CN113711944B - Sow estrus identification method, device and system - Google Patents
Sow estrus identification method, device and system Download PDFInfo
- Publication number
- CN113711944B CN113711944B CN202110996071.1A CN202110996071A CN113711944B CN 113711944 B CN113711944 B CN 113711944B CN 202110996071 A CN202110996071 A CN 202110996071A CN 113711944 B CN113711944 B CN 113711944B
- Authority
- CN
- China
- Prior art keywords
- probability
- sow
- oestrus
- estrus
- ultrasonic
- 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.)
- Active
Links
- 230000012173 estrus Effects 0.000 title claims abstract description 323
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000000694 effects Effects 0.000 claims abstract description 51
- 230000005906 menstruation Effects 0.000 claims abstract description 26
- 239000000178 monomer Substances 0.000 claims description 4
- 238000002604 ultrasonography Methods 0.000 claims 1
- 238000010801 machine learning Methods 0.000 abstract description 12
- 230000008859 change Effects 0.000 abstract description 11
- 230000036544 posture Effects 0.000 description 75
- 238000013473 artificial intelligence Methods 0.000 description 13
- 238000012545 processing Methods 0.000 description 13
- 238000009395 breeding Methods 0.000 description 11
- 230000001488 breeding effect Effects 0.000 description 11
- 230000006399 behavior Effects 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000002776 aggregation Effects 0.000 description 6
- 238000004220 aggregation Methods 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 6
- 241000282887 Suidae Species 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 5
- 238000001914 filtration Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 241000282898 Sus scrofa Species 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 230000013011 mating Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 239000000523 sample Substances 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000002354 daily effect Effects 0.000 description 2
- 230000006806 disease prevention Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000006651 lactation Effects 0.000 description 2
- 230000033001 locomotion Effects 0.000 description 2
- 230000005541 medical transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 210000003905 vulva Anatomy 0.000 description 2
- 208000007407 African swine fever Diseases 0.000 description 1
- 206010052904 Musculoskeletal stiffness Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000000712 assembly Effects 0.000 description 1
- 238000000429 assembly Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 244000144972 livestock Species 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 230000016087 ovulation Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001850 reproductive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- 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)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a method, a device and a system for identifying oestrus of sows, which are characterized in that ultrasonic emission signals and corresponding ultrasonic reflection signals are received; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state. According to the method, the posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through the model obtained through machine learning.
Description
Technical Field
The invention relates to the field of livestock breeding, in particular to a method, a device and a system for identifying oestrus of sows.
Background
In the modern pig raising production, sows mainly play a breeding role, are the source of the whole pig raising production, and the reproductive capacity of the sows is an important production benefit index. The estrus identification is a technology for judging the mating time of sows, is the first key link in the whole production cycle of the sows, and is directly related to whether the sows can be normally mated, whether the sows are born after mating and the final farrowing quantity after the sows are born.
At present, the traditional artificial estrus identification method is mainly adopted in China, is a high-requirement technology, and needs experienced professional breeders to determine the breeding time through the state expression of sows. In conclusion, the sow estrus identification needs to consume large manpower and material resources to achieve high registration rate through artificial identification. With the further expansion of the breeding scale, the short supply of the professional estrus appraisers becomes a significant problem in the modern pig-raising industry.
Therefore, the research on the low-cost and high-efficiency automatic estrus identification technology is a problem to be solved by those skilled in the art.
Disclosure of Invention
The invention aims to provide a method, a device and a system for identifying the estrus of a sow, and aims to solve the problems of high labor cost and low identification accuracy rate of the estrus identification of the sow in the prior art.
In order to solve the technical problem, the invention provides a sow estrus identification method, which comprises the following steps:
receiving an ultrasonic transmitting signal and a corresponding ultrasonic reflecting signal;
determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal;
determining sow posture information according to the ultrasonic wave menstruation duration;
and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the oestrus state of the sow.
Optionally, in the method for identifying oestrus of sows, the receiving ultrasonic emission signals and corresponding ultrasonic reflection signals includes:
receiving a plurality of ultrasonic emission signals and corresponding ultrasonic reflection signals; the ultrasonic detector corresponding to the ultrasonic reflection signal is arranged above the single column and transmits ultrasonic waves downwards;
correspondingly, determining the corresponding ultrasonic wave passing time length according to the plurality of ultrasonic wave reflection signals;
determining the shortest ultrasonic wave passing time as a target time;
correspondingly, the sow posture information is determined according to the target duration.
Optionally, in the method for identifying oestrus of sows, the determining posture information of the sows according to the length of the ultrasonic wave menstruation includes:
determining daytime attitude characteristic information and night attitude characteristic information according to the ultrasonic wave passing time and an acquisition time tag corresponding to the ultrasonic wave passing time;
determining the activity information of the sow according to the daytime posture characteristic information and the night posture characteristic information corresponding to the target date;
correspondingly, the sow activity information corresponding to the target date is input into the pre-trained estrus state model to obtain the sow estrus state.
Optionally, in the method for identifying sow oestrus, the inputting of the posture information of the sow corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state includes:
inputting the sow posture information of a complete day into a pre-trained oestrus state model to obtain oestrus probability information; the estrus probability information comprises probability of no estrus, recorded as probability A, probability of a day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C, probability of estrus in the afternoon and recorded as probability D, and probability of four conditions in total;
and determining the situation with the maximum probability as the oestrus state of the sow according to the oestrus probability information.
Optionally, in the sow estrus identification method, when the sow estrus state corresponding to the first day after the sow is weaned and slaughtered is not estrus, acquiring the estrus probability information corresponding to the first day after the sow is weaned and slaughtered;
judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to the oestrus probability information corresponding to the first day after the sows are weaned and turned;
and when the difference values do not exceed the first threshold value, determining that the oestrus state of the sows corresponding to the first day after the sows are weaned and turned is oestrous.
Optionally, in the method for identifying sow estrus, when the sow estrus state corresponding to a target date is morning estrus or afternoon estrus but the sow estrus state corresponding to the day before the target date is not estrus, acquiring estrus probability information corresponding to the target date;
judging whether the probability C and the probability D both exceed the probability A and the probability B according to the estrus probability information corresponding to the target date;
and when at least one of the C probability and the D probability does not exceed the A probability and the B probability, determining the condition corresponding to the larger one of the A probability and the B probability as the sow oestrus state.
