CN115968810A - Identification method and identification system for sick and epidemic pigs - Google Patents
Identification method and identification system for sick and epidemic pigs Download PDFInfo
- Publication number
- CN115968810A CN115968810A CN202310275810.7A CN202310275810A CN115968810A CN 115968810 A CN115968810 A CN 115968810A CN 202310275810 A CN202310275810 A CN 202310275810A CN 115968810 A CN115968810 A CN 115968810A
- Authority
- CN
- China
- Prior art keywords
- video
- data
- area
- live
- electronic tag
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/70—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in livestock or poultry
Abstract
The invention discloses a method and a system for identifying sick and epidemic live pigs. The identification method comprises the steps of collecting a first video K in a first breeding area, identifying abnormal behavior information of live pigs, and transferring the live pigs with the abnormal behavior information of the live pigs to a second breeding area. And the video acquisition unit in the second breeding area acquires a second video P, a plurality of body temperature identification areas of the second video P are identified according to the position relation between the electronic tag and the body temperature identification areas, second data are generated, and the second data are sent to the second transceiver. The identification method of the sick and epidemic live pig realizes the sharing of the RFID radio frequency data and the video data, and the body temperature identification area of the live pig is quickly identified by the aid of the first transceiver and the second transceiver for assisting in positioning the electronic tag. Whether the live pigs suffer from the diseases and the epidemics is judged by two screening modes, the inspection result is more accurate, and the breeding management personnel can judge the disease and the epidemic states of the live pigs in real time conveniently.
Description
Technical Field
The invention relates to an animal epidemic identification technology, in particular to an identification method and an identification system for an epidemic live pig.
Background
The modern breeding industry tends to the development trend of automation, integration of things and scale. In a modern pig breeding mode, establishing a traceable standardized breeding management system is particularly important, and scientific quarantine inspection is an important link for solving the problem of insufficient labor resources, improving pork food safety and guaranteeing the economy of a farm in pig breeding. In the prior art, automatic quarantine of sick and epidemic live pigs mainly takes identification of cough sound of live pigs as a main basis, for example, CN107799114A discloses a method and a system for identifying cough sound of pigs. The method has reference significance for inspecting the sick and epidemic pigs, but the cough serving as the only variable of the sick and epidemic pigs is relatively single, and the misjudgment rate is high. In the prior art, quarantine of sick and epidemic live pigs can be carried out by means of animal physical signs, environmental sensing, health data monitoring and the like, for example, CN114463701A discloses a method for collecting pig farm data according to multiple sensors and providing monitoring and early warning for animal breeding data after fusion. The method has the advantages of wide applicability and high cost investment, and the actual implementation needs to be adjusted by considering the characteristics of different cultured animals. The breeding data of the live pigs are more difficult to obtain than other livestock and poultry, so that the prior art needs to be further improved.
Disclosure of Invention
Aiming at the problems, the invention discloses a method for identifying sick and epidemic live pigs. According to the identification method, abnormal feeding, abnormal drinking and abnormal behaviors of the live pigs are identified according to optical sensing video data, and the abnormal live pigs are subjected to body temperature inspection in a body temperature identification area after primary screening to judge whether the body temperature of the abnormal live pigs is increased. The sick and epidemic live pigs are accurately judged through twice screening, and the abnormal behavior information and the body temperature of the live pigs are shared in real time based on the RFID radio frequency sensing network. Furthermore, the invention also discloses an identification system for realizing the identification method of the sick and epidemic live pigs.
