JP7450561B2 - Pig breeding support device, pig breeding support method, and pig breeding support program - Google Patents

Description

The present invention relates to a pig breeding support device, a pig breeding support method, and a pig breeding support program.

When pigs are raised in groups in pens, the reality is that once the pigs are placed in the pen, the pens and the pigs are not washed very often until they grow up and are removed from the pen. Therefore, a screen-like floor surface is sometimes used to prevent the pen from getting dirty as much as possible. When a screen-like floor surface is used, hard stool with an appropriate amount of moisture excreted by healthy pigs falls through the gaps, and even if some of it remains on the screen, it quickly dries into small pieces. , since it still falls through the gap, the hygiene of the floor surface is maintained to a certain extent. However, when sick pigs in poor physical condition pass loose stools or diarrhea containing a large amount of water, it adheres to the cage and stains the floor surface. On such a dirty floor surface, the body surface of even healthy pigs becomes soiled due to movements such as lying down or lying down, or contact with other pigs in the pen.

Until now, zookeepers had to regularly observe the pens to determine if they were dirty. In addition, diseased pigs have been detected by attaching various sensors to pigs and monitoring the sensor output (for example, see Patent Document 1).

Japanese Patent Application Publication No. 2017-201930

If the number of dirty pigs with dirty body surfaces increases in the pen, it is assumed that there are sick pigs in the pen. If such a situation is observed, it is desirable to take measures such as adjusting feed, administering medicine, and removing sick pigs. Additionally, if the number of soiled pigs in the pen decreases, it can be confirmed that the measures taken were effective. However, it required a great deal of effort for keepers to continuously observe a large number of pigs.

The present invention was made to solve such problems, and it is possible to increase or decrease the number of pigs with dirty body surfaces among the pigs reared in groups in pens without requiring excessive cost and labor. The purpose of the present invention is to provide a pig rearing support device, etc. that can notify the pig keepers without any warnings.

A pig rearing support device according to a first aspect of the present invention includes an acquisition unit that acquires image data of an image captured by a camera installed toward a pen in which pigs are reared in groups, and an image acquired by the acquisition unit. The apparatus includes an evaluation section that evaluates stains appearing on the body surface of a pig from data images, and an output section that outputs change information indicating changes over time in the evaluation results by the evaluation section.

A pig rearing support method according to a second aspect of the present invention includes an acquisition step of acquiring image data of an image captured by a camera installed toward a pen in which pigs are reared in groups; The present invention includes an evaluation step for evaluating stains appearing on a pig's body surface from an image of image data, and an output step for outputting change information indicating a change over time in the evaluation result from the evaluation step.

A pig rearing support program according to a third aspect of the present invention includes an acquisition step of acquiring image data of an image captured by a camera installed toward a pen in which pigs are reared in groups; A computer is caused to execute an evaluation step of evaluating stains appearing on a pig's body surface from an image of the image data, and an output step of outputting change information indicating a change over time in the evaluation result of the evaluation step.

The present invention provides a pig rearing support device, etc. that can notify a breeder of the increase or decrease in the number of soiled pigs with dirty body surfaces among the pigs reared in a group in a pen without requiring excessive cost or labor. can do.

FIG. 1 is a diagram showing an overall picture of a pig farming environment in which the pig breeding support device according to the present embodiment is adopted. It is a diagram showing the hardware configuration of a pig breeding support device and peripheral devices. It is a figure explaining the evaluation target part of each pig. FIG. 2 is a diagram illustrating the concept of a method for determining whether a pig is contaminated or not. It is a figure explaining the detection procedure of the soiled pig using a learning model. It is a figure which shows an example of an evaluation log. It is a figure which shows the example of a display of the caretaker terminal which received change information. It is a figure which shows another example of a display of the zookeeper's terminal which received change information. FIG. 2 is a flow diagram illustrating a processing procedure of a calculation unit.

The present invention will be described below through embodiments of the invention, but the claimed invention is not limited to the following embodiments. Furthermore, not all of the configurations described in the embodiments are essential as means for solving the problem.

