An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement

By collecting heat map data in the animal body temperature intelligent detection system for animal husbandry and veterinary medicine, obtaining temperature contribution values ​​and posture characteristics, and adjusting sampling time intervals, the problems of inaccurate detection and waste of resources in the existing technology are solved, and the animal body temperature detection is improved Accuracy and efficiency.

CN118844954BActive Publication Date: 2024-11-29JIANGSU JINSHENG GOAT BREEDING TECH DEV CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411336627.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-25
Publication Date
2024-11-29
Estimated Expiration
2044-09-25

AI Technical Summary

Technical Problem

In the prior art, the sampling interval for animal body temperature detection is unreasonable, resulting in inaccurate detection, and continuous monitoring consumes a lot of resources, making it difficult to maintain real-time and accuracy.

Method used

Design an intelligent animal temperature detection system for animal husbandry and veterinary medicine based on temperature measurement. Thermal map is collected through the data acquisition module, and the temperature contribution value, change value and difference value are obtained. Combined with attitude characteristics and fluctuation values, the sampling time interval is adjusted to improve detection accuracy. sex.

Benefits of technology

Accurate detection of animal body temperature is achieved, resource waste is reduced, detection efficiency and accuracy is improved, and temperature abnormalities are detected in a timely manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118844954B_ABST
    Figure CN118844954B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of temperature measurement technology, and in particular to an intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement. The system collects a thermogram of a single animal; obtains the overall temperature of the single animal according to the temperature field in the thermogram; obtains the possible value of temperature anomaly, the temperature change value and the temperature difference value according to the overall temperature; obtains the temperature detection value according to the possible value of temperature anomaly, the temperature change value and the temperature difference value, and performs abnormal detection on the animal according to the temperature detection value; obtains posture features according to key points of the body in the thermogram, obtains data accuracy indicators according to the posture features, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value of the posture features, and adjusts the sampling time interval according to the data accuracy index. The present invention collects the temperature information of the animal in a centralized manner by adjusting a reasonable sampling time interval, thereby improving the accuracy and efficiency of abnormal detection on the animal.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and in particular to an intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement. Background Art

[0002] When animals suffer from some diseases, they will be accompanied by symptoms of fever. Therefore, the body temperature of animals becomes an important physiological indicator to measure whether the animals are healthy. The traditional method of measuring animal body temperature is to obtain the temperature of animals by manually contacting the animals. This method is easy to cause animals to overreact, and it requires a lot of manpower and material resources and has low detection efficiency. In the existing method, the temperature of animals is detected by infrared thermal imaging technology. As a non-contact and non-destructive temperature measurement technology, infrared thermal imaging technology has the advantages of high temperature measurement accuracy, wide range, fast and high sensitivity, and has been widely used, and will not cause stress reactions in animals. In the existing method, the temperature of each part of the animal is obtained by infrared thermal imaging technology, and then the temperature of the whole animal is judged abnormally. In the prior art, in order to ensure the real-time nature of abnormal monitoring, it is necessary to continuously monitor the animals, but continuous monitoring will result in a large amount of monitoring costs. If a fixed sampling time interval is set to reduce the monitoring cost, it will be easy to lose information with strong reference due to changes in the posture of the animal, resulting in inaccurate abnormal analysis results. Summary of the invention

[0003] In order to solve the technical problem of inaccurate animal abnormality detection caused by unreasonable sampling time interval setting, the purpose of the present invention is to provide an animal body temperature intelligent detection system for animal husbandry and veterinary medicine based on temperature measurement. The technical solution adopted is as follows:

[0004] An embodiment of the present invention provides an animal body temperature intelligent detection system for animal husbandry and veterinary medicine based on temperature measurement, the system comprising the following:

[0005] A data acquisition module, used to collect a thermal map of a single animal at a preset sampling time interval according to a preset sampling period and a preset sampling frequency; the thermal map contains body key point information;

[0006] A temperature anomaly possible value acquisition module is used to acquire at least two temperature fields in the thermal map according to the temperature distribution, and acquire the temperature contribution value of each temperature field according to the number and category of key points of the body in the temperature field; acquire the overall temperature of the animal according to the average temperature value and temperature contribution value of each temperature field; and acquire the temperature anomaly possible value of the animal according to the overall temperature;

[0007] A temperature change value acquisition module is used to obtain the temperature change value at the time to be detected based on the difference between the overall temperature of a single animal at the time to be detected and other times in the historical data;

[0008] The temperature difference value acquisition module calculates the difference in overall temperature between each individual animal and all other similar animals at the time of detection to obtain the temperature difference value of the corresponding individual animal;

[0009] An anomaly detection module, used to obtain a temperature detection value according to a possible temperature anomaly value, a temperature change value and a temperature difference value of a single animal at the time to be detected, and perform an anomaly detection on the single animal according to the temperature detection value;

[0010] The time interval adjustment module is used to obtain the posture characteristics at each moment according to all the key points of the body in the thermal map, and obtain the data accuracy index of a single animal at each moment according to the posture characteristics at each moment, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value of the posture characteristics in the preset adjustment time period before the corresponding moment, and adjust the sampling time interval of the next sampling time period according to all the data accuracy indicators in each sampling time period.

[0011] Furthermore, the method for obtaining the temperature contribution value includes:

[0012] Different preset marking values ​​are set for different categories of body key points;

[0013] The result of accumulating the preset marking values ​​of all key points of the body in each temperature field is used as the overall marking value of the corresponding temperature field;

[0014] Obtaining the sum of the number of key points of the body in each temperature field and a preset constant as the overall number of the corresponding temperature field;

[0015] The ratio of the overall mark value to the overall quantity of each temperature field is used as the temperature contribution value of the corresponding temperature field.

[0016] Furthermore, the method for obtaining the overall temperature includes:

[0017] The ratio of each temperature contribution value in the thermal diagram to the sum of all temperature contribution values ​​is used as the temperature weight of the corresponding temperature field;

[0018] According to the temperature value of each pixel point in the temperature field, the average temperature value of each temperature field is obtained;

[0019] The product of the temperature weight and the average temperature value of each temperature field is calculated as the thermal value of the corresponding temperature field, and the result of accumulating all the thermal values ​​in the thermal map is taken as the overall temperature.

