An intelligent monitoring and management system for pig breeding based on internet of things technology

Abstract

The application discloses a kind of intelligent monitoring management system of pig breeding based on internet of things technology belongs to intelligent breeding technical field, this intelligent monitoring management system of pig breeding is cooperated with identity recognition module by video acquisition module, accurate and fast identification is carried out to pig in the video image of pig shown in terminal display device, so as to facilitate staff to identify the identity of pig in pig house;Compared with the way that weighing device is set in the place that pig must pass in activity to measure the weight of pig, the application can realize the contactless pig weight measurement and record, so as to improve the detection frequency and efficiency of pig weight, reduce the time cost and labor cost of pig weight detection.

Description

Technical Field

[0001] This invention belongs to the field of intelligent farming technology, specifically, it relates to an intelligent monitoring and management system for pig farming based on Internet of Things (IoT) technology. Background Technology

[0002] my country is a major pork consumer, and pig farming is a vital sector of its agriculture. It plays a crucial role in ensuring a safe supply of meat. my country's pig farming industry is transitioning from traditional to modern methods, with significant changes occurring in breeding models, regional distribution, production methods, and production capacity. This transformation presents numerous challenges, making intelligent pig farming a promising future direction.

[0003] The application of IoT technology in pig farming has greatly reduced the difficulty of data collection, the labor intensity of staff, and the labor costs of farms. Furthermore, the more timely and widespread acquisition of data enhances the potential for scaling up farming operations. However, current technologies for monitoring pig data in pig farms rely solely on information transmitted through electronic ear tags to analyze the status of each pig, failing to identify the pigs using cameras. This prevents staff from directly understanding the growth status of each pig, hindering their ability to control the farming results. Moreover, weight measurement requires the installation of card readers and weighing devices in the pigs' usual passageways, resulting in low efficiency and infrequent measurement. To address these issues, this invention provides the following technical solution. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent monitoring and management system for pig farming based on Internet of Things (IoT) technology, which solves the problem that in the existing technology, pigs carrying electronic ear tags can only be identified by handheld card readers and other devices. Therefore, the weight of pigs during the farming process depends on weighing devices, which is inefficient.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A smart monitoring and management system for pig farming based on Internet of Things (IoT) technology includes:

[0007] The identification module includes an electronic ear tag attached to the pig and a card reader capable of reading information from the electronic ear tag;

[0008] The identification module can collect observational parameters such as the pig's body temperature, location, and step count data;

[0009] When setting up, three or more card readers are distributed throughout the pigsty;

[0010] The video acquisition module is configured to acquire video image information inside the pig house.

[0011] The alarm module is configured to send alarm information.

[0012] The deep learning module is configured to calculate the weight of the live pig.

[0013] The method for calculating the weight of the live pig by the deep learning module comprises the following steps:

[0014] S1. The identity recognition module and the video acquisition module are used to locate the live pigs in the pig house, and the identity information of each live pig in the image information acquired by the video acquisition module is obtained.

[0015] The method for locating is as follows:

[0016] The relationship between the field strength and the distance between the electronic ear tag and the card reader is obtained.

[0017] The distance between the electronic ear tag and each card reader is obtained according to the relationship between the field strength and the distance between the electronic ear tag and each card reader and the detected field strength relationship between the electronic ear tag and each card reader.

[0018] The coordinate position (a, b) of the live pig corresponding to the electronic ear tag is obtained according to the distance between the electronic ear tag and each card reader, and the coordinate is marked as a first coordinate.

[0019] The image information in the pig house is obtained by the video acquisition module.

[0020] The coordinate position (a1, b1) of each live pig in the pig house is obtained according to the position of each live pig in the image information, and the coordinate is marked as a second coordinate.

[0021] For a first coordinate, the intersection coefficient c between the first coordinate and each second coordinate is calculated according to the formula The second coordinate corresponding to the smallest intersection coefficient c is selected as the merged coordinate of the first coordinate.

