Livestock management device, livestock management method and program

Abstract

To provide a technique that can prevent the occurrence of the stand-up difficult state of livestock.SOLUTION: A livestock management device according to one aspect of the present disclosure includes: a first determination unit that determines whether or not the state of livestock is in the laid-down state on the basis of a first photographed image obtained by imaging the livestock; a second determination unit that determines whether or not the elapsed time representing the duration of the laid-down state is equal to or greater than a prescribed first threshold when the state of the livestock is determined to be the laid-down state; and an output unit that outputs a predetermined stimulus to shift the state of the livestock from the laid-down state to the lying-on state to the livestock when the elapsed time is determined to be equal to or greater than the first threshold.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a livestock management device, a livestock management method, and a program. [Background technology]

[0002] It is known that fattening cattle are prone to developing standing difficulty (or standing inability) in the later stages of fattening. If standing difficulty is left untreated, the cattle may die from suffocation or other causes, so fattening cattle farms patrol the barns to check whether the cattle are experiencing standing difficulty. In response to this, there is known a technology that estimates whether the cattle are experiencing standing difficulty from the sensor values ​​of an acceleration sensor attached to the cattle, and notifies the user when standing difficulty is estimated (for example, Patent Document 1, etc.). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2019 / 039118 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the conventional technology estimates when a cow has difficulty standing up, and is unable to prevent the cow from becoming unable to stand up.

[0005] The present disclosure has been made in consideration of the above points, and provides a technology that can prevent livestock from having difficulty standing up. [Means for solving the problem]

[0006] A livestock management device according to one embodiment of the present disclosure includes a first judgment unit that judges whether the state of the livestock is lying down based on a first captured image of the livestock, a second judgment unit that, if the state of the livestock is judged to be lying down, judges whether an elapsed time representing the time the livestock has been lying down is equal to or greater than a predetermined first threshold, and an output unit that, if the elapsed time is judged to be equal to or greater than the first threshold, outputs a predetermined stimulus to the livestock to transition the state of the livestock from the lying down state to a prone state. [Effects of the Invention]

[0007] To provide a technology that can prevent livestock from having difficulty standing up. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram showing an example of the overall configuration of a livestock management system according to an embodiment of the present invention; [Figure 2] FIG. 1 is a diagram showing an example of a cow lying down and lying down. [Figure 3] FIG. 2 is a diagram illustrating an example of the functional configuration of a livestock management processing unit according to the present embodiment. [Figure 4] 10 is a flowchart showing an example of livestock management processing according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] An embodiment of the present invention will be described below. In the following embodiment, a cow is assumed as an example of livestock, and a livestock management system 1 will be described that prevents cows from having difficulty standing up and can notify a user when the cow has difficulty standing up or when there is a risk of the cow having difficulty standing up. Here, the difficulty in standing up refers to, for example, a state in which livestock such as a cow is struggling with its legs pointed upward. If the cow continues to have difficulty standing up, gas can build up in the first stomach of the cow, causing it to suffocate and die.

[0010] However, cows are an example of livestock, and livestock to which the following embodiments can be applied are not limited to cows.

[0011] <Overall configuration example of livestock management system 1> An example of the overall configuration of a livestock management system 1 according to this embodiment is shown in Fig. 1. As shown in Fig. 1, the livestock management system 1 according to this embodiment includes a livestock management apparatus 10, one or more imaging devices 20, one or more speakers 30, one or more GW devices 40, and one or more user terminals 50. The livestock management apparatus 10, each GW device 40, and each user terminal 50 are communicatively connected via a communication network 60 including, for example, the Internet.

[0012] The livestock management device 10 prevents the cow 70 from having difficulty standing up, based on photographed image data of the cow 70 captured by the photographing device 20. That is, the livestock management device 10 determines from the photographed image data whether the cow 70 has been in a state known as lying down for a certain period of time, and if the cow 70 has been in a lying down state for a certain period of time, instructs the speaker 30 to output a predetermined sound (or, without being limited to sound, it may be any sound such as a mechanical sound). This is because if the cow continues to lie down for a long period of time and its legs begin to point towards the ceiling, it may fall into a state where it has difficulty standing up. Furthermore, it has been confirmed by the inventors' findings that outputting sound from the speaker 30 will cause the cow to transition from a lying down state to a state known as prone, and that there is no possibility of the cow falling into a state where it has difficulty standing up when in a prone state.