[0020] Furthermore, the method for obtaining the temperature change value includes:

[0021] taking all overall temperatures in a preset temperature analysis period before the time when the single animal is to be detected as the first historical temperature;

[0022] Acquire the target position in a preset periodic sampling time period at the time to be detected, and acquire the overall temperature corresponding to the target position in a preset number of the periodic sampling time periods as the second historical temperature;

[0023] Calculate the difference between the overall temperature of a single animal at the time to be detected and each first historical temperature as a first difference, and use the result of accumulating the first differences as a first change value at the time to be detected;

[0024] The difference between the overall temperature at each moment and the overall temperature at the previous moment is taken as an adjacent difference; the adjacent difference of the overall temperature at the moment to be detected is obtained as a target adjacent difference, the adjacent difference of each first historical temperature is obtained as a historical adjacent difference, the difference between the target adjacent difference and each of the historical adjacent differences is calculated as a second difference, and the result of accumulating the second differences is taken as a second change value at the moment to be detected;

[0025] Calculate the difference between the overall temperature of a single animal at the time to be detected and each second historical temperature as a third difference, and use the result of accumulating the third differences as a third change value at the time to be detected;

[0026] The temperature change value of a single animal at the time of detection is obtained according to the first change value, the second change value and the third change value, and the first change value, the second change value and the third change value are all positively correlated with the temperature change value.

[0027] Furthermore, the temperature difference value acquisition method includes:

[0028] The result of accumulating the difference in overall temperature between each individual animal and all other similar animals at the time of detection is taken as the temperature difference value of the corresponding individual animal.

[0029] Furthermore, the abnormality detection of a single animal according to the temperature detection value includes:

[0030] The possible value of temperature anomaly, temperature change value and temperature difference value at the time to be detected are all positively correlated with the temperature detection value at the corresponding time;

[0031] A temperature detection value threshold is set. When the temperature detection value is greater than the temperature detection value threshold, it is determined that the corresponding single animal has a temperature abnormality.

[0032] Furthermore, the method for acquiring the posture feature includes:

[0033] The result of adding the preset label values ​​of all body key points in the heat map is used as the posture feature.

[0034] Furthermore, the method for obtaining the data accuracy index includes:

[0035] Setting a preset adjustment time period, and obtaining the variance of the posture feature within the preset adjustment time period as the posture fluctuation value of the animal at the time of detection;

[0036] Obtain the difference between the temperature change value at each moment and the temperature at the previous moment as the adjacent deviation degree at the corresponding moment;

[0037] The product of the posture feature and the possible value of the temperature anomaly at each moment is obtained as the credible possible value of the temperature anomaly at the corresponding moment;

[0038] The data accuracy index at the corresponding moment is obtained according to the credible temperature anomaly possible value, the temperature change value, the adjacent deviation degree and the posture fluctuation value at each moment; the credible temperature anomaly possible value, the temperature change value, the adjacent deviation degree and the posture fluctuation value are all positively correlated with the data accuracy index.

[0039] Further, adjusting the sampling time interval of the next sampling time period according to all data accuracy indicators in each sampling time period includes:

[0040] Calculate the mean of the data accuracy index in each sampling time period as the target data accuracy index, set the data accuracy index interval, when the target data accuracy index is within the data accuracy index interval, the sampling time interval does not need to be adjusted; when the target data accuracy index is less than the minimum value within the data accuracy index interval, increase the sampling time interval by a preset time step; when the data accuracy index is greater than the maximum value within the data accuracy index interval, shorten the sampling time interval by a preset time step.

[0041] Furthermore, the method for obtaining at least two temperature fields in a thermodynamic map according to temperature distribution includes:

[0042] The DBSCAN density clustering algorithm is used to cluster the thermal map according to the temperature distribution to obtain at least two clusters, one cluster being one temperature field.

[0043] The present invention has the following beneficial effects:

[0044] 1. The present invention obtains the posture characteristics at each moment based on the information of the key points of the body of a single animal. The temperature at the key points of the body can accurately reflect the temperature of various parts of the animal. Therefore, the more key points of the body, the more accurate the overall temperature of the animal obtained at the corresponding moment, and then the posture fluctuation value of the animal in a preset time period is obtained according to the posture characteristics, and the posture of the animal in the preset time period is determined, thereby avoiding the interference of different postures on temperature detection; according to the posture characteristics at each moment, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value in the corresponding preset time period, the data accuracy index of the animal at the corresponding moment is obtained, and the time interval for obtaining data is determined according to the data accuracy index, thereby avoiding the waste of resources caused by continuous monitoring of data, and at the same time ensuring the accuracy of the obtained data; according to the collected accurate temperature data, the possible value of temperature anomaly, the temperature change value and the temperature difference value at each moment are obtained, and then the temperature detection value at the corresponding moment is obtained, and the animals that may have temperature abnormalities are accurately and efficiently determined, and more detailed manual detection is performed on the animals that may have temperature abnormalities, thereby improving the efficiency of abnormal detection of animals.

[0045] 2. The temperature contribution value of each temperature field is obtained according to the key points of the body set in the temperature field, and the temperature proportion of each temperature field is determined, so that the overall temperature obtained according to the average temperature value and the temperature contribution value of the temperature field is more consistent with the actual temperature of the animal; that is, the possible value of temperature anomaly, temperature change value and temperature difference value obtained according to the overall temperature have a stronger reference value, which can make the final abnormality detection result more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A structural block diagram of an intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, characteristics and effects of an intelligent temperature detection system for animal husbandry and veterinary medicine based on temperature measurement proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0049] 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.

[0050] The specific scheme of the intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement provided by the present invention is described in detail below with reference to the accompanying drawings.

[0051] See also Figure 1 , which shows a structural block diagram of an animal body temperature intelligent detection system for animal husbandry and veterinary medicine based on temperature measurement provided by an embodiment of the present invention. The system includes: a data acquisition module 10, a temperature abnormality possible value acquisition module 20, a temperature change value acquisition module 30, a temperature difference value acquisition module 40, an abnormality detection module 50, and a time interval adjustment module 60.

[0052] The data acquisition module 10 is used to collect the thermal map of a single animal at a preset sampling time interval according to a preset sampling period and a preset sampling frequency; the thermal map contains body key point information.

