Pig activity amount measuring method, non-temporary computer-readable storage medium in which pig activity amount measuring program is stored, and pig activity amount measuring system

Abstract

To detect a pig that has an abnormality in the growth such as damage to health out of a plurality of pigs grouped and raised in the same pigpen, and take measures such as separating the pig from other pigs easily.SOLUTION: A user gives an instruction to start pig activity amount measurement using an input user interface, performs a guiding operation to pigs reared in a pigpen, and continuously captures moving images of the pig moving due to the guiding operation using a camera unit. A pig identification unit of a computer device identifies a pig individual included in the moving image, imparts a temporary ID to the pig individual and tracks the pig individual. An activity amount calculation unit calculates an activity amount, which is a moving distance, a moving speed, or an acceleration within a predetermined time of the pig being tracked, A reference activity amount calculation unit calculates a reference activity amount. An abnormal pig extraction unit compares at least the activity amount of the pig being tracked with the reference activity amount, and extracts an abnormal pig with an abnormal activity amount.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present disclosure relates to a method for measuring the activity level of pigs, a non-transitory computer-readable storage medium storing a program for measuring the activity level of pigs, and a system for measuring the activity level of pigs.

Background Art

[0002] Conventionally, a system for monitoring animals kept in captivity using a camera or the like has been known. In the following prior art document (Patent Document 1), a system for identifying pigs using a two-dimensional code or the like readable attached to an ear tag for pigs, and using a deep neural network, by image analysis at the time of feeding, a bounding box is drawn around the animal detected from the video stream in the first frame, in the second frame, using the speed and acceleration of the animal and the size of the bounding box, determining which animal is most likely to be associated with each animal bounding box in the previous frame, generating the average speed of a subgroup of animals, that animals with severe illness should be culled, etc. are described.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the pig farming industry, a plurality of pigs are grouped and kept in the same pigsty. Among such a plurality of pigs, pigs with abnormalities in growth such as those causing harm to health need to be dealt with such as being isolated from other pigs.

[0005] In the invention described in Patent Document 1, monitoring is performed using the video of the movement of animals in a specific direction that occurs when moving from a specific breeding area to a specific breeding area, and livestock are counted. At that time, although it is described that animals with serious illnesses should be withdrawn, there is no teaching about specific methods, and it can be said that there is no disclosure that a person skilled in the art can implement without difficulty.

[0006] Therefore, at least one aspect of the problem of the present disclosure is to easily detect pigs with abnormalities in growth such as those that are harmful to health from a plurality of pigs grouped and raised in the same pigsty in the pig farming industry, and to take measures such as isolating them from other pigs. In addition, for those skilled in the art who can read from the embodiments and their descriptions with characteristics of the present disclosure described in the specification, drawings, etc. of the present disclosure, problems that are obvious can also be problems that the divided inventions should solve when a divisional application is filed based on the present disclosure.

Means for Solving the Problems

[0007] To achieve the above object, a method for measuring the activity level of pigs uses a pig activity level measurement system including at least a camera unit capable of shooting videos and a computer device with an input user interface and an output user interface, without the need to attach sensors to the pig body. The method for calculating the activity level of a plurality of pigs raised in a pigsty by a user includes a measurement start instruction step in which the user instructs the start of measuring the activity level of pigs using the input user interface, a movement induction step in which the user induces the movement of pigs by performing an induction operation on the pigs raised in the pigsty, a continuous shooting step in which the user continuously shoots videos of the pigs moving due to the induction operation using the camera unit, a pig identification step in which the pig identification unit of the computer device identifies the individual pigs included in the videos, a pig tracking step in which the pig tracking unit of the computer device assigns a temporary ID to the individual pigs identified in the pig identification step and tracks them, an activity level calculation step in which the activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of the tracked pigs within a predetermined time, using the videos of the pigs being tracked in the pig tracking step, a reference activity level calculation step in which the reference activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of the plurality of pigs being tracked in the pig tracking step within a predetermined time, and calculates the reference activity level using the activity levels of the plurality of pigs, and an abnormal pig extraction step in which the abnormal pig extraction unit of the computer device compares at least the activity level of the pigs being tracked in the pig tracking step with the reference activity level and extracts abnormal pigs with abnormal activity levels.

Advantages of the Invention

[0008] According to the present disclosure, it is possible to easily detect and isolate pigs with abnormalities in growth, such as those harming their health, from a plurality of pigs grouped and raised in the same pigsty.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Mode for Carrying Out the Invention

[0010] In the pig farming industry, multiple pigs are grouped and raised in the same pigsty. A pigsty often refers to the unit of the smallest breeding area divided by fences or the like inside a pig shed on a farm. Generally, multiple pigs grouped in the same pigsty are grouped as a homogeneous and uniform group with the same growth stage, similar age, the same breed, and the same livestock type (production roles in pig farming such as fattening pigs, sows, etc.). If there is a pig in such a same pigsty that is harming its health, such as being sick, it is necessary to discover it early, isolate it, and take measures.