The invention purpose of this application can be realized through the following technical scheme:
a method for identifying sick and epidemic pigs comprises the following steps:
step 1: respectively arranging a first breeding area and a second breeding area in the breeding areas, feeding n live pigs in the first breeding area, marking a unique electronic tag for each live pig, and storing specific information for identifying the live pigs by using the electronic tags;
step 2: obtaining a first video K of a first breeding area, marking n pigs in the first video K according to the position coordinates of the electronic tag, extracting a plurality of feature points, and constructing a first feature point set A = { a } by the feature points 1 ,a 2 ,...,a n }；
And step 3: calculating the live pig liveness based on the first video K, identifying abnormal behavior information of the live pig according to the live pig liveness, entering a step 4 if the abnormal behavior information exists, and otherwise, entering a step 2;
and 4, step 4: generating m pieces of first data according to the m pieces of abnormal behavior information, wherein m is less than or equal to n, sending the first data to a first transceiver in a first breeding area, transmitting the first data to an electronic tag corresponding to a live pig by the first transceiver, and distributing m live pigs to a second breeding area;
and 5: obtaining a second video P of a second breeding area, marking m pigs in the second video P according to the position coordinates of the electronic tag, extracting a plurality of feature points, and constructing a second feature point set B = { B } by the feature points 1 ,b 2 ,...,b n }；
Step 6: determining a position relation between any electronic tag and a live pig body temperature identification area based on a first video K, and extracting body temperature data T of a live pig corresponding to the electronic tag based on a second video P, wherein the position relation is a weighted directed distance H (A, B) = max (H (A, B), H (B, A)) of a first feature point set A and a second feature point set B, and H (A, B) is the directed distance from A to B
H (B, A) is the directed distance from B to A,
；
and 7: if at least one piece of body temperature data T is larger than the threshold value, entering a step 8, otherwise, entering a step 5;
and 8: generating u pieces of second data according to the u pieces of body temperature data T larger than the threshold, wherein u is less than or equal to m, and sending the second data to a second transceiver in a second breeding area;
and step 9: and the second transceiver transmits the second data to the corresponding electronic tag, and the electronic tag alarm lamp is normally on and displays the body temperature data T at the same time.
In the invention, in step 2, contour parameters of live pigs in the first video K are identified, the position coordinates of the body temperature identification area are determined according to the contour parameters, and then the position relation between the electronic tag and the body temperature identification area is established.
In the invention, in step 5, a plurality of body temperature identification areas of the second video P are identified according to the position relationship, and the maximum temperature of the body temperature identification areas is the body temperature data T of the live pig.
In the present invention, the first video K is optical image data, and the second video P is infrared image data.
In the invention, in step 6, a first frame S and a second frame S which are adjacent to each other in the first video K are extracted 0 Positioning the live pig and extracting the action information of the live pig, and the second frame S 1 Selecting a plurality of objects matched with the live pig characteristics, marking the objects by adopting a rectangular frame, and identifying the live pig objects in the plurality of objects.
In the invention, in step 3, a feeding area, a drinking area and an active area are defined in the first culture area, and feeding abnormality, drinking abnormality and behavior abnormality in the first culture area are identified based on a convolutional neural network.
An identification system for realizing the identification method of the sick and epidemic live pig comprises the following steps:
the electronic tag is fixed on the back of the pig ear;
the optical imaging unit is used for acquiring a first video K of the first culture area;
the infrared imaging unit is used for acquiring a second video P of the second culture area;
the control center calculates the live pig activity degree based on the first video K and extracts body temperature data T based on the second video P;
the first transceiver device is used for determining the position coordinates of n electronic tags in the first breeding area and sending first data to m electronic tags;
and the second transceiver determines the position coordinates of the m electronic tags in the second culture area and sends second data to the u electronic tags.
In the invention, the first transceiver and the second transceiver are RF devices, and the electronic tag comprises an RFID sensor, a display, an alarm lamp and a register.
In the invention, the identification system also comprises a terminal device, and the terminal device receives the first data or the second data of the control center and reads the specific information of the live pig from the electronic tag.
The implementation of the identification method and the identification system of the sick and epidemic live pigs has the advantages that: the method is characterized in that live pigs with abnormal feeding, drinking and behaviors are screened in the first breeding area through optical sensing video data, the temperature of a live pig body temperature identification area is detected for the live pigs with abnormal behavior information in the second breeding area through infrared heat sensing video data, whether the live pigs suffer from diseases or not is judged through two screening modes, the inspection result is more accurate, and the misdetection rate is lower. In addition, the identification method of the sick and epidemic live pigs realizes the sharing of the RFID radio frequency sensing network and quarantine data, and the live pigs are positioned by the aid of the first transceiver and the second transceiver, so that the body temperature identification areas of the live pigs are identified quickly, and the breeding management personnel can judge the sick and epidemic states of the live pigs in real time conveniently.