[0053] Specifically, all the embodiments of the present invention take pigs as an example, analyze the temperature of the pigs, realize abnormal detection of the pigs, and indirectly achieve the purpose of epidemic prevention in the pig breeding process.

[0054] During the diagnosis and treatment of pigs, experienced veterinary workers can determine whether the pig has a fever by touching the base of the pig's ear, the tip of the nose, the skin temperature, and the chest fossa behind the left elbow. However, direct contact can easily induce a stress response in the pig, and the temperature of the pig is greatly affected by human subjectivity, and there is a situation where the temperature is inaccurate. In order to accurately measure the temperature of the pig, the embodiment of the present invention uses an infrared thermal imager to obtain an infrared thermal image of a single pig raised separately, and denoises the infrared thermal image. The denoising process is a prior art and will not be described in detail here. Use an industrial camera to obtain an image of a single pig at the same time at the same angle as the infrared thermal image, grayscale and denoise the image, and obtain a grayscale image of a single pig. Grayscale and denoising are prior art and will not be described in detail here.

[0055] Since the temperature of different parts of the pig's body is different, when obtaining the temperature of the pig based on the infrared thermal image, it is necessary to determine the position of the pig in the infrared thermal image, and then obtain the overall temperature of the pig based on the temperature of each part of the pig's body. The temperature change of the pig is not only related to the pig's physical health, but also to the time of temperature detection of the pig. Therefore, when performing temperature detection, it is necessary to mark the acquisition time for each infrared thermal image and grayscale image obtained. At the same time, an infrared thermal imager and an industrial camera are used to obtain the infrared thermal image and grayscale image of the same pig at the same angle at the same position. At this time, the acquisition time marked on the infrared thermal image and the grayscale image is the same, that is, each image corresponds to a moment, so as to better analyze the possibility of pig abnormality through temperature changes. In order to efficiently analyze pig abnormalities, the embodiment of the present invention sets the initial preset sampling time interval to 1 hour, the preset sampling time period to 10 minutes, and the preset sampling frequency to once per minute, that is, 10 minutes of data are collected every hour, and infrared thermal images and grayscale images are collected once every 1 minute within 10 minutes of concentrated collection; the implementer can set the initial preset sampling time interval, preset sampling time period and preset sampling frequency according to actual conditions.

[0056] In order to obtain the position of the pig in the infrared thermal image, the embodiment of the present invention obtains the contour area of ​​the pig in the grayscale image and marks it through the semantic segmentation network, and uses the body key point neural network to mark the body key points of the pig in the grayscale image after the contour area is marked; the grayscale image after the body key point marking is intersected with the infrared thermal image of the same pig at the same time, and the infrared thermal image corresponding to the pig is determined as the thermal map of the pig, and the body key points marked in the thermal map. In the embodiment of the present invention, the nose, limbs and trunk of the pig are used as body key points. The implementer can set different parts of the animal as body key points according to the type of animal and set them according to the actual situation.

[0057] The semantic segmentation network of the actual example of the present invention uses a U-net neural network, the input is a grayscale image; the output is the contour area of ​​the pig; the U-net neural network is trained and labeled in the following way: the contour area of ​​the pig is marked as 1, and other areas are marked as 0; the loss function of the U-net neural network is a cross entropy loss function. The U-net neural network is a well-known technology and will not be described in detail here.

[0058] In another embodiment of the present invention, the pig contour region in the grayscale image can also be obtained by threshold segmentation and edge detection. The specific algorithm content is a technical means well known to those skilled in the art and will not be described in detail here. In other embodiments of the present invention, other region acquisition methods can be selected, which are not limited here.

[0059] The temperature anomaly possible value acquisition module 20 is used to obtain at least two temperature fields in the thermal map according to the temperature distribution, and obtain the temperature contribution value of each temperature field according to the number and category of key points of the body in the temperature field; obtain the overall temperature of the animal according to the average temperature value and temperature contribution value of each temperature field; and obtain the temperature anomaly possible value of the animal according to the overall temperature.

[0060] Specifically, because the temperature of each part of the pig's body is different, in order to accurately obtain the overall temperature of the pig, the embodiment of the present invention obtains the temperature of each area of ​​the pig's body, determines the temperature proportion of each area according to the number of key points of the body in each area, and then determines the overall temperature of the pig according to the temperature of each area. The specific method for obtaining the overall temperature of the pig is as follows:

[0061] (1) Obtain the temperature field of each area.

[0062] Preferably, the method for obtaining the temperature field is: clustering the thermal map using the DBSCAN density clustering algorithm according to the temperature distribution to obtain at least two clusters, one cluster being a temperature field. The DBSCAN density clustering algorithm is a well-known technology and will not be described in detail here.

[0063] As an example, the temperature of the pixel points of each part of a single pig can be determined according to the heat map, and the DBSCAN density clustering algorithm is used to cluster the heat map according to the size of the temperature, wherein the embodiment of the present invention sets the initial radius of the DBSCAN density clustering algorithm to 3, and the minimum number of sample points in the neighborhood to 3, which can be set by the implementer according to the actual situation. When clustering using the DBSCAN density clustering algorithm, if the radius is too small, the smaller the cluster cluster obtained, the smaller the area corresponding to the cluster cluster, that is, the temperature field, resulting in too many temperature fields obtained, making the overall temperature of the pig inaccurate and complex to obtain; if the radius is too large, the larger the cluster cluster obtained, the larger the area corresponding to the cluster cluster, that is, the temperature field, resulting in too few temperature fields obtained, making the overall temperature of the pig inaccurate. Therefore, the embodiment of the present invention uses a DBSCAN density clustering algorithm with an adaptive clustering radius, and achieves the optimal clustering effect by continuously changing the clustering radius, so that the cluster cluster obtained by the DBSCAN density clustering algorithm, that is, the temperature field, is more consistent with the actual distribution of temperature, and then the overall temperature obtained according to the temperature field is more accurate.

[0064] (2) Obtain the temperature contribution value of the temperature field.

