CN111294565A - Intelligent pig raising monitoring method and management terminal - Google Patents
Intelligent pig raising monitoring method and management terminal Download PDFInfo
- Publication number
- CN111294565A CN111294565A CN202010159849.9A CN202010159849A CN111294565A CN 111294565 A CN111294565 A CN 111294565A CN 202010159849 A CN202010159849 A CN 202010159849A CN 111294565 A CN111294565 A CN 111294565A
- Authority
- CN
- China
- Prior art keywords
- pig
- live
- live pig
- individual
- behavior
- 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.)
- Pending
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 214
- 238000000034 method Methods 0.000 title claims abstract description 82
- 241000282887 Suidae Species 0.000 claims abstract description 87
- 238000009395 breeding Methods 0.000 claims abstract description 25
- 230000001488 breeding effect Effects 0.000 claims abstract description 25
- 241000282898 Sus scrofa Species 0.000 claims description 795
- 230000006399 behavior Effects 0.000 claims description 178
- 238000012545 processing Methods 0.000 claims description 67
- 230000003068 static effect Effects 0.000 claims description 35
- 235000005809 Carpobrotus aequilaterus Nutrition 0.000 claims description 27
- 235000004550 Disphyma australe Nutrition 0.000 claims description 27
- 244000187801 Carpobrotus edulis Species 0.000 claims description 25
- 238000004590 computer program Methods 0.000 claims description 22
- 238000004458 analytical method Methods 0.000 claims description 18
- 230000037237 body shape Effects 0.000 claims description 8
- 239000012634 fragment Substances 0.000 claims description 8
- 238000003062 neural network model Methods 0.000 claims description 8
- 238000013310 pig model Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 7
- 238000007493 shaping process Methods 0.000 claims description 5
- 230000007774 longterm Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 241000200554 Disphyma crassifolium Species 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000004634 feeding behavior Effects 0.000 description 2
- 235000015277 pork Nutrition 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000036760 body temperature Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009313 farming Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 235000013372 meat Nutrition 0.000 description 1
- 238000003307 slaughter Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01K—ANIMAL HUSBANDRY; AVICULTURE; APICULTURE; PISCICULTURE; FISHING; REARING OR BREEDING ANIMALS, NOT OTHERWISE PROVIDED FOR; NEW BREEDS OF ANIMALS
- A01K29/00—Other apparatus for animal husbandry
- A01K29/005—Monitoring or measuring activity, e.g. detecting heat or mating
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Environmental Sciences (AREA)
- Biophysics (AREA)
- Animal Husbandry (AREA)
- Biodiversity & Conservation Biology (AREA)
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an intelligent pig-raising monitoring method and a management terminal, which are used for acquiring a live pig monitoring video of live pigs in a breeding farm in the breeding process; identifying the walking track and the performance behavior of each live pig in the live pig monitoring video; and obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig. The invention carries out real-time monitoring on the video of the live pigs in the breeding process, so that farmers can observe the growth condition of the live pigs in real time to judge the health state of the live pigs; meanwhile, the walking track and the performance behavior of each live pig are automatically analyzed according to the monitored video system, so that the health state of each live pig is obtained, and intelligent real-time health monitoring is automatically provided for farmers.
Description
Technical Field
The invention relates to the technical field of intelligent breeding, in particular to an intelligent pig-raising monitoring method and a management terminal.
Background
With the rapid development of domestic economy and the increase of disposable income of people, the consumption of meat by people also increases year by year. Among them, pork is one of the most common animal food on dining tables.
The annual increase in pork consumption presents a significant challenge to the traditional live pig farming industry. The raising cost of the live pigs gradually increases, and the profit of many traditional pig farmers is in a downward trend due to factors such as piglet cost, feed cost for feeding the live pigs and the like, labor cost, live pig marketing time and the like. Therefore, artificial intelligence begins to enter the field of live pig breeding. The current intelligent pig raising field mainly includes: establishing a set of files for each pig by using the intelligent ear tags, thereby recording data of processes such as breeding, quarantine, slaughtering and harmlessness; by means of a sensing technology, physiological data of each pig are monitored and sensed in real time by means of AI and big data, health monitoring of the pigs is achieved, epidemic situation early warning, productivity prediction and the like are achieved through data analysis; real-time monitoring is carried out through videos so as to realize large-scale pig raising monitoring; however, the existing health monitoring is based on physiological data, usually body temperature data, and the judgment basis is single and the judgment result is one-sided.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the intelligent pig raising monitoring method and the intelligent pig raising management terminal are provided, so that more intelligent real-time health monitoring is carried out on the pig raising process.
In order to solve the technical problems, the invention adopts the technical scheme that:
an intelligent pig raising monitoring method comprises the following steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
In order to solve the technical problem, the invention adopts another technical scheme as follows:
an intelligent pig raising monitoring management terminal comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the computer program to realize the following steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
The invention has the beneficial effects that: an intelligent pig-raising monitoring method and a management terminal monitor live pigs in real time in a breeding process, so that farmers can observe the growth conditions of the live pigs in real time to judge the health states of the live pigs; meanwhile, the walking track and the performance behavior of each live pig are automatically analyzed according to the monitored video system, so that the health state of each live pig is obtained, and intelligent real-time health monitoring is automatically provided for farmers.
Drawings
FIG. 1 is a schematic flow chart of an intelligent pig-raising monitoring method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an intelligent pig raising monitoring management terminal according to an embodiment of the invention.
Description of reference numerals:
1. an intelligent pig raising monitoring management terminal; 2. a processor; 3. a memory; 4. a first camera; 5. a second camera; 6. and a third camera.
Detailed Description
In order to explain technical contents, achieved objects, and effects of the present invention in detail, the following description is made with reference to the accompanying drawings in combination with the embodiments.
Referring to fig. 1, an intelligent pig raising monitoring method includes the steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
From the above description, the beneficial effects of the present invention are: the live pigs are monitored in real time in a video mode in the breeding process, so that farmers can observe the growth conditions of the live pigs in real time to judge the health states of the live pigs; meanwhile, the walking track and the performance behavior of each live pig are automatically analyzed according to the monitored video system, so that the health state of each live pig is obtained, and intelligent real-time health monitoring is automatically provided for farmers.
Further, the step S2 specifically includes the following steps:
s21, extracting N frames of live pig monitoring images with intervals of a first preset time from the live pig monitoring video to obtain a track generation image set, extracting M frames of continuous live pig monitoring images from the live pig monitoring video to obtain a behavior analysis image set, dividing each frame of the live pig monitoring images in the track generation image set and the behavior analysis image set into X individual pig images corresponding to the X live pigs one by one, recording the relative position information of each individual pig image in the live pig monitoring images and the relative time information of each live pig monitoring image in the live pig monitoring video, placing the X individual pig images in the same frame into the same pig colony data set to obtain N pig colony data sets corresponding to the track generation image set and executing the step S22 on the pig colony data sets, and obtaining M of said swine population data sets corresponding to said behavior analysis image set and performing step S23 thereon, said N being greater than or equal to 1 and said M being greater than or equal to 16;
s22, carrying out pig individual identification on X individual pig images in N individual pig group data sets corresponding to the track generation image set in sequence to obtain pig individual identity information corresponding to each individual pig image, classifying the N individual pig images with the same pig individual identity information to obtain a first individual pig data set corresponding to the pig individual identity information, and generating the walking track of each live pig according to the relative position information and the relative time information of each individual pig image in the first individual pig data set;
s23, carrying out pig individual identification and limb form identification on X individual pig images in M individual pig group data sets corresponding to the behavior analysis image set in sequence to obtain pig individual identity information and current limb form corresponding to each individual pig image, classifying the M individual pig images with the same pig individual identity information to obtain a second individual pig data set corresponding to the pig individual identity information, and generating the performance behavior of each pig according to the relative position information, the relative time information and the current limb form of each individual pig image in the second individual pig data set.