Optionally, in the method for identifying sow estrus, when the sow estrus state corresponding to a target date is the day before estrus and the sow estrus state corresponding to the day before the target date is also the day before estrus, acquiring the estrus probability information corresponding to the target date;
and determining the condition that the greater one of the C probability and the D probability corresponds to the sow oestrus state according to the oestrus probability information corresponding to the target date.
A sow estrus identification device comprising:
the receiving module is used for receiving the ultrasonic reflection signal;
the travelling module is used for determining the travelling time of the ultrasonic waves according to the ultrasonic reflection signals;
the behavior module is used for determining the posture information of the sow according to the ultrasonic wave menstruation duration;
and the state module is used for inputting the sow posture information corresponding to the target time period into the pre-trained estrus state model to obtain the sow estrus state.
A sow oestrus identification system comprises an ultrasonic detector and a data processor;
the ultrasonic detector is used for transmitting target ultrasonic waves to the sow to be detected and receiving the target ultrasonic waves reflected by the sow to be detected;
the data processor is used for receiving ultrasonic wave transmitting signals and ultrasonic wave reflecting signals corresponding to the target ultrasonic waves; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state.
Optionally, in the sow estrus identification system, a plurality of ultrasonic detectors are arranged in the sow estrus identification system corresponding to a single solitary fence;
the ultrasonic detector is arranged above the monomer column and emits ultrasonic waves downwards;
the plurality of ultrasonic detectors are arranged at equal intervals along a perpendicular bisector perpendicular to the long axis direction of the horizontal projection of the single fence.
The sow estrus identification method provided by the invention receives ultrasonic emission signals and corresponding ultrasonic reflection signals; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state. The posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through a model obtained through machine learning. The invention also provides a sow estrus identification device and system with the beneficial effects.
Drawings
In order to more clearly illustrate the embodiments or technical solutions of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained based on these drawings without creative efforts.
FIG. 1 is a schematic flow chart of an embodiment of the method for identifying estrus in sows according to the present invention;
FIG. 2 is a schematic flow chart of another embodiment of the method for identifying estrus in sows provided by the present invention;
FIG. 3 is a schematic flow chart of another embodiment of the method for identifying estrus in sows according to the present invention;
fig. 4 is a schematic structural diagram of an embodiment of the sow estrus identification device provided by the invention;
FIG. 5 is a schematic structural diagram of an embodiment of the system for identifying oestrus in sows according to the present invention;
fig. 6 is a schematic structural diagram of another embodiment of the sow estrus identification system provided by the invention.
Detailed Description
The current oestrus identification mainly adopts a method of manual observation and boar oestrus induction, namely, one part of people pulls the boar to pass through before the sow column, the head of the boar can be effectively contacted with the head of the sow, the other part of people is positioned behind the column, and whether the sow oestrus is comprehensively judged by observing the phenomena of sow vulva state, sow back stiffness, sow ear erection, sow back static and motionless pressing, vulva temperature rise and the like. In the method, the working personnel are required to have abundant experience of condition finding, and the conditions of missed judgment and misjudgment are reduced; in addition, the ovulation time cannot be accurately judged by the method, the estrus of the sows can last for 2-4 days under the common condition, the sows need to be bred for multiple times to ensure the registration rate, and the production efficiency is low. With the further expansion of the breeding scale, the great demand of professional estrus appraisers becomes a major problem in the modern pig-raising industry.
In addition, modern breeding is generally a large-scale intensive breeding mode, which requires high isolation of a breeding environment, avoids the contact of pigs with external people or objects as much as possible, and prevents disease transmission, especially African swine fever which is a disease easy to spread. The existing estrus identification mode needs the personnel to contact with the sow for many times and closely contact with the boar, which is not beneficial to disease prevention and control.
The existing automatic estrus identification equipment mainly comprises the following devices: 1. the wearable ear tag/neck ring and other devices monitor the state of the sow through a built-in temperature sensor, a built-in motion sensor and the like; 2. the fixed video monitoring equipment judges whether the sow oestrus occurs or not by detecting the behavior posture of the sow; 3. whether the sow is in estrus or not is judged by measuring the vaginal resistance of the sow. Wearable equipment is often used in a large-circle breeding mode, is difficult to apply under the condition of single-column breeding, and is easy to damage; fixed video monitoring equipment needs to continuously process video information in real time, has high requirements on software and hardware, has high equipment cost and is difficult to popularize on a large scale; the vaginal resistance measurement method has the risk of disease transmission among different pigs, and is not beneficial to disease prevention and control.
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It should be apparent that the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The core of the invention is to provide a sow estrus identification method, the flow diagram of one specific embodiment of which is shown in figure 1, and the specific embodiment is called as the first specific embodiment, and comprises the following steps:
s101: and receiving the ultrasonic emission signal and the corresponding ultrasonic reflection signal.
The ultrasonic emission signal and the ultrasonic reflection signal are signals actively emitted by the ultrasonic detector and signals reflected back after the signals contact with the obstacles, and the obstacles are sows to be detected in an ideal state.
S102: and determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal.
And obtaining the time length of the ultrasonic wave passing by using the time information of the transmitted ultrasonic wave and the time information corresponding to the received reflected ultrasonic wave.
S103: and determining the posture information of the sow according to the menstruation duration of the ultrasonic waves.
In this step, can pass through it is long when the ultrasonic wave menstruation, distinguish the sow gesture, mainly divide into active state and state of sleeping, wherein active state is including the gesture of standing and dog seat gesture, and the state of sleeping calmly includes the state of lying on one's side and lying prone.
For example, the ultrasonic detector can be arranged above the monoblock fence to emit ultrasonic waves downwards, if the distance between the sow and the ultrasonic detector is shortened, the sow is in a standing posture, and otherwise the sow is in a lying state; the ultrasonic detectors can also form an array on the side surface of the single fence, can draw the general shape of the sow so as to judge the posture, and can select other modes for confirming the posture of the sow through ultrasonic waves according to actual conditions, and the invention is not limited herein.
S104: and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state.