FIG. 1 is a diagram showing an overall picture of a pig farming environment in which a pig breeding support device according to the present embodiment is adopted. The pig farm includes a plurality of pens 301 separated by walls and fences. A plurality of pigs 302 (for example, about 10 pigs) are housed in each pen 301 and raised in groups. Note that the number of pigs 302 raised in each pen 301 can be adjusted depending on the breed of pigs 302, the breeding environment, etc. The floor surface of the pen 301 has a screen-like structure, and is devised so that excrement, feed scraps, etc. discharged by the pigs 302 fall through the gaps 301a, thereby maintaining the hygiene of the floor surface. The excrement, urine, and the like that have fallen from the gap 301a are mainly collected during cleaning performed when the entire housed pig 302 is replaced.

A camera unit 210 for observing the housed pigs 302 is installed in each pen 301. The camera unit 210 is installed, for example, by being suspended from the ceiling, facing the pen 301 to be observed so as to be able to take an image of the entire pen 301 as viewed from above. The camera unit 210 converts the captured image into image data and transmits it to the server 100 via the network 200. Specifically, a wireless unit 230 installed within the facility is connected to the network 200, and the camera unit 210 can transmit image data to the server 100 by establishing wireless communication with the wireless unit 230. can. Of course, communication may be established through a wired connection. Note that the network 200 that connects the camera unit 210 and the server 100 may use the Internet or an intranet, or if the management facility where the server 100 is installed is located in a pig farm, short-range wireless communication may be used. May be adopted.

A keeper who takes care of pig 302 may have a keeper terminal 220. The zookeeper terminal 220 is, for example, a tablet terminal or a smartphone, and can exchange various information with the server 100 via the wireless unit 230 and the network 200. For example, the breeder can input the breeding record into the breeder terminal 220 and transfer it to the server 100, or can call up the breeding record stored in the server 100. Further, when the breeder terminal 220 receives change information to be described later, it displays information regarding a soiled pig whose body surface is dirty.

A server 100 as a pig breeding support device is installed in the management facility. Server 100 is connected to network 200 . The server 100 sequentially acquires image data from the camera unit 210 installed for each pen 301, and evaluates stains appearing on the body surface of the pig 302 based on the image data. The server 100 can display the results on the display monitor 150 in response to requests from managers and zookeepers. The display monitor 150 is, for example, a monitor equipped with a liquid crystal panel. Further, the server 100 is connected to an input device 160 that receives operations from a manager or a caretaker. The input device 160 includes a keyboard, a mouse, a touch panel superimposed on the display surface of the display monitor 150, and the like.

Now, when pigs 302 are raised in a group in a pen 301, the reality is that once the pigs 302 are put into the pen 301, the pen 301 and the pigs 302 are not cleaned much until they grow up and are removed from the pen 301. be. Therefore, as described above, the floor surface of the pen 301 is shaped like a screen to prevent the pen 301 from getting dirty as much as possible. When a screen-like floor surface is adopted, hard stool with an appropriate amount of moisture discharged from a healthy pig 302 falls through the gap 301a, and even if some of it remains on the screen, it dries quickly. Since the particles break into small pieces and fall through the gap 301a, the hygiene of the floor surface is maintained to a certain extent. However, when the unwell pig 302 excretes loose stool or diarrhea containing a large amount of water, it adheres to the cage and stains the floor surface. On such a dirty floor surface, the body surface of even a healthy pig 302 becomes soiled due to movements such as lying down or lying down, or contact with other pigs in the pen.

In other words, the increase or decrease in the number of soiled pigs with dirty body surfaces is an indicator of changes in the presence of diseased pigs. If the number of soiled pigs increases in the pen, it is assumed that there are sick pigs in the pen, or that the number of diseased pigs in the pen is increasing. If such a situation is observed, it is recommended that the breeder take measures such as adjusting the feed, administering medicine, and removing the sick pig. On the other hand, if the number of dirty pigs in the pen decreases, it can be confirmed that the measures taken were effective. However, it was a labor-intensive task for keepers to continuously observe a large number of pigs. In particular, in relatively large-scale pig farming where multiple pens are installed in a pigsty, the burden is even greater. The pig rearing support device in this embodiment supports rearing work by detecting soiled pigs in the pen 301 and notifying the breeder of the increase or decrease in the number of soiled pigs.

Keepers can save a lot of effort by simply being informed of changes in soiled pigs over time. For example, it is possible to reduce the frequency of observing the pen or the time spent gazing at it. Even inexperienced keepers can easily take action by searching for diseased pigs only in that pen, if they know in which pen there has been a sudden increase or decrease in the number of contaminated pigs.

FIG. 2 is a diagram showing the hardware configuration of the server 100 and peripheral devices as a pig breeding support device. The server 100 can be connected to the display monitor 150, input device 160, camera unit 210, and zookeeper terminal 220, as described above.