As can be seen from the above description, the position information of the live pig is obtained every first preset time to generate the walking track of the individual pig; identifying the current limb form of each individual pig within continuous time, and matching with corresponding position information to identify the performance of the live pig; the image frames with different time requirements are identified according to the walking track and the expression behavior, so that the accuracy of the identification result is ensured, and the calculated amount in the identification process is reduced; meanwhile, the form of the limb of the individual pig is used as a real-time variable quantity, the walking track is used as a long-term variable quantity, and the live pig is subjected to more accurate health monitoring by combining the real-time change condition and the long-term change condition of the live pig.
Further, the step S2 specifically includes the following steps:
s21', starting to judge whether a clear pig face exists from a first frame of live pig monitoring image in the live pig monitoring video until a first live pig monitoring image including the clear pig face is obtained, dividing the first live pig monitoring image including the clear pig face into a first pig individual image including the clear pig face, carrying out pig individual identification on the first pig individual image to obtain first pig individual identity information corresponding to the first pig individual image, carrying out target tracking on a first live pig corresponding to the first pig individual identity information so as to mark the first live pig in each frame of live pig monitoring image in the live pig monitoring video, and then continuing to judge whether the clear pig face exists so as to mark each live pig in each frame of live pig monitoring image in the live pig monitoring video;
s22', connecting the relative position information of each live pig in each frame of live pig monitoring image in the live pig monitoring video according to the time sequence of the live pig monitoring video to obtain the walking track of each live pig;
s23', extracting M continuous live pig monitoring images from the live pig monitoring video, dividing each extracted live pig monitoring image into X individual pig images corresponding to the X live pigs one by one, recording the marks of the live pigs corresponding to the individual pig images, the relative position information of the marks on the live pig monitoring images and the relative time information of the live pig monitoring images in the live pig monitoring video, carrying out limb shape identification on each individual pig image to obtain the current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same marks to obtain a second individual pig data set corresponding to the marks, and carrying out body shape identification according to the relative position information of each individual pig image in the second individual pig data set, And generating the performance behavior of each live pig according to the relative time information and the current limb form, wherein M is greater than or equal to 16.
According to the description, the individual pigs are tracked and locked in real time, so that the position information of each live pig in each frame of live pig monitoring image is quickly locked, different live pigs are distinguished through marks, and therefore identity recognition is not needed during subsequent track generation and behavior recognition, the calculated amount of a system is reduced, the calculating speed is improved, and the walking track and the expression behavior can be obtained in real time.
Further, the step S1 is preceded by the following steps:
acquiring entry videos shot by the same live pig by first cameras respectively positioned at different positions of an entrance in a farm, wherein the entry videos comprise static segments of the same live pig in a standing state and behavior segments during different performance behaviors, the positions of the first cameras at least comprise the front, the left, the right, the back and the upper of the live pig, and each behavior segment corresponds to one performance behavior respectively;
extracting a plurality of pig individual pictures of the live pig in a standing state from the static fragment, and reconstructing and shaping a preset pig individual three-dimensional model according to the plurality of pig individual pictures to obtain a live pig three-dimensional model corresponding to the live pig, wherein the plurality of pig individual pictures of the live pig in the standing state at least comprise a live pig static front picture, a live pig static left side picture, a live pig static right side picture, a live pig static rear view picture and a live pig static top view picture;
sequentially acquiring a live pig behavior front picture, a live pig behavior left picture, a live pig behavior right picture, a live pig behavior rear view picture and a live pig behavior overlooking picture which are positioned at the same time point for behavior clips shot in different directions of the same performance behavior according to a time axis sequence, and simultaneously processing the live pig three-dimensional model at the current time point according to the live pig behavior front picture, the live pig behavior left picture, the live pig behavior right picture, the live pig behavior rear view picture and the live pig behavior overlooking picture at the same time point and recording the processing process of the live pig three-dimensional model to obtain the processing characteristics of each performance behavior on the live pig three-dimensional model;
the step S1 specifically includes the following steps:
acquiring live pig monitoring videos sent by a second camera positioned above a pigsty and a third camera positioned in front of the pigsty in the farm, wherein the front of the pigsty is a position on a pig feeding groove in the pigsty, which is as high as the live pigs;
the step S2 of "recognizing the performance behavior of each live pig in the live pig monitoring video" specifically includes the following steps:
extracting individual pig top-view pictures from the live pig monitoring video sent by the second camera, and extracting individual pig front-view pictures from the live pig monitoring video sent by the third camera to obtain individual pig top-view pictures of each live pig in the live pig monitoring video and individual pig front-view pictures corresponding to a time axis;
and processing the three-dimensional pig model corresponding to each live pig in real time according to the pig individual overlooking picture and the pig individual front picture corresponding to each live pig to obtain a real-time processing process of each live pig, and obtaining the performance of each live pig according to the similarity between the real-time processing process of each live pig and the processing characteristics of each performance.
As can be seen from the above description, the video of the standing live pig is shot in all directions by the plurality of first cameras, so as to establish a three-dimensional model conforming to the actual situation of the live pig; the method is characterized in that videos of live pigs under different performances are shot in an omnibearing manner through a plurality of first cameras to simulate three-dimensional changes of the live pigs under the different performances, wherein the three-dimensional changes are changes of body shapes of the live pigs in the whole process of finishing one performance in a standing state, such as feeding, and the most obvious characteristic is that the live pigs need to bend down the heads of the pigs and keep a certain time to finish feeding. In the real-time monitoring process, at least one upper top view and at least one front view are needed, the top views can avoid crowding of live pigs and can not obtain effective information of each live pig, and at least a body shape contour map of the live pigs can be obtained; the third camera positioned on the pig feeding groove in the pigsty and at the position with the same height as the live pigs can clearly see the pig faces and the limb shapes of the live pigs in the feeding behavior in the feeding process of the live pigs, and can better acquire the limb shapes of the live pigs in other expressive behaviors so as to acquire the limb shapes of each live pig in the real-time breeding process as accurately as possible; the obtained front picture and the overlook picture are used for processing the three-dimensional pig model, the judgment processing process is compared with the processing characteristics recorded before, so that the performance of the current live pig is judged, namely the body change of a certain behavior of the live pig can be better reflected through three-dimensional simulation, and the performance of the live pig can be better identified under the condition that the complete limb form of each live pig is difficult to obtain.