[0013] Furthermore, when a state of difficulty in standing up occurs or there is a risk of a state of difficulty in standing up occurring, the livestock management device 10 notifies the user terminal 50 that a state of difficulty in standing up has occurred or that there is a risk of such a state occurring. The livestock management device 10 is realized, for example, by a PC (personal computer), a general-purpose server, etc.

[0014] The photographing device 20 photographs the cattle 70 being fattened in the cowshed 80, and transmits the photographed image data generated by the photographing to the livestock management apparatus 10 via the GW device 40. The photographing device 20 can be installed anywhere in the cowshed 80 as long as it is possible to photograph the cattle 70, but it is preferable to install it in a location where it can photograph the cattle 70 from the side (for example, a pillar or wall of the cowshed 80). However, this installation location is one example and is not limited to this. For example, the photographing device 20 may be installed in a location where it can photograph the cattle 70 from above, such as the ceiling of the cowshed 80.

[0015] The captured image data generated by the image capture device 20 may be data representing a still image or data representing a moving image. When the captured image data is data representing a still image, the captured image at time index t representing the capture time is denoted by x t Then, the captured image data is x t On the other hand, if the captured image data is data representing a video, the captured image data is expressed as {x t |t∈[t s ,t e ]} where t s and t e are the time indexes indicating the start and end times of shooting when the captured image data is generated. Note that if the captured image data is data representing a video, x t is also called a frame image or simply a frame. Hereinafter, the time index will also be simply called "time."

[0016] The speaker 30 outputs sound in accordance with a sound output instruction received from the livestock management apparatus 10 via the GW device 40. The speaker 30 can be installed anywhere as long as it is possible to output sound to the cows 70 in the cowshed 80. As an example, the speaker 30 is installed on a pillar, wall, ceiling, etc. of the cowshed 80.

[0017] The GW device 40 receives photographed image data from the photographing device 20 and transmits the photographed image data to the livestock management device 10. The GW device 40 also receives audio output instructions from the livestock management device 10 and transmits the audio output instructions to the speaker 30. The GW device 40 is realized by, for example, a communication device called a gateway device or an IoT (Internet of Things) gateway.

[0018] The user terminal 50 is a terminal of any kind used by a user (e.g., a ranch operator) who manages the cows 70. The user terminal 50 receives various notifications (notifications that a cow has difficulty standing up or that there is a risk of this happening) from the livestock management device 10, and presents the notification contents to the user. The user terminal 50 is realized, for example, by a PC, a smartphone, a tablet terminal, a wearable device, etc.

[0019] The livestock management apparatus 10 according to this embodiment includes a livestock management processing section 100 and a storage section 200. The livestock management processing section 100 is realized, for example, by a processor such as a CPU (Central Processing Unit) executing one or more programs installed in the livestock management apparatus 10. The storage section 200 is realized, for example, by various storage devices such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), or a flash memory.

[0020] The livestock management processing unit 100 stores the photographed image data received from the photographing device 20 in the memory unit 200 via the GW device 40. The livestock management processing unit 100 also executes the livestock management process described below using the photographed image data stored in the memory unit 200. The livestock management process executes processes to prevent the occurrence of a state in which the cow 70 has difficulty standing up and to notify the cow that a state in which the cow has difficulty standing up has occurred or that there is a risk of this occurring. A detailed example of the functional configuration of the livestock management processing unit 100 will be described later.

[0021] The storage unit 200 stores the captured image data. In addition to the captured image data, the storage unit 200 also stores various data necessary for executing the livestock management process (for example, elapsed time, the number of audio outputs, and thresholds used to determine the state of the cows 70, which will be described later).