The server 100 mainly includes a calculation section 110, a storage section 120, and a communication unit 130. The calculation unit 110 is a processor (CPU: Central Processing Unit) that controls the server 100 and executes programs. The processor may be configured to cooperate with an arithmetic processing chip such as an ASIC (Application Specific Integrated Circuit) or a GPU (Graphics Processing Unit). The calculation unit 110 reads out the pig rearing support program stored in the storage unit 120 and executes various processes related to support for pig rearing.

The storage unit 120 is a nonvolatile storage medium, and is configured by, for example, an HDD (Hard Disk Drive). In addition to programs for controlling and processing the server 100, the storage unit 120 can store various parameter values, functions, display element data, lookup tables, etc. used for control and calculation. The storage unit 120 particularly stores an evaluation neural network 121 and an evaluation log 122. When the evaluation neural network 121 receives an image captured by the camera unit 210, it outputs the probability that each pig present in the image is a contaminated pig. The evaluation log 122 is a record file that mainly records the number of detected soiled pigs over time. Each of these will be specifically explained later. Note that the storage unit 120 may be configured with a plurality of pieces of hardware; for example, the storage medium that stores the program and the storage medium that stores the evaluation neural network 121 may be configured with separate hardware.

The communication unit 130 includes, for example, a LAN unit, and transmits an imaging control signal generated by the calculation unit 110 to the camera unit 210 via the network 200, and sends image data sent from the camera unit 210 to the calculation unit 110. or hand it over. It also relays data exchange executed between the zookeeper terminal 220 and the calculation unit 110. Note that the communication unit 130 can also relay data and control signals to and from other external devices. For example, it can also be used to import update data for the pig breeding support program and the evaluation neural network 121 from an external server.

The calculation unit 110 also plays a role as a functional calculation unit that executes various calculations according to the processing instructed by the pig breeding support program. The calculation unit 110 can function as an acquisition unit 111, an evaluation unit 112, and an output unit 113. The acquisition unit 111 mainly acquires image data of an image captured by the camera unit 210 and delivers it to the evaluation unit 112. The evaluation unit 112 mainly evaluates stains appearing on the pig's body surface from the image of the image data received from the acquisition unit 111. The output unit 113 outputs change information indicating a change in the evaluation result by the evaluation unit 112 over time.

FIG. 3 is a diagram illustrating the evaluation target parts of each pig. The pig rearing support device in this embodiment evaluates the soiling appearing on the body surface of the pig 302, and sorts the pigs to be evaluated into either soiled pigs or non-fouled pigs. When evaluating dirt that adheres to the body surface of the diseased pig or other pigs due to loose stools or diarrhea from the diseased pig, it is better to evaluate soil that adheres to the body surface of the diseased pig or other pigs, rather than areas on the body surface where dirt repeatedly adheres and falls off over a short period of time. , it is preferable to evaluate areas that are naturally difficult to stain and once soiled are difficult to remove. Therefore, in this embodiment, the target areas for evaluating staining are limited to the occiput, shoulders, back, and flanks, as shown in the figure. These areas do not get dirty much during the normal behavior of pigs under normal circumstances, so it can be said that these areas are areas where the dirt on the floor surface is more reflected. The regions of the occipital region, shoulders, back region, and flank region are each defined in advance with respect to the outline of the pig 302. Specifically, it can be determined relatively with ears, groin, tail, etc. as a reference. In this embodiment, these four surface areas are the target areas for evaluation, but the evaluation may be performed for at least one area or a combination of two or more areas.

In this embodiment, instead of directly evaluating the dirt on the floor surface, the dirt appearing on the body surface of the pig 302 being raised is evaluated. This is because the floor surface of the pen 301 has been trampled by the many pigs 302 housed there, and has been trampled by feed and water, causing it to become dirty and wiped frequently, indicating the presence of diseased pigs. This is because it is difficult to use as evidence. On the other hand, if soiling is continuously evaluated on parts of the body surface of the pig 302 that are not often soiled, changes in floor soiling caused by the stool of the sick pig can be appropriately detected. That is, once such dirt adheres to the evaluation target site of the pig 302, it often remains there until it dries, becomes small pieces, and scatters. This change corresponds well to the change in soiling caused by the feces of sick pigs adhering to the floor surface even if there is no trampling, that is, it has a strong correlation with the change in the presence of diseased pigs. Therefore, in the present embodiment, the stains appearing on the body surface of the reared pig 302 are continuously evaluated, and changes in the evaluation results over time are output as change information.