Further, the live pigs comprise healthy pig individuals and pig individuals with different types of attacks;
the step of obtaining the processing characteristics of each performance behavior on the live pig three-dimensional model specifically comprises the following steps:
taking the processing process and the walking track of each live pig as input parameters, taking the health state and the performance behavior of each live pig as output parameters, and taking the processing process, the walking track, the health state and the performance behavior of each live pig marked manually as a training set of a neural network model to obtain a live pig health recognition model;
in the step S2, "the performance of each live pig is obtained according to the degree of similarity between the real-time processing procedure of each live pig and the processing characteristics of each performance" and the step S3 is replaced with the following steps:
and respectively inputting the real-time processing process and the walking track of each live pig into the live pig health recognition model to obtain the performance behavior and the health state of each live pig.
According to the description, based on big data, live pigs regardless of health or diseases are subjected to data collection, are distinguished through manual marking, and are placed into a neural network model as a training set to obtain a live pig health recognition model, the live pig health recognition model can automatically extract feature data of the live pigs under different expressive behaviors and feature data of the live pigs under different health conditions under the same expressive behavior, and accordingly the expressive behaviors and health states of the live pigs can be rapidly and accurately recognized according to a real-time received processing process.
Referring to fig. 2, an intelligent pig raising monitoring management terminal includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor executes the computer program to implement the following steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
From the above description, the beneficial effects of the present invention are: the live pigs are monitored in real time in a video mode in the breeding process, so that farmers can observe the growth conditions of the live pigs in real time to judge the health states of the live pigs; meanwhile, the walking track and the performance behavior of each live pig are automatically analyzed according to the monitored video system, so that the health state of each live pig is obtained, and intelligent real-time health monitoring is automatically provided for farmers.
Further, the step S2 of the computer program executed by the processor specifically includes the following steps:
s21, extracting N frames of live pig monitoring images with intervals of a first preset time from the live pig monitoring video to obtain a track generation image set, extracting M frames of continuous live pig monitoring images from the live pig monitoring video to obtain a behavior analysis image set, dividing each frame of the live pig monitoring images in the track generation image set and the behavior analysis image set into X individual pig images corresponding to the X live pigs one by one, recording the relative position information of each individual pig image in the live pig monitoring images and the relative time information of each live pig monitoring image in the live pig monitoring video, placing the X individual pig images in the same frame into the same pig colony data set to obtain N pig colony data sets corresponding to the track generation image set and executing the step S22 on the pig colony data sets, and obtaining M of said swine population data sets corresponding to said behavior analysis image set and performing step S23 thereon, said N being greater than or equal to 1 and said M being greater than or equal to 16;
s22, carrying out pig individual identification on X individual pig images in N individual pig group data sets corresponding to the track generation image set in sequence to obtain pig individual identity information corresponding to each individual pig image, classifying the N individual pig images with the same pig individual identity information to obtain a first individual pig data set corresponding to the pig individual identity information, and generating the walking track of each live pig according to the relative position information and the relative time information of each individual pig image in the first individual pig data set;
s23, carrying out pig individual identification and limb form identification on X individual pig images in M individual pig group data sets corresponding to the behavior analysis image set in sequence to obtain pig individual identity information and current limb form corresponding to each individual pig image, classifying the M individual pig images with the same pig individual identity information to obtain a second individual pig data set corresponding to the pig individual identity information, and generating the performance behavior of each pig according to the relative position information, the relative time information and the current limb form of each individual pig image in the second individual pig data set.
As can be seen from the above description, the position information of the live pig is obtained every first preset time to generate the walking track of the individual pig; identifying the current limb form of each individual pig within continuous time, and matching with corresponding position information to identify the performance of the live pig; the image frames with different time requirements are identified according to the walking track and the expression behavior, so that the accuracy of the identification result is ensured, and the calculated amount in the identification process is reduced; meanwhile, the form of the limb of the individual pig is used as a real-time variable quantity, the walking track is used as a long-term variable quantity, and the live pig is subjected to more accurate health monitoring by combining the real-time change condition and the long-term change condition of the live pig.
Further, the step S2 of the computer program executed by the processor specifically includes the following steps:
s21', starting to judge whether a clear pig face exists from a first frame of live pig monitoring image in the live pig monitoring video until a first live pig monitoring image including the clear pig face is obtained, dividing the first live pig monitoring image including the clear pig face into a first pig individual image including the clear pig face, carrying out pig individual identification on the first pig individual image to obtain first pig individual identity information corresponding to the first pig individual image, carrying out target tracking on a first live pig corresponding to the first pig individual identity information so as to mark the first live pig in each frame of live pig monitoring image in the live pig monitoring video, and then continuing to judge whether the clear pig face exists so as to mark each live pig in each frame of live pig monitoring image in the live pig monitoring video;
s22', connecting the relative position information of each live pig in each frame of live pig monitoring image in the live pig monitoring video according to the time sequence of the live pig monitoring video to obtain the walking track of each live pig;
s23', extracting M continuous live pig monitoring images from the live pig monitoring video, dividing each extracted live pig monitoring image into X individual pig images corresponding to the X live pigs one by one, recording the marks of the live pigs corresponding to the individual pig images, the relative position information of the marks on the live pig monitoring images and the relative time information of the live pig monitoring images in the live pig monitoring video, carrying out limb shape identification on each individual pig image to obtain the current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same marks to obtain a second individual pig data set corresponding to the marks, and carrying out body shape identification according to the relative position information of each individual pig image in the second individual pig data set, And generating the performance behavior of each live pig according to the relative time information and the current limb form, wherein M is greater than or equal to 16.
According to the description, the individual pigs are tracked and locked in real time, so that the position information of each live pig in each frame of live pig monitoring image is quickly locked, different live pigs are distinguished through marks, and therefore identity recognition is not needed during subsequent track generation and behavior recognition, the calculated amount of a system is reduced, the calculating speed is improved, and the walking track and the expression behavior can be obtained in real time.