The Machine learning model used in the invention can be a Support Vector Machine (SVM), and certainly, other classification models can be adopted in practical application, taking the SVM as an example, and an analysis method of the sow posture information corresponding to the target time period is given as follows:
(1) According to the posture distribution of the pigs in the target time period, obtaining Activity characteristics such as active state holding time st, active times sn, maximum active time st _ max and the like in the time period to form a characteristic vector A (Activity) = [ st, sn, st _ max ]; the Activity Feature of the target time period is AF (Activity Feature);
(2) After the Activity Feature AF of the target time period is extracted, comparing the Activity Feature AF with the Activity Feature PAF (Previous Activity Feature) of the Previous time period to obtain an Activity Difference value AD (Activity Difference) = AF-PAF;
(3) And (3) forming an estrus prediction vector X = [ AF, AD ] by using the extracted activity characteristics AF and AD, wherein X is a floating point type array with the length of 24, and inputting the vector into a trained machine learning model to predict the state of the sow.
In addition, when the machine learning model is trained, the characteristic extraction mode is consistent with the steps, and the model label is given by a sentiment searching person at the first production line; an estrus prediction model required by the system can be trained by acquiring data offline and recording the estrus labels.
Of course, in the target time period, the more intensive the acquisition of the sow posture information, that is, the shorter the acquisition interval, the more precise and comprehensive the obtained change of the sow posture and the more realistic the obtained data of the sow activity amount, and therefore, preferably, the time interval for determining the sow posture information ranges from 1 second to 5 seconds, inclusive.
Furthermore, after the ultrasonic wave passing time length is obtained, the ultrasonic wave passing time length is subjected to smooth filtering, and abnormal values are eliminated.
If a plurality of ultrasonic detectors exist in a single body column in the system, data of the plurality of ultrasonic detectors can be integrated to obtain behavior information corresponding to the sow in the column.
The step of inputting the sow posture information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state comprises the following steps:
inputting the posture information of the sows in one complete day into a pre-trained oestrus state model to obtain oestrus probability information; the estrus probability information comprises probability of estrus failure, recorded as probability A, probability of being in the day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C and probability of estrus in the afternoon, recorded as probability D, and probability of four conditions in total;
and determining the situation with the maximum probability as the oestrus state of the sow according to the oestrus probability information.
The oestrus states of the sows are divided into four types in the preferred embodiment, namely, the sows are not oestrous, are in oestrus before one day, are oestrus in the morning of the day and are oestrus in the afternoon of the day, the sows can be conveniently scheduled and arranged by the farm workers in advance under four conditions, the management efficiency of the farm is greatly improved, a certain oestrus prediction function is achieved, the oestrus missing period is avoided, and the breeding opportunity is wasted.
Continuing the four classifications of the oestrus states, providing a preferred implementation mode, and when the oestrus state of the sow corresponding to the first day after the sow is weaned and turned is not oestrous, acquiring oestrus probability information corresponding to the first day after the sow is weaned and turned;
judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to the oestrus probability information corresponding to the first day after the sows are weaned and turned;
and when the difference values do not exceed the first threshold value, determining that the oestrus state of the sow corresponding to the first day after the sow is weaned and turned into a stall is not oestrous.
The sow can be put into a pigsty after the lactation period is finished, at the moment, because the sow just finishes the lactation on the piglet and the probability of directly entering the estrus is very low, when the estrus state of the sow corresponding to the first day after the sow is weaned and transferred is not in estrus, whether the sow is in the estrus is required to be further judged, the judging method is to respectively compare the probabilities of three conditions of the day before estrus, the morning estrus in the day and the afternoon of the day with the probability of not estrus, if the probabilities of the three conditions are not too much higher than the probability of not estrus (if the probabilities are within 10%), the prediction is judged to be in the misjudgment, the prediction result is changed into the estrus, and therefore the prediction accuracy is greatly improved, and resource waste is avoided.
Continuing the four categories of the oestrus states, and providing a preferred embodiment, when the oestrus state of the sow corresponding to the target date is morning oestrus or afternoon oestrus but the oestrus state of the sow corresponding to the day before the target date is not oestrus, acquiring oestrus probability information corresponding to the target date;
judging whether the probability C and the probability D both exceed the probability A and the probability B according to the estrus probability information corresponding to the target date;
and when at least one of the C probability and the D probability does not exceed the A probability and the B probability, determining the condition corresponding to the larger one of the A probability and the B probability as the sow oestrus state.
In the above four cases, the cases of directly judging the oestrus of the sows are C and D, namely the morning oestrus and the afternoon oestrus, and the continuous days of judging whether the sows oestrus in the ideal case are from the oestrus failure (a) to the day before oestrus (B), then to the morning oestrus (C) or the afternoon oestrus (D), the case of skipping B is directly from a to C/D, which is a rare case, and may be system misjudgment, so further verification is needed.
In the preferred embodiment, the logic is followed that if the sow oestrus on the same day, the activity of the sow on the same day should be large, and it is unlikely that the sow is in oestrus in the morning, in the afternoon, or in the morning, and oestrus ends in the afternoon, so that if the predicted oestrus on the same day in the morning or in the afternoon is true, the probability of oestrus on the other half of the day should be much higher than the probabilities of the cases a and B, and therefore the probabilities of the cases C and D are examined, and both probabilities are determined to be higher than the probabilities of the cases a and B, and then the original prediction result is determined to be correct, otherwise, the sow is not oestrus at all on the same day, the original prediction result is determined to be incorrect, and the final prediction result with high probability is selected from the cases a and B, and the prediction accuracy is improved again.
Continuing the four categories of the oestrus states, and providing a preferred embodiment, when the oestrus state of the sow corresponding to a target date is the day before oestrus and the oestrus state of the sow corresponding to the day before the target date is also the day before oestrus, acquiring oestrus probability information corresponding to the target date;
and determining the condition that the greater one of the C probability and the D probability corresponds to the sow oestrus state according to the oestrus probability information corresponding to the target date.
Of course, if the result of the prediction is output for two consecutive days, the result of the 'day before oestrus' is output, the prediction is also out of tolerance approximately, the prediction result of the next day is modified, the sow is acquiescently oestrous on the next day, and the sow with high probability is selected from the condition C and the condition D to serve as the final prediction result of the next day, so that the sow oestrus is prevented from being missed, and the oestrus utilization rate of the sow is improved.
The invention aims to automatically monitor the oestrus state of sows through a non-contact ultrasonic signal, and has the following advantages:
1. the artificial intelligence algorithm can effectively help the inexperienced breeders to check the estrus of the sows, so that the estrus utilization rate of the sows is improved;
2. the non-contact and non-interference sow monitoring can be realized by utilizing ultrasonic signals, which is not only beneficial to biosafety prevention and control, but also does not influence the daily activities of pigs;
3. the ultrasonic equipment is relatively low in cost and can be popularized on a large scale.