Drawings
FIG. 1 is a flow chart of the method for identifying sick and epidemic pigs of the present invention;
FIG. 2 is a simplified diagram of a video frame of a first video K of the present invention;
FIG. 3 is a schematic diagram illustrating the principle of the process of extracting each pixel point of the first video K according to the present invention;
FIG. 4 is a schematic diagram of extracting pig contour parameters from a first video K according to the present invention;
FIG. 5 is a schematic diagram of the present invention for extracting the live pig body temperature data from the second video P;
FIG. 6 is a schematic diagram of a body temperature identification zone determined from pig profile parameters according to the present invention;
fig. 7 is a hardware block diagram of the epidemic pig identification system of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
The method is characterized in that whether diseases exist or not is judged through abnormal behaviors of live pigs in live pig breeding, wherein the abnormal behaviors of the live pigs specifically comprise short eating time, more or less drinking times and low swinery activity. Because the abnormal behavior of the live pig can be used as a reference basis for the epidemic disease of the live pig, the body temperature of the live pig is judged through infrared imaging under the condition that the live pig is suspected to be infected with the epidemic disease, and the accuracy of automatic quarantine can be improved by determining whether the live pig is infected with the disease. The invention integrates the judgment process into video acquisition and radio frequency sensing positioning to construct an integrated identification method of the sick and epidemic live pigs.
Example one
Referring to fig. 1, the method for identifying a sick and epidemic live pig, which is described in detail in this embodiment, includes the following steps:
step 1: the method comprises the steps that a first breeding area and a second breeding area are respectively arranged in a breeding area, n pigs are fed in the first breeding area, a unique electronic tag is marked for each pig, and the electronic tag stores specific information for identifying the pigs. In a specific embodiment, a plurality of independent live pig breeding areas can be defined according to the live pig captive breeding proportion. N live pigs are fed in the first breeding area, a unique electronic tag is distributed to any one live pig, and the electronic tag stores specific information for identifying the live pigs. The n live pigs are imprinted with marks on the backs, and the marks are sorted according to letters or numbers and used for identifying the identities of the live pigs. The unique specific information of the live pig corresponds to the identification one by one, and the video acquisition unit stores an original image to a training set.
In this embodiment, at least one video capture unit is arranged in the first culture area, and the video capture unit is a single module formed by two independent cameras. In first breed district, video acquisition unit is arranged in first breed district central point and is put directly over, and the height is with covering whole live pig home range as the standard.
Step 2: and acquiring a first video K of the first breeding area, and marking n live pigs in the first video K according to the position coordinates of the electronic tag. The first video K is optical image data, a plurality of feature points are extracted, and a first feature point set A = { a } is constructed by the feature points 1 ,a 2 ,...,a n }。
Fig. 2 is a simplified diagram of one video frame of the first video K. In the embodiment, the same live pig between two adjacent frames in the first video K is randomly extracted for association, namely an initial frame S 0 Positioning the live pig i, extracting the action information of the live pig i, and comparing the action information with an initial frame S 0 Next adjacent frame S 1 Selecting a plurality of live pig objects matched with the characteristics of the live pigs, marking the live pig objects by adopting a rectangular frame, and in the positioning process, carrying out initial frame S 0 Each live pig in its next adjacent frame S 1 There is at least one live pig corresponding to it. The data processing module is provided with a plurality of independent storage areas, when the video acquisition unit positions a live pig, the acquired first video K is allocated with a unique code in each independent storage area, and the code is used for identifying the live pig object tracked by the first video K in the storage area.
And 3, step 3: calculating the live pig liveness based on the first video K, and identifying abnormal behavior information of the live pig according to the live pig liveness. And if the abnormal behavior information exists, entering the step 4, otherwise, entering the step 2. The abnormal behavior information comprises abnormal feeding, abnormal drinking and abnormal behavior. The specific identification method is described with reference to example two.
And 4, step 4: and generating m pieces of first data according to the m pieces of abnormal behavior information, wherein m is less than or equal to n, sending the first data to a first transceiver in the first breeding area, transmitting the first data to an electronic tag corresponding to the live pig by the first transceiver, and distributing the m live pigs to a second breeding area. After the electronic tag receives the first data, the electronic tag registers '1' in a register, and when the count in the register of the electronic tag is greater than or equal to '2', the abnormal behavior information of the live pigs is identified, and the m live pigs are distributed to a second breeding area.