[0022] The merged coordinates of each first coordinate are calculated and obtained, and the live pig at the position corresponding to the merged coordinate is matched with the first coordinate.

[0023] S2. The card reader acquires a group of first coordinates every preset time t1, and acquires a group of second coordinates after acquiring the frame image by the video acquisition module, and the identity information of each live pig in the frame image is obtained according to the method in step S1.

[0024] The weight of each live pig in the frame image is obtained by the deep learning module.

[0025] S3, for a pig, the body weight obtained by the deep learning module in a period is sequentially marked as z1, z2, …, zn, n is the number of body weight data obtained by the deep learning module in the corresponding period of the pig;

[0026] The real body weight zt of the corresponding pig is calculated according to the set of body weight data.

[0027] As a further scheme of the present application, the pig house is divided into several breeding areas, and each breeding area corresponds to three or more card readers.

[0028] As a further scheme of the present application, the monitoring management system further comprises a sensor module for detecting environmental parameters in the pig house;

[0029] The environmental parameters in the pig house are monitored by the sensor module;

[0030] When the corresponding environmental parameter value is greater than the preset threshold value, the alarm module sends an alarm information;

[0031] The environmental parameters include temperature, humidity, ammonia concentration and hydrogen sulfide concentration.

[0032] As a further scheme of the present application, when the number of intersection coefficients c satisfying c≤cy corresponding to a first coordinate is two or more, the pig corresponding to the first coordinate is not subjected to subsequent processing;

[0033] Wherein cy is a preset value.

[0034] As a further scheme of the present application, the calculation method of the real body weight zt is:

[0035] According to the formula The standard deviation value C of the set of data z1 to zn is calculated;

[0036] When C<C1 is established, zp is taken as the real body weight zt of the corresponding pig in the period;

[0037] On the contrary, when C≥C1 is established, the corresponding zi value is sequentially deleted in the order of |zi-zp| from large to small until C<C1 is established, and the average value of the corresponding zi value which is not deleted is taken as the real body weight zt of the corresponding pig in the period;

[0038] Wherein C1 is a preset value, 1≤i≤n, and zp=(z1+z2+…+zn) / n.

[0039] As a further scheme of the present application, the deep learning module is further used for identifying pigs with abnormal body weight, and the identification method comprises the following steps:

[0040] For a pig, the real body weight of the pig in the past k periods is sequentially marked as u1, u2, …, uk in time sequence;

[0041] A body weight change curve Q corresponding to the pig is generated in a two-dimensional rectangular coordinate system with time as the horizontal coordinate and the real body weight as the vertical coordinate;

[0042] The change curve of the pig is divided into several growth intervals in time sequence, and the body weight change amount h of each pig in each growth interval is obtained, and when h≤hy, it is considered that the state of the corresponding pig in the corresponding growth interval is abnormal, and the alarm module sends an alarm information to prompt the staff;

[0043] Wherein hy is a preset value, and the corresponding hy is different in different growth intervals.

[0044] The beneficial effects of the present application are:

[0045] 1、The present application realizes accurate and rapid identification of pigs in the video image of the pigs displayed in the terminal display device through the cooperation of the video acquisition module and the identity recognition module, thereby facilitating the staff to identify the identity of the pigs in the pig house;

[0046] 2、Compared with the method of setting a weighing device at the place where the pig must pass through to measure the weight of the pig, the present application can realize non-contact pig weight measurement and recording, thereby improving the detection frequency and efficiency of the pig weight, and reducing the time cost and labor cost of pig weight detection;

[0047] 3、The present application realizes accurate identification and positioning of pigs with abnormal weight by real-time detection and monitoring of the weight change of the pigs, thereby realizing early detection and early solution, and reducing the loss of the farm caused by abnormal conditions. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0049] A pig breeding intelligent monitoring and management system based on Internet of Things technology, comprising:

