be understood that the device provided in the embodiment of the present invention is used to execute the above-mentioned method for detecting abnormal poultry activity behavior based on the intelligent monitoring system, and thus can achieve the same effect as the above-mentioned implementation method.

[0145] When an integrated unit is employed, the device may include a processing module and a storage module. When the device is applied to a device, the processing module can be used to control and manage the device's operations. The storage module can be used to support the device in executing program code, etc. The processing module can be a processor or controller that can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the present disclosure. The processor can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processor (DSP) and a microprocessor, etc. The storage module can be a memory.

[0146] In addition, the device provided in the embodiment of the present invention can be specifically a chip, component or module, and the chip may include a connected processor and memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute the poultry activity behavior abnormality detection method based on the intelligent monitoring system provided in the above embodiment.

[0147] An embodiment of the present invention also provides a computer-readable storage medium, which stores computer program code. When the computer program code is run on a computer, the computer executes the above-mentioned related method steps to implement the poultry activity behavior abnormality detection method based on the intelligent monitoring system provided by the above embodiment.

[0148] An embodiment of the present invention further provides a computer program product. When the computer program product is run on a computer, the computer is caused to execute the above-mentioned related steps to implement the method for detecting abnormal poultry activity behavior based on an intelligent monitoring system provided in the above embodiment.

[0149] Among them, the device, computer-readable storage medium, computer program product or chip provided in the embodiments of the present invention are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here. Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In the embodiments provided by the present invention, it should be understood that the disclosed device and method can be implemented in other ways.

[0150] The device embodiments described above are merely illustrative. For example, the division into modules or units represents only one logical functional division. Actual implementations may employ alternative divisions, such as combining or integrating multiple units or components into another device, or omitting or disabling certain features. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through an interface, or indirect coupling or communication connection between devices or units may be electrical, mechanical, or otherwise.

[0151] It should also be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or terminal device comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or terminal device comprising the element.

[0152] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0153] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0154] The above content is only a specific implementation method of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for detecting abnormal poultry behavior based on an intelligent monitoring system, characterized in that: The method comprises the following steps: Obtain chicken house surveillance video and corresponding surveillance images; The energy deficiency of each chicken is determined according to the aggregation of chickens and the travel distance of the chickens in the chicken house monitoring video; wherein, the method for obtaining the energy deficiency is: clustering the chickens in each frame of the monitoring image in the chicken house monitoring video to obtain multiple clusters; taking any chicken as the target chicken, and taking the number of chickens in the cluster where the target chicken is located in each frame of the monitoring image as the first tendency of the target chicken in each frame of the monitoring image; calculating the negative correlation mapping value of the distance between the target chicken and the cluster center in each frame of the monitoring image as the second tendency of the target chicken in each frame of the monitoring image; calculating the product of the first tendency and the second tendency of the target chicken in each frame of the monitoring image as the single-frame tendency; taking the average of the single-frame tendencies of the target chicken in all monitoring images of the chicken house monitoring video as the aggregation tendency of the target chicken; combining the travel distance and aggregation tendency of each chicken in the chicken house monitoring video to determine the energy deficiency of each chicken; The abnormal morphological changes of the chickens in the monitoring images at different times are used to determine the abnormality of the chickens' morphology. The number of times the chickens were pecked at the anus is determined based on the head-tail contact time and the movement speed of the chickens. The disease risk of the chickens is determined by combining the abnormality of the chickens' morphology and the number of times they were pecked at the anus. Determine the flocking tendency of the chickens based on the stability of the flocking group where the chickens gather; The chicken's hidden danger manifestation degree is determined based on the chicken's energy deficiency degree, disease risk degree, and flocking tendency degree; and an abnormal warning is issued for the chicken's behavioral activities based on the chicken's hidden danger manifestation degree.

2. The method for detecting abnormal poultry activity behavior based on an intelligent monitoring system according to claim 1, characterized in that: The step of determining the abnormality of the chicken's morphology based on abnormal changes in the morphology of the chicken's outline in the monitoring images of the chicken at different times includes: Extract the head vector and tail vector of each chicken in the chicken silhouette; The degree of morphological abnormality of the chicken when it has a hunched head and arched back is determined based on the angle formed by the head vector and the tail vector.

3. The method for detecting abnormal poultry behavior based on an intelligent monitoring system according to claim 2, characterized in that: Determining the number of times a chicken is pecked in the anus based on the head-tail contact moments of different chickens and the movement speeds of the chickens includes: Taking any chicken as the target chicken, determine the number of contact interactions when the tail vector of the target chicken comes into contact with the head vector of other chickens; Obtain the angle corresponding to the head vector of the target chicken in two adjacent frames of monitoring images after contact occurs, as well as the moving speed of the target chicken after contact occurs; Determining the chicken's anal pecking performance based on the angle and the movement rate; The number of times the anal pecking performance level in the contact interaction number is greater than the preset normal range is regarded as the number of times the chicken has been anally pecked.

4. The method for detecting abnormal poultry activity behavior based on an intelligent monitoring system according to claim 1, characterized in that: The combination of the abnormality of the chicken's morphology and the number of anal pecking behaviors is used to determine the chicken's disease risk, including: The disease risk of the chickens is determined based on the abnormality of the chickens' morphology, the number of times they have been pecked at the anus, and the severity of the pecks.

5. The method for detecting abnormal poultry behavior based on an intelligent monitoring system according to claim 1, characterized in that: The method of determining the flocking tendency of chickens based on the stability of the flocking group in which the chickens gather includes: Cluster the surveillance images of the chicken house surveillance video to determine the cluster where each chicken is located; Take any chicken as the target chicken and obtain the number of times other chickens are in the same cluster with the target chicken as the number of co-existences; The other birds that have been in the same household with the target bird for more than the preset stability threshold are regarded as stable birds of the target bird; The mean value of the ratio of stable chickens to unstable chickens in the cluster where the target chicken is located in different monitoring images is obtained as the gregarious tendency degree of the target chicken.

6. The method for detecting abnormal poultry behavior based on an intelligent monitoring system according to claim 1, characterized in that: The step of determining the chicken's hidden danger manifestation degree by combining the chicken's energy deficiency degree, disease risk degree, and flocking tendency degree includes: Determining the disease compliance of the chicken based on the energy deficiency degree and the disease risk degree; The hidden danger manifestation degree of the chickens is determined by combining the symptom compliance and the flocking tendency of the chickens.

7. The method for detecting abnormal poultry behavior based on an intelligent monitoring system according to claim 1, characterized in that: According to the degree of hidden danger performance of the chickens, abnormal early warning of the chickens' behavior activities is provided, including: Calculate the normalized value of the difference between the chicken's hidden danger performance level and the preset hidden danger threshold as the activity warning level; when the activity warning level is within the preset normal range, the chicken is judged to be normal; When the activity warning level is within the preset abnormal range, it is determined that the chicken has abnormal activity.

8. The method for detecting abnormal poultry behavior based on an intelligent monitoring system according to claim 1, characterized in that: The method of combining the travel distance and aggregation tendency of each chicken in the chicken house monitoring video to determine the energy deficiency of each chicken includes: The ratio of each chicken's aggregation tendency and its travel distance is used as the energy deficiency of each chicken.

Images