[0022] The overall configuration of the livestock management system 1 shown in Figure 1 is an example and is not limited to this. For example, Figure 1 shows one image capture device 20, one speaker 30, one GW device 40, and one user terminal 50, but there may be multiple image capture devices 20, multiple speakers 30, multiple GW devices 40, and multiple user terminals 50. Furthermore, for example, if the image capture device 20 and the speaker 30 can communicate with the livestock management apparatus 10 without going through the GW device 40, the GW device 40 may not be present. Furthermore, for example, the image capture device 20 and the speaker 30 may be configured as a single unit.

[0023] <Lying down and lying down> An example of the lying and prone states of a cow 70 is shown in Figure 2. Both the lying and prone states refer to states in which the cow 70 is lying down, but the lying state refers to a state in which the cow 70 is lying on its side with its limbs stretched out (Figure 2(a)), and the prone state refers to a state in which the cow 70 has its knees lowered and its chest raised (Figure 2(b)).

[0024] If the cow 70 remains lying down for a long time and its legs begin to point toward the ceiling, it may become difficult for the cow 70 to stand up. Therefore, if the cow 70 remains lying down for a certain period of time, there is a risk that the cow 70 will have difficulty standing up. On the other hand, if the cow 70 remains lying down, there is no risk that the cow 70 will have difficulty standing up. The inventors have confirmed that it is possible to transition the cow 70 from a lying down state to a prone state by outputting a sound (for example, a sound such as "Heyyy") from the speaker 30. Therefore, if the cow 70 remains lying down for a certain period of time, it is possible to prevent the cow 70 from having difficulty standing up by outputting a sound from the speaker 30.

[0025] <Example of functional configuration of livestock management processing unit 100> An example of the functional configuration of the livestock management processing unit 100 according to this embodiment is shown in Fig. 3. As shown in Fig. 3, the livestock management processing unit 100 according to this embodiment includes an acquisition unit 101, a state determination unit 102, an elapsed time determination unit 103, an update unit 104, an output instruction unit 105, a count determination unit 106, and a notification unit 107.

[0026] The acquisition unit 101 acquires the captured image x at the latest time t. t from the storage unit 200. However, the acquisition unit 101 acquires, for example, the captured images x t ,x t-1 ,···,x t-N may be acquired from the storage unit 200. Here, N is a hyperparameter indicating how many times in the past captured images are to be acquired, and is set to a predetermined integer of 1 or greater.

[0027] Hereinafter, the captured image acquired by the acquisition unit 101 is referred to as x t' Let (t′∈T), where T is a time set of captured images acquired by the acquisition unit 101, and where t is the latest time, T may be T={t} or T={t, t−1, . . . , tN}.

[0028] The state determination unit 102 determines whether the captured image x acquired by the acquisition unit 101 is t' (t′∈T) to determine the state of the cow 70. That is, the state determination unit 102 determines the state of the cow 70 by using the captured image x t' Using (t'∈T), it is determined whether the state of the cow 70 is lying down, lying prone, or having difficulty standing up.

[0029] For example, the state determination unit 102 determines whether the captured image x t' (t'∈T) is input and the captured image x t' The state determination unit 102 determines whether the cow 70 is lying down or not by using a first classifier that classifies the state of the cow 70 in the image into a lying down state or other states. t' (t'∈T) is input and the captured image xt' Whether the cow 70 is lying down or not is determined by a second classifier that classifies the state of the cow 70 in the photographed image x t' (t'∈T) is input and the captured image x t' A third classifier classifies the state of the cow 70 inside as either a state of difficulty standing up or a state other than that, and determines whether the cow 70 is in a state of difficulty standing up.

[0030] The first classifier is created by an existing machine learning method, for example, using as training data data that associates photographed images of a cow with labels that indicate whether the cow is in a lying state or in another state. Similarly, the second classifier is created by an existing machine learning method, for example, using as training data data that associates photographed images of a cow with labels that indicate whether the cow is in a lying state or in another state. Similarly, the third classifier is created by an existing machine learning method, for example, using as training data data that associates photographed images of a cow with labels that indicate whether the cow is in a state of difficulty standing up or in another state.