Further, the processor is respectively connected with a first camera positioned at an inlet of the farm in different directions, a second camera positioned above a swinery in the farm and a third camera positioned in front of the swinery in the farm, and the directions of the first cameras at least comprise the front, the left, the right, the back and the upper parts of the live pigs;
the processor further comprises the following steps before executing the step S1 of the computer program:
acquiring an entrance video shot by the first camera for the same live pig, wherein the entrance video comprises a static segment of the same live pig in a standing state and behavior segments during different expression behaviors, and each behavior segment corresponds to one of the expression behaviors;
extracting a plurality of pig individual pictures of the live pig in a standing state from the static fragment, and reconstructing and shaping a preset pig individual three-dimensional model according to the plurality of pig individual pictures to obtain a live pig three-dimensional model corresponding to the live pig, wherein the plurality of pig individual pictures of the live pig in the standing state at least comprise a live pig static front picture, a live pig static left side picture, a live pig static right side picture, a live pig static rear view picture and a live pig static top view picture;
sequentially acquiring a live pig behavior front picture, a live pig behavior left picture, a live pig behavior right picture, a live pig behavior rear view picture and a live pig behavior overlooking picture which are positioned at the same time point for behavior clips shot in different directions of the same performance behavior according to a time axis sequence, and simultaneously processing the live pig three-dimensional model at the current time point according to the live pig behavior front picture, the live pig behavior left picture, the live pig behavior right picture, the live pig behavior rear view picture and the live pig behavior overlooking picture at the same time point and recording the processing process of the live pig three-dimensional model to obtain the processing characteristics of each performance behavior on the live pig three-dimensional model;
when the processor executes the step S1 of the computer program, the following steps are specifically implemented:
acquiring live pig monitoring videos sent by the second camera and the third camera, wherein the front of the pigsty is a position on a pig feeding groove in the pigsty, which is as high as the live pig;
the processor, when executing the "identifying the performance behavior of each live pig in the live pig monitoring video" in the step S2 of the computer program, specifically implements the following steps:
extracting individual pig top-view pictures from the live pig monitoring video sent by the second camera, and extracting individual pig front-view pictures from the live pig monitoring video sent by the third camera to obtain individual pig top-view pictures of each live pig in the live pig monitoring video and individual pig front-view pictures corresponding to a time axis;
and processing the three-dimensional pig model corresponding to each live pig in real time according to the pig individual overlooking picture and the pig individual front picture corresponding to each live pig to obtain a real-time processing process of each live pig, and obtaining the performance of each live pig according to the similarity between the real-time processing process of each live pig and the processing characteristics of each performance.
As can be seen from the above description, the video of the standing live pig is shot in all directions by the plurality of first cameras, so as to establish a three-dimensional model conforming to the actual situation of the live pig; the method is characterized in that videos of live pigs under different performances are shot in an omnibearing manner through a plurality of first cameras to simulate three-dimensional changes of the live pigs under the different performances, wherein the three-dimensional changes are changes of body shapes of the live pigs in the whole process of finishing one performance in a standing state, such as feeding, and the most obvious characteristic is that the live pigs need to bend down the heads of the pigs and keep a certain time to finish feeding. In the real-time monitoring process, at least one upper top view and at least one front view are needed, the top views can avoid crowding of live pigs and can not obtain effective information of each live pig, and at least a body shape contour map of the live pigs can be obtained; the third camera positioned on the pig feeding groove in the pigsty and at the position with the same height as the live pigs can clearly see the pig faces and the limb shapes of the live pigs in the feeding behavior in the feeding process of the live pigs, and can better acquire the limb shapes of the live pigs in other expressive behaviors so as to acquire the limb shapes of each live pig in the real-time breeding process as accurately as possible; the obtained front picture and the overlook picture are used for processing the three-dimensional pig model, the judgment processing process is compared with the processing characteristics recorded before, so that the performance of the current live pig is judged, namely the body change of a certain behavior of the live pig can be better reflected through three-dimensional simulation, and the performance of the live pig can be better identified under the condition that the complete limb form of each live pig is difficult to obtain.
Further, the live pigs comprise healthy pig individuals and pig individuals with different types of attacks;
when the processor executes the computer program to obtain the processing characteristics of each performance behavior on the three-dimensional live pig model, the following steps are specifically realized:
taking the processing process and the walking track of each live pig as input parameters, taking the health state and the performance behavior of each live pig as output parameters, and taking the processing process, the walking track, the health state and the performance behavior of each live pig marked manually as a training set of a neural network model to obtain a live pig health recognition model;
the processor executes the computer program in step S2 of obtaining the performance of each live pig according to the degree of similarity between the real-time processing procedure of each live pig and the processing characteristics of each performance, and step S3 of replacing the steps with the following steps:
and respectively inputting the real-time processing process and the walking track of each live pig into the live pig health recognition model to obtain the performance behavior and the health state of each live pig.
According to the description, based on big data, live pigs regardless of health or diseases are subjected to data collection, are distinguished through manual marking, and are placed into a neural network model as a training set to obtain a live pig health recognition model, the live pig health recognition model can automatically extract feature data of the live pigs under different expressive behaviors and feature data of the live pigs under different health conditions under the same expressive behavior, and accordingly the expressive behaviors and health states of the live pigs can be rapidly and accurately recognized according to a real-time received processing process.
Referring to fig. 1, a first embodiment of the present invention is:
an intelligent pig raising monitoring method comprises the following steps:
s1: a live pig monitoring video of live pigs during the breeding process in the farm is obtained, in this embodiment. A live pig monitoring video can be obtained in real time by installing a camera;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
in this embodiment, step S2 specifically includes the following steps:
s21, extracting N frames of live pig monitoring images with intervals of a first preset time from a live pig monitoring video to obtain a track generating image set, extracting M frames of continuous live pig monitoring images from the live pig monitoring video to obtain a behavior analyzing image set, dividing each frame of live pig monitoring image in the track generating image set and the behavior analyzing image set into X individual pig images in one-to-one correspondence with the X live pigs, recording the relative position information of each individual pig image in the live pig monitoring image and the relative time information of each live pig monitoring image in the live pig monitoring video, putting the X individual pig images in the same frame into the same pig colony data set, obtaining N pig colony data sets corresponding to the track generating image set and executing the step S22 on the pig colony data sets, obtaining M pig colony data sets corresponding to the behavior analyzing image set and executing the step S23 on the pig colony data sets, n is greater than or equal to 1, M is greater than or equal to 16;
s22, carrying out individual pig identification on X individual pig images in N individual pig group data sets corresponding to the track generation image set in sequence to obtain individual pig identity information corresponding to each individual pig image, classifying the N individual pig images with the same individual pig identity information to obtain a first individual pig data set corresponding to the individual pig identity information, and generating the walking track of each live pig according to the relative position information and the relative time information of each individual pig image in the first individual pig data set;
s23, carrying out pig individual identification and limb shape identification on X individual pig images in M individual pig group data sets corresponding to the behavior analysis image set in sequence to obtain pig individual identity information and current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same individual pig identity information to obtain a second individual pig data set corresponding to the individual pig identity information, and generating the performance behavior of each live pig according to the relative position information, the relative time information and the current limb shape of each individual pig image in the second individual pig data set, wherein the relative position changes in the running process of the live pig, so that the performance behavior takes the relative position information into consideration, and the judgment of the result is more accurate.
If the live pig monitoring videos in the period of time are analyzed at intervals, N is related at intervals, for example, 10-minute videos, and if the first preset time is 1S, N is 600; if the track is generated in real time, one frame of picture can be extracted to obtain the position information of the picture each time the real-time data is received, and then the picture is connected with the position information before and after the real-time data is received in series to generate the track in real time; for the performance behaviors, M can select 48 frames, namely 3 seconds, and whether the live pig monitoring video in the period of time is analyzed at intervals or calculated in real time, the continuous 48 frames, namely the images in the 3S period are needed for identification, and the performance analysis can be performed only once in one minute in consideration of the problem of identification calculation amount.