The sow estrus identification method provided by the invention receives ultrasonic emission signals and corresponding ultrasonic reflection signals; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state. The posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through a model obtained through machine learning.
On the basis of the first embodiment, a plurality of ultrasonic detectors are further provided, and a target ultrasonic wave is screened out from signals of the plurality of ultrasonic detectors, so as to obtain a second embodiment, a flowchart of which is shown in fig. 2, and includes:
s201: receiving a plurality of ultrasonic emission signals and corresponding ultrasonic reflection signals; the ultrasonic detector corresponding to the ultrasonic reflection signal is arranged above the monomer column and transmits ultrasonic waves downwards.
S202: and determining the corresponding ultrasonic wave passing time length according to the plurality of ultrasonic wave reflection signals.
S203: and determining the shortest ultrasonic wave passing time length as the target time length.
S204: and determining the posture information of the sow according to the target duration.
S205: and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state.
In this specific embodiment, in combination with the setting mode of the ultrasonic probe, the processing mode of the ultrasonic signal is further limited, specifically, when the ultrasonic wave is emitted downward above the solitary fence, because the position of the sow in the solitary fence is not fixed, there may be a part of the ultrasonic wave emitted by the ultrasonic probe that does not contact the sow, but strikes the ground and reflects back, and this kind of ultrasonic wave cannot reflect the posture information of the sow, and therefore belongs to an invalid signal.
Preferably, a measurement receiving threshold value can be set, namely, the longest effective distance (namely, the longest menstruation time) is set, and the ultrasonic waves which are lower than the longest menstruation time indicate that the corresponding ultrasonic waves are applied to the sow instead of the ground, so that the ultrasonic waves can be collected into effective time signals, the positions of the ultrasonic detectors corresponding to the effective time signals are further combined, the judgment result of the sow posture with high precision can be obtained, and the measurement accuracy is greatly improved.
On the basis of the first specific embodiment, the accuracy of evaluating the activity of sows is further improved, and a third specific embodiment is obtained, wherein a flow diagram of the third specific embodiment is shown in fig. 3, and comprises the following steps:
s301: and receiving the ultrasonic emission signal and the corresponding ultrasonic reflection signal.
S302: and determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal.
S303: and determining daytime attitude characteristic information and night attitude characteristic information according to the ultrasonic wave passing time and the acquisition time tag corresponding to the ultrasonic wave passing time.
S304: and determining the activity information of the sow according to the daytime posture characteristic information and the night posture characteristic information corresponding to the target date.
S305: and inputting the sow activity information corresponding to the target date into a pre-trained estrus state model to obtain the sow estrus state.
In the specific embodiment, the information of the oestrus state of the sow is generated by taking the day as a cycle, and the characteristics of the activity amounts of the day and the night in the past day are respectively extracted at a fixed time point (such as 6 am) every day, wherein the duration of the day is 16 hours (namely 6 am to 10 pm), and the duration of the night is 8 hours (10 pm to 6 next morning); in the daytime, sows are artificially interfered, and activities of artificial participation such as feeding, condition checking, mating and the like exist; therefore, the activity of the sow in the daytime is higher than that of the sow without interference, and if the sow is judged to be oestrous according to the fixed activity, the misjudgment caused by artificial influence is high in probability.
Therefore, in the present embodiment, the daily activity of the sow is divided into the daytime activity (i.e. the daytime posture characteristic information) and the nighttime activity (i.e. the nighttime posture characteristic information), the basal activity of the sow is determined according to the nighttime activity of the same sow, and then whether the daytime activity is increased due to human influence or increased due to the estrus is determined according to the basal activity of the sow, so that the difference between pigs is eliminated, and the probability of misjudgment of estrus caused by the difference between the basal activity of different sows and the difference between the degrees of human influence is reduced.
Further, the determining the sow activity amount information according to the daytime posture characteristic information and the night posture characteristic information corresponding to the target date comprises:
and dividing the daytime attitude characteristic information and the night attitude characteristic information to obtain the sow activity information.
The division operation is simple, the calculation is convenient, and the effect of eliminating the difference of different sows by taking the undisturbed state attitude characteristic information as the standard is realized.
In the following, the sow estrus identification device provided by the embodiment of the present invention is introduced, and the sow estrus identification device described below and the sow estrus identification method described above can be referred to correspondingly.
Fig. 4 is a block diagram illustrating a structure of a sow estrus identification apparatus according to an embodiment of the present invention, and referring to fig. 4, the sow estrus identification apparatus may include:
a receiving module 100, configured to receive an ultrasonic reflection signal;
a menstruation module 200, configured to determine a menstruation duration of the ultrasonic wave according to the ultrasonic reflection signal;
the behavior module 300 is used for determining the posture information of the sow according to the menstruation duration of the ultrasonic waves;
and the state module 400 is used for inputting the sow posture information corresponding to the target time period into the pre-trained estrus state model to obtain the sow estrus state.
As a preferred embodiment, the receiving module 100 includes:
the multi-source receiving unit is used for receiving a plurality of ultrasonic transmitting signals and corresponding ultrasonic reflecting signals; the ultrasonic detector corresponding to the ultrasonic reflection signal is arranged above the monomer column and transmits ultrasonic waves downwards;
accordingly, the traversing module 200 comprises:
the multi-source passing unit is used for determining corresponding ultrasonic wave passing time according to the ultrasonic wave reflection signals;
the short-taking unit is used for determining the shortest ultrasonic wave passing time length as a target time length;
accordingly, the behavior module 300 includes:
and the target behavior unit is used for determining the posture information of the sow according to the target duration.
As a preferred embodiment, the behavior module 300 includes:
the day and night label unit is used for determining daytime attitude characteristic information and night attitude characteristic information according to the ultrasonic wave menstruation duration and the acquisition time label corresponding to the ultrasonic wave menstruation duration;
and the relative activity unit is used for determining the activity information of the sow according to the daytime posture characteristic information and the night posture characteristic information corresponding to the target date.