[0031] However, using the first classifier, the second classifier, and the third classifier is just one example, and other methods may be used to determine the state of the cow 70. For example, the state determination unit 102 may use an image processing technique such as pattern matching to determine whether the state of the cow 70 is lying down, lying prone, or having difficulty standing up.

[0032] The elapsed time determination unit 103 determines whether the time during which the state determination unit 102 has continuously determined that the state of the cow 70 is in a lying state (hereinafter referred to as the "elapsed time in a lying state" or simply "elapsed time") is equal to or greater than a predetermined threshold value.

[0033] The update unit 104 updates the time elapsed since the lying state, and updates the "number of audio outputs" which indicates the number of times that the output instruction unit 105 has transmitted an audio output instruction.

[0034] When the elapsed time determination unit 103 determines that the elapsed time is equal to or greater than a predetermined threshold, the output instruction unit 105 transmits a voice output instruction to the speaker 30 via the GW device 40.

[0035] The number determining unit 106 determines whether the number of times the sound is output is equal to or greater than a predetermined threshold value.

[0036] When the state determination unit 102 determines that the cow 70 is in a state of difficulty standing up, the notification unit 107 notifies the user terminal 50 of an alert (hereinafter also referred to as a "difficulty standing up alert") indicating that the cow 70 has developed a state of difficulty standing up. In addition, when the number of times determination unit 106 determines that the number of times audio is output is equal to or greater than a predetermined threshold, the notification unit 107 notifies the user terminal 50 of an alert (hereinafter also referred to as a "long-term lying down alert") indicating that there is a risk of the cow 70 developing a state of difficulty standing up.

[0037] <Livestock Management> The livestock management process according to this embodiment will be described below with reference to Fig. 4. In the following, it is assumed that the elapsed time of lying state and the number of times sound is output are both initialized to 0. Note that every time photographed image data is received from the photographing device 20 via the GW device 40, the photographed image data is saved in the storage unit 200.

[0038] The acquisition unit 101 of the livestock management processing unit 100 acquires the captured image x t' (t'∈T) is obtained from the storage unit 200 (step S101).

[0039] Next, the state determination unit 102 of the livestock management processing unit 100 determines whether the captured image x obtained in step S101 is t' Using (t'εT), it is determined whether the cow 70 is in a lying state (step S102).

[0040] If it is not determined that the animal is in a lying state in step S102 (NO in step S103), the update unit 104 of the livestock management processing unit 100 initializes the time elapsed in the lying state to 0 (step S104).

[0041] On the other hand, if it is determined in step S102 that the cow 70 is in a lying state (YES in step S103), the update unit 104 of the livestock management processing unit 100 updates the elapsed time of the lying state (step S105). In other words, if the difference between the current time and the time when the elapsed time of the lying state was last initialized or updated is Δt, the update unit 104 updates the elapsed time of the lying state by (elapsed time of the lying state) ← (elapsed time of the lying state) + Δt. In this way, the time that the cow 70 has been in a lying state is obtained as the "elapsed time of the lying state."

[0042] Next, the elapsed time determination unit 103 of the livestock management processing unit 100 determines whether the elapsed time of the lying state is equal to or greater than a predetermined threshold value th1 (step S106). The threshold value th1 can be set to any appropriate value. For example, the threshold value th1 may be set to about several tens of minutes, or about 2 to 3 hours.

[0043] If it is not determined in step S106 above that the elapsed time in the lying state is equal to or greater than the threshold value th1 (NO in step S106), the livestock management processing section 100 returns to step S101, whereby step S101 and subsequent steps are executed again.

[0044] On the other hand, if it is determined in step S106 above that the elapsed time in the lying state is equal to or greater than the threshold value th1 (YES in step S106), the output instruction unit 105 of the livestock management processing unit 100 sends a voice output instruction to the speaker 30 via the GW device 40 (step S107). This causes a predetermined voice (for example, a voice such as "Heyyy") to be output from the speaker 30, and it is expected that the state of the cow 70 will transition from the lying state to the prone state.

[0045] Next, the update unit 104 of the livestock management processing unit 100 updates the number of times of voice output (step S108). That is, the update unit 104 updates the number of times of voice output by setting the number of times of voice output←the number of times of voice output+1.