And S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
Referring to fig. 1, the second embodiment of the present invention is:
an intelligent pig raising monitoring method is characterized in that on the basis of the first embodiment, the step S2 in the first embodiment is replaced by the following steps:
s21', starting to judge whether a clear pig face exists from a first frame of pig monitoring image in a pig monitoring video until a first pig monitoring image comprising the clear pig face is obtained, dividing the first pig monitoring image comprising the clear pig face into a first pig individual image comprising the clear pig face, carrying out pig individual identification on the first pig individual image to obtain first pig individual identity information corresponding to the first pig individual image, carrying out target tracking on a first pig corresponding to the first pig individual identity information so as to mark the first pig in each frame of pig monitoring image in the pig monitoring video, continuing to judge whether the clear pig face exists, so as to mark each pig in each frame of pig monitoring image in the pig monitoring video, namely, identifying each pig at the beginning, and then carrying out target locking on each pig, therefore, each subsequent frame of image does not need to identify the live pigs to distinguish each live pig, so that the walking track and the performance behavior can be obtained in real time;
s22', connecting the relative position information of each live pig in each frame of live pig monitoring image in the live pig monitoring video according to the time sequence of the live pig monitoring video to obtain the walking track of each live pig;
s23', extracting M continuous live pig monitoring images from the live pig monitoring video, dividing each extracted live pig monitoring image into X individual pig images corresponding to the X live pigs one by one, recording the marks of the live pigs corresponding to each individual pig image, the relative position information marked on the live pig monitoring images and the relative time information of each pig monitoring image in the live pig monitoring video, carrying out limb shape recognition on each individual pig image to obtain the current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same marks to obtain a second individual pig data set corresponding to the marks, and generating the performance of each live pig according to the relative position information, the relative time information and the current limb shape of each individual pig image in the second individual pig data set, m is greater than or equal to 16.
Referring to fig. 1, a third embodiment of the present invention is:
an intelligent pig raising monitoring method, on the basis of the first embodiment, before the step S1, further includes the following steps:
the method comprises the steps that entrance videos shot by first cameras located at different positions of an entrance of a farm for the same live pig are obtained, the entrance videos comprise static fragments of the same live pig in a standing state and behavior fragments during different performance behaviors, the position of each first camera at least comprises the front, the left, the right, the rear and the upper of the live pig, each behavior fragment corresponds to one performance behavior, and the live pig comprises a healthy pig individual and pig individuals with different types of attacks;
extracting a plurality of pig individual pictures of a standing live pig from the static fragment, and reconstructing and shaping a preset pig individual three-dimensional model according to the plurality of pig individual pictures to obtain a live pig three-dimensional model corresponding to the live pig, wherein the plurality of pig individual pictures of the standing live pig at least comprise a live pig static front picture, a live pig static left side picture, a live pig static right side picture, a live pig static rear view picture and a live pig static overlooking picture;
sequentially acquiring a live pig behavior front picture, a live pig behavior left picture, a live pig behavior right picture, a live pig behavior rear view picture and a live pig behavior overlooking picture at the same time point according to a time axis sequence for behavior clips which show the same behavior and are shot in different directions, and simultaneously processing and recording a live pig three-dimensional model at the current time point according to the live pig behavior front picture, the live pig behavior left picture, the live pig behavior right picture, the live pig behavior rear view picture and the live pig behavior overlooking picture at the same time point;
taking the processing process and the walking track of each live pig as input parameters, taking the health state and the performance behavior of each live pig as output parameters, and taking the processing process, the walking track, the health state and the performance behavior of each live pig marked manually as a training set of a neural network model to obtain a live pig health recognition model;
in this embodiment, step S1 specifically includes the following steps:
acquiring live pig monitoring videos sent by a second camera positioned above the pigsty and a third camera positioned in front of the pigsty in the farm, wherein the front of the pigsty is a position on a feeding trough in the pigsty, which is as high as the live pigs;
step S2 and step S3 are replaced with the following steps:
identifying the walking track of each live pig in the live pig monitoring video;
extracting a pig individual overlooking picture from the live pig monitoring video sent by the second camera, and extracting a pig individual front picture from the live pig monitoring video sent by the third camera to obtain the pig individual overlooking picture of each live pig in the live pig monitoring video and the pig individual front picture corresponding to the time axis;
processing the live pig three-dimensional model corresponding to the live pig in real time according to the pig individual overlooking picture and the pig individual front picture corresponding to each live pig to obtain the real-time processing process of each live pig
And respectively inputting the real-time processing process and the walking track of each live pig into the live pig health recognition model to obtain the performance behavior and the health state of each live pig.
Referring to fig. 2, a fourth embodiment of the present invention is:
an intelligent pig raising monitoring management terminal 1 comprises a memory 3, a processor 2 and a computer program which is stored on the memory 3 and can run on the processor 2, wherein the processor 2 is respectively connected with a first camera 4 positioned at an inlet of a farm in different directions, a second camera 5 positioned above a swinery in the farm and a third camera 6 positioned in front of the swinery in the farm, and the directions of the first camera 4 at least comprise the front, the left side, the right side, the back and the upper parts of live pigs; wherein, the processor 2 implements the steps in the first embodiment, the second embodiment or the third embodiment when executing the computer program.
In conclusion, the intelligent pig-raising monitoring method and the intelligent pig-raising management terminal provided by the invention have the advantages that the live pigs are monitored in real time in a video mode in the breeding process, so that farmers can observe the growth conditions of the live pigs in real time to judge the health states of the live pigs; meanwhile, the identity of each individual pig in each frame of image is distinguished by identifying each frame of image after being segmented or marking the individual pigs after being tracked and locked in real time, so that walking tracks are obtained according to position changes in time sequence, and expressive behaviors are obtained according to limb form changes in time sequence, namely, the health state of each pig is obtained by combining real-time change conditions and long-time change conditions of the pigs, and more intelligent, more accurate and more real-time and rapid health monitoring is automatically provided for farmers. Meanwhile, the body change of the live pigs during a certain behavior is better reflected by establishing the three-dimensional model, so that the performance behaviors of the live pigs can be better identified under the condition that the complete limb form of each live pig is difficult to obtain. The live pig health recognition model established based on the big data and the neural network model can automatically extract the feature data of the live pig under different expressive behaviors and the feature data of the live pig under different health conditions under the same expressive behavior, so that the expressive behavior and the health state of the current live pig can be quickly and accurately recognized according to the processing process received in real time.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent changes made by using the contents of the present specification and the drawings, or applied directly or indirectly to the related technical fields, are included in the scope of the present invention.
Claims (10)
1. An intelligent pig raising monitoring method is characterized by comprising the following steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
2. The intelligent pig-raising monitoring method according to claim 1, wherein the step S2 specifically comprises the following steps:
s21, extracting N frames of live pig monitoring images with intervals of a first preset time from the live pig monitoring video to obtain a track generation image set, extracting M frames of continuous live pig monitoring images from the live pig monitoring video to obtain a behavior analysis image set, dividing each frame of the live pig monitoring images in the track generation image set and the behavior analysis image set into X individual pig images corresponding to the X live pigs one by one, recording the relative position information of each individual pig image in the live pig monitoring images and the relative time information of each live pig monitoring image in the live pig monitoring video, placing the X individual pig images in the same frame into the same pig colony data set to obtain N pig colony data sets corresponding to the track generation image set and executing the step S22 on the pig colony data sets, and obtaining M of said swine population data sets corresponding to said behavior analysis image set and performing step S23 thereon, said N being greater than or equal to 1 and said M being greater than or equal to 16;
s22, carrying out pig individual identification on X individual pig images in N individual pig group data sets corresponding to the track generation image set in sequence to obtain pig individual identity information corresponding to each individual pig image, classifying the N individual pig images with the same pig individual identity information to obtain a first individual pig data set corresponding to the pig individual identity information, and generating the walking track of each live pig according to the relative position information and the relative time information of each individual pig image in the first individual pig data set;
s23, carrying out pig individual identification and limb form identification on X individual pig images in M individual pig group data sets corresponding to the behavior analysis image set in sequence to obtain pig individual identity information and current limb form corresponding to each individual pig image, classifying the M individual pig images with the same pig individual identity information to obtain a second individual pig data set corresponding to the pig individual identity information, and generating the performance behavior of each pig according to the relative position information, the relative time information and the current limb form of each individual pig image in the second individual pig data set.