And 5: and acquiring a second video P of the second breeding area, and marking m live pigs in the second video P according to the position coordinates of the electronic tag. The second video P is infrared image data, a plurality of feature points are extracted, and a second feature point set B = { B } is constructed by the feature points 1 ,b 2 ,...,b n }。
Step 6: determining a position relation between any electronic tag and a live pig body temperature identification area based on a first video K, and extracting body temperature data T of a live pig corresponding to the electronic tag based on a second video P, wherein the position relation is a weighted directed distance H (A, B) = max (H (A, B), H (B, A)) of a first feature point set A and a second feature point set B, and H (A, B) is a directed distance from A to B
H (B, A) is the directed distance from B to A,
the specific steps of the algorithm can be described with reference to the third embodiment. The body temperature identification area is a position where the body temperature of the live pig is abnormal when diseases and epidemics occur, such as the abdomen, the ear back and the like.
In the embodiment, a key frame containing complete live pig contour parameters in the first video K is extracted, and a live pig contour parameter foreground image is obtained based on SIFT feature point detection. In order to quickly identify the live pig contour, a rectangular frame in which the live pig contour is located needs to be preset. For example, live pig contour data having specificity information a001 includes a deflection angle θ, a center R, a length L, and a width D. And R is a model of the gravity center position vector coordinate of the live pig under a two-dimensional plane coordinate system. Setting a rectangular frame with color identification to mark the live pig contour, and recording live pig contour data information in a quadruple mode at the moment t: y (a 001, t) = (R, θ, L, D). The external length of the corresponding rectangular frame is (L plus delta L), delta L = L-L, and L is the body length of the live pig. The external width of the corresponding rectangular frame is (D +. DELTA.d), and the ratio of < DELTA.d = D-D. d is the body width of the live pig. The corresponding rectangular frame deflection angle α = θ; corresponding rectangular box center O = R. The rectangular frame data information is updated as follows: t (a 0001, T) = (O, α, (L + Δl), (D + Δd)).
And 7: the m electronic tags receive first data sent by the first transceiver, if at least one body temperature data T is larger than a threshold value, the step 8 is carried out, otherwise, the step 5 is carried out, the threshold value is the maximum value of the normal body temperature interval of the live pigs, and the threshold value is 39.5 ℃.
And 8: and generating u pieces of second data according to the u pieces of body temperature data T larger than the threshold, wherein u is less than or equal to m, and sending the second data to a second transceiver in the second breeding area.
And step 9: and the second transceiver transmits the second data to the corresponding electronic tag, and the electronic tag alarm lamp is normally on and displays the body temperature data T at the same time. In this embodiment, the first transceiver and the second transceiver each include a counter, and count "1" when the temperature T of the live pig body temperature identification area is greater than the threshold, reject the electronic tag request and return NACK when the counter is "0", otherwise receive the electronic tag request and read the specific information of the electronic tag, and return ACK.
Example two
In this embodiment, a method for identifying abnormal behavior information of a live pig is described in detail, and the process of identifying abnormal behavior information of a live pig includes three processes of sample training, live pig tracking, and live pig identification.
In this embodiment, the sample training selects a first video K of a live pig in a first breeding area for a period of time, generates an XML document of tag data for an image of any one frame, encodes features in abnormal behavior information of the live pig, and trains a detection network based on the encoding result.
The live pig tracking method includes the steps of generating a candidate area based on RP according to a first breeding area, inputting a first video K to obtain a plurality of feature maps, marking candidate objects with probability larger than 50% through a rectangular frame according to the RP in the feature maps, screening the candidate objects with probability 200 before the marking, predicting the peripheral positions of the rectangular frame of the candidate objects based on a trained detection network, and finally marking all the existing live pig objects on the feature maps.
In this embodiment, the live pig identification includes the contents of three modules, namely, feeding action, drinking action and behavior action. A feeding area, a drinking area and an active area are defined in the first breeding area, and the feeding time is judged by respectively identifying the head action of the pigs in the feeding area through a video acquisition unit; identifying the head action of the pig in the drinking area to judge the drinking times; and identifying the liveness of the live pigs in the activity area and judging whether the abnormal behaviors exist. The identification method of the feeding time of the live pigs in the feeding area comprises the following steps:
step 101: in a feeding period, the data processing module acquires a frame of image of any sequence in the first video K and identifies the head and the trough of a live pig in a feeding area;
step 102: marking the head characteristics of the live pigs in the feeding area, and marking the head of the live pigs in the feeding area and a rectangular frame through the rectangular frame;
step 103: calling current time information, if the current time is in a food taking time interval, entering a step 4, otherwise, entering a step 1;
step 104: if the rectangular frame mark of the head of the live pig is intersected with the rectangular frame of the trough, recording the frame of image, and storing the frame of image in a cache database, otherwise, deleting the frame of image;
step 105: and calculating the number of the continuous multiple frame images, and acquiring the feeding time of the pig to generate first data.