[0050] A sensor module comprising a temperature sensor, a humidity sensor, an ammonia concentration sensor, a hydrogen sulfide concentration sensor, etc., for detecting relevant environmental parameters in the pig house;

[0051] An identity recognition module, including an electronic ear tag fixed on the pig body and a card reader capable of reading information in the electronic ear tag;

[0052] The identity recognition module is capable of collecting observation parameter information of the live pig, such as body temperature, position and step counting data;

[0053] The electronic ear tag also stores identity information of the live pig;

[0054] The identity information includes a serial number and a birth date of the live pig;

[0055] The electronic ear tag transmits the observation parameter information once every preset time t1, and in an embodiment of the present application, t1 is 1 s;

[0056] In the setting, three or more card readers are distributed in the pig house;

[0057] In an embodiment of the present application, the pig house can be divided into several breeding areas, and each breeding area corresponds to three or more card readers, which can reduce the ranging range requirement of the card reader and improve the monitoring accuracy;

[0058] A video acquisition module for acquiring video image information inside the pig house;

[0059] An alarm module for sending alarm information;

[0060] A deep learning module for receiving video image information transmitted by the video acquisition module, obtaining parameters such as body length and height of the live pig according to a machine vision algorithm, and estimating the body weight of the live pig according to the parameters such as body length and height of the live pig;

[0061] The image analysis technology is used for live pig weight detection, which is a prior art, and thus will not be described here;

[0062] In addition, the deep learning module is also used for identifying live pigs with abnormal body weight;

[0063] The method for calculating the body weight of the live pig by the deep learning module includes the following steps:

[0064] S1, monitoring environmental parameters such as temperature, humidity, ammonia concentration and hydrogen sulfide concentration in the pig house by the sensor module;

[0065] When the corresponding environmental parameter value is greater than a preset threshold value, the alarm module sends an alarm information to prompt the staff to check the corresponding environmental parameter;

[0066] S2, positioning the live pigs in the pig house by the identity recognition module and the video acquisition module, and obtaining identity information of each live pig in the image information collected by the video acquisition module;

[0067] The method for positioning specifically comprises:

[0068] obtaining the relationship between the field strength and the distance between the electronic ear tag and the card reader;

[0069] obtaining the distance between the electronic ear tag and each card reader according to the relationship between the field strength and the distance between the electronic ear tag and each card reader and the detected field strength relationship between the electronic ear tag and each card reader;

[0070] obtaining the coordinate position (a, b) of the corresponding pig of the electronic ear tag according to the distance between the electronic ear tag and each card reader, and marking the coordinate as a first coordinate;

[0071] obtaining image information in the pig house through a video acquisition module;

[0072] obtaining the coordinate position (a1, b1) of each pig in the pig house according to the position of each pig in the image information, and marking the coordinate as a second coordinate;

[0073] for a first coordinate, the intersection coefficient c between the first coordinate and each second coordinate is calculated according to the formula , and the second coordinate corresponding to the smallest intersection coefficient c is selected as the merged coordinate of the first coordinate;

[0074] the merged coordinate of each first coordinate is calculated and obtained, and the pig at the corresponding position of the merged coordinate is matched with the first coordinate, so as to complete the positioning of the pig in the image information collected by the video acquisition unit and obtain the identity information of each pig in the image information;

[0075] in an embodiment of the present application, when for a first coordinate, the number of intersection coefficients c satisfying c≤cy corresponding to the first coordinate is two or more than two, the pig corresponding to the first coordinate is not subjected to subsequent processing;

[0076] wherein cy is a preset value;

[0077] due to the high density of pig breeding, several pigs often crowd together, considering the positioning accuracy, this type of pig is not subjected to subsequent processing in one frame image, which can improve the accuracy of subsequent pig weight measurement;