[0065] Preferably, the method for obtaining the temperature contribution value is: setting different preset marking values ​​for different categories of body key points; accumulating the preset marking values ​​of all body key points in each temperature field as the overall marking value of the corresponding temperature field; obtaining the sum of the number of body key points in each temperature field and a preset constant as the overall number of the corresponding temperature field; and taking the ratio of the overall marking value of each temperature field to the overall number as the temperature contribution value of the corresponding temperature field.

[0066] As an example, it can be known from the data acquisition module 10 that the pig's body key points are six body key points, namely the nose, trunk and limbs. The body key point neural network of the embodiment of the present invention uses the OpenPose network model to mark the pig's body key points. The OpenPose network model is a well-known algorithm and will not be described in detail here. The preset marking value corresponding to each limb of the pig is set to 3, the preset marking value corresponding to the nose is set to 4, and the preset marking value corresponding to the trunk is set to 5. The body key points of the pig in the thermal map are marked by the body key point neural network. In another embodiment of the present invention, the pig's body key points can also be marked by PDM and DCNN. The specific algorithm content is a technical means well known to those skilled in the art and will not be described in detail here. Other key point marking methods can be selected in other embodiments of the present invention, which are not limited here. Arbitrarily select a temperature field as the target temperature field, and obtain the temperature contribution value of the target temperature field according to the preset marking value and number of the body key points. The formula is:

[0067]

[0068] In the formula, is the temperature contribution value of the target temperature field; S is the overall mark value of the target temperature field; x is the number of key points of the body in the target temperature field; To preset a constant and avoid the denominator being 0, the embodiment of the present invention is set to 0.01.

[0069] It should be noted that the more key points of the body in the target temperature field, The larger it is, the more important the target temperature field is. When the overall temperature is obtained, the larger the temperature in the target temperature field is. When there is no body key point in the target temperature field, the temperature of the target temperature field is not considered, because the key point area of ​​the pig's body can represent the parts of the pig with different temperatures as a whole. Therefore, only the temperature field where the key point of the body is located is considered.

[0070] According to the method for obtaining the temperature contribution value of the target temperature field, the temperature contribution value of each temperature field is obtained.

[0071] (3) Obtain the overall temperature.

[0072] Preferably, the method for obtaining the overall temperature is: taking the ratio of each temperature contribution value in the thermal map to the sum of all temperature contribution values ​​as the temperature weight of the corresponding temperature field; obtaining the average temperature value of each temperature field according to the temperature value of each pixel in the temperature field; calculating the product of the temperature weight of each temperature field and the average temperature value as the thermal value of the corresponding temperature field, and taking the result of accumulating all thermal values ​​in the thermal map as the overall temperature.

[0073] As an example, a heat map is randomly selected as the target heat map. According to the temperature contribution value and average temperature value corresponding to each temperature field in the target heat map, the formula for obtaining the overall temperature T of the pig corresponding to the target heat map is:

[0074]

[0075] Where T is the overall temperature of the pig corresponding to the target thermal map; n is the number of temperature fields in the target thermal map; is the temperature contribution value of the i-th temperature field; is the number of pixels in the i-th temperature field; is the temperature value of the jth pixel in the ith temperature field.

[0076] It should be noted that the temperature weight The larger the value, the more body key points there are in the ith temperature field, the greater the temperature proportion, and the larger T is. When there is no body key point in the ith temperature field, the temperature weight is 0, ignoring the temperature in the i-th temperature field; the average temperature value The larger it is, the higher the temperature in the i-th temperature field, and the larger T is. Note that the higher the temperature of the temperature field containing the key points of the body, the larger T is. Therefore, the larger T is, the higher the temperature of the temperature field containing the key points of the body, the higher the overall temperature of the pig corresponding to the target thermal map, and the more likely the pig corresponding to the target thermal map is to be abnormal.

[0077] According to the method of obtaining the overall temperature of the pig corresponding to the target thermal map, the overall temperature of the pig corresponding to each thermal map is obtained.

[0078] By analyzing the overall temperature of each pig at each moment, the pigs that may be abnormal are promptly identified, and a preliminary abnormality index can be directly obtained based on the overall temperature. In one embodiment of the present invention, the difference between the overall temperature and the preset standard temperature is normalized as the possible value of temperature abnormality.

[0079] The normal overall temperature of a pig is between 38 and 39.5 degrees. In the embodiment of the present invention, 39 degrees is used as the preset standard temperature, and the implementer can set it according to the actual situation. The formula for obtaining the temperature abnormality possible value R based on the overall temperature and the preset standard temperature is:

[0080]

[0081] Where, R is the possible value of temperature anomaly; T is the overall temperature; 39 is the preset standard temperature; exp is an exponential function with the natural constant e as the base; is the absolute value function.

[0082] It should be noted that The larger the value is, the greater the difference between the overall temperature and the preset standard temperature is, the greater the possibility that the overall temperature of the pig is abnormal at the corresponding moment, and the greater the possibility that the pig is abnormal, and the larger R is; therefore, It is positively correlated with R. In other embodiments of the present invention, other pairs may also be selected The normalization method, such as function transformation, maximum and minimum normalization, Sigmoid function and other normalization methods, is not limited here.

[0083] In another embodiment of the present invention, a multi-threshold segmentation method may also be used, and a fixed abnormal possibility value is preset in the interval between each threshold value, and the corresponding abnormal possibility value is determined by determining the position of the threshold interval where the overall temperature is located. In other embodiments of the present invention, a neural network or other method may also be used to obtain the abnormal possibility value, which will not be repeated here.

[0084] The temperature change value acquisition module 30 is used to obtain the temperature change value at the time to be detected based on the difference in the overall temperature of a single animal at the time to be detected and at other times in the historical data.