[0011] To determine whether a pig is suffering from a disease, a veterinary examination is required. However, it is difficult to respond quickly unless the farm has a resident veterinarian. Therefore, the first person to detect abnormalities will be the workers who work regularly on the farm. Experienced workers can discover a pig suffering from a disease or a pig showing signs of it among multiple raised pigs and isolate it from the group. However, in recent years, there is a shortage of people in the entire agricultural industry, especially in the livestock industry, and the presence of such experienced workers is particularly rare. Also, even if such experienced workers are on the farm, it is a burden to patrol the entire farm at a high frequency.

[0012] Therefore, there is a need for a method that does not rely on experienced workers, requires no experience, and is as quantitative as possible to identify pigs that are harmful to health. The following will describe a method for measuring the activity level of pigs, a non-transitory computer-readable storage medium storing a program for measuring the activity level of pigs, and a system for measuring the activity level of pigs, which can easily detect and isolate pigs with abnormalities in growth, such as those harmful to health, from a plurality of pigs grouped and raised in the same pigsty.

[0013] <Configuration> FIG. 1 is a network diagram showing the configuration of the pig activity level measurement system of the present disclosure. The pig activity level measurement system 1 may include a first computer device 10, a second computer device 20, and a camera unit 30 connected to a network NW. The first computer device 10 and the second computer device 20 may be computer devices as described with reference to FIG. 2 below. As will be described later, if the first computer device 10 alone has sufficient computing power, the second computer device 20 may not be provided. Also, depending on the required computing power, a third or more computer device may be provided.

[0014] FIG. 2 is a diagram showing an example of the computer device 10 that can constitute the first computer device 10 and the second computer device 20. The computer device 10 includes a communication interface 11, an input user interface 12, an output user interface 13, a processor 14, and a storage device 15.

[0015] The computer device 100 can execute a predetermined process through the cooperation of software and hardware resources by executing, by the processor 14, an application program based on the application data stored in the storage device 15. The storage device 15 may store operating system data necessary for the computer device 10 to function as a general-purpose computer, and the operating system functions using the operating system data.

[0016] Such a computer device 10 may be any type of electronic device, such as a desktop computer, a laptop computer, a portable or mobile device, a camera, a mobile phone, a smartphone, a tablet computer, a television, a wearable device (display glass or goggles, a head-mounted display (HMD), a wristwatch, a headset, an armband, etc.), a virtual reality (VR) and / or augmented reality (AR) compatible device, a personal digital assistant, etc.

[0017] The computer device 10 may be connected so as to be accessible by wired or wireless communication to a local database similar to the storage device 15 or another storage device. The storage device 15 may be any processor-readable storage medium suitable for storing instructions executed by a processor, such as a random access memory (RAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, etc., which is provided to be accessible by the processor 230. The storage device 15 may be arranged separately from the processor 14 and / or integrated with the processor 14.

[0018] Also, any software may be stored in any suitable other auxiliary storage device, secondary storage device, or temporary or non-temporary computer-readable storage medium. Additionally, any type of storage device (magnetic disk, optical disk, magnetic tape, or other tangible medium) may be regarded as a storage device.

[0019] The input user interface 12 and the output user interface 13 may be hardware devices for the user to input and output information in relation to the computer device 10 and / or other computer devices. Specifically, input devices that may constitute the input user interface 12 may include a keyboard, a mouse, one or more touch panel sensors, a microphone, etc. Similarly, output devices that may constitute the output user interface 13 may include a display, a monitor, a printer, a speaker, etc.

[0020] The communication interface 11 can communicate with other electronic devices corresponding to various types of electrical communication lines such as wired connections and wireless connections. For example, communication can be performed via a wide area network connection through an optical fiber network or a digital telephone line, a local wireless connection, short-range wireless communication, or a satellite-based position information system.

[0021] The processor 14 can be one or more of any type of computer processing element, such as an integrated circuit or a controller that executes processor operations by a central processing unit (CPU), etc. For example, the processor 14 may be one or more single-core processors. Also, the processor 14 may be one or more multi-core processors having a plurality of independent processing units. The processor 14 may also include register memory for temporarily storing executed instructions and related data, as well as cache memory for temporarily storing recently used instructions and data. When processing is performed using multiple processors, it is not necessary for the same processor to perform all the processing.

[0022] A computer system may adopt a cluster configuration in which a plurality of computers are grouped and connected by a network. In that case, the same computer system may be installed in multiple locations. The specific installation locations and connection methods of these computer devices are not important, and they may be located outside the country where the user and the farm are located. Such a group of computer devices may be treated as one cloud computing resource distributed across various data centers.

[0023] Also, when a mobile terminal such as a smartphone used by the user has sufficient processing power, the terminal can be used as a single small computer to perform all computational processing. It is also possible to have this mobile terminal undertake part of the computational processing as part of a "cluster" or "edge computing" environment grouped in combination with a plurality of other computers.