FIG. 4 is a diagram illustrating the concept of a method for determining whether a pig is contaminated or not. In the present embodiment, it is evaluated whether or not the pig 302 is a soiled pig based on the area of the region recognized as soiled relative to the area of the region to be evaluated on the body surface of the pig 302. However, since the pig 302 housed in the pen 301 can take various postures with respect to the orientation of the camera unit 210, it is not always possible to observe the entire region to be evaluated. For example, in an image of a pig 302 taken almost from the side as shown in FIG. , only the front side of the dorsal ridge can be observed.

Even in such a case, in this embodiment, if the observed target area 304 occupies a proportion equal to or higher than a preset threshold value with respect to the entire area surrounded by the outline of the pig 302, , subject to evaluation. The threshold ratio of the target area 304 to the entire area may be set for each target area 304. For example, since the shoulder region is originally narrower than the back region, it is preferable to set the ratio of the shoulder region to be smaller than the ratio of the back region for each threshold value to be evaluated. Then, when it is determined that the area is to be evaluated, the evaluation unit 112 calculates the ratio of the soiled area 303 to the area of the target area 304, and if the ratio is equal to or higher than the standard ratio for determining soiling, the evaluation unit 112 calculates the ratio of the soiled area 303 to the area of the target area 304. pigs are evaluated as soiled pigs, and those with less than the standard percentage are evaluated as non-fouling pigs.

FIG. 4B shows a pig 302 viewed from above, and the entire target region 304 set as the back can be observed. When the entire area can be observed in this way, the evaluation can be made by simply calculating the ratio of the soiled area 303 to the area of the target area 304 without considering the relationship between the outline of the pig 302 and the target area 304. Note that the actual body surface of the pig 302 is a three-dimensional curved surface, but when evaluating using images as in this embodiment, the area as a two-dimensional plane should be calculated based on the outline of each region. Bye. In addition, whether or not it is a soiled area is evaluated based on, for example, the average hue of the pig's body surface, which is preset, and the average hue of stains caused by loose stool, etc., which is also preset. Judgment is made based on the hue of a certain area.

The process of determining whether the pig 302 shown in the captured image is a contaminated pig can be realized by using the analytical method described above, but in this embodiment, the evaluation unit 112 uses the evaluation neural network 121. This series of processing is executed using FIG. 5 is a diagram illustrating a procedure for detecting soiled pigs using the evaluation neural network 121, which is a learning model.

The evaluation neural network 121 learns by giving a considerable number of teacher image data in which a rectangular area in which a pig is photographed is specified for an overhead image of a pen, and a correct answer as to whether each pig is a soiled pig or not is associated. This is a learning model created in advance using supervised learning. The teacher image data is created, for example, by the operator drawing a rectangle around each pig in the bird's-eye view image, determining whether each pig is a soiled pig, and giving the correct answer. The evaluation neural network 121, which is a learning model generated in this way, is incorporated into the server 100, which is a pig breeding support device, and is made available for use.

For example, if a pig farm is provided with eight pens 301 to be observed, image data is sequentially sent to the server 100 from eight camera units 210 that provide an bird's-eye view of each pen 301. The acquisition unit 111 receives the image data and passes it to the evaluation unit 112. The evaluation unit 112 sequentially inputs images of the image data received from the acquisition unit 111 to the evaluation neural network 121 read from the storage unit 120.

Each time an image is input, the evaluation neural network 121 defines an evaluation area by surrounding the pig in the image with a rectangle, and outputs the probability that the pig is a soiled pig for each determined evaluation area. Specifically, for example, if seven pigs 302 are reflected in the image as shown in the figure, seven evaluation areas pig01 to pig07 are defined, and the pigs 302 appearing in each evaluation area are soiled pigs. The probabilities P ₀₁ to P ₀₇ are sequentially output.

If the output probability is greater than or equal to a preset threshold, the evaluation unit 112 determines that the pig 302 appearing in the evaluation area is a soiled pig; if the output probability is less than the threshold, the evaluation unit 112 determines that the pig 302 appearing in the evaluation area is a soiled pig. 302 is determined to be a non-contaminated pig. Note that the threshold value is set to a value that is adjusted in advance through trial and error, depending on the output accuracy of the learning model, the degree of soiling recognized as a soiled pig, and the like. The evaluation unit 112 totals the number of contaminated pigs and the number of non-contaminated pigs determined in this way, and records it in the evaluation log 122. The evaluation unit 112 repeats this series of processes for each image received from the acquisition unit 111.