[0085] Specifically, the overall temperature abnormality of the pig is a continuous and long-term change. Therefore, it is necessary to detect the overall temperature of the pig within a certain period of time to accurately determine whether the overall temperature of the pig is abnormal. Analyze the overall temperature changes of the pig within a certain period of time and compare them with the overall temperature in the historical data to obtain the temperature change value at each moment. The specific method for obtaining the temperature change value is as follows:

[0086] Preferably, the method for obtaining the temperature change value is as follows: taking all the overall temperatures in the preset temperature analysis time period before the time when a single animal is to be detected as the first historical temperature; obtaining the target position in the preset periodic sampling time period at the time when the animal is to be detected, and obtaining the overall temperature corresponding to the target position in a preset number of periodic sampling time periods as the second historical temperature; calculating the difference between the overall temperature of the single animal at the time when the animal is to be detected and each of the first historical temperatures as the first difference, and taking the result of accumulating the first differences as the first change value at the time when the animal is to be detected; taking the difference between the overall temperature at each moment and the overall temperature at the adjacent previous moment as the adjacent difference; obtaining the overall temperature at the time when the animal is to be detected The adjacent difference of the first historical temperature is taken as the target adjacent difference, the adjacent difference of each first historical temperature is obtained as the historical adjacent difference, the difference between the target adjacent difference and each historical adjacent difference is calculated as the second difference, and the result of accumulating the second differences is taken as the second change value at the time to be detected; the difference between the overall temperature of a single animal at the time to be detected and each second historical temperature is calculated as the third difference, and the result of accumulating the third differences is taken as the third change value at the time to be detected; the temperature change value of a single animal at the time to be detected is obtained according to the first change value, the second change value and the third change value, and the first change value, the second change value and the third change value are all positively correlated with the temperature change value.

[0087] As an example, a pig is randomly selected as a reference pig. The specific method for obtaining the temperature change value of the reference pig at the current moment is as follows: obtain all the overall temperatures of the reference pig before the current moment in one day as the first historical temperature, that is, the length of the preset temperature analysis time period is the length of all time before the current moment in the day. Take one day as the periodic sampling time period, 0 o'clock as the starting point of data collection, and 24 o'clock as the end point of data collection. Obtain the overall temperature corresponding to the target position of the 30 periodic sampling time periods before the current moment of the reference pig, that is, the second historical temperature, that is, the 30 periodic sampling time periods contain 30 days of data, and the target position is the same time position corresponding to the current moment in each day. If the current moment is 12 o'clock, then all 12 o'clock in the 30 periodic sampling time periods are the target positions. Calculate the difference between the overall temperature of the reference pig at the current moment and each first historical temperature, that is, the first difference, and use the result of the accumulation of the first difference as the first change value of the reference pig at the current moment. Obtain the difference between the overall temperature of the reference pig at the current moment and the overall temperature at the previous moment, that is, the target adjacent difference of the reference pig, and obtain the adjacent difference of each first historical temperature, that is, the historical adjacent difference, wherein the adjacent difference of the first first historical temperature is the difference between the first first historical temperature and the second first historical temperature. Calculate the difference between the target adjacent difference and each historical adjacent difference, that is, the second difference, and use the result of accumulating the second difference as the second change value of the reference pig at the current moment. Calculate the difference between the overall temperature of the reference pig at the current moment and each second historical temperature, that is, the third difference, and use the result of accumulating the third difference as the third change value of the reference pig at the current moment; the formula for obtaining the temperature change value W of the reference pig at the current moment according to the first change value, the second change value and the third change value is:

[0088]

[0089] Where W is the temperature change value of the reference pig at the current moment; u is the number of the first historical temperature or the number of adjacent differences corresponding to the reference pig; T is the overall temperature of the reference pig at the current moment; is the kth first historical temperature; is the target adjacent difference of the reference pig at the current moment; is the kth historical adjacent difference; n is the number of the second historical temperature; is the kth second historical temperature; is the absolute value function.

[0090] It should be noted that the first difference The larger the value, the greater the difference between the overall temperature of the reference pig at the current moment and the corresponding first historical temperature. The more likely there is an abnormality in the overall temperature of the reference pig at the current moment, and the larger W is; the second difference The larger the value, the more inconsistent the target adjacent differences of the reference pig are with the historical adjacent differences, indicating that the overall temperature of the reference pig at the current moment may have mutated. The larger the value, the greater the third difference. The larger the value is, the greater the difference between the overall temperature of the reference pig at the current moment and the overall temperature at the historical target position is, the more likely the overall temperature of the reference pig at the current moment is abnormal, and the larger W is; therefore, the larger W is, the more likely the overall temperature of the reference pig at the current moment is abnormal, and the greater the possibility that the reference pig's body is abnormal. In the embodiment of the present invention, the sum of the first change value, the second change value, and the third change value is used as the temperature change value, which is positively correlated with the temperature change value. In another embodiment of the present invention, the product of the first change value, the second change value, and the third change value can be used as the temperature change value, so that the first change value, the second change value, and the third change value are always positively correlated with the temperature change value. In other embodiments of the present invention, other basic mathematical operations can be selected to form a positive correlation, which is not limited here.

[0091] According to the method of obtaining the temperature change value of the reference pig at the current moment, the temperature change value of each pig at the time to be tested is obtained.

[0092] The temperature difference value acquisition module 40 calculates the difference in overall temperature between each individual animal and all other similar animals at the time to be detected to obtain the temperature difference value of the corresponding individual animal.

[0093] Specifically, the abnormality of the overall temperature information obtained by the current pig is determined by obtaining the overall temperature information of other pigs at the time to be detected. That is, the greater the difference between the overall temperature information of the current pig and the overall temperature information of other pigs, that is, the greater the temperature difference value, the more likely the overall temperature of the current pig is to be abnormal.

[0094] Preferably, the method for obtaining the temperature difference value is: the result of accumulating the difference in overall temperature between each individual animal and all other similar animals at the time of detection is used as the temperature difference value corresponding to the individual animal.

[0095] As an example, taking the reference pig in the temperature change value acquisition module 30 as an example, the formula for obtaining the temperature difference value E of the reference pig at the current moment is:

[0096]

[0097] Where, E is the temperature difference value of the reference pig at the current moment; T is the overall temperature of the reference pig at the current moment; o is the number of other pigs at the current moment; is the overall temperature of the ath other pig; is the absolute value function.

[0098] It should be noted that The larger the value is, the greater the difference in overall temperature between the reference pig and the ath other pig at the current moment, which means that the overall temperature of the reference pig is more abnormal and the larger E is; therefore, the larger the value is, the more likely it is that the reference pig is abnormal at the current moment.

[0099] According to the method of obtaining the temperature difference value of the reference pig at the current moment, the temperature difference value of each pig at the time to be tested is obtained.