In this embodiment, the frame images in the buffer database correspond to unique specific information of the live pigs, the buffer database records the frame images within a fixed time, and when the frame images recorded in the buffer database are not read in a feeding period, the buffer frame image data are automatically cleared.
In the embodiment, the method for identifying the drinking time of the live pigs in the drinking areas comprises the following steps:
step 201: in a feeding period, the data processing module acquires a frame of image of any sequence in the first video K and identifies the head and the sink of a live pig in a drinking area;
step 202: marking the head characteristics of the live pigs in the drinking area, and marking the heads and the rectangular frames of the live pigs in the drinking area through the rectangular frames;
step 203: calling current time information, if the current time is in a food taking time interval, entering a step 4, otherwise, entering a step 1;
step 204: if the rectangular frame mark of the head of the live pig is intersected with the rectangular frame of the water tank, recording the frame of image, storing the frame of image in a cache database, and otherwise, deleting the frame of image;
step 205: and calculating the number of continuous multiple frame images, and acquiring the drinking time of the pigs to generate first data.
In this embodiment, the method for determining the liveness η of a live pig in an activity area includes the following steps:
step 301: at time t n The data processing module works, acquires a frame image of any sequence in the first video K, and extracts the information of each pixel point in the frame image;
step 302: any pixel point information is marked as Y (x) tn ，y tn ) Wherein x is tn Is the coordinate of the horizontal axis of the pixel point, y tn Is the coordinate of the longitudinal axis of the pixel, x tn ,y tn ∈N*；
Step 303: at time t n+1 The data processing module works, acquires a frame image of any sequence in the first video K, and extracts the information of each pixel point in the frame image;
step 304: any pixel point information is marked as Y (x) tn+1 ，y tn+1 ) Wherein x is tn+1 Is the coordinate of the horizontal axis of the pixel point, y tn+1 Is the coordinate of the longitudinal axis of the pixel, x tn ,y tn ∈N*；
Step 305: definition of t n+1 The liveness of the live pigs is determined in time.
In this embodiment, the data processing module extracts a frame image of any sequence in the first video K, and extracts information of each pixel point in the frame image as shown in fig. 3, at t n At any moment, if a certain pixel point information Y (x) tn ，y tn ) If the pixel point is not live pig activity, the pixel point is a first breeding area background; if a certain pixel point information Y (x) tn ，y tn ) And if the pixel point is not equal to 1, the live pig activity exists in the pixel point, and the pixel point is a foreground live pig image. Further, it is possible to provideAt t, at n+1 At a certain time, a certain pixel point Y (x) tn ，y tn ) And if the numerical value is changed from '0' to '1', recording the effective activity of the live pig with 1 pixel point.
EXAMPLE III
The embodiment further discloses a method for extracting the body temperature data T of the live pig corresponding to the electronic tag in the step 6.
Step 601: performing standard graying processing on the key frame of the first video K, and extracting edge feature points to obtain a feature point set A = { a } of live pig contour data 1 ,a 2 ,...,a n As shown in fig. 4. According to the physiological structure of the live pig, a plurality of radial auxiliary rays extending inwards from the characteristic point set A are generated according to the deflection angle beta, and the intersection point of the radial auxiliary rays is determined as a point set V (x) of the position coordinates of the body temperature identification area t ,y t ) The abdominal region shown in fig. 5 is a body temperature identification region.
Step 602: according to the first transceiver, the coordinate point set B of the electronic tag of the live pig in the first video K can be determined, so that the point set V (x) is calculated t ,y t ) And the position relation with the coordinate point set B. The position relationship of this embodiment is the directed distance h (V, B) of the point set,
。
step 603: extracting a key frame containing the whole live pig in the second video P, and extracting a plurality of point sets V of high-temperature intervals in the key frame i (x t ,y t ) And i is the number of high temperature intervals. And simultaneously extracting a coordinate point set B' of the live pig electronic tag of the second video P.
Step 604: set of computation points V i (x t ,y t ) Directed distance h (V) from coordinate point set B i And B'). Determining a directed distance h (V) with a minimum deviation of the directed distances h (V, B) i And B'), the corresponding high temperature zone is a body temperature identification zone, as shown in fig. 6.