3. The intelligent pig-raising monitoring method according to claim 1, wherein the step S2 specifically comprises the following steps:
s21', starting to judge whether a clear pig face exists from a first frame of live pig monitoring image in the live pig monitoring video until a first live pig monitoring image including the clear pig face is obtained, dividing the first live pig monitoring image including the clear pig face into a first pig individual image including the clear pig face, carrying out pig individual identification on the first pig individual image to obtain first pig individual identity information corresponding to the first pig individual image, carrying out target tracking on a first live pig corresponding to the first pig individual identity information so as to mark the first live pig in each frame of live pig monitoring image in the live pig monitoring video, and then continuing to judge whether the clear pig face exists so as to mark each live pig in each frame of live pig monitoring image in the live pig monitoring video;
s22', connecting the relative position information of each live pig in each frame of live pig monitoring image in the live pig monitoring video according to the time sequence of the live pig monitoring video to obtain the walking track of each live pig;
s23', extracting M continuous live pig monitoring images from the live pig monitoring video, dividing each extracted live pig monitoring image into X individual pig images corresponding to the X live pigs one by one, recording the marks of the live pigs corresponding to the individual pig images, the relative position information of the marks on the live pig monitoring images and the relative time information of the live pig monitoring images in the live pig monitoring video, carrying out limb shape identification on each individual pig image to obtain the current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same marks to obtain a second individual pig data set corresponding to the marks, and carrying out body shape identification according to the relative position information of each individual pig image in the second individual pig data set, And generating the performance behavior of each live pig according to the relative time information and the current limb form, wherein M is greater than or equal to 16.
4. The intelligent pig-raising monitoring method according to claim 1, characterized in that the step S1 is preceded by the following steps:
acquiring entry videos shot by the same live pig by first cameras respectively positioned at different positions of an entrance in a farm, wherein the entry videos comprise static segments of the same live pig in a standing state and behavior segments during different performance behaviors, the positions of the first cameras at least comprise the front, the left, the right, the back and the upper of the live pig, and each behavior segment corresponds to one performance behavior respectively;
extracting a plurality of pig individual pictures of the live pig in a standing state from the static fragment, and reconstructing and shaping a preset pig individual three-dimensional model according to the plurality of pig individual pictures to obtain a live pig three-dimensional model corresponding to the live pig, wherein the plurality of pig individual pictures of the live pig in the standing state at least comprise a live pig static front picture, a live pig static left side picture, a live pig static right side picture, a live pig static rear view picture and a live pig static top view picture;
sequentially acquiring a live pig behavior front picture, a live pig behavior left picture, a live pig behavior right picture, a live pig behavior rear view picture and a live pig behavior overlooking picture which are positioned at the same time point for behavior clips shot in different directions of the same performance behavior according to a time axis sequence, and simultaneously processing the live pig three-dimensional model at the current time point according to the live pig behavior front picture, the live pig behavior left picture, the live pig behavior right picture, the live pig behavior rear view picture and the live pig behavior overlooking picture at the same time point and recording the processing process of the live pig three-dimensional model to obtain the processing characteristics of each performance behavior on the live pig three-dimensional model;
the step S1 specifically includes the following steps:
acquiring live pig monitoring videos sent by a second camera positioned above a pigsty and a third camera positioned in front of the pigsty in the farm, wherein the front of the pigsty is a position on a pig feeding groove in the pigsty, which is as high as the live pigs;
the step S2 of "recognizing the performance behavior of each live pig in the live pig monitoring video" specifically includes the following steps:
extracting individual pig top-view pictures from the live pig monitoring video sent by the second camera, and extracting individual pig front-view pictures from the live pig monitoring video sent by the third camera to obtain individual pig top-view pictures of each live pig in the live pig monitoring video and individual pig front-view pictures corresponding to a time axis;
and processing the three-dimensional pig model corresponding to each live pig in real time according to the pig individual overlooking picture and the pig individual front picture corresponding to each live pig to obtain a real-time processing process of each live pig, and obtaining the performance of each live pig according to the similarity between the real-time processing process of each live pig and the processing characteristics of each performance.
5. The intelligent pig-raising monitoring method according to claim 4, wherein the pig comprises healthy pig individuals and pig individuals with different types of attacks;
the step of obtaining the processing characteristics of each performance behavior on the live pig three-dimensional model specifically comprises the following steps:
taking the processing process and the walking track of each live pig as input parameters, taking the health state and the performance behavior of each live pig as output parameters, and taking the processing process, the walking track, the health state and the performance behavior of each live pig marked manually as a training set of a neural network model to obtain a live pig health recognition model;
in the step S2, "the performance of each live pig is obtained according to the degree of similarity between the real-time processing procedure of each live pig and the processing characteristics of each performance" and the step S3 is replaced with the following steps:
and respectively inputting the real-time processing process and the walking track of each live pig into the live pig health recognition model to obtain the performance behavior and the health state of each live pig.
6. An intelligent pig raising monitoring management terminal, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, and is characterized in that the processor executes the computer program to realize the following steps:
s1: acquiring a live pig monitoring video of live pigs in the breeding process in a breeding farm;
s2, identifying the walking track and the performance behavior of each live pig in the live pig monitoring video;
and S3, obtaining the health state of each live pig according to the walking track and the performance behavior of each live pig.
7. The intelligent pig-raising monitoring management terminal according to claim 6, wherein the processor executing the step S2 of the computer program specifically comprises the following steps:
s21, extracting N frames of live pig monitoring images with intervals of a first preset time from the live pig monitoring video to obtain a track generation image set, extracting M frames of continuous live pig monitoring images from the live pig monitoring video to obtain a behavior analysis image set, dividing each frame of the live pig monitoring images in the track generation image set and the behavior analysis image set into X individual pig images corresponding to the X live pigs one by one, recording the relative position information of each individual pig image in the live pig monitoring images and the relative time information of each live pig monitoring image in the live pig monitoring video, placing the X individual pig images in the same frame into the same pig colony data set to obtain N pig colony data sets corresponding to the track generation image set and executing the step S22 on the pig colony data sets, and obtaining M of said swine population data sets corresponding to said behavior analysis image set and performing step S23 thereon, said N being greater than or equal to 1 and said M being greater than or equal to 16;
s22, carrying out pig individual identification on X individual pig images in N individual pig group data sets corresponding to the track generation image set in sequence to obtain pig individual identity information corresponding to each individual pig image, classifying the N individual pig images with the same pig individual identity information to obtain a first individual pig data set corresponding to the pig individual identity information, and generating the walking track of each live pig according to the relative position information and the relative time information of each individual pig image in the first individual pig data set;
s23, carrying out pig individual identification and limb form identification on X individual pig images in M individual pig group data sets corresponding to the behavior analysis image set in sequence to obtain pig individual identity information and current limb form corresponding to each individual pig image, classifying the M individual pig images with the same pig individual identity information to obtain a second individual pig data set corresponding to the pig individual identity information, and generating the performance behavior of each pig according to the relative position information, the relative time information and the current limb form of each individual pig image in the second individual pig data set.