[0100] The abnormality detection module 50 is used to obtain a temperature detection value according to the possible temperature abnormality value, temperature change value and temperature difference value of a single animal at the time to be detected, and perform abnormality detection on the single animal according to the temperature detection value.

[0101] Preferably, the method for performing abnormality detection on a single animal based on the temperature detection value is as follows: the possible temperature abnormality value, the temperature change value and the temperature difference value at the time to be detected are all positively correlated with the temperature detection value at the corresponding time; a temperature detection value threshold is set, and when the temperature detection value is greater than the temperature detection value threshold, it is determined that the corresponding single animal has a temperature abnormality.

[0102] As an example, taking the reference pig in the temperature change value acquisition module 30 as an example, the formula for obtaining the temperature detection value Y of the reference pig at the current moment is:

[0103]

[0104] Wherein, Y is the temperature detection value of the reference pig at the current moment; W is the temperature change value of the reference pig at the current moment; R is the possible value of the temperature abnormality of the reference pig at the current moment; and E is the temperature difference value of the reference pig at the current moment.

[0105] It should be noted that the larger the W is, the greater the change in the overall temperature of the reference pig at the current moment is, the more likely the overall temperature of the reference pig at the current moment is abnormal, and the larger the Y is; the larger the R is, the greater the change in the overall temperature of the reference pig at the current moment is, the greater the possibility of abnormality, and the larger the Y is; the larger the E is, the greater the difference between the overall temperature of the reference pig at the current moment and the overall temperature of other pigs, the more likely the overall temperature of the reference pig at the current moment is abnormal, and the larger the Y is; therefore, the larger the Y is, the more likely the overall temperature of the reference pig at the current moment is abnormal, and the greater the possibility that the reference pig at the current moment is abnormal. In the embodiment of the present invention, the product of the temperature abnormality possible value, the temperature change value and the temperature difference value is used as the temperature detection value, which is positively correlated with the temperature detection value. In another embodiment of the present invention, the addition result of the temperature abnormality possible value, the temperature change value and the temperature difference value can be used as the temperature detection value, so that the temperature abnormality possible value, the temperature change value and the temperature difference value are always positively correlated with the temperature detection value. In other embodiments of the present invention, other basic mathematical operations can be selected to form a positive correlation, which is not limited here.

[0106] According to the method of obtaining the temperature detection value of the reference pig at the current moment, the temperature detection value of each pig at the time to be detected is obtained.

[0107] The embodiment of the present invention sets the temperature detection value threshold to 0.7, and the implementer can set the size of the temperature detection value threshold according to the actual situation; the temperature detection value is normalized to obtain the normalized temperature detection value. When the normalized temperature detection value is greater than the temperature detection value threshold, it is determined that the overall temperature of the corresponding pig is abnormal, and manual inspection is required to avoid problems with the pig and thus infection or death.

[0108] The time interval adjustment module 60 is used to obtain the posture characteristics at each moment according to all the key points of the body in the thermal map, and obtain the data accuracy index of a single animal at each moment according to the posture characteristics at each moment, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value of the posture characteristics in the preset adjustment time period before the corresponding moment, and adjust the sampling time interval of the next sampling time period according to all the data accuracy indicators in each sampling time period.

[0109] Specifically, if the overall temperature of the pig is continuously detected, and an abnormal analysis is performed on the overall temperature of each pig at each moment, it will cause a large amount of waste of manpower and material resources. Therefore, the embodiment of the present invention uses an intermittent centralized data collection method to obtain the overall temperature data of the pig, and then analyze whether the overall temperature is abnormal. When using the intermittent centralized data collection method, the sampling time interval between each sampling time period must be set reasonably. A reasonable sampling time interval can ensure the accuracy of the acquired data. If the sampling time interval is set unreasonably, the posture of the pig may be inconsistent within the sampling time period, which may easily lead to inaccurate centralized data collection. In order to ensure that the acquired data is more accurate and the analysis of the abnormal temperature of the pig is more practical, the preset sampling time interval that is initially set is adjusted to make the sampling time interval more reasonable. The specific method for adjusting the initially set sampling time interval and obtaining a reasonable sampling time interval is as follows:

[0110] (1) Obtain posture features.

[0111] Preferably, the method for obtaining the posture feature is: taking the result of adding the preset marking values ​​of all the key points of the body in the heat map as the posture feature.

[0112] As an example, taking the thermogram of the reference pig at the current moment in the temperature change value acquisition module 30 as an example, the result of adding the preset mark values ​​of all body key points in the thermogram of the reference pig at the current moment is obtained as the posture feature of the reference pig at the current moment.

[0113] According to the method of obtaining the posture features of the reference pig at the current moment, the posture features of each pig at each moment are obtained.

[0114] (2) Obtain data accuracy indicators.

[0115] Preferably, the method for obtaining the data accuracy index is: setting a preset adjustment time period, obtaining the variance of the posture characteristics within the preset adjustment time period as the posture fluctuation value of the animal at the time to be detected; obtaining the difference in temperature change value between each moment and the adjacent previous moment as the adjacent deviation degree at the corresponding moment; obtaining the product of the posture characteristics at each moment and the possible value of the temperature anomaly as the credible possible value of the temperature anomaly at the corresponding moment; obtaining the data accuracy index at the corresponding moment according to the credible possible value of the temperature anomaly, the temperature change value, the adjacent deviation degree and the posture fluctuation value at each moment; the credible possible value of the temperature anomaly, the temperature change value, the adjacent deviation degree and the posture fluctuation value are all positively correlated with the data accuracy index.