The evaluation neural network 121 consists of two neural networks: a first neural network that finds pigs from an overhead image and determines an evaluation area, and a second neural network that outputs the probability that a pig appearing in the evaluation area is a contaminated pig. may be configured. In this case, the first neural network is trained using teacher image data in which correct evaluation areas are given for images showing a plurality of pigs. In addition, the second neural network uses teacher image data in which a correct answer is given to an image of a single pig that should be evaluated as a soiled pig, and an incorrect answer is given to an image that should be evaluated as a non-stained pig. learned by.

FIG. 6 is a diagram schematically showing an example of the evaluation log 122. For example, one evaluation log 122 is created for each pen 301 as shown in the figure. In this embodiment, the camera unit 210 captures images for each pen 301 several times at a set time once a day, and performs the processing described using FIG. 5 on each captured image. is assumed.

The evaluation log 122 for each pen 301 includes a pen number, the number of animals accommodated, and a log list. The pen number indicates which pen the log belongs to, such as "first pen". The number of pigs accommodated indicates the number of pigs 302 housed in the pen. The log list includes the observation date, the number of contaminated pigs, the number of non-contaminated pigs, and the change from the previous day. In the observation date column, a character string representing the passage of time is written in accordance with the observation cycle, such as "1st day", for example. In the column of number of soiled pigs, the total number of pigs 302 evaluated as soiled pigs on each observation date is described. In the column of the number of non-fouling pigs, the total number of pigs 302 evaluated as non-fouling pigs on each observation date is described. The column of change from the previous day describes the increase or decrease in the number of soiled pigs on the day of observation compared to the previous day.

Note that as long as the evaluation log 122 records at least the number of soiled pigs on each observation day, it is possible to grasp the increase or decrease in the number of soiled pigs over time, so it is not necessary that the number of non-fouled pigs or the comparison from the previous day is written. On the other hand, if the number of non-contaminated pigs on each observation day is recorded, caretakers, etc. can check how many pigs were overlapped in the evaluation by referring to the total with the number of soiled pigs and the number of pigs housed. Alternatively, it is possible to estimate how many animals remained unevaluated.

Further, the items to be written in the evaluation log 122 are not limited to these, and additional information may be further written. For example, when a breeder removes a pig 302 suspected of having a disease from the pen 301, an item indicating the number of pigs removed may be added to the log list as a record regarding the removed pigs. .

Note that the number of times and cycle at which the camera unit 210 executes imaging, or the number and cycle at which the acquisition unit 111 acquires image data from the camera unit 210, depends on the evaluation accuracy by the evaluation unit 112 and the number of times the pig 302 accommodated in the pen 301 Adjustments will be made as appropriate depending on the number and other circumstances. In this embodiment, individual identification of the pigs 302 housed in the pen 301 is not performed, so if images are repeatedly acquired and evaluated during one observation period, the same individual may be counted twice. However, such overlap can be ignored to the extent that the overall trend of increase or decrease in contaminated pigs can be grasped. In other words, if adjustments are made taking into consideration the number of pigs 302 housed in the pen 301, the angle of view of the camera unit 210, the time of day when the pigs 302 move around inside the pen 301, etc., there will be some overlap or adverse evaluation. Even if there are pigs 302 that are not contaminated, the overall trend of increase or decrease in the number of soiled pigs can be grasped.

For example, when the calculation unit 110 receives a request signal from the caretaker terminal 220 or a display instruction from the input device 160, the output unit 113 outputs change information indicating a change in the evaluation result by the evaluation unit 112 over time. The information is output to the employee terminal 220 and the display monitor 150. FIG. 7 is a diagram showing a display example of the zookeeper terminal 220 that has received the change information.

As described above, when the output unit 113 outputs the change information, the zookeeper terminal 220 receives the change information and displays the contents on the display panel. The change information output by the output unit 113 is the evaluation log 122 in this embodiment. As described above, the evaluation log 122 includes a record of the number of soiled pigs for each observation day, and therefore is information indicating changes in the evaluation results by the evaluation unit 112 over time. The change information is not limited to the format of the evaluation log 122, but may be in any format as long as it directly or indirectly includes a change in the evaluation result by the evaluation unit 112 over time.