[0078] S3, the card reader obtains a group of first coordinates every preset time t1, simultaneously obtains a group of second coordinates through the video acquisition module after obtaining a frame image, and obtains the identity information of each pig in the frame image according to the method in step S2;

[0079] the weight of each pig in the frame image is obtained through a deep learning module;

[0080] S4, for a pig, the body weight obtained by the deep learning module in a period is sequentially marked as z1, z2,..., zn, n is the number of body weight data obtained by the deep learning module in the corresponding period for the pig;

[0081] In an embodiment of the present application, a period is one day;

[0082] According to the formula The standard deviation value C of the z1 to zn data is calculated;

[0083] When C < C1 is established, zp is taken as the true body weight zt of the corresponding pig in the period;

[0084] Conversely, when C >= C1 is established, the corresponding zi value is sequentially deleted in the order of |zi-zp| from large to small until C < C1 is established, and the average value of the corresponding zi value which is not deleted is taken as the true body weight zt of the corresponding pig in the period;

[0085] In an embodiment of the present application, if the number of deleted zi values in the process of calculating the true body weight zt of the pig is greater than the preset value b1, it is considered that the system is abnormal, and the alarm module issues an alarm information;

[0086] Wherein C1 is a preset value, 1 <= i <= n, zp = (z1+z2+,...+zn) / n;

[0087] The present application realizes accurate and rapid identification of pigs in the video image of pigs displayed in the terminal display device by cooperating the video acquisition module with the identity recognition module, so as to facilitate the staff to identify the identity of the pigs in the pig house;

[0088] In addition, compared with the way of measuring the weight of the pig by setting a weighing device at the place where the pig must pass, the present application can realize contactless pig weight measurement and recording, thereby improving the detection frequency and efficiency of the pig weight, and reducing the time cost and labor cost of pig weight detection;

[0089] S5, for a pig, the true body weight of the pig in the past k periods is sequentially marked as u1, u2,..., uk in time sequence;

[0090] The body weight change curve Q of the corresponding pig is generated in a two-dimensional rectangular coordinate system with time as the horizontal coordinate and the true body weight as the vertical coordinate;

[0091] The change curve of the pig is divided into several growth intervals in time sequence, and the body weight change amount h of each pig in each growth interval is obtained, when h <= hy, it is considered that the state of the corresponding pig in the corresponding growth interval is abnormal, and the alarm module issues an alarm information to prompt the staff;

[0092] Wherein hy is a preset value, and in different growth intervals, the corresponding hy is different;

[0093] As a further scheme of the present application, the alarm module can also identify the live pigs with abnormal state in the terminal display device through the video acquisition module, so as to facilitate the staff to identify the live pigs with abnormal state.

[0094] The present application can accurately identify and locate the live pigs with abnormal weight by real-time detection and monitoring of the weight change of the live pigs, so as to realize early discovery and early solution, and reduce the loss of the farm caused by abnormal conditions.

[0095] In the description of the specification, the description of the reference terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are contained in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0096] The above is only an example and description of the present application, and those skilled in the art can make various modifications or supplements or use similar ways to replace the described specific embodiments, as long as they do not deviate from the invention or exceed the scope defined by the present claims, which shall belong to the protection scope of the present application.