[0046] Next, the acquisition unit 101 of the livestock management processing unit 100 acquires the captured image x t' (t'∈T) is obtained from the storage unit 200 (step S109).

[0047] Next, the state determination unit 102 of the livestock management processing unit 100 determines whether the captured image x obtained in step S109 is t' Using (t'εT), it is determined whether the cow 70 is lying down or not (step S110).

[0048] If it is determined in step S110 that the cow is in a prone state (YES in step S111), the livestock management processing section 100 proceeds to step S104. As a result, the time elapsed in the lying state is initialized to 0. This means that the sound output from the speaker 30 has caused the cow 70 to transition from a lying state to a prone state, preventing the cow 70 from having difficulty standing up.

[0049] On the other hand, if it is not determined in step S110 that the animal is in a prone state (NO in step S111), the state determination unit 102 of the livestock management processing unit 100 determines whether the animal is in a prone state based on the photographed image x acquired in step S109. t' Using (t′∈T), it is determined whether the cow 70 is in a state where it is difficult to stand up (step S112).

[0050] If it is determined in step S112 above that the cow 70 is in a state of difficulty standing up (YES in step S113), the notification unit 107 of the livestock management processing unit 100 notifies the user terminal 50 of a difficulty standing up alert (step S114). This allows the user to know that the cow 70 has developed difficulty standing up.

[0051] On the other hand, if it is not determined in step S112 that the animal is in a state of difficulty standing up (NO in step S113), the count determination unit 106 of the livestock management processing unit 100 determines whether the number of times the sound is output is equal to or greater than a predetermined threshold value th2 (step S115). The threshold value th2 can be set to any appropriate value. For example, the threshold value th2 can be set to about several times.

[0052] If it is not determined in step S115 above that the number of times the sound is output is equal to or greater than the threshold value th2 (NO in step S115), the livestock management processing section 100 returns to step S107. As a result, the predetermined sound is output from the speaker 30 again, and it is expected that the state of the cow 70 will transition from the lying state to the prone state.

[0053] On the other hand, if it is determined in step S115 above that the number of voice outputs is equal to or greater than the threshold value th2 (YES in step S115), the notification unit 107 of the livestock management processing unit 100 notifies the user terminal 50 of a long-time lying down alert (step S116). This allows the user to know that the cow 70 has been lying down for a long time and there is a risk that it may have difficulty standing up.

[0054] <Modification> In the above embodiment, a predetermined sound is output from the speaker 30 to cause the state of the cow 70 to transition from a recumbent state to a prone state, but any stimulus other than sound or voice may be given to the cow 70 as long as it is a stimulus that can cause the state of the cow 70 to transition from a recumbent state to a prone state. For example, the cow 70 may be irradiated with light, subjected to a light vibration, or given a light electrical stimulus.

[0055] Also, various combinations of stimuli such as sound, voice, light, light vibration, light electrical stimulation, etc. may be given to the cow 70. In particular, the combination of stimuli may be changed according to an increase in the number of times a stimulus has been given (corresponding to the number of times a sound has been output in the above embodiment). For example, the number of combinations of stimuli may be increased according to an increase in the number of times a stimulus has been given, such that one type of stimulus (e.g., voice) is given when the number of times a stimulus has been given is 0, two types of stimuli (e.g., voice and light) are given when the number of times a stimulus has been given is 1, and three types of stimuli (e.g., voice, light, and vibration) are given when the number of times a stimulus has been given is 2.

[0056] Furthermore, the degree of stimulation (stimulus intensity) may be increased as the number of times the stimulation is applied increases, for example, the volume of a sound or voice may be increased as the number of times the stimulation is applied increases.

[0057] <Summary> As described above, the livestock management system 1 according to this embodiment is able to determine the lying state of livestock using photographed image data of the livestock, and if the lying state continues for a certain period of time, it is able to provide the livestock with a stimulus to transition from the lying state to a prone state, thereby making it possible to prevent the livestock from having difficulty standing up due to the state continuing for a long period of time.