The zookeeper terminal 220 processes the evaluation result by the evaluation unit 112 in a manner that allows the change over time to be read, and displays the change information on the display panel. Specifically, the title (in the example in the figure, ``Number of contaminated pigs''), the pen number (in the example in the figure, ``1st pen''), and the number of pigs housed in that pen (in the example in the figure, ``8''). ), an indoor map (in the illustrated example, the location of the "first pen" is shaded so that the position of the "first pen" is recognized) regarding the pen arrangement of the pig farm is displayed.

It also displays the numerical values in the log list in a graph. Here, the number of contaminated and non-contaminated pigs on each observation day is shown in a bar graph. Keepers can check trends in the number of contaminated pigs using graphs like this, so they can see whether the number of contaminated pigs is increasing or decreasing, and how steeply the change is occurring. It is possible to understand each pen. Additionally, the number of uncontaminated pigs and the number of pigs housed are also displayed, allowing zookeepers to estimate how many pigs have been evaluated twice, or how many pigs have remained unevaluated. . Note that when the zookeeper performs, for example, a flick operation on the zookeeper terminal 220, change information for the next pen (the second pen in the illustrated example) is displayed in the same manner.

FIG. 8 is a diagram showing another display example of the zookeeper terminal 220 that has received the change information. The display example shown in FIG. 8 is similar to the display example shown in FIG. Includes indoor map for placement. Furthermore, as a numerical value in the log list, the change in the number of soiled pigs on each observation day from the previous day is shown by a line graph. Using graphs like this, keepers can check trends in the number of contaminated pigs, so they can see whether the number of contaminated pigs is increasing or decreasing, and how steeply the change is occurring. can be determined for each pen. When the zookeeper terminal 220 receives a flicking motion or the like, it displays the next pen change information in the same manner.

Next, the processing procedure of the pig rearing support method using the server 100 will be explained. FIG. 9 is a flow diagram illustrating the processing procedure of the calculation unit 110. The flow starts when an instruction to start observation is received.

In step S101, the acquisition unit 111 checks the current time and determines whether the current time is a preset observation timing. If it is determined that it is observation timing, the process proceeds to step S102, and if it is determined that it is not observation timing, the process proceeds to step S105. The observation timing is set in conjunction with the number of times or period at which the camera unit 210 performs imaging, or the number or period at which the acquisition unit 111 acquires image data from the camera unit 210.

Proceeding to step S102, the acquisition unit 111 acquires image data of img _n from the camera unit 210 of the n-th pen via the communication unit 130. For example, if n=1, image data of img ₁ is acquired from the camera unit 210 directed toward the first pen. The acquisition unit 111 delivers the acquired image data to the evaluation unit 112. n is incremented each time step S102 is executed. The acquisition unit 111 delivers the acquired image data to the evaluation unit 112.

In step S103, the evaluation unit 112 inputs the image img _n of the received image data to the evaluation neural network 121, as described using FIG. Then, from the output, the probability that each pig appearing in the image img _n is a soiled pig is obtained, and each pig to be evaluated is classified as either a soiled pig or a non-stained pig using a preset threshold value. Then, the process proceeds to step S104, and the result is recorded in the evaluation log 122 of the n-th pen. When the recording to the evaluation log 122 is completed, the process advances to step S105.

After proceeding to step S105, the output unit 113 checks whether a request signal requesting output of change information is received from the zookeeper terminal 220 or the like. If the request signal has been received, the process advances to step S106; if the request signal has not been received, the process advances to step S107. If the process proceeds to step S106, the output unit 113 reads out the evaluation log 122 from the storage unit 120 as change information, and outputs it to the caretaker terminal 220 or the like that issued the request signal. After that, the process advances to step S107.

After proceeding to step S107, the arithmetic unit 110 checks whether an instruction to end has been received. The termination instruction is generated when a termination operation is received from the administrator via the input device 160 or when conditions preset in the program are met. If it is determined that the termination instruction has not been received, the process returns to step S101. When it is determined that an instruction to end has been received, the series of processing is ended.

In the embodiment described above, one camera unit 210 is installed for each pen 301, but a camera unit that provides a bird's-eye view of a plurality of pens 301 may be installed. In that case, the acquisition unit 111 may divide the image acquired from the camera unit along the boundaries of each pen 301 and sequentially deliver each divided image to the evaluation unit 112. Conversely, a plurality of camera units 210 may be installed for one pen 301. For example, by installing a camera unit 210 that images the pig 302 inside the pen 301 from the side, false detection of a soiled pig can be reduced. Furthermore, if a plurality of camera units 210 are installed for one pen 301, there will be more opportunities to capture images of pigs moving around within the pen 301, so that soiled pigs can be detected with higher accuracy.