Claims

1. An intelligent monitoring and management system for pig breeding based on Internet of Things technology, characterized in that, The utility model relates to a pig house environment parameter monitoring system, including: Identity recognition module, including fixed on the pig's electronic ear tag and the card reader that can read the information in electronic ear tag; The identity recognition module can collect the body temperature, position and step counting data of live pig; When setting, three or more than three card readers are distributed and set in the pig house; Video acquisition module, for collecting the video image information inside the pig house; Alarm module, for sending alarm information; Deep learning module, for calculating the weight of live pig; The method for calculating the weight of live pig by the deep learning module includes the following steps: S1, through the identity recognition module and video acquisition module, the live pig in the pig house is positioned, and the identity information of each live pig in the image information collected by the video acquisition module is obtained; The positioning method is as follows: Obtain the relationship between the field strength and the distance between the electronic ear tag and the card reader; According to the relationship between the field strength and the distance between the electronic ear tag and the card reader and the detected field strength relationship between the electronic ear tag and each card reader, the distance between the electronic ear tag and each card reader is obtained; According to the distance between the electronic ear tag and each card reader, the coordinate position (a, b) of the live pig corresponding to the electronic ear tag is obtained, and the coordinate is marked as the first coordinate; Through the video acquisition module, the image information in the pig house is obtained; According to the position of each live pig in the image information, the coordinate position (a1, b1) of each live pig in the pig house is obtained, and the coordinate is marked as the second coordinate; For a first coordinate, the intersection coefficient c between the first coordinate and each second coordinate is calculated according to the formula The second coordinate corresponding to the minimum intersection coefficient c is selected as the merged coordinate of the first coordinate. The combined coordinates of each first coordinate are calculated, and the live pig at the corresponding position of the combined coordinate is matched with the first coordinate; S2, the card reader obtains a group of first coordinates every preset time t1, and obtains a group of second coordinates after obtaining the frame image through the video acquisition module, and the identity information of each live pig in the frame image is obtained according to the method in step S1; Through the deep learning module, the weight of each live pig in the frame image is obtained; S3, for a live pig, the weight obtained by the deep learning module in a period is sequentially marked as z1, z2, …, zn, n is the number of weight data of the live pig obtained by the deep learning module in the corresponding period; According to the group of weight data, the real weight zt of the corresponding live pig is calculated; The calculation method of the real weight zt is as follows: The standard deviation value C of the set of data z1 to zn is calculated according to the formula C = sqrt(∑(z - μ)2 / n) When C < C1 is established, zp is taken as the real weight zt of the corresponding live pig in the period; Otherwise, when C >= C1 is established, the corresponding zi value is deleted in the order of |zi-zp| from large to small until C < C1 is established, and the average value of the corresponding zi value which is not deleted is taken as the real weight zt of the corresponding live pig in the period; Wherein C1 is a preset value, 1 <= i <= n, zp=(z1+z2+、…、+zn) / n. 2.The intelligent monitoring and management system for pig breeding based on Internet of Things technology according to claim 1, characterized in that, The pig house is divided into several breeding areas, and each breeding area corresponds to three or more than three card readers. 3.The intelligent monitoring and management system for pig breeding based on Internet of Things technology according to claim 1, characterized in that, It also includes a sensor module for detecting environmental parameters in the pig house; The environmental parameters in the pig house are monitored through the sensor module; When the corresponding environmental parameter value is greater than the preset threshold value, the alarm module sends alarm information; The environmental parameters include temperature, humidity, ammonia concentration and hydrogen sulfide concentration.

4. The intelligent monitoring and management system for pig breeding based on Internet of Things technology according to claim 1, characterized in that, When the number of intersection coefficients c satisfying c≤cy corresponding to a first coordinate is two or more, the pig corresponding to the first coordinate is not subjected to subsequent processing. cy is a preset value. 5.The intelligent monitoring and management system for pig breeding based on Internet of Things technology according to claim 1, characterized in that, The deep learning module is further configured to identify pigs with abnormal body weight, and the identification method comprises the following steps: For a pig, the real body weight of the pig in the past k periods is sequentially labeled as u1, u2, …, uk in time sequence; A body weight change curve Q of the pig is generated in a two-dimensional rectangular coordinate system with time as the horizontal coordinate and real body weight as the vertical coordinate; The change curve of the pig is divided into several growth intervals in time sequence, and the body weight change amount h of each pig in each growth interval is obtained. When h≤hy, it is considered that the state of the corresponding pig in the corresponding growth interval is abnormal, and the alarm module sends an alarm information. hy is a preset value, and hy is different in different growth intervals.

Images