[0058] Furthermore, the livestock management system 1 according to this embodiment can notify the user when a state of difficulty in standing up occurs or when there is a risk of this occurring. This makes it possible to prevent the death of livestock due to a state of difficulty in standing up even when stimulation is not enough to transition the livestock from a lying state to a prone state.

[0059] Unlike conventional techniques, the livestock management system 1 according to this embodiment does not require the attachment of sensor devices such as acceleration sensors to livestock. This eliminates the need to attach a sensor device to each livestock, reducing the user's workload, especially when there are a large number of livestock. Another advantage is that it prevents situations such as the sensor device falling off due to the livestock's movements, making it impossible to estimate the standing difficulty state.

[0060] The present invention is not limited to the above-described specifically disclosed embodiments, and various modifications, changes, and combinations with known technologies are possible without departing from the scope of the claims. [Explanation of symbols]

[0061] 1 Livestock Management System 10 Livestock management equipment 20 Imaging equipment 30 speakers 40 GW equipment 50 User Terminals 60 Communication Network 70 cow 80 Cowshed 100 Livestock Management Department 101 Acquisition Department 102 Status determination unit 103 Elapsed time determination unit 104 Update section 105 Output instruction section 106 Number of times judgment unit 107 Notification Department 200 Storage section

Claims

1. a first determination unit that determines whether or not the state of the livestock is lying down based on a first photographed image of the livestock; a second determination unit that, when the state of the livestock is determined to be the recumbent state, determines whether or not an elapsed time representing a duration of the recumbent state is equal to or greater than a predetermined first threshold value; an output unit that outputs one or more types of predetermined stimuli to the livestock to transition the state of the livestock from the lying state to a prone state when it is determined that the elapsed time is equal to or greater than the first threshold value; and The output unit A livestock management device that increases at least one of the number of types of stimuli to be output to the livestock and the intensity of the stimuli in response to an increase in the number of times the stimuli have been output to the livestock.

2. a third determination unit that determines whether the state of the livestock is the lying state based on a second photographed image of the livestock after the stimulus has been output to the livestock; a fourth determination unit that determines whether the state of the livestock is in a state where it is difficult to stand up based on the second photographed image when the state of the livestock is not determined to be the lying state; a notification unit that, when the state of the livestock is determined to be the standing difficulty state, notifies a user of a first alert indicating that the livestock is in the standing difficulty state; 2. The livestock management device according to claim 1, further comprising:

3. a fifth determination unit that determines whether or not the number of times the stimulus is output is equal to or greater than a predetermined second threshold value when the state of the livestock is not determined to be the standing difficulty state; The notification unit The livestock management device of claim 2, wherein if it is determined that the number of times the stimulus is output is greater than or equal to the second threshold, a second alert is sent to the user indicating that the livestock is at risk of becoming unable to stand up.

4. a first determination step of determining whether or not the livestock is in a lying state based on a first photographed image of the livestock; a second determination step of determining whether an elapsed time representing a duration of the lying state of the livestock is equal to or greater than a predetermined first threshold value when the state of the livestock is determined to be the lying state; an output step of outputting one or more types of predetermined stimuli to the livestock to transition the state of the livestock from the lying state to a prone state when it is determined that the elapsed time is equal to or greater than the first threshold value; The computer executes The output procedure includes: A livestock management method, wherein at least one of the number of types of stimuli to be output to the livestock and the intensity of the stimuli is increased in accordance with an increase in the number of times the stimuli have been output to the livestock.

5. a first determination step of determining whether or not the livestock is in a lying state based on a first photographed image of the livestock; a second determination step of determining whether an elapsed time representing a duration of the lying state of the livestock is equal to or greater than a predetermined first threshold value when the state of the livestock is determined to be the lying state; an output step of outputting one or more types of predetermined stimuli to the livestock to transition the state of the livestock from the lying state to a prone state when it is determined that the elapsed time is equal to or greater than the first threshold value; on the computer, The output procedure includes: A program that increases at least one of the number of types of stimuli to be output to the livestock and the intensity of the stimuli in response to an increase in the number of times the stimuli have been output to the livestock.

Images