8. The intelligent pig-raising monitoring management terminal according to claim 6, wherein the processor executing the step S2 of the computer program specifically comprises the following steps:
s21', starting to judge whether a clear pig face exists from a first frame of live pig monitoring image in the live pig monitoring video until a first live pig monitoring image including the clear pig face is obtained, dividing the first live pig monitoring image including the clear pig face into a first pig individual image including the clear pig face, carrying out pig individual identification on the first pig individual image to obtain first pig individual identity information corresponding to the first pig individual image, carrying out target tracking on a first live pig corresponding to the first pig individual identity information so as to mark the first live pig in each frame of live pig monitoring image in the live pig monitoring video, and then continuing to judge whether the clear pig face exists so as to mark each live pig in each frame of live pig monitoring image in the live pig monitoring video;
s22', connecting the relative position information of each live pig in each frame of live pig monitoring image in the live pig monitoring video according to the time sequence of the live pig monitoring video to obtain the walking track of each live pig;
s23', extracting M continuous live pig monitoring images from the live pig monitoring video, dividing each extracted live pig monitoring image into X individual pig images corresponding to the X live pigs one by one, recording the marks of the live pigs corresponding to the individual pig images, the relative position information of the marks on the live pig monitoring images and the relative time information of the live pig monitoring images in the live pig monitoring video, carrying out limb shape identification on each individual pig image to obtain the current limb shape corresponding to each individual pig image, classifying the M individual pig images with the same marks to obtain a second individual pig data set corresponding to the marks, and carrying out body shape identification according to the relative position information of each individual pig image in the second individual pig data set, And generating the performance behavior of each live pig according to the relative time information and the current limb form, wherein M is greater than or equal to 16.
9. The intelligent pig-raising monitoring management terminal according to claim 6, wherein the processor is respectively connected with a first camera positioned at an entrance of a farm in different directions, a second camera positioned above a pigsty in the farm and a third camera positioned in front of the pigsty in the farm, and the directions of the first cameras at least comprise the front, the left, the right, the back and the upper of the live pig;
the processor further comprises the following steps before executing the step S1 of the computer program:
acquiring an entrance video shot by the first camera for the same live pig, wherein the entrance video comprises a static segment of the same live pig in a standing state and behavior segments during different expression behaviors, and each behavior segment corresponds to one of the expression behaviors;
extracting a plurality of pig individual pictures of the live pig in a standing state from the static fragment, and reconstructing and shaping a preset pig individual three-dimensional model according to the plurality of pig individual pictures to obtain a live pig three-dimensional model corresponding to the live pig, wherein the plurality of pig individual pictures of the live pig in the standing state at least comprise a live pig static front picture, a live pig static left side picture, a live pig static right side picture, a live pig static rear view picture and a live pig static top view picture;
sequentially acquiring a live pig behavior front picture, a live pig behavior left picture, a live pig behavior right picture, a live pig behavior rear view picture and a live pig behavior overlooking picture which are positioned at the same time point for behavior clips shot in different directions of the same performance behavior according to a time axis sequence, and simultaneously processing the live pig three-dimensional model at the current time point according to the live pig behavior front picture, the live pig behavior left picture, the live pig behavior right picture, the live pig behavior rear view picture and the live pig behavior overlooking picture at the same time point and recording the processing process of the live pig three-dimensional model to obtain the processing characteristics of each performance behavior on the live pig three-dimensional model;
when the processor executes the step S1 of the computer program, the following steps are specifically implemented:
acquiring live pig monitoring videos sent by the second camera and the third camera, wherein the front of the pigsty is a position on a pig feeding groove in the pigsty, which is as high as the live pig;
the processor, when executing the "identifying the performance behavior of each live pig in the live pig monitoring video" in the step S2 of the computer program, specifically implements the following steps:
extracting individual pig top-view pictures from the live pig monitoring video sent by the second camera, and extracting individual pig front-view pictures from the live pig monitoring video sent by the third camera to obtain individual pig top-view pictures of each live pig in the live pig monitoring video and individual pig front-view pictures corresponding to a time axis;
and processing the three-dimensional pig model corresponding to each live pig in real time according to the pig individual overlooking picture and the pig individual front picture corresponding to each live pig to obtain a real-time processing process of each live pig, and obtaining the performance of each live pig according to the similarity between the real-time processing process of each live pig and the processing characteristics of each performance.