As a preferred embodiment, the status module 400 includes:
the situation probability unit is used for inputting the posture information of the sows in one complete day into the pre-trained estrus state model to obtain estrus probability information; the estrus probability information comprises probability of estrus failure, recorded as probability A, probability of being in the day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C and probability of estrus in the afternoon, recorded as probability D, and probability of four conditions in total;
and the oestrus classification unit is used for determining the situation with the maximum probability as the oestrus state of the sow according to the oestrus probability information.
As a preferred embodiment, the method further comprises:
the sow oestrus control system comprises a weaning turning acquisition module, a control module and a control module, wherein the weaning turning acquisition module is used for acquiring oestrus probability information corresponding to a first day after a sow is weaned and turned when the oestrus state of the sow corresponding to the first day after the sow is weaned and turned is not oestrus;
the inspection error module is used for judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to oestrus probability information corresponding to the first day after the sows are weaned and turned;
and the oestrus failure confirming module is used for determining that the oestrus state of the sow corresponding to the first day after the sow is weaned and turned into a stall is oestrus failure when the difference values do not exceed the first threshold value.
As a preferred embodiment, the method further comprises:
the non-transition acquisition module is used for acquiring oestrus probability information corresponding to a target date when the oestrus state of the sow corresponding to the target date is in the morning or afternoon but the oestrus state of the sow corresponding to the day before the target date is in the absence of oestrus;
the probability confirmation module is used for judging whether the probability C and the probability D both exceed the probability A and the probability B according to the oestrus probability information corresponding to the target date;
and the non-transition misjudgment module is used for determining the situation corresponding to the larger one of the A probability and the B probability as the oestrus state of the sow when at least one of the C probability and the D probability does not exceed the A probability and the B probability.
As a preferred embodiment, the method further comprises:
the repeated transition acquisition module is used for acquiring oestrus probability information corresponding to a target date when the oestrus state of the sows corresponding to the target date is the day before oestrus and the oestrus state of the sows corresponding to the day before the target date is the day before oestrus;
and the oestrus determining module is used for determining the condition corresponding to the larger one of the probability C and the probability D as the oestrus state of the sow according to the oestrus probability information corresponding to the target date.
The sow estrus identification device of the embodiment is used for realizing the above-mentioned sow estrus identification method, and therefore specific embodiments in the sow estrus identification device can be seen in the above-mentioned embodiment parts of the sow estrus identification method, for example, the receiving module 100, the menstruation module 200, the behavior module 300, and the status module 400 are respectively used for realizing steps S101, S102, S103, and S104 in the above-mentioned sow estrus identification method, so that the specific embodiments thereof can refer to descriptions of corresponding partial embodiments, and are not repeated herein.
The sow estrus identification device provided by the invention is used for receiving ultrasonic reflection signals through the receiving module; the travelling module is used for determining the travelling time of the ultrasonic waves according to the ultrasonic reflection signals; the behavior module is used for determining the posture information of the sow according to the ultrasonic wave menstruation duration; and the state module is used for inputting the sow posture information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state. According to the sow estrus monitoring method, the posture of the sow can be determined through the retracing time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the estrus state or not can be inferred through a model obtained through machine learning.
A sow estrus identification apparatus comprising:
a memory for storing a computer program;
a processor for implementing the steps of the sow oestrus identification method as described in any one of the above when the computer program is executed. The sow estrus identification method provided by the invention receives ultrasonic emission signals and corresponding ultrasonic reflection signals; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state. The posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through a model obtained through machine learning.
A computer-readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method of oestrus identification of sows as claimed in any one of the preceding claims. The sow estrus identification method provided by the invention receives ultrasonic emission signals and corresponding ultrasonic reflection signals; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state. The posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through a model obtained through machine learning.
The invention also provides a sow estrus identification system with the beneficial effects, the structural schematic diagram of which is shown in fig. 5 and is called as a fifth specific implementation mode, and the sow estrus identification system comprises an ultrasonic detector 10 and a data processor 20;
the ultrasonic detector 10 is used for transmitting target ultrasonic waves to the sow to be detected and receiving the target ultrasonic waves reflected by the sow to be detected;
the data processor 20 is configured to receive an ultrasonic emission signal and an ultrasonic reflection signal corresponding to the target ultrasonic wave; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the sow oestrus state.
The sow estrus identification system provided by the invention comprises an ultrasonic detector 10 and a data processor 20; the ultrasonic detector 10 is used for transmitting target ultrasonic waves to the sow to be detected and receiving the target ultrasonic waves reflected by the sow to be detected; the data processor 20 is configured to receive an ultrasonic emission signal and an ultrasonic reflection signal corresponding to the target ultrasonic wave; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining sow posture information according to the ultrasonic wave menstruation duration; and inputting the sow posture information corresponding to the target time period into a pre-trained oestrus state model to obtain the oestrus state of the sow. The posture of the sow can be determined through the turning-back time of the ultrasonic signals, the activity of the sow (namely the posture information of the sow) can be obtained through the posture change of the sow at different times, and then whether the sow is in the oestrus state or not can be inferred through a model obtained through machine learning.
Preferably, the sow estrus identification system is provided with a plurality of ultrasonic detectors 10 corresponding to a single monoblock fence;
the ultrasonic detector 10 is arranged above the single column and emits ultrasonic waves downwards;
the plurality of ultrasonic detectors 10 are arranged at equal intervals along a perpendicular bisector perpendicular to the long axis direction of the horizontal projection of the single fence.
Referring to fig. 6, fig. 6 is a top view of the preferred embodiment, because the size of the solitary column is not much different from that of the sow, but the sow has more activity space in the minor axis direction compared with the front-back movement in the major axis direction, in the preferred embodiment, a plurality of ultrasonic detectors 10 are arranged at intervals along the perpendicular bisector perpendicular to the major axis direction of the solitary column, the solitary column can be covered by a small number of ultrasonic detectors 10, the ultrasonic signals reflected back by the hit sow can be guaranteed, and the working stability of the system can be improved.