Furthermore, in the present embodiment, the evaluation unit 112 performs an evaluation to classify whether the pig 302 to be evaluated is a soiled pig or a non-stained pig, but the evaluation unit 112 evaluates the soiling that appears on the pig's body surface from the image. The evaluation method is not limited to this. Instead of counting the number of pigs by classifying them into soiled pigs and non-fouled pigs, for example, the ratio of the total sum of soiled areas to the total sum of target areas 304 may be used as an index for evaluating soiling.

Furthermore, in this embodiment, a case is assumed in which a plurality of pens 301 are provided in one pigsty, but pens 301 provided across a plurality of pigsty may be observed. Further, the output destination of the change information is not limited to the zookeeper terminal 220. For example, one display panel connected to the server 100 via the wireless unit 230 and the network 200 may be installed adjacent to each pen 301, and change information may be output to the display panel. In this case, the output unit 113 may output change information only to the evaluation log 122 of the fifth pen to the fifth pen, for example. Furthermore, the output unit 113 may output the change information to the display monitor 150 connected to the server 100. In this case, it is preferable to convert the image into a display image displayed on the display monitor 150 and output it. The output unit 113 may output the change information not only when an output request is received, but also at a preset time or cycle, for example. Further, change information may be output every time the evaluation log 122 is updated.

Further, in the present embodiment described above, a case has been described in which the server 100 functions as a pig breeding support device, but the hardware configuration is not limited to this. If the mobile terminal described as the breeder terminal 220 performs the same processing as the server 100, the mobile terminal can function as a pig breeding support device. Furthermore, for example, if a part of the processing of the server 100 is configured to be carried out by the breeder terminal 220, a system in which the server 100 and the breeder terminal 220 cooperate can become a pig breeding support device.

DESCRIPTION OF SYMBOLS 100... Server, 110... Arithmetic unit, 111... Acquisition unit, 112... Evaluation unit, 113... Output unit, 120... Storage unit, 121... Neural network for evaluation, 122... Evaluation log, 130... Communication unit, 150... Display monitor , 160...Input device, 200...Network, 210...Camera unit, 220...Keeper terminal, 230...Wireless unit, 301...Pen, 302...Pig, 303...Dirty area, 304...Target area

Claims (7)

an acquisition unit that acquires image data of an image captured by a camera installed toward a pen in which pigs are reared in groups;
an evaluation unit that evaluates stains appearing on the body surface of the pig from the image of the image data acquired by the acquisition unit;
A pig breeding support device comprising: an output section that outputs change information indicating a change over time in the evaluation result by the evaluation section. The pig breeding support device according to claim 1, wherein the evaluation unit evaluates the staining by limiting the target to at least one region of the pig's body surface, including the shoulder, flank, back, and occiput. The evaluation unit evaluates whether the pig is a soiled pig based on the area of the region recognized as soiled with respect to the area of the body surface to be evaluated,
The pig breeding support device according to claim 1 or 2, wherein the output unit totals and outputs the number of soiled pigs evaluated by the evaluation unit at each preset time. The pig according to any one of claims 1 to 3, wherein the evaluation unit evaluates soiling of the pig to be evaluated using a learning model learned using a teacher image showing a pig recognized as being soiled. Breeding support equipment. The pig breeding support device according to any one of claims 1 to 4, wherein the output unit adds a record regarding the pigs that have left the pen to the change information and outputs the resultant information. an acquisition step of acquiring image data of an image taken by a camera installed toward a pen in which pigs are reared in groups;
an evaluation step of evaluating stains appearing on the body surface of the pig from the image of the image data acquired in the acquisition step;
A pig breeding support method comprising: an outputting step of outputting change information indicating a change over time in the evaluation result obtained in the evaluation step. an acquisition step of acquiring image data of an image taken by a camera installed toward a pen in which pigs are reared in groups;
an evaluation step of evaluating stains appearing on the body surface of the pig from the image of the image data acquired in the acquisition step;
A pig breeding support program that causes a computer to execute an output step of outputting change information indicating a change over time in the evaluation result obtained in the evaluation step.

Images