10. The intelligent pig-raising monitoring management terminal according to claim 9, wherein the pig comprises healthy pig individuals and pig individuals with different types of attacks;
when the processor executes the computer program to obtain the processing characteristics of each performance behavior on the three-dimensional live pig model, the following steps are specifically realized:
taking the processing process and the walking track of each live pig as input parameters, taking the health state and the performance behavior of each live pig as output parameters, and taking the processing process, the walking track, the health state and the performance behavior of each live pig marked manually as a training set of a neural network model to obtain a live pig health recognition model;
the processor executes the computer program in step S2 of obtaining the performance of each live pig according to the degree of similarity between the real-time processing procedure of each live pig and the processing characteristics of each performance, and step S3 of replacing the steps with the following steps:
and respectively inputting the real-time processing process and the walking track of each live pig into the live pig health recognition model to obtain the performance behavior and the health state of each live pig.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010159849.9A CN111294565A (en) | 2020-03-10 | 2020-03-10 | Intelligent pig raising monitoring method and management terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010159849.9A CN111294565A (en) | 2020-03-10 | 2020-03-10 | Intelligent pig raising monitoring method and management terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111294565A true CN111294565A (en) | 2020-06-16 |
Family
ID=71022715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010159849.9A Pending CN111294565A (en) | 2020-03-10 | 2020-03-10 | Intelligent pig raising monitoring method and management terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111294565A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111667037A (en) * | 2020-06-24 | 2020-09-15 | 山西凯永养殖有限公司 | Large-area epidemic situation detection system for indoor pig farm |
| CN111783745A (en) * | 2020-08-06 | 2020-10-16 | 珠海南方利洋水产科技有限公司 | Fish health judgment method and device applied to pond culture and computer-readable storage medium |
| CN112257608A (en) * | 2020-10-23 | 2021-01-22 | 成都希盟泰克科技发展有限公司 | Yak breeding health state monitoring method |
| CN112801118A (en) * | 2021-02-26 | 2021-05-14 | 潘志乐 | Pork pig slaughtering benefit evaluation system and method based on artificial intelligence and big data |
| CN113068657A (en) * | 2021-03-31 | 2021-07-06 | 重庆市畜牧技术推广总站 | Intelligent efficient pig raising method and system |
| CN113507490A (en) * | 2021-04-30 | 2021-10-15 | 华中农业大学 | Multi-source information fusion live pig full-chain intelligent monitoring method and device |
| CN113632749A (en) * | 2021-09-07 | 2021-11-12 | 华中农业大学 | Pig physical state monitoring method based on animal behaviors |
| CN113678786A (en) * | 2021-08-19 | 2021-11-23 | 陆荣清 | Ecological breeding method for improving disease resistance of live pigs |
| WO2022172232A1 (en) * | 2021-02-12 | 2022-08-18 | Mechatronik Zehentner Eu | Method and device for tracking animals |
| CN115250952A (en) * | 2022-08-18 | 2022-11-01 | 深圳进化动力数码科技有限公司 | Live pig health monitoring method, device, equipment and storage medium |
| CN115936401A (en) * | 2023-01-08 | 2023-04-07 | 宁夏亿博丰担保品管理有限公司 | Data supervision and processing method and system for living mortgage |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080314325A1 (en) * | 2007-06-22 | 2008-12-25 | Hempstead David W | Method and apparatus for improved monitoring and managing of livestock |
| CN104008367A (en) * | 2014-05-08 | 2014-08-27 | 中国农业大学 | Automatic fattening pig behavior analyzing system and method based on computer vision |
| CN107437068A (en) * | 2017-07-13 | 2017-12-05 | 江苏大学 | Pig individual discrimination method based on Gabor direction histograms and pig chaeta hair pattern |
| 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 |
| CN110163103A (en) * | 2019-04-18 | 2019-08-23 | 中国农业大学 | A kind of live pig Activity recognition method and apparatus based on video image |
-
2020
- 2020-03-10 CN CN202010159849.9A patent/CN111294565A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080314325A1 (en) * | 2007-06-22 | 2008-12-25 | Hempstead David W | Method and apparatus for improved monitoring and managing of livestock |
| CN104008367A (en) * | 2014-05-08 | 2014-08-27 | 中国农业大学 | Automatic fattening pig behavior analyzing system and method based on computer vision |
| CN107437068A (en) * | 2017-07-13 | 2017-12-05 | 江苏大学 | Pig individual discrimination method based on Gabor direction histograms and pig chaeta hair pattern |
| 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 |
| CN110163103A (en) * | 2019-04-18 | 2019-08-23 | 中国农业大学 | A kind of live pig Activity recognition method and apparatus based on video image |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111667037A (en) * | 2020-06-24 | 2020-09-15 | 山西凯永养殖有限公司 | Large-area epidemic situation detection system for indoor pig farm |
| CN111783745A (en) * | 2020-08-06 | 2020-10-16 | 珠海南方利洋水产科技有限公司 | Fish health judgment method and device applied to pond culture and computer-readable storage medium |
| CN112257608A (en) * | 2020-10-23 | 2021-01-22 | 成都希盟泰克科技发展有限公司 | Yak breeding health state monitoring method |
| WO2022172232A1 (en) * | 2021-02-12 | 2022-08-18 | Mechatronik Zehentner Eu | Method and device for tracking animals |
| CN112801118A (en) * | 2021-02-26 | 2021-05-14 | 潘志乐 | Pork pig slaughtering benefit evaluation system and method based on artificial intelligence and big data |
| CN112801118B (en) * | 2021-02-26 | 2024-02-02 | 潘志乐 | Pork pig marketing benefit evaluation system and method based on artificial intelligence and big data |
| CN113068657B (en) * | 2021-03-31 | 2022-04-08 | 重庆市畜牧技术推广总站 | Intelligent efficient pig raising method and system |
| CN113068657A (en) * | 2021-03-31 | 2021-07-06 | 重庆市畜牧技术推广总站 | Intelligent efficient pig raising method and system |
| CN113507490A (en) * | 2021-04-30 | 2021-10-15 | 华中农业大学 | Multi-source information fusion live pig full-chain intelligent monitoring method and device |
| CN113678786A (en) * | 2021-08-19 | 2021-11-23 | 陆荣清 | Ecological breeding method for improving disease resistance of live pigs |
| CN113632749A (en) * | 2021-09-07 | 2021-11-12 | 华中农业大学 | Pig physical state monitoring method based on animal behaviors |
| CN115250952A (en) * | 2022-08-18 | 2022-11-01 | 深圳进化动力数码科技有限公司 | Live pig health monitoring method, device, equipment and storage medium |
| CN115936401A (en) * | 2023-01-08 | 2023-04-07 | 宁夏亿博丰担保品管理有限公司 | Data supervision and processing method and system for living mortgage |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111294565A (en) | Intelligent pig raising monitoring method and management terminal | |
| Yang et al. | Feeding behavior recognition for group-housed pigs with the Faster R-CNN | |
| Fernandes et al. | A novel automated system to acquire biometric and morphological measurements and predict body weight of pigs via 3D computer vision | |
| CN109255297B (en) | Animal state monitoring method, terminal device, storage medium and electronic device | |
| TW201539357A (en) | Livestock identification system and method | |
| Tscharke et al. | A brief review of the application of machine vision in livestock behaviour analysis | |
| CN116935327B (en) | Aquaculture monitoring method, device, equipment and storage medium based on AI vision | |
| Tscharke et al. | Review of methods to determine weight and size of livestock from images | |
| CN111297367A (en) | Animal state monitoring method and device, electronic equipment and storage medium | |
| Chen et al. | A kinetic energy model based on machine vision for recognition of aggressive behaviours among group-housed pigs | |
| CN113785783A (en) | Livestock grouping system and method | |
| CN112528823B (en) | Method and system for analyzing batcharybus movement behavior based on key frame detection and semantic component segmentation | |
| CN113068657B (en) | Intelligent efficient pig raising method and system | |
| CN114155377A (en) | Poultry self-adaptive feeding method based on artificial intelligence and growth cycle analysis | |
| CN113516139A (en) | Data processing method, device, equipment and storage medium | |
| CN113057598A (en) | Meat quality grading method and system for animal living body slaughterless | |
| CN110414369B (en) | Cow face training method and device | |
| CN115830078A (en) | Live pig multi-target tracking and behavior recognition method, computer equipment and storage medium | |
| Ojukwu et al. | Development of a computer vision system to detect inactivity in group-housed pigs | |
| CN115439789A (en) | Intelligent identification method and identification system for life state of silkworm | |
| Avanzato et al. | Dairy cow behavior recognition using computer vision techniques and CNN networks | |
| CN111476118B (en) | Animal behavior automatic identification method and device | |
| CN114070862A (en) | Livestock raising control method and system based on Internet of things and readable storage medium | |
| CN116596167B (en) | Livestock breeding scheme optimization method based on data deep fusion processing | |
| CN118135500B (en) | Intelligent monitoring platform and monitoring method for livestock breeding |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200616 |
|
| RJ01 | Rejection of invention patent application after publication |