Example four
Referring to fig. 7, the identification system for implementing the identification method of sick and epidemic pigs of the present invention comprises: the device comprises an electronic tag fixed on the back of a pig ear, an optical imaging unit, an infrared imaging unit, a control center, a first transceiver, a second transceiver and a terminal device. The optical imaging unit is used for acquiring a first video K of the first culture area. The infrared imaging unit is used for acquiring a second video P of the second culture area. The control center calculates the live pig liveness based on the first video K and extracts body temperature data T based on the second video P; the first transceiver determines the position coordinates of n electronic tags in the first breeding area, and sends first data to m electronic tags.
The first transceiver and the second transceiver are RF devices, and the electronic tag comprises an RFID sensor, a display, an alarm lamp and a register. And the second transceiver determines the position coordinates of the m electronic tags in the second breeding area, and sends second data to the u electronic tags. And the terminal equipment receives the first data or the second data of the control center and reads the specific information of the live pigs from the electronic tags. In the present invention, the first transceiver device and the second transceiver device may also be used to locate the electronic tag, so as to determine the position of the corresponding live pig in the first video K or the second video P. When the body temperature identification area of the live pig is determined, the electronic tag is matched with the profile data of the live pig.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and should not be taken as limiting the scope of the present invention, which is intended to cover any modifications, equivalents and improvements made within the spirit and scope of the present invention.
Claims (9)
1. A method for identifying sick and epidemic live pigs is characterized by comprising the following steps:
step 1: respectively arranging a first breeding area and a second breeding area in the breeding areas, feeding n pigs in the first breeding area, marking a unique electronic tag for each pig, and storing specific information for identifying the pigs by the electronic tag;
step 2: obtaining a first video K of a first breeding area, marking n live pigs in the first video K according to position coordinates of an electronic tag, extracting a plurality of feature points, and constructing a first feature point set A = { a } by the plurality of feature points 1 ,a 2 ,...,a n }；
And step 3: calculating the live pig liveness based on the first video K, identifying abnormal behavior information of the live pig according to the live pig liveness, entering a step 4 if the abnormal behavior information exists, and otherwise, entering a step 2;
and 4, step 4: generating m pieces of first data according to the m pieces of abnormal behavior information, wherein m is less than or equal to n, sending the first data to a first transceiver in a first breeding area, transmitting the first data to an electronic tag corresponding to a live pig by the first transceiver, and distributing m live pigs to a second breeding area;
and 5: obtaining a second video P of a second breeding area, marking m pigs in the second video P according to the position coordinates of the electronic tag, extracting a plurality of feature points, and constructing a second feature point set B = { B } by the plurality of feature points 1 ,b 2 ,...,b n }；
Step 6: determining a position relation between any electronic tag and a live pig body temperature identification area based on a first video K, and extracting body temperature data T of a live pig corresponding to the electronic tag based on a second video P, wherein the position relation is a weighted directed distance H (A, B) = max (H (A, B), H (B, A)) of a first feature point set A and a second feature point set B, and H (A, B) is a directed distance from A to B
H (B, A) is the directed distance from B to A,
；
and 7: if at least one piece of body temperature data T is larger than the threshold value, entering a step 8, otherwise, entering a step 5;
and step 8: generating u pieces of second data according to u pieces of body temperature data T larger than the threshold, wherein u is less than or equal to m, and sending the second data to a second transceiver in a second breeding area;
and step 9: and the second transceiver transmits the second data to the corresponding electronic tag, and the electronic tag alarm lamp is normally on and displays the body temperature data T at the same time.
2. The method according to claim 1, wherein in step 2, contour parameters of the live pigs in the first video K are identified, the position coordinates of the body temperature identification area are determined according to the contour parameters, and the position relationship between the electronic tag and the body temperature identification area is established.
3. The method according to claim 2, wherein in step 5, the plurality of body temperature identification areas of the second video P are identified according to the position relationship, and the maximum temperature of the body temperature identification area is the body temperature data T of the live pig.
4. The method according to claim 1, wherein the first video K is optical image data, and the second video P is infrared image data.
5. The method according to claim 1, wherein in step 6, the first frame S and the second frame S adjacent to each other in the first video K are extracted 0 Positioning the live pig and extracting the action information of the live pig, and the second frame S 1 Selecting a plurality of objects matched with the live pig characteristics, marking the objects by adopting a rectangular frame, and identifying the live pig objects in the plurality of objects.