[0024] Returning to FIG. 1 again, a configuration example of the pig weight measurement system of the present disclosure will be described using the above configuration. As described above, the first computer device 10 and the second computer device 20 are selected according to the required computing resources. If the computing power is sufficient, one computer device can be used. If there is a shortage, the computing resources of a plurality of computer devices connected by the network NW can be utilized. Also, it is not limited to being used only as computing resources, and it may be used as data storage.

[0025] <Program> FIG. 3 is a functional block diagram for explaining the configuration of a pig activity measurement program P100 used in the pig activity measurement method of the present disclosure. The functions described here are realized by the cooperation of software and hardware resources by executing an application program based on the application data stored in the storage device 15 described above by the processor 14.

[0026] When an activity measurement start signal is transmitted according to an instruction from a user via the input user interface 12, the shooting instruction unit P110 transmits a shooting instruction signal to the camera unit 30 to continuously shoot the pigs housed in the pigsty by the camera unit 30. The user who transmits such an instruction is, for example, a worker working on a farm and can regularly perform the activity measurement of pigs as a daily task. As a more specific example, although not particularly shown, the measurement start signal can be transmitted by pressing a measurement start button displayed on the screen of a smartphone having a touch panel.

[0027] Here, an induction operation performed on the pigs housed in the pigsty, which is the operation of a human worker and does not depend on the processing by the pig activity measurement program P100, will also be described.

[0028] For example, the worker enters the inside of the pigsty at the same time as the start of shooting the video described later and walks along its outer periphery to encourage the pigs to move. At this time, since there are also pigs that avoid movement and do not move much, movement can be encouraged by approaching them or directly touching them. In some cases, more stimulating behaviors such as making sounds or clapping hands may be performed, but there is also a risk of being attacked, such as by the pigs coming towards the worker.

[0029] There are also risks associated with such induction behaviors, and considering that the results of measurement may vary due to the induction behaviors, as shown in FIG. 3, the shooting instruction unit P110 may teach the induction assist data from the output user interface 13 of the computer device 10. More specifically, it is the display of information such as the elapsed time since the measurement start instruction, instructions such as "Please move", and the timer function. By this, quantitative and homogeneous induction behaviors can be performed.

[0030] Based on the shooting instruction signal transmitted from the shooting instruction unit P110, the camera unit 30 starts continuously shooting the pigs raised in the pigsty. Then, the captured video is sequentially and streamed to the computer device so as to be provided for the processing of the pig activity measurement program P100. That is, the shooting of the video is started in conjunction with the measurement start signal, and the video data will be used for measuring the activity amount of the pigs.

[0031] The pig identification unit P120 has a function of identifying a pig region coordinate group from the pixels or point clouds included in the image by using the video acquired from the camera unit 30. Briefly speaking, it can find the pigs included in the image.

[0032] Here, among the plurality of pixels or point clouds included in the image, those indicating the presence of pigs are defined as the pig region coordinate group. The pig region coordinate group can be identified by using known image analysis techniques such as semantic segmentation and instance segmentation. Since the pigs in the pigsty often appear overlapping in the image, it is preferable to use instance segmentation to identify them one by one.

[0033] When identifying the pig region coordinate group by using the video, it is common to use techniques such as machine learning and deep learning, such as feature extraction and classification. Even a general instance segmentation model has the ability to recognize and segment the shape of an object from visual features such as the contour, texture, and shape of the object. However, in order to avoid misrecognition, additional learning and reinforcement learning may be performed by using a dataset containing pig images. Also, the pigs may be identified by using a pose model specialized for quadruped animals, not limited to pigs.

[0034] Figure 4 shows an example of the identification of a group of pig coordinates. In this figure, it is identified that there is a pig in the area surrounded by the bounding box. This image is only an illustrative diagram. All pigs that can be identified, including those other than the pigs surrounded by the bounding box, are identified. That is, even if the bounding box is not shown as an image on the output user interface 13, the identification can be performed internally.

[0035] Returning to FIG. 3 again, a temporary ID is assigned to the pigs identified by the pig identification unit P120. Pigs raised on farms that utilize ICT may be assigned individual IDs and managed. Regarding such individual IDs of pigs, there are also technologies for individually authenticating using, for example, two-dimensional codes or labels carried on ear tags. However, the temporary ID mentioned here is a provisional ID different from such individual IDs. The reason is that in activity measurement, more advanced technology is required for the identification of individual IDs, which increases the cost, and in this disclosure, it is sufficient if abnormal pigs can be found from a homogeneous group. In addition, since pigs move around and it becomes difficult to track them when they go out of the field of view and are lost from sight, and a new ID is assigned when they reappear within the screen again, the temporary ID can be operated more efficiently.

[0036] In this way, the video with the temporary ID assigned to the identified pigs will include the data of the pigs to which the temporary ID is assigned and are to be identified. The pig tracking unit P130 has a function of continuously and automatically tracking the pigs to which the temporary ID is assigned and are to be identified. As a result, the video will include the data of the pigs to which the temporary ID is assigned and are being tracked.

[0037] The video including the data of the pigs to which the temporary ID is assigned and are being tracked is input to the activity amount calculation unit P140, the reference activity amount calculation unit P150, the aspect ratio calculation unit P160, the reference aspect ratio calculation unit P170, and the weight estimation unit (not shown in this figure) as data for calculating the activity amount, the reference activity amount, the aspect ratio, the reference aspect ratio, and the weight of the pigs (not shown in this figure), which are indicators for extracting abnormal pigs from a homogeneous group.

[0038] The activity amount calculation unit P140 has a function of calculating the activity amount, which is the moving distance, moving speed, or acceleration of the pig being tracked within a predetermined time, using a video including the pig data with temporary ID assigned.

[0039] Here, regarding the moving distance of the pig, the representative coordinates such as the center of gravity or the center of the bounding box (temporarily referred to as the center coordinates) are taken as the distance moved on the image. That is, the unit of distance is the distance between coordinates on the image. For a two-dimensional image, it is the difference between coordinate a1(x1, y1) and coordinate a2(x2, y2). Such a difference in coordinates becomes the moving distance. And the moving speed is obtained by dividing the moving distance by time such as the number of frames or seconds. The acceleration is the derivative of the moving speed with respect to time. In this way, the activity amount calculation unit P140 calculates the activity amount of the individual pig being tracked.

[0040] The calculated activity amount of the individual may be transmitted to the user via the output user interface 13. FIG. 5 shows an image presenting the activity amount of the pig presented on the display as an example of the output user interface 13 of the computer device 10. In this figure, the identified, tracked pig with a temporary ID assigned and its activity amount are shown. Specifically, for the pig with the temporary ID "pig:1", it is displayed adjacent to the bounding box that the activity amount "Speed" is "0.3", and similarly, for the pig with the temporary ID "pig:0", it is displayed that the activity amount "Speed" is "0.0".

[0041] The reasons for using such speed and acceleration as indicators for extracting candidates for pigs that may be abnormal, such as having diseases, are as follows.

[0042] Generally, pigs for meat consumption tend to have a decreasing speed and acceleration during daily exercise as their weight increases with age as they grow. On the other hand, as the weight increases and the shipment approaches, they become more stable in terms of health and less likely to contract diseases. That is, if they have a certain age and weight, there is no problem with having an absolutely low speed and acceleration.

[0043] However, young piglets immediately after weaning, etc., have high speeds and accelerations during daily exercise, but are in a state where they are prone to diseases. When they contract a disease, their activity level decreases, and so do their speed and acceleration. Then, if the activity level is low despite their young age, there is a possibility that they are suffering from a disease. This becomes prominent when compared with pigs of similar ages in the same generation.

[0044] Thus, the weights of pigs immediately after weaning and pigs immediately before shipment are different, and it is not appropriate to compare them as the same group. For this reason, on farms, pigs of similar nature, basically pigs of similar ages, are grouped, lot-by-loted, and managed by being raised in the same pigsty, etc. In this way, one pigsty is suitable as a population of homogenized pigs for comparison purposes.

[0045] From the above, in order to compare with pigs raised in the same pigsty, similarly, the reference activity amount calculation unit P150 also calculates the activity amounts of a plurality of pigs that can all be tracked. Then, using the activity amounts of these plurality of trackable pigs, a reference activity amount is calculated.

[0046] Regarding the reference activity amount, it can be arbitrarily set to various values using the activity amounts of a plurality of trackable pigs. Specifically, it is also possible to compare with the average value or median value of the activity amounts of a plurality of pigs in the same pigsty. However, if all those below the average value are regarded as abnormal, the number extracted will be too large. Therefore, it is conceivable to set the reference activity amount so as to extract individuals below or less than the first quartile of the activity amount, or simply individuals below or less than an arbitrary ratio with respect to the whole. Specifically, it can also be set such that the activity amount is below or less than 10th percentile, 5th percentile of the whole, or below or less than 2 standard deviations in a normal distribution, so-called 2σ or less.

[0047] By inputting the activity amount of the individual calculated as described above and the reference activity amount as data, the abnormal pig extraction unit P180 has a function of extracting abnormal pigs from among a plurality of temporarily assigned ID pigs to be tracked. Specifically, the activity amount of the individual may be compared with the reference activity amount, and if the activity amount of the individual is lower than or equal to the reference activity amount, it may be extracted as an abnormal pig.

[0048] The extracted abnormal pigs are presented to the user by the output user interface 13. Specifically, although not particularly shown, on the screen of a computer device such as a smartphone held by the user, among the pigs surrounded by a plurality of bounding boxes being tracked, it is possible to make a display indicating that it is an abnormal pig, such as changing the color of the box.

[0049] The user can recognize which pigs are abnormal pigs via the output user interface 13. To present more specific convenience, while looking at the screen of the smartphone in the actual pigsty, it is possible to confirm which pigs are abnormal pigs and attach a label as an abnormal pig candidate to that pig. As a method of attaching a label as an abnormal pig candidate, it is non-invasive to mark the body surface, skin, etc. of the pig with a color spray or the like, and it is also preferable from the perspective of animal welfare.

[0050] The pigs labeled as abnormal pig candidates in this way can be isolated for diagnosis by experts such as veterinarians at a later date or on the same day. If it is determined to be sick after diagnosis, measures such as isolation or treatment can be taken. Also, it is considered that prompt isolation may be able to prevent the infection of diseases to other pigs in some cases.

[0051] Also, a case where the abnormal pig extraction unit P180 extracts abnormal pigs using not only the activity amount but also the aspect ratio will be described.

[0052] The aspect ratio refers to the ratio when the "vertical" is the length from the tip of the head to the tail of the pig's body, and the "horizontal" is the length of the widest part of the body in a direction perpendicular to the "vertical". Such an aspect ratio is accurately obtained from an image of a stationary pig, but as reference information only, the vertical and horizontal can also be determined from the pixels constituting the image of the pig being tracked in the video or from the shape of the bounding box. For example, the maximum value of the horizontal width and vertical height of the pig area coordinate group may be used as the vertical and horizontal, or the vertical height and horizontal width of the bounding box may be used as the vertical and horizontal.

[0053] Such an aspect ratio represents the body shape of the pig. Even if the aspect ratio is defined as vertical / horizontal (vertical divided by horizontal) or horizontal / vertical (horizontal divided by vertical), if it is significantly large or small, it can be said that the pig is thin. Therefore, by considering it together with the activity level, it can be effectively utilized to extract abnormal pigs. That is, a thin pig has an aspect ratio larger or smaller than the reference value. And even though it is thin, a pig with a low activity level is highly likely to have some abnormality such as a disease.

[0054] Regarding the aspect ratio as well, similar to the activity level, the reference aspect ratio calculation unit P170 calculates the aspect ratios of a plurality of pigs being tracked using the videos of the plurality of pigs, and uses these to calculate the reference aspect ratio. The method for this is the same as the method by which the above-mentioned reference activity level calculation unit P150 calculates the reference activity level, so the explanation is omitted.

[0055] Also, the estimated weight of the pig may be used to indicate the body shape of the pig. Specifically, instead of the aspect ratio calculation unit, a pig weight estimation unit (not shown) is provided in the pig activity measurement program P100 to estimate the weight of the pig.

[0056] To estimate the weight of a pig, a weight estimation model that estimates Weight (body weight) using the length L (mm) and width W (mm) of the pig derived from a video as parameters may be used. This weight estimation model is a pre-trained model that has been previously trained on the relationship between the length L (mm), width W (mm), and Weight (body weight) through machine learning such as deep learning. The training of such a model, for example, constructs a neural network that takes the length L (mm) and width W (mm) as input features and outputs Weight (body weight), and uses training data in which the length L (mm), width W (mm), and Weight (body weight) are paired for the constructed NN to perform training so that the estimation accuracy of the weight is improved, thereby obtaining a pre-trained model.

[0057] In addition, the pig weight calculation formula may be estimated by inputting parameters that are a plurality of explanatory variables such as the horizontal width, vertical width, height, etc. of the pig, for example, using a regression formula based on multivariate analysis. As a non-limiting example, the length L (mm) and width W (mm) are shown. For example, assuming that the weight is estimated using a multiple regression formula such as Weight (body weight) = AL + BW + C (A and B are coefficients, and C is a constant term), L and W may represent the length L (mm) and width W (mm).

[0058] Real values such as the horizontal width, vertical width, and height can be calculated by the conversion ratio between pixels and actual distances, which is derived using the internal parameters of the camera such as the focal length of the camera, the field of view angle of one pixel, the baseline distance (parallax between the left and right cameras), and the size of the image sensor, based on the number of pixels. Thereby, the weight estimation unit can obtain the estimated weight of the pig using the video of the pig.

[0059] The obtained estimated weight may be used as an index when extracting abnormal pigs, similar to the aspect ratio. Similarly, a reference weight for comparison may also be estimated.

[0060] Next, a method for estimating the weight of pigs using the pig weight estimation system of the present disclosure will be described with reference to the flowchart shown in FIG.*. Among the following steps S101 to S108, step S102 is executed by a human user, and step S103 is executed using the camera unit 30. Steps S101 and subsequent steps S104 to S108 are executed by a program of any computer device. Note that when the camera unit 30 is built into a computer device or incorporated as a component that substantially integrally processes instructions through communication, it is also possible to execute them with a consistent program from S101 to S108.

[0061] In step S101, the user uses the input user interface 12 to instruct the start of measuring the activity amount of the pig (measurement start instruction step).

[0062] In step S102, the user induces the movement of the pig by performing an inducing action on the pig housed in the pigsty (movement induction step).

[0063] In step S103, the camera unit 30 continuously captures a video of the pig moving due to the inducing action of the user (continuous capture step).

[0064] In step S104, the pig identification unit P120 identifies the pig individual included in the video (pig identification step).

[0065] In step S105, the pig tracking unit P130 assigns a temporary ID to the pig individual identified in step S104 and tracks it (pig tracking step).

[0066] In step S106, at least the activity amount calculation unit P140 calculates the activity amount, which is the moving distance, moving speed, or acceleration of the tracked pig within a predetermined time, using the video of the pig being tracked in step S105 (activity amount calculation step).

[0067] In step S107, at least the reference activity amount calculation unit P150 calculates the activity amount, which is the moving distance, moving speed, or acceleration of a plurality of pigs being tracked in step S105 within a predetermined time, and calculates the reference activity amount using the activity amounts of the plurality of pigs (reference activity amount calculation step).

[0068] In step S108, the abnormal pig extraction unit P180 compares the activity amount of the pigs being tracked in S105 or calculated in S106 with the reference activity amount calculated in step S107, and extracts the abnormal pigs with abnormal activity amounts (abnormal pig extraction step). Thereby, among the plurality of pigs grouped and raised in the same pigsty, it is possible to easily detect pigs with abnormalities in growth such as being harmful to health and take measures such as isolating them from other pigs.

[0069] As described above, the method for measuring the activity amount of pigs according to the present disclosure can be executed. As one configuration example of the present disclosure, a video camera unit capable of continuously shooting a video of a pig moving in a pigsty as the camera unit 30, an AR device as the first computer device 10, and a server computer as the second computer device 20 can be mentioned. For example, the AR device includes a head-mounted display or a glasses-type display, and the camera unit 30 is a camera unit integrally attached to a head-mounted display connected to a computer device.

[0070] FIG. 7 is an example of the pig activity amount measurement system according to the present disclosure. As shown in FIG. 7, a video camera component built into a portable computer device or a smartphone as the camera unit 30, and a computer device or a smartphone as the first computer device 10 can be mentioned.

[0071] FIG. 8 is also an example of the pig activity measurement system of the present disclosure. A video camera fixedly arranged at a position where all pigs in the pigsty can be imaged is provided on a farm as the camera unit 30, a laptop computer as the first computer device 10, and a server computer as the second computer device 20. The camera unit 30 is fixed at a position where all pigs in the pigsty can be imaged. The user can check from a display connected so as to be able to receive the presentation of abnormal pigs in the farm.

[0072] The above is an example, and all configuration examples combining the above-described respective configurations can be implemented.

[0073] <Supplementary Note> The matters described in each of the above embodiments are appended below. (Supplementary Note 1) A method for measuring the activity amount of pigs, which uses a pig activity measurement system including at least a camera unit capable of shooting videos and a computer device having an input user interface and an output user interface, without the need to attach a sensor to the livestock body, and calculates the activity amount of a plurality of pigs raised in a pigsty by a user, a measurement start instruction step in which the user instructs the start of measuring the activity amount of pigs using the input user interface; a movement induction step of inducing the movement of pigs by performing an induction operation on the pigs raised in the pigsty by the user; a continuous shooting step in which the camera unit continuously shoots a video of the pigs moving due to the induction operation; a pig identification step in which the pig identification unit of the computer device identifies the pig individuals included in the video; a pig tracking step in which the pig tracking unit of the computer device assigns a temporary ID to the pig individuals identified in the pig identification step and tracks them; an activity amount calculation step in which the activity amount calculation unit of the computer device calculates the activity amount, which is the moving distance, moving speed, or acceleration of the pigs being tracked within a predetermined time, using the video of the pigs being tracked in the pig tracking step. A reference activity amount calculation step in which a reference activity amount calculation unit of the computer device calculates an activity amount that is the moving distance, moving speed, or acceleration of a plurality of pigs being tracked in the pig tracking step within a predetermined time, and calculates a reference activity amount using the activity amounts of the plurality of pigs; An abnormal pig extraction step in which an abnormal pig extraction unit of the computer device compares at least the activity amount of the pigs being tracked in the pig tracking step with the reference activity amount, and extracts abnormal pigs with abnormal activity amounts. A method for measuring the activity amount of pigs, including this step. (Appendix 2) An aspect ratio calculation step in which an aspect ratio calculation unit of the computer device calculates the aspect ratio of the pigs being tracked in the pig tracking step using the video of the pigs being tracked in the pig tracking step; A reference aspect ratio calculation step in which a reference aspect ratio calculation unit of the computer device calculates the aspect ratios of a plurality of pigs being tracked in the pig tracking step using the videos of the plurality of pigs being tracked in the pig tracking step, and calculates a reference aspect ratio using the aspect ratios of the plurality of pigs. The method further includes this step. In the abnormal pig extraction step, an abnormal pig extraction unit of the computer device compares at least the activity amount and aspect ratio of the pigs being tracked in the pig tracking step with the reference activity amount and the reference aspect ratio, and extracts abnormal pigs with abnormal activity amounts and aspect ratios. The method for measuring the activity amount of pigs according to Appendix 1. (Appendix 3) A weight estimation step in which a weight estimation unit of the computer device estimates the weight of the pigs being tracked in the pig tracking step using the video of the pigs being tracked in the pig tracking step; A reference weight calculation step in which a reference weight calculation unit of the computer device estimates the weights of a plurality of pigs being tracked in the pig tracking step using the videos of the plurality of pigs being tracked in the pig tracking step, and calculates a reference weight using the weights of the plurality of pigs. The method further includes this step. In the abnormal pig extraction step, the abnormal pig extraction unit of the computer device compares at least the activity level and body weight of the pigs being tracked in the pig tracking step with the reference activity level and the reference body weight, and extracts abnormal pigs with abnormal activity levels and body weights. The method for measuring the activity level of pigs according to Supplementary Note 1. (Supplementary Note 4) A pig activity level measurement program for measuring the activity levels of a plurality of pigs housed in a pigsty using a pig activity level measurement system including at least a camera unit capable of shooting videos and a computer device equipped with an input user interface and an output user interface, without the need to attach sensors to the livestock body, When a measurement start instruction signal for instructing the start of pig activity level measurement is received from the input user interface, a continuous shooting step of continuously shooting a video of the pigs by the camera unit, A pig identification step in which the pig identification unit of the computer device identifies the pig individuals included in the video, A pig tracking step in which the pig tracking unit of the computer device assigns a temporary ID to the pig individuals identified in the pig identification step and tracks them, An activity level calculation step in which the activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of the pigs being tracked within a predetermined time, using the video of the pigs being tracked in the pig tracking step, A reference activity level calculation step in which the reference activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of a plurality of pigs being tracked in the pig tracking step within a predetermined time, and calculates a reference activity level using the activity levels of the plurality of pigs, An abnormal pig extraction step in which the abnormal pig extraction unit of the computer device compares at least the activity level of the pigs being tracked in the pig tracking step with the reference activity level and extracts abnormal pigs with abnormal activity levels. A non-temporary computer-readable storage medium storing a pig activity level measurement program for causing a computer device to execute the steps. (Supplementary Note 5) A pig activity level measurement system that does not require attaching sensors to the livestock body, At least one computer device, A camera unit capable of continuously photographing videos of pigs moving in a pigsty, An input user interface of the computer device that transmits a measurement start signal instructing the start of measuring the activity amount of pigs to the computer device, A pig identification unit of the computer device that identifies pig individuals included in the video obtained from the camera unit, A pig tracking unit of the computer device that assigns a temporary ID to the identified pig individuals and tracks them, Using the video of the tracked pigs, calculate the activity amount, which is the moving distance, moving speed, or acceleration of the tracked pigs within a predetermined time, and Calculate the activity amount, which is the moving distance, moving speed, or acceleration of the plurality of tracked pigs within a predetermined time, and calculate a reference activity amount using the activity amounts of the plurality of pigs. An activity amount calculation unit, An abnormal pig extraction unit that extracts abnormal pigs with abnormal activity amounts from the tracked pigs by comparing them with the reference activity amount, An output user interface of the computer device that presents the abnormal pigs extracted by the abnormal pig extraction unit. A pig activity amount measurement system including the above. (Appendix 6) The camera unit is fixed at a position where all pigs in the pigsty can be imaged, The output user interface is a display connected so as to be able to receive the presentation of the abnormal pigs within the farm. The pig activity amount measurement system according to Appendix 5. (Appendix 7) The output user interface is a display connected so as to be able to receive the presentation of the abnormal pigs within the farm, The camera unit is a camera unit integrally attached to the display connected to the computer device. The pig activity amount measurement system according to Appendix 5.

Explanation of symbols

[0074] 1 Pig activity amount measurement system 10 First computer device 20 Second computer device 30 Camera unit 11 Communication interface 12 Input user interface 13 Output user interface 14 Processor 15 Storage device P100 Pig activity measurement program P110 Shooting instruction section P120 Pig identification section P130 Pig tracking section P140 Activity amount calculation section P150 Reference activity amount calculation section P160 Aspect ratio calculation section P170 Reference aspect ratio calculation section P180 Abnormal pig extraction section

Claims

1. A method for measuring the activity level of pigs, which uses a pig activity level measurement system including at least a video camera unit capable of shooting videos and a computer device equipped with an input user interface and an output user interface, without the need to attach sensors to the pig body. The method includes: A measurement start instruction step in which the user instructs the start of pig activity level measurement using the input user interface; An exercise induction step in which the user induces the movement of pigs by performing an induction action on the pigs raised in the pigsty; A continuous shooting step in which the camera unit continuously shoots videos of the pigs moving due to the induction action; A pig identification step in which the pig identification unit of the computer device identifies the pig individuals included in the videos; A pig tracking step in which the pig tracking unit of the computer device assigns a temporary ID to the pig individuals identified in the pig identification step and tracks them; An activity level calculation step in which the activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of the tracked pigs within a predetermined time, using the videos of the pigs being tracked in the pig tracking step; A reference activity level calculation step in which the reference activity level calculation unit of the computer device calculates the activity level, which is the moving distance, moving speed, or acceleration of the plurality of pigs being tracked in the pig tracking step within a predetermined time, and calculates the reference activity level using the activity levels of the plurality of pigs; An abnormal pig extraction step in which the abnormal pig extraction unit of the computer device compares at least the activity level of the pigs being tracked in the pig tracking step with the reference activity level and extracts the abnormal pigs with abnormal activity levels.

2. An aspect ratio calculation step in which the aspect ratio calculation unit of the computer device calculates the aspect ratio of the pigs being tracked in the pig tracking step using the videos of the pigs being tracked in the pig tracking step; A reference aspect ratio calculation step in which the reference aspect ratio calculation unit of the computer device calculates the aspect ratios of the plurality of pigs being tracked in the pig tracking step using the videos of the plurality of pigs being tracked in the pig tracking step, and calculates the reference aspect ratio using the aspect ratios of the plurality of pigs. In the abnormal pig extraction step, the abnormal pig extraction unit of the computer device compares at least the activity amount and aspect ratio of the pigs being tracked in the pig tracking step with the reference activity amount and the reference aspect ratio, and extracts abnormal pigs with abnormal activity amounts and aspect ratios. The method for measuring the activity amount of pigs according to claim 1.

3. A weight estimation step in which the weight estimation unit of the computer device estimates the weight of the pigs being tracked in the pig tracking step using the videos of the pigs being tracked in the pig tracking step, A reference weight calculation step in which the reference weight calculation unit of the computer device estimates the weights of a plurality of pigs being tracked in the pig tracking step using the videos of the plurality of pigs being tracked in the pig tracking step, and calculates a reference weight using the weights of the plurality of pigs. The method further includes: In the abnormal pig extraction step, the abnormal pig extraction unit of the computer device compares at least the activity amount and weight of the pigs being tracked in the pig tracking step with the reference activity amount and the reference weight, and extracts abnormal pigs with abnormal activity amounts and weights. The method for measuring the activity amount of pigs according to claim 1.

4. A program for measuring the activity amount of pigs in a pigsty using a pig activity amount measurement system including at least a video - shootable camera unit and a computer device having an input user interface and an output user interface, without the need to attach sensors to the livestock body. A continuous shooting step of causing the camera unit to continuously shoot videos of pigs when a measurement start instruction signal for instructing the start of measuring the activity amount of pigs is received from the input user interface. A pig identification step in which the pig identification unit of the computer device identifies pig individuals included in the video. A pig tracking step in which the pig tracking unit of the computer device assigns a temporary ID to the pig individuals identified in the pig identification step and tracks them. An activity amount calculation step in which the activity amount calculation unit of the computer device calculates the activity amount, which is the moving distance, moving speed, or acceleration of the pigs being tracked within a predetermined time, using the videos of the pigs being tracked in the pig tracking step. A reference activity amount calculation step in which a reference activity amount calculation unit of the computer device calculates an activity amount that is a moving distance, a moving speed, or an acceleration of a plurality of pigs being tracked in the pig tracking step, and calculates a reference activity amount using the activity amounts of the plurality of pigs; An abnormal pig extraction step in which an abnormal pig extraction unit of the computer device compares at least the activity amount of the pigs being tracked in the pig tracking step with the reference activity amount and extracts abnormal pigs with abnormal activity amounts. A non-transitory computer-readable storage medium storing a pig activity amount measurement program for causing a computer device to execute the steps.

5. A pig activity amount measurement system that does not require attaching a sensor to a livestock body, comprising: At least one computer device; A camera unit capable of continuously photographing videos of pigs moving in a pigsty; An input user interface of the computer device that transmits a measurement start signal instructing the start of pig activity amount measurement to the computer device; A pig identification unit of the computer device that identifies pig individuals included in the video acquired from the camera unit; A pig tracking unit of the computer device that assigns a temporary ID to the identified pig individuals and tracks them; Using the video of the pigs being tracked, calculates an activity amount that is a moving distance, a moving speed, or an acceleration of the pigs being tracked within a predetermined time, and An activity amount calculation unit that calculates an activity amount that is a moving distance, a moving speed, or an acceleration of a plurality of pigs being tracked within a predetermined time, and calculates a reference activity amount using the activity amounts of the plurality of pigs; An abnormal pig extraction unit that extracts abnormal pigs with abnormal activity amounts from the pigs being tracked by comparing with the reference activity amount; An output user interface of the computer device that presents the abnormal pigs extracted by the abnormal pig extraction unit. A pig activity amount measurement system comprising the above components.

6. The camera unit is fixed at a position where all pigs in the pigsty can be imaged. The output user interface is a display connected so as to be able to receive the presentation of the abnormal pigs within the farm. The pig activity amount measurement system according to Claim 5.

7. The output user interface is a display connected so as to be able to receive the presentation of the abnormal pigs within the farm, The pig activity measurement system according to claim 5, wherein the camera unit is a camera unit integrally attached to the display connected to the computer device.

Images