[0116] As an example, taking the reference pig in the temperature change value acquisition module 30 as an example, the method for obtaining the data accuracy index of the reference pig at the current moment is as follows: determine the sampling time period of the centralized data collection where the reference pig is at the current moment as the first target time period, and take the sampling time period of the centralized data collection that is adjacent to the first target time period as the second target time period; take the time period between the current moment of the reference pig and the first moment in the second target time period as the preset adjustment time period corresponding to the current moment of the reference pig, and obtain the variance of all posture features in the preset adjustment time period corresponding to the current moment of the reference pig as the posture fluctuation value of the reference pig at the current moment. The length of the preset adjustment time period is the sampling time period of two adjacent centralized data collection. According to the posture fluctuation value of the preset adjustment time period, the posture change of the reference pig at the current moment can be judged. The larger the posture fluctuation value, the greater the posture change of the reference pig. The smaller the posture fluctuation value, the less activity of the reference pig, and the smaller the posture change. If the length of the preset adjustment time period is too long, it may cause the reference pig's posture change to be inaccurate, which is greatly affected by the posture of a long time ago, resulting in too large a posture fluctuation value; if the length of the preset adjustment time period is too short, it may cause the change of the reference pig's posture to not be reflected, resulting in inaccurate acquisition of the temperature information of the reference pig. Obtain the absolute value of the difference between the temperature change value of the reference pig at the current moment and the temperature change value at the previous moment, that is, the adjacent deviation degree of the reference pig at the current moment; obtain the product of the posture feature of the reference pig at the current moment and the possible value of the temperature anomaly, that is, the credible temperature anomaly possible value of the reference pig at the current moment; according to the credible temperature anomaly possible value, temperature change value, adjacent deviation degree and posture fluctuation value of the reference pig at the current moment, the formula for obtaining the data accuracy index P of the reference pig at the current moment is:

[0117]

[0118] Where, P is the data accuracy index of the reference pig at the current moment; B is the posture feature of the reference pig at the current moment; R is the possible value of the temperature abnormality of the reference pig at the current moment; W is the temperature change value of the reference pig at the current moment; The adjacent deviation degree of the reference pig at the current moment; It is the posture fluctuation value of the reference pig at the current moment.

[0119] It should be noted that the credible temperature anomaly may be valued at B The larger the R is, the higher the credibility of the possible value of temperature anomaly is, the more accurate the temperature information is, and the more likely the overall temperature of the reference pig at the current moment is abnormal. The larger the P is, the greater the change in the overall temperature of the reference pig at the current moment is. The larger the value, the more obvious the change in the overall temperature of the reference pig at the current moment is. The sampling time interval of intermittent centralized data collection should be shortened to facilitate the determination of the time period when the temperature anomaly occurs. At this time, the larger the value, the greater the value. The larger the value is, the greater the change in the reference pig's posture in the preset adjustment time period corresponding to the reference pig at the current moment is, and the larger P is; therefore, the larger P is, the more likely the overall temperature of the reference pig at the current moment is abnormal. In order to obtain data more accurately, it is necessary to shorten the sampling time interval and obtain temperature information when the reference pig's posture is relatively consistent; at the same time, it is necessary to consider that when P is too small, it means that the sampling time interval is too small, the obtained data has hardly changed, and the abnormal temperature detection is meaningless, and the sampling time interval needs to be expanded. In the embodiment of the present invention, the product of posture characteristics, possible value of temperature anomaly, temperature change value, adjacent deviation degree and posture fluctuation value is used as a data accuracy index, which is positively correlated with the data accuracy index. In another embodiment of the present invention, the addition result of posture characteristics, possible value of temperature anomaly, temperature change value, adjacent deviation degree and posture fluctuation value can be used as a data accuracy index, so that posture characteristics, possible value of temperature anomaly, temperature change value, adjacent deviation degree and posture fluctuation value are always positively correlated with the data accuracy index. In other embodiments of the present invention, other basic mathematical operations can be selected to form a positive correlation, which is not limited here.

[0120] According to the method of obtaining the data accuracy index of the reference pig at the current moment, the data accuracy index of each pig at each moment is obtained.

[0121] (3) Adjust the sampling time interval.

[0122] Preferably, the method for adjusting the sampling time interval is: calculating the mean of the data accuracy index within each sampling time period as the target data accuracy index, setting the data accuracy index interval, when the target data accuracy index is within the data accuracy index interval, the sampling time interval does not need to be adjusted; when the target data accuracy index is less than the minimum value within the data accuracy index interval, the sampling time interval is increased by a preset time step; when the data accuracy index is greater than the maximum value within the data accuracy index interval, the sampling time interval is shortened by a preset time step.

[0123] The embodiment of the present invention takes the average of the data accuracy indicators corresponding to the 10 data in each centralized data sampling period as the target data accuracy indicator of the corresponding sampling period, normalizes the target data accuracy indicator, and obtains the normalized target data accuracy indicator; the embodiment of the present invention sets the data accuracy indicator interval to When the normalized target data accuracy index is within the data accuracy index interval, the next sampling time interval does not need to be adjusted; when the normalized target data accuracy index is less than the minimum value within the data accuracy index interval, the next sampling time interval increases by a preset time step; when the normalized target data accuracy index is greater than the maximum value within the data accuracy index interval, the next sampling time interval decreases by a preset time step. In this embodiment of the present invention, the preset time step is set to 10 minutes, and the implementer can set the size of the data accuracy index interval and the preset time step according to the actual situation.

[0124] Determine the reasonable sampling time interval between sampling time periods to make data acquisition more accurate and efficient, and to detect temperature anomalies in pigs more accurate and efficient.

[0125] So far, the present invention is completed.

[0126] In summary, the embodiments of the present invention collect a thermogram of a single animal; obtain the overall temperature of the single animal according to the temperature field in the thermogram; obtain the possible value of temperature anomaly, the temperature change value and the temperature difference value according to the overall temperature; obtain the temperature detection value according to the possible value of temperature anomaly, the temperature change value and the temperature difference value, and perform abnormality detection on the animal according to the temperature detection value; obtain posture features according to the key points of the body in the thermogram, obtain data accuracy indicators according to the posture features, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value of the posture features, and adjust the sampling time interval according to the data accuracy indicator. The present invention collects the temperature information of animals in a centralized manner by adjusting a reasonable sampling time interval, thereby improving the accuracy and efficiency of abnormality detection on animals.

[0127] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0128] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An intelligent animal temperature detection system for animal husbandry and veterinary medicine based on temperature measurement, It is characterized in that The system comprises: A data acquisition module, used to collect a thermal map of a single animal at a preset sampling time interval according to a preset sampling period and a preset sampling frequency; the thermal map contains body key point information; A temperature anomaly possible value acquisition module is used to acquire at least two temperature fields in the thermal map according to the temperature distribution, and acquire the temperature contribution value of each temperature field according to the number and category of key points of the body in the temperature field; acquire the overall temperature of the animal according to the average temperature value and temperature contribution value of each temperature field; and acquire the temperature anomaly possible value of the animal according to the overall temperature; A temperature change value acquisition module is used to obtain the temperature change value at the time to be detected based on the difference between the overall temperature of a single animal at the time to be detected and other times in the historical data; The temperature difference value acquisition module calculates the difference in overall temperature between each individual animal and all other similar animals at the time of detection to obtain the temperature difference value of the corresponding individual animal; An anomaly detection module, used to obtain a temperature detection value according to a possible temperature anomaly value, a temperature change value and a temperature difference value of a single animal at the time to be detected, and perform an anomaly detection on the single animal according to the temperature detection value; The time interval adjustment module is used to obtain the posture characteristics at each moment according to all the key points of the body in the thermal map, and obtain the data accuracy index of a single animal at each moment according to the posture characteristics at each moment, the possible value of temperature anomaly, the temperature change value and the posture fluctuation value of the posture characteristics in the preset adjustment time period before the corresponding moment, and adjust the sampling time interval of the next sampling time period according to all the data accuracy indicators in each sampling time period.

2. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The method for obtaining the temperature contribution value comprises: Different preset marking values ​​are set for different categories of body key points; The result of accumulating the preset marking values ​​of all key points of the body in each temperature field is used as the overall marking value of the corresponding temperature field; Obtaining the sum of the number of key points of the body in each temperature field and a preset constant as the overall number of the corresponding temperature field; The ratio of the overall mark value to the overall quantity of each temperature field is used as the temperature contribution value of the corresponding temperature field.

3. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The method for obtaining the overall temperature comprises: The ratio of each temperature contribution value in the thermal diagram to the sum of all temperature contribution values ​​is used as the temperature weight of the corresponding temperature field; According to the temperature value of each pixel in the temperature field, the average temperature value of each temperature field is obtained; The product of the temperature weight and the average temperature value of each temperature field is calculated as the thermal value of the corresponding temperature field, and the result of accumulating all the thermal values ​​in the thermal map is taken as the overall temperature.

4. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The method for obtaining the temperature change value includes: taking all overall temperatures in a preset temperature analysis period before the time when the single animal is to be detected as the first historical temperature; Acquire the target position in a preset periodic sampling time period at the time to be detected, and acquire the overall temperature corresponding to the target position in a preset number of the periodic sampling time periods as the second historical temperature; Calculate the difference between the overall temperature of a single animal at the time to be detected and each first historical temperature as a first difference, and use the result of accumulating the first differences as a first change value at the time to be detected; The difference between the overall temperature at each moment and the overall temperature at the previous moment is taken as an adjacent difference; the adjacent difference of the overall temperature at the moment to be detected is obtained as a target adjacent difference, the adjacent difference of each first historical temperature is obtained as a historical adjacent difference, the difference between the target adjacent difference and each of the historical adjacent differences is calculated as a second difference, and the result of accumulating the second differences is taken as a second change value at the moment to be detected; Calculate the difference between the overall temperature of a single animal at the time to be detected and each second historical temperature as a third difference, and use the result of accumulating the third differences as a third change value at the time to be detected; The temperature change value of a single animal at the time of detection is obtained according to the first change value, the second change value and the third change value, and the first change value, the second change value and the third change value are all positively correlated with the temperature change value.

5. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The temperature difference value acquisition method comprises: The result of accumulating the difference in overall temperature between each individual animal and all other similar animals at the time of detection is taken as the temperature difference value of the corresponding individual animal.

6. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The abnormality detection of a single animal according to the temperature detection value comprises: The possible value of temperature anomaly, temperature change value and temperature difference value at the time to be detected are all positively correlated with the temperature detection value at the corresponding time; A temperature detection value threshold is set. When the temperature detection value is greater than the temperature detection value threshold, it is determined that the corresponding single animal has a temperature abnormality.

7. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 2, It is characterized in that The method for acquiring the posture feature comprises: The result of adding the preset label values ​​of all body key points in the heat map is used as the posture feature.

8. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The method for obtaining the data accuracy indicator includes: Setting a preset adjustment time period, and obtaining the variance of the posture feature within the preset adjustment time period as the posture fluctuation value of the animal at the time of detection; Obtain the difference between the temperature change value at each moment and the temperature at the previous moment as the adjacent deviation degree at the corresponding moment; The product of the posture feature and the possible value of the temperature anomaly at each moment is obtained as the credible possible value of the temperature anomaly at the corresponding moment; The data accuracy index at the corresponding moment is obtained according to the credible temperature anomaly possible value, the temperature change value, the adjacent deviation degree and the posture fluctuation value at each moment; the credible temperature anomaly possible value, the temperature change value, the adjacent deviation degree and the posture fluctuation value are all positively correlated with the data accuracy index.

9. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The step of adjusting the sampling time interval of the next sampling time period according to all data accuracy indicators in each sampling time period includes: Calculate the mean of the data accuracy index in each sampling time period as the target data accuracy index, set the data accuracy index interval, when the target data accuracy index is within the data accuracy index interval, the sampling time interval does not need to be adjusted; when the target data accuracy index is less than the minimum value within the data accuracy index interval, increase the sampling time interval by a preset time step; when the data accuracy index is greater than the maximum value within the data accuracy index interval, shorten the sampling time interval by a preset time step.

10. An intelligent animal body temperature detection system for animal husbandry and veterinary medicine based on temperature measurement as claimed in claim 1, It is characterized in that The method for obtaining at least two temperature fields in a thermodynamic map according to temperature distribution includes: The DBSCAN density clustering algorithm is used to cluster the thermal map according to the temperature distribution to obtain at least two clusters, one cluster being one temperature field.

Citation Information

Patent Citations

  • Livestock animal fence entry detection method based on gun-ball linkage and BIM

    CN111612645A

  • Livestock body temperature image processing method based on artificial intelligence and active gun-ball linkage

    CN111833331A

  • Beef cattle temperature anomaly detection method

    CN117173640A

  • Method and device for predicting future thermographic measurement data of animals

    WO2014063686A1

  • Livestock and poultry body temperature monitoring system and method with wearable sensors and infrared cameras

    WO2024001791A1