6. The method according to claim 1, wherein in step 3, a feeding area, a drinking area and an active area are defined in the first breeding area, and feeding abnormality, drinking abnormality and behavior abnormality in the first breeding area are identified based on a convolutional neural network.
7. An identification system for realizing the method for identifying a sick and epidemic pig according to claim 1, comprising:
the electronic tag is fixed on the back of the pig ear;
the optical imaging unit is used for acquiring a first video K of the first culture area;
the infrared imaging unit is used for acquiring a second video P of the second culture area;
the control center calculates the live pig liveness based on the first video K and extracts body temperature data T based on the second video P;
the first transceiver device is used for determining the position coordinates of n electronic tags in the first breeding area and sending first data to m electronic tags;
and the second transceiver determines the position coordinates of the m electronic tags in the second culture area and sends second data to the u electronic tags.
8. The identification system of claim 7, wherein the first and second transceiver devices are RF radio frequency devices and the electronic tag comprises an RFID sensor, a display, an alarm light, and a register.
9. The identification system according to claim 7, characterized in that the identification system further comprises a terminal device, the terminal device receiving the first data or the second data of the control center and reading the pig specific information from the electronic tag.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310275810.7A CN115968810B (en) | 2023-03-21 | 2023-03-21 | Identification method and identification system for epidemic pigs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310275810.7A CN115968810B (en) | 2023-03-21 | 2023-03-21 | Identification method and identification system for epidemic pigs |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115968810A true CN115968810A (en) | 2023-04-18 |
| CN115968810B CN115968810B (en) | 2023-06-27 |
Family
ID=85974527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310275810.7A Active CN115968810B (en) | 2023-03-21 | 2023-03-21 | Identification method and identification system for epidemic pigs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115968810B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150282457A1 (en) * | 2014-04-08 | 2015-10-08 | Medisim, Ltd. | Cattle monitoring for illness |
| CN108990831A (en) * | 2018-06-22 | 2018-12-14 | 成都睿畜电子科技有限公司 | A kind of animal health monitoring method and system |
| CN110402840A (en) * | 2019-07-25 | 2019-11-05 | 深圳市阿龙电子有限公司 | A kind of live pig monitoring terminal and live pig monitoring system based on image recognition |
| CN110495405A (en) * | 2019-07-12 | 2019-11-26 | 中国农业大学 | A kind of pig-breeding monitoring node, system and method |
| US20210289746A1 (en) * | 2019-03-22 | 2021-09-23 | Arizona Board Of Regents On Behalf Of Arizone State University | Systems, methods, and apparatuses for implementing real time beef cattle monitoring utilizing radio-frequency identification (rfid) based technologies |
| CN113869130A (en) * | 2021-08-31 | 2021-12-31 | 江苏师范大学 | Binocular vision intelligent detection method and system |
| CN114155216A (en) * | 2021-11-26 | 2022-03-08 | 北京小龙潜行科技有限公司 | Pig temperature detection method and device |
| KR20220039373A (en) * | 2020-09-22 | 2022-03-29 | 주식회사 핀텔 | Smart pigsty system and pigsty monitoring method using the same |
| CN114287357A (en) * | 2021-12-31 | 2022-04-08 | 双胞胎（集团）股份有限公司 | Intelligent live pig breeding service platform |
| CN114639017A (en) * | 2022-03-04 | 2022-06-17 | 华南农业大学 | Pig health monitoring method and system based on sound and body temperature |
| US20220284725A1 (en) * | 2019-08-21 | 2022-09-08 | Dairymaster | A method and apparatus for determining the identity of an animal of a herd of animals |
| CN115683355A (en) * | 2022-10-21 | 2023-02-03 | 李毅成 | Pig farm body temperature monitoring system based on Internet of things |
-
2023
- 2023-03-21 CN CN202310275810.7A patent/CN115968810B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150282457A1 (en) * | 2014-04-08 | 2015-10-08 | Medisim, Ltd. | Cattle monitoring for illness |
| CN108990831A (en) * | 2018-06-22 | 2018-12-14 | 成都睿畜电子科技有限公司 | A kind of animal health monitoring method and system |