The invention provides a specific implementation mode, and the intelligent estrus identification system built by the invention comprises four components, namely: the system comprises a distance acquisition component, a data processing component, an AI (Artificial Intelligence) calculation component and a time sequence aggregation component, wherein the four components work in series. The distance acquisition component continuously monitors the distance change for 24 hours and transmits the distance change to the data processing component; the data processing component is used for extracting activity characteristics and then transmitting the characteristics to the AI computing component; the AI computing component predicts whether the sow is oestrous according to the trained model and transfers the prediction result to the time sequence aggregation component; and the time sequence aggregation component judges whether the sow is oestrous or not according to the state change condition of a plurality of continuous days. The specific working principle flow of each component is as follows:
1. the distance acquisition assembly consists of array type ultrasonic equipment, and each sow of the system is provided with two ultrasonic probes for ensuring the accuracy of distance measurement and preventing the influence caused by the limited beam angle of the ultrasonic equipment;
2. the distance acquisition component transmits the data to the data processing component through the serial port. The data processing component is a single-board computing terminal with certain computing power and is used for controlling data acquisition, processing and uploading. One data processing assembly is connected with 20 data acquisition assemblies, the distance values acquired by each ultrasonic device are sequentially read in a polling mode, the data of each assembly are stored in the local of the processor for backup and calculation at intervals (such as 3 minutes), and meanwhile, the original data are uploaded to a remote data transfer station for storage, so that research and development personnel can call the data conveniently; taking out the data volume of each device in the past day from the local area every time when the oestrus is predicted (such as 6 am), calculating a characteristic vector required by oestrus identification, and uploading the characteristic vector to an AI (artificial intelligence) calculation component;
3. the data processing process comprises smooth filtering, attitude judgment, data integration, activity statistics and feature extraction;
(1) The smooth filtering part adopts a median filtering mode to remove abnormal values;
(2) In the posture judging process, according to the distance values acquired by the ultrasonic equipment, the differences of the distance values of the sows in different postures are combined, and the sow at the time point under the ultrasonic sensor is judged to be in a lying state or a standing state, wherein the standing state is represented by 0, and the lying state is represented by 1;
(3) In the data integration process, attitude data obtained by a plurality of sensors of the same column are integrated, and as long as one sensor judges that the pig is only in a standing state, the sow is judged to be in an active state at the time point;
(4) The activity statistical process is to count the parameters of the number of times, duration and the like of the sows in the active state in a specified time interval, and is used for quantifying the activity condition of the sows;
(5) And in the characteristic extraction process, the statistical parameters of the activity are subjected to summation, subtraction, ratio and other processing, and all the parameters are combined to obtain a characteristic vector.
4. And the data processing component uploads the original data and the feature vector to the AI computing component in a wireless transmission mode. The AI computing component is usually a server platform, on one hand, the received characteristic values are input into an offline trained machine learning model to obtain the oestrus judgment result and the possibility of each category of the current day, and on the other hand, the received original data are forwarded to be transmitted to the time sequence aggregation component;
if the performance of the data processing component is better, AI calculation can be carried out, oestrus judgment can be directly finished on the data processing component, at the moment, the AI calculation component only finishes data cleaning and data forwarding work, namely, the single board calculation terminal directly finishes the output of the machine learning model, then a model output result is uploaded to an AI calculation server, at the moment, the AI calculation server only needs to judge the validity of data, and valid data is uploaded to a data console.
5. The time sequence aggregation component is usually completed in a data center, on one hand, the time sequence aggregation component plays a role in data storage, on the other hand, simple logic operation is carried out, and a final prediction result is judged according to certain time sequence logic.
The specific working principle and other preferred schemes of the sow oestrus verification system are referred to the above parts of the sow oestrus identification method and device, and are not described herein again.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
It should be noted that, in the present specification, relational terms such as first and second, and the like are used only for distinguishing one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional like elements in a process, method, article, or apparatus that comprises the element.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The sow estrus identification method, the device and the system provided by the invention are introduced in detail. The principles and embodiments of the present invention have been described herein using specific examples, which are presented only to assist in understanding the method and its core concepts of the present invention. It should be noted that, for those skilled in the art, without departing from the principle of the present invention, it is possible to make various improvements and modifications to the present invention, and those improvements and modifications also fall within the scope of the claims of the present invention.
Claims (8)
1. A sow oestrus identification method is characterized by comprising the following steps:
receiving an ultrasonic transmitting signal and a corresponding ultrasonic reflecting signal;
determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal;
determining sow posture information according to the ultrasonic wave menstruation duration;
inputting sow posture information corresponding to a target time period into a pre-trained oestrus state model to obtain a sow oestrus state;
the step of inputting the sow attitude information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state comprises the following steps:
inputting the posture information of the sows in one complete day into a pre-trained oestrus state model to obtain oestrus probability information; the estrus probability information comprises probability of no estrus, recorded as probability A, probability of a day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C, probability of estrus in the afternoon and recorded as probability D, and probability of four conditions in total;
according to the oestrus probability information, determining the situation with the maximum probability as the oestrus state of the sow;
when the oestrus state of the sows corresponding to the first day after the sows are weaned and turned is not oestrus, acquiring oestrus probability information corresponding to the first day after the sows are weaned and turned;
judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to the oestrus probability information corresponding to the first day after the sows are weaned and turned;
and when the difference values do not exceed the first threshold value, determining that the oestrus state of the sow corresponding to the first day after the sow is weaned and turned into a stall is not oestrous.
2. The method of claim 1 wherein said receiving an ultrasonic emission signal and a corresponding ultrasonic reflection signal comprises:
receiving a plurality of ultrasonic emission signals and corresponding ultrasonic reflection signals; the ultrasonic detector corresponding to the ultrasonic reflection signal is arranged above the monomer column and transmits ultrasonic waves downwards;
correspondingly, determining the corresponding ultrasonic wave passing time length according to the plurality of ultrasonic wave reflection signals;
determining the shortest ultrasonic wave passing time as a target time;
correspondingly, the posture information of the sow is determined according to the target duration.
3. The sow estrus identification method of claim 1, wherein said determining sow posture information based on said ultrasound menstruation duration comprises:
determining daytime attitude characteristic information and night attitude characteristic information according to the ultrasonic wave passing time and an acquisition time tag corresponding to the ultrasonic wave passing time;
determining sow activity information according to the posture characteristic information of the day and the posture characteristic information of the night corresponding to the target date;
correspondingly, the sow activity information corresponding to the target date is input into the pre-trained oestrus state model to obtain the sow oestrus state.
4. The sow estrus identification method of claim 1, wherein when the sow estrus state corresponding to a target date is morning estrus or afternoon estrus but the sow estrus state corresponding to the day before the target date is not estrus, acquiring the estrus probability information corresponding to the target date;
judging whether the probability C and the probability D both exceed the probability A and the probability B according to the estrus probability information corresponding to the target date;
and when at least one of the C probability and the D probability does not exceed the A probability and the B probability, determining the condition corresponding to the larger one of the A probability and the B probability as the sow oestrus state.
5. The sow estrus identification method of claim 1, wherein when the sow estrus state corresponding to a target date is the day before estrus and the sow estrus state corresponding to the day before the target date is also the day before estrus, acquiring the estrus probability information corresponding to the target date;
and determining the condition that the greater one of the probability C and the probability D corresponds to the oestrus state of the sow according to the oestrus probability information corresponding to the target date.
6. A sow estrus identification device, comprising:
the receiving module is used for receiving the ultrasonic reflection signal;
the travelling module is used for determining the travelling time of the ultrasonic waves according to the ultrasonic reflection signals;
the behavior module is used for determining the posture information of the sow according to the ultrasonic wave menstruation duration;
the state module is used for inputting the sow posture information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state;
the step of inputting the sow posture information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state comprises the following steps:
inputting the posture information of the sows in one complete day into a pre-trained oestrus state model to obtain oestrus probability information; the estrus probability information comprises probability of no estrus, recorded as probability A, probability of a day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C, probability of estrus in the afternoon and recorded as probability D, and probability of four conditions in total;
according to the oestrus probability information, determining the situation with the maximum probability as the oestrus state of the sow;
when the oestrus state of the sows corresponding to the first day after the sows are weaned and turned is not oestrous, acquiring oestrus probability information corresponding to the first day after the sows are weaned and turned;
judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to the oestrus probability information corresponding to the first day after the sows are weaned and turned;
and when the difference values do not exceed the first threshold value, determining that the oestrus state of the sows corresponding to the first day after the sows are weaned and turned is oestrous.
7. A sow oestrus identification system is characterized by comprising an ultrasonic detector and a data processor;
the ultrasonic detector is used for transmitting target ultrasonic waves to the sow to be detected and receiving the target ultrasonic waves reflected by the sow to be detected;
the data processor is used for receiving ultrasonic wave transmitting signals and ultrasonic wave reflecting signals corresponding to the target ultrasonic waves; determining the ultrasonic wave passing time length according to the ultrasonic wave transmitting signal and the ultrasonic wave reflecting signal; determining the posture information of the sow according to the ultrasonic wave menstruation duration;
inputting sow posture information corresponding to a target time period into a pre-trained oestrus state model to obtain a sow oestrus state;
the step of inputting the sow attitude information corresponding to the target time period into the pre-trained oestrus state model to obtain the sow oestrus state comprises the following steps:
inputting the sow posture information of a complete day into a pre-trained oestrus state model to obtain oestrus probability information; the estrus probability information comprises probability of estrus failure, recorded as probability A, probability of being in the day before estrus, recorded as probability B, probability of estrus in the morning, recorded as probability C and probability of estrus in the afternoon, recorded as probability D, and probability of four conditions in total;
according to the oestrus probability information, determining the situation with the maximum probability as the oestrus state of the sow;
when the oestrus state of the sows corresponding to the first day after the sows are weaned and turned is not oestrous, acquiring oestrus probability information corresponding to the first day after the sows are weaned and turned;
judging whether the difference values of the probability A and the probability B, the probability C and the probability D respectively exceed a first threshold value or not according to the oestrus probability information corresponding to the first day after the sows are weaned and turned;
and when the difference values do not exceed the first threshold value, determining that the oestrus state of the sow corresponding to the first day after the sow is weaned and turned into a stall is not oestrous.
8. The sow estrus identification system of claim 7 wherein said sow estrus identification system is provided with a plurality of said ultrasonic detectors corresponding to a single pigsty;
the ultrasonic detector is arranged above the single column and transmits ultrasonic waves downwards;
the plurality of ultrasonic detectors are arranged at equal intervals along a perpendicular bisector perpendicular to the long axis direction of the horizontal projection of the single fence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110996071.1A CN113711944B (en) | 2021-08-27 | 2021-08-27 | Sow estrus identification method, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110996071.1A CN113711944B (en) | 2021-08-27 | 2021-08-27 | Sow estrus identification method, device and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113711944A CN113711944A (en) | 2021-11-30 |
| CN113711944B true CN113711944B (en) | 2023-03-03 |
Family
ID=78678526
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110996071.1A Active CN113711944B (en) | 2021-08-27 | 2021-08-27 | Sow estrus identification method, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113711944B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114586701A (en) * | 2022-04-15 | 2022-06-07 | 东南大学 | Milk cow oestrus prediction device based on body temperature and exercise amount data |
| CN114793948B (en) * | 2022-05-10 | 2024-03-12 | 重庆市六九畜牧科技股份有限公司 | Sow oestrus intelligent management system |
| CN114747537B (en) * | 2022-05-10 | 2024-01-26 | 重庆市六九畜牧科技股份有限公司 | Sow oestrus management method |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103767802A (en) * | 2014-01-16 | 2014-05-07 | 诠懋实业有限公司 | Sow mating management method |
| CN106037685A (en) * | 2016-07-07 | 2016-10-26 | 王凡 | Intelligent analysis system for sow oestrous diagnosis |
| CN208002799U (en) * | 2017-09-29 | 2018-10-26 | 汉王科技股份有限公司 | Sitting posture detection device |
| CN108717523A (en) * | 2018-04-26 | 2018-10-30 | 华南农业大学 | Oestrus of sow behavioral value method based on machine vision |
| CN109042514A (en) * | 2018-08-15 | 2018-12-21 | 广西农垦永新畜牧集团新兴有限公司 | A method of improving wean oestrus of sow rate |
| CN109984054A (en) * | 2019-04-19 | 2019-07-09 | 广州影子科技有限公司 | Oestrous detection method, oestrous detection device and oestrous detection system |
| CN110839557A (en) * | 2019-10-16 | 2020-02-28 | 北京海益同展信息科技有限公司 | Sow oestrus monitoring method, device and system, electronic equipment and storage medium |
| CN111034642A (en) * | 2019-11-25 | 2020-04-21 | 秒针信息技术有限公司 | Method and device for managing female livestock to be bred |
| CN111914685A (en) * | 2020-07-14 | 2020-11-10 | 北京小龙潜行科技有限公司 | Sow oestrus detection method and device, electronic equipment and storage medium |
| CN112568140A (en) * | 2019-09-29 | 2021-03-30 | 牧原食品股份有限公司 | Device for inducing oestrus of sows |
| CN112586368A (en) * | 2020-12-31 | 2021-04-02 | 重庆市六九畜牧科技股份有限公司 | Boar lures feelings fence pig house with sow monitoring |
| CN113016657A (en) * | 2021-03-05 | 2021-06-25 | 河南牧原智能科技有限公司 | Pigsty sow oestrus identification system and application method thereof |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ITMI20071072A1 (en) * | 2007-05-25 | 2008-11-26 | Lps Electronics S R L | EQUIPMENT FOR THE AUTOMATIC DETECTION OF THE HEAT OF A SHROUD. |
| US9044297B2 (en) * | 2011-03-17 | 2015-06-02 | Technologies Holdings Corp. | System and method for estrus detection using real-time location |
| KR102089165B1 (en) * | 2018-07-19 | 2020-03-13 | 신상운 | A system that recognizes objects and recognizes motion by diverging ultrasonic waves |
| CN109258508B (en) * | 2018-09-26 | 2020-11-10 | 深圳市倍适沃智能设备有限公司 | Sow oestrus analysis method and device, terminal and computer readable storage medium |
| KR102307327B1 (en) * | 2019-04-05 | 2021-09-30 | 주)엠트리센 | Electronic Sow Management Apparatus |
| JP7162749B2 (en) * | 2019-09-06 | 2022-10-28 | 日本ハム株式会社 | Estrus determination device for sows, estrus determination method for sows, and estrus determination program for sows |
| CN113179963A (en) * | 2021-03-31 | 2021-07-30 | 广东广垦畜牧工程研究院有限公司 | Intelligent management and accurate feeding system and method for farm |
-
2021
- 2021-08-27 CN CN202110996071.1A patent/CN113711944B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103767802A (en) * | 2014-01-16 | 2014-05-07 | 诠懋实业有限公司 | Sow mating management method |
| CN106037685A (en) * | 2016-07-07 | 2016-10-26 | 王凡 | Intelligent analysis system for sow oestrous diagnosis |
| CN208002799U (en) * | 2017-09-29 | 2018-10-26 | 汉王科技股份有限公司 | Sitting posture detection device |
| CN108717523A (en) * | 2018-04-26 | 2018-10-30 | 华南农业大学 | Oestrus of sow behavioral value method based on machine vision |
| CN109042514A (en) * | 2018-08-15 | 2018-12-21 | 广西农垦永新畜牧集团新兴有限公司 | A method of improving wean oestrus of sow rate |
| CN109984054A (en) * | 2019-04-19 | 2019-07-09 | 广州影子科技有限公司 | Oestrous detection method, oestrous detection device and oestrous detection system |
| CN112568140A (en) * | 2019-09-29 | 2021-03-30 | 牧原食品股份有限公司 | Device for inducing oestrus of sows |
| CN110839557A (en) * | 2019-10-16 | 2020-02-28 | 北京海益同展信息科技有限公司 | Sow oestrus monitoring method, device and system, electronic equipment and storage medium |
| CN111034642A (en) * | 2019-11-25 | 2020-04-21 | 秒针信息技术有限公司 | Method and device for managing female livestock to be bred |
| CN111914685A (en) * | 2020-07-14 | 2020-11-10 | 北京小龙潜行科技有限公司 | Sow oestrus detection method and device, electronic equipment and storage medium |
| CN112586368A (en) * | 2020-12-31 | 2021-04-02 | 重庆市六九畜牧科技股份有限公司 | Boar lures feelings fence pig house with sow monitoring |
| CN113016657A (en) * | 2021-03-05 | 2021-06-25 | 河南牧原智能科技有限公司 | Pigsty sow oestrus identification system and application method thereof |
Non-Patent Citations (2)
| Title |
|---|
| 不同处理方案对肉牛同期发情率和情期受胎率的影响研究;张长兴;《黑龙江畜牧兽医》;20190831(第15期);第73-75页 * |
| 猪人工授精之发情鉴定;徐建党;《养猪》;20121231(第6期);第33-34页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113711944A (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113711944B (en) | Sow estrus identification method, device and system | |
| Mayo et al. | Automated estrous detection using multiple commercial precision dairy monitoring technologies in synchronized dairy cows | |
| CN109618961B (en) | Intelligent feeding system and method for livestock | |
| KR102307327B1 (en) | Electronic Sow Management Apparatus | |
| WO2019218602A1 (en) | Internet of things system for cow behavior and physiological characteristics, and operation method therefor | |
| Maselyne et al. | Online warning systems for individual fattening pigs based on their feeding pattern | |
| US20190183436A1 (en) | Information processing device, information processing system, and information processing method | |
| CA3093190A1 (en) | System and method for determining animal behavioral phenotypes | |
| KR20190030877A (en) | Livestock specification management system based on livestock knowledge information | |
| CN110741962A (en) | Oestrus and health monitoring method based on cow behavior recognition algorithm | |
| CN104200623A (en) | Automatic animal behavior information digitization acquisition system and method | |
| CN112568141A (en) | Supervision system for preventing diseases of pigs | |
| CN205593626U (en) | Livestock -raising automatic weighing comparison system | |
| WO2005070326A1 (en) | System and process for determining whether an animal is in oestrus | |
| Higaki et al. | An attempt at estrus detection in cattle by continuous measurements of ventral tail base surface temperature with supervised machine learning | |
| US11096375B2 (en) | Smart cattle reproduction management and digital farm to market transparency metrics | |
| CA2854344C (en) | Infrared thermography and behaviour information for identification of biolically important states in animals | |
| CN112201345A (en) | Method for analyzing cattle diseases based on motion sensor | |
| Tahsin | Development of a propeller P8X 32A based wireless biosensor system for cattle health monitoring and disease detection | |
| JP2002233261A (en) | System for increasing efficiency of feeding control of cattle with cpu | |
| KR102461183B1 (en) | Apparatus for detecting cattle estrus | |
| CN215602584U (en) | Milk cow individual weighs and daily ration information real-time supervision device that adopts | |
| Xu | Detecting estrus in sows using a robotic imaging system and neural networks | |
| CN117034971A (en) | Livestock inspection system and method | |
| Pukrongta et al. | Cattle Collars with a Low-Cost LPWAN-Based System for Cattle Estrous Monitoring |
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 |