| US20210289746A1 (en) * | 2019-03-22 | 2021-09-23 | Arizona Board Of Regents On Behalf Of Arizone State University | Systems, methods, and apparatuses for implementing real time beef cattle monitoring utilizing radio-frequency identification (rfid) based technologies |
| CN110495405A (en) * | 2019-07-12 | 2019-11-26 | 中国农业大学 | A kind of pig-breeding monitoring node, system and method |
| CN110402840A (en) * | 2019-07-25 | 2019-11-05 | 深圳市阿龙电子有限公司 | A kind of live pig monitoring terminal and live pig monitoring system based on image recognition |
| US20220284725A1 (en) * | 2019-08-21 | 2022-09-08 | Dairymaster | A method and apparatus for determining the identity of an animal of a herd of animals |
| KR20220039373A (en) * | 2020-09-22 | 2022-03-29 | 주식회사 핀텔 | Smart pigsty system and pigsty monitoring method using the same |
| CN113869130A (en) * | 2021-08-31 | 2021-12-31 | 江苏师范大学 | Binocular vision intelligent detection method and system |
| CN114155216A (en) * | 2021-11-26 | 2022-03-08 | 北京小龙潜行科技有限公司 | Pig temperature detection method and device |
| CN114287357A (en) * | 2021-12-31 | 2022-04-08 | 双胞胎（集团）股份有限公司 | Intelligent live pig breeding service platform |
| CN114639017A (en) * | 2022-03-04 | 2022-06-17 | 华南农业大学 | Pig health monitoring method and system based on sound and body temperature |
| CN115683355A (en) * | 2022-10-21 | 2023-02-03 | 李毅成 | Pig farm body temperature monitoring system based on Internet of things |
Non-Patent Citations (2)
| Title |
|---|
| 任春江;陶松兵;吴学栋;: "规模化生猪养殖场的疫病防护机器人应用研究", 科技视界, no. 19, pages 254 - 256 * |
| 杨威;俞守华;: "基于机器视觉的圈养豪猪检测与基本行为识别方法研究", 福建农业学报, vol. 32, no. 09, pages 1021 - 1025 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115968810B (en) | 2023-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6203238B2 (en) | Livestock management system | |
| CN108990831A (en) | A kind of animal health monitoring method and system | |
| MX2015006775A (en) | Systems and methods for predicting the outcome of a state of a subject. | |
| US20210192736A1 (en) | System for high performance, ai-based dairy herd management and disease detection | |
| CN113662530B (en) | Pig physiological growth state monitoring and early warning method | |
| KR102506029B1 (en) | Apparatus and method for monitoring growing progress of livestock individual based on image | |
| US20210216758A1 (en) | Animal information management system and animal information management method | |
| CN111767794A (en) | Cage-rearing poultry abnormal behavior detection method and detection system based on machine vision | |
| US20230260327A1 (en) | Autonomous livestock monitoring | |
| KR102522239B1 (en) | Apparatus and method for analyzing feeding behavior of livestock based on image | |
| CN112101333A (en) | Smart cattle farm monitoring and identifying method and device based on deep learning | |
| CN115830490A (en) | Multi-target tracking and behavior statistical method for herd health pigs | |
| CN111797831A (en) | BIM and artificial intelligence based parallel abnormality detection method for poultry feeding | |
| Doornweerd et al. | Passive radio frequency identification and video tracking for the determination of location and movement of broilers | |
| González et al. | Real-time monitoring of poultry activity in breeding farms | |
| CN115968810A (en) | Identification method and identification system for sick and epidemic pigs | |
| JP2011135786A (en) | Automatic weight selector and program | |
| CA3230401A1 (en) | Systems and methods for the automated monitoring of animal physiological conditions and for the prediction of animal phenotypes and health outcomes | |
| CN113068657B (en) | Intelligent efficient pig raising method and system | |
| CN115457468A (en) | Intelligent livestock monitoring method and system for large grassland | |
| Alon et al. | Machine vision-based automatic lamb identification and drinking activity in a commercial farm | |
| CN109002791B (en) | System and method for automatically tracking rumination behavior of dairy cow based on video | |
| Bastiaansen et al. | Continuous real-time cow identification by reading ear tags from live-stream video | |
| Labrecque et al. | Real-time individual pig tracking and behavioural metrics collection with affordable security cameras | |
| Xiong et al. | Estimating body weight and body condition score of mature beef cows using depth images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |