Method for monitoring chicken coop

Abstract

A movement device (1), to which an environment sensor (61) and a camera (62) are attached according to the height of each of a plurality of cage rows (90R), is moved back and forth along the front face of a cage row layer (90L) in the direction of extension of the cage rows (90R), and environment information acquired by the environment sensor (61) and an image acquired by the camera (62) are stored in a storage device in association with cage information for identifying each cage.

Description

How to monitor your chicken coop

[0001] The present invention relates to a method for monitoring the interior of a poultry house where birds are kept in cages.

[0002] In large-scale poultry houses, a large number of birds are raised in cage row layers, where rows of cages are arranged horizontally and stacked in multiple levels. It is desirable for a large number of birds to be raised in the same environment, but when the poultry house is large, the environment tends to vary depending on the location within the house.

[0003] For example, a typical chicken coop is long and narrow in the direction of the cage rows, with a large air inlet on one of a pair of walls perpendicular to the longitudinal direction, and an exhaust fan on the other wall. The exhaust fan expels air from inside the coop to the outside, and air is drawn into the coop from the outside through the air inlet, providing forced ventilation. In such a coop, air flows longitudinally. During this process, the air is warmed by the body heat of the birds, resulting in a large temperature difference between upstream and downstream.

[0004] Therefore, the present applicant has proposed that the temperature inside the chicken coop be adjusted by providing multiple auxiliary air inlets on the side walls of the chicken coop (side walls parallel to the longitudinal direction) and by installing temperature sensors near each of the auxiliary air inlets and opening and closing the auxiliary air inlets based on the detected temperature (see, for example, Patent Document 1).

[0005] However, in this conventional technology, the installation points of temperature sensors were limited, making it difficult to regulate the temperature in large poultry houses with large temperature differences. While it would be ideal to install temperature sensors in multiple locations, there are limitations in terms of cost. Therefore, even in large poultry houses, there is a demand for a more accurate understanding of the temperature environment. Furthermore, the applicant believes that it is desirable to accurately understand factors other than air temperature when it comes to the rearing environment of poultry.

[0006] Meanwhile, in poultry houses, it is also necessary to understand the condition of the birds themselves. For example, if dead birds are left behind, the sanitary environment inside the cages will deteriorate, which may have a negative impact on living birds. Furthermore, egg-producing poultry facilities typically install conveyors along rows of cages, with the cage floors inclined so that eggs can roll onto the conveyor under their own weight and be collected. Therefore, if the bodies of dead birds become an obstacle and prevent the eggs from being transferred out, and after some time the eggs roll onto the conveyor and are collected, old eggs may be mixed in with fresh eggs. Because fresh and old eggs are indistinguishable by appearance, this situation must be avoided.

[0007] In large chicken coops, a large number of birds can be raised by a small number of workers by mechanizing feeding, watering, egg collection, and manure removal. Therefore, even if a bird dies in a cage, it is difficult for the workers to notice. If workers try to make regular rounds to check on the condition of the birds, the large number of cages means that it is a heavy burden on the workers and takes a lot of time.

[0008] Therefore, the present applicant has proposed a method of detecting dead birds by attaching a camera to a device that moves within a poultry house (see Patent Documents 2 and 3). In the invention of Patent Document 2, a camera is attached to a device that moves on a path between layers of cage rows. In the invention of Patent Document 3, rails are provided above each layer of cage rows, and a camera is attached to a device that moves along these rails. By determining whether the birds are alive or dead based on images captured by the camera as the device moves, dead birds can be detected with less labor burden on the worker.

[0009] In this way, it is important to understand both the rearing environment of the birds and the condition of the birds themselves. If we can accurately understand both, it will be possible to examine whether there is a correlation between the rearing environment and the deterioration of the birds' condition, such as death.

[0010] Patent No. 3598212 International Publication No. 2023 / 037397 International Publication No. 2023 / 195063

[0011] As described above, an object of the present invention is to provide a method for monitoring a chicken coop that can accurately monitor both the rearing environment of the birds in the chicken coop and the condition of the birds.

[0012] In order to solve the above-mentioned problems, the monitoring method for a poultry house according to the present invention (hereinafter simply referred to as the "monitoring method") is "a monitoring method for a poultry house in which birds are raised using cage row layers in which cage rows, each of which has a plurality of cages arranged side by side horizontally, are stacked in multiple tiers, the method comprising: using a moving device to which an environmental sensor and a camera are attached; the moving device reciprocates along the front of the cage row layer in the direction in which the cage rows extend; one or more types of environmental sensors are attached to the moving device according to the height of each of the multiple tiers of cage rows; the camera acquires still images or moving images as images, and one or more cameras are attached to the moving device according to the height of each of the multiple tiers of cage rows; as the moving device moves, at least one piece of environmental information is acquired for each of the cages by the environmental sensor, and at least one image is acquired for each of the cages by the camera; and the environmental information and the image are stored in a storage device in association with cage information that identifies each cage."

[0013] In addition to the above configuration, the monitoring method of the present invention can be configured as follows: "using a determination device that determines the occurrence of an abnormality based on at least one of the environmental information and the image, and when it is determined that an abnormality has occurred, combining the environmental information and the image associated with the same cage information and storing them in a storage device."

[0014] In addition to the above configuration, the monitoring method of the present invention can be configured as follows: "As the camera, in addition to a camera that captures images of the inside of the cages belonging to each of the cage rows, a camera that captures images of the inside of a feeding trough arranged along the front of each of the cage rows is used."

[0015] In addition to the above configuration, the monitoring method according to the present invention can be configured such that "a microphone is attached to the mobile device, and sounds made by the birds are collected by the microphone."

[0016] In addition to the above configuration, the monitoring method of the present invention can be configured such that "the moving device includes a slider that moves along the opening edge of at least one of the feeding troughs arranged along the front face of each of the cage rows."

[0017] According to the present invention, a method for monitoring a poultry house can be provided that can accurately monitor both the rearing environment of the birds in the poultry house and the condition of the birds.

[0018] Fig. 1 is a side view of a moving device used in the monitoring method of this embodiment, arranged on a cage row layer. Fig. 2 is a perspective view of the moving device of Fig. 1. Fig. 3(a) is a perspective view of the A1 range, Fig. 3(b) is a perspective view of the B1 range, and Fig. 3(c) is a perspective view of the C1 range. Fig. 3(a) is a side view of a moving device different from the moving device of Fig. 1, arranged on a cage row layer. Fig. 4 is a perspective view of the moving device of Fig. 4. Fig. 6 is a perspective view of the D range.

[0019] A monitoring method according to a specific embodiment of the present invention will now be described with reference to the drawings. The monitoring method of this embodiment is implemented in a poultry house where chickens are raised. A plurality of cage row layers 90L are arranged in the poultry house. Each cage row layer 90L is a structure in which cage rows 90R, each of which has a large number of cages arranged horizontally, are stacked in multiple levels. A long feeding trough 85 is arranged horizontally along each cage row 90R.

[0020] In the cage row 90R and the cage row layer 90L, the side on which the feed trough 85 is arranged is referred to as the "front," the opposite side as the "rear," and the direction from the front to the rear is referred to as the "front-to-rear direction." In a typical poultry house, two cage row layers 90L are held in a frame with their rears facing each other (not shown). The frame is supported by struts 80 that stand on the installation surface. The cage row layers 90L are supported by the struts 80 so that the lowest cage row 90R is higher than the installation surface.

[0021] The feeding troughs 85 are supported from below by trough supports 86 attached to the frame, and are positioned slightly higher than the floor 91 of the cage. This allows birds to stick their heads out of the cage to peck at the food supplied inside the feeding trough 85. An egg collection conveyor (not shown) is provided below each feeding trough 85. The floor 91 is inclined downward toward the egg collection conveyor. This allows eggs laid by the birds in the cage to roll over the floor 91 due to their own weight and be collected by the egg collection conveyor. A tray may be provided instead of an egg collection conveyor.

[0022] The monitoring system in which the monitoring method is used is configured by adding a moving device 1, a slide drive mechanism, an environmental sensor 61, a camera 62, a control device, and a determination device to the cage row layer 90L configured as described above.

[0023] The moving device 1 is arranged for each cage row layer 90L and is a device that reciprocates along the front surface of the cage row layer 90L in the direction in which the cage row 90R extends (horizontal direction). As shown in Figures 1 to 3, the moving device 1 is mainly composed of a moving body 10, a slider 11, and guide members (first guide member 20, second guide member 30). The moving body 10 is a bar that extends vertically.

[0024] The slider 11 slides along the opening edge of the feed trough 85. In this document, when simply referring to the "opening edge" of the feed trough 85, it refers to the opening edge that is farther away from the cage row 90R. In this embodiment, the slider 11 slides along the opening edge of the feed trough 85 that is arranged along the second-highest cage row 90R out of the multiple feed troughs 85. The second-highest feed trough 85 is covered from above with a reinforcing cover 85e to increase the mechanical strength of the opening edge (see FIG. 3(a)). The reinforcing cover 85e is a long member with an angular inverted U-shaped cross section and is attached over the entire length of the feed trough 85.

[0025] Specifically, the slider 11 is a long member extending in the direction of the cage row 90R, and is provided with a pair of rollers 11r at each end of its length. These rollers 11r are free rollers that can rotate freely around a vertical axis. The distance between the pair of rollers 11r is approximately equal to the thickness of the opening edge of the feed trough 85 when the reinforcing cover 85e is in place. This allows the pair of rollers 11r to roll along the opening edge of the feed trough 85 while sandwiching the opening edge from the front and back (see Figure 3(a)). The sides of the slider 11 are fixed to the mobile body 10, and the mobile body 10 and the slider move together.

[0026] The guide members include a first guide member 20 that guides the movement of the moving device 1 along the third feeding trough 85 from the top, and a second guide member 30 that guides the movement of the moving device 1 along the bottom feeding trough 85. The first guide member 20 and the second guide member 30 are each equipped with rollers 20r, 30r that roll along the opening edge of the feeding trough 85.

[0027] Specifically, the first guide member 20 has an intermediate section 21 that extends in the same direction as the cage row 90R, and inclined sections 22 that bend and extend from both ends of the intermediate section 21, with rollers 20r at the tip of each inclined section 22. These rollers 20r are free rollers that can rotate freely around a vertical axis. The intermediate section 21 is fixed to the movable body 10. The length and angle of the inclined sections 22 are set so that the rollers 20r roll while abutting against the opening edge of the feeding trough 85 from the front (see FIG. 3(b)).

[0028] The second guide member 30 protrudes from the movable body 10 toward the feeding trough 85 and is equipped with two rollers 30r. These rollers 30r are spaced apart in the front-to-rear direction, with the distance between them being approximately equal to the thickness of the opening edge of the feeding trough 85. The two rollers 30r can roll along the opening edge of the feeding trough 85 while sandwiching the opening edge from the front and rear (see FIG. 3(c)).

[0029] The moving device 1 further includes a blower 50 and a duct 52 connected to the blower 50. The blower 50 is connected to the slider 11 by a connector 54 (see FIG. 2). The duct 52 is a vertically extending pipe, and the upper end of the duct 52 is connected to the blower 50 by a curved connecting pipe 51. The duct 52 is connected to the moving body 10 by multiple connectors 62b. With this configuration, the blower 50 and duct 52 move integrally with the slider and the moving body 10. The duct 52 has openings 55 corresponding to the height of each cage row 90R. The height and direction of these openings are set so that the air compressed by the blower 50 is blown toward the space in front of the cage. This air blowing reduces the risk of dust adhering to the camera 62 or floating in the air affecting the images captured by the camera 62.

[0030] The slide drive mechanism is a mechanism that drives the reciprocating motion of the moving device 1 and includes a drive pulley 41, a driven pulley 42, a wire 40, and a motor 45. The wire 40 forms an endless loop with a connecting plate 40p interposed therebetween. This endless wire 40 is wound around the drive pulley 41 and the driven pulley 42 and stretched in the horizontal direction. The output shaft of the motor 45 is connected to the rotation shaft of the drive pulley 41. The connecting plate 40p is connected to the moving body 10 by a connector 44. Although not shown, the motor 45, drive pulley 41, and driven pulley 42 are supported by the frame of the cage row layer 90L.

[0031] One or more cameras 62 are attached to the mobile body 10 directly or via brackets, depending on the height of each of the multiple cage rows 90R. Figure 2 illustrates an example in which the connector 62b connecting the duct 52 to the mobile body 10 also serves as a bracket for attaching the camera 62 to the mobile body 10. The camera 62 can be of any type that captures only still images, only moving images, or both still and moving images. The camera 62 may be equipped with a light that irradiates light toward the subject, or a separate light may be provided adjacent to the camera 62.

[0032] The environmental sensors 61 may be temperature sensors, humidity sensors, carbon dioxide gas sensors, ammonia gas sensors, wind speed sensors, wind direction sensors, or illuminance sensors. One or more types of environmental sensors 61 are attached to the mobile body 10 directly or via brackets according to the height of each of the multiple cage rows 90R. For simplicity, the illustration shows one environmental sensor 61 attached to each cage row height, but in this embodiment, multiple types of environmental sensors 61 are attached within the same cage row 90R height range.

[0033] A control device (not shown) is attached to the mobile main body 10. The control device is composed of a computer that mainly includes a storage device consisting of a main storage device and an auxiliary storage device, a processor, and a communication device that communicates with the determination device. The storage device stores a program that causes the computer to function as a movement control means, a program that causes the computer to function as an imaging control means, and a program that causes the computer to function as a sensing control means.

[0034] The movement control means controls the slide drive mechanism to move the mobile device 1. For example, when the environmental sensor 61 acquires environmental information and the camera 62 acquires an image, the movement control means controls the slide drive mechanism to temporarily stop the mobile device 1, or controls the movement of the mobile device 1 to move at a slow speed.

[0035] The imaging control means controls the operation of the camera 62 to acquire images. The sensing control means controls the operation of the environmental sensor 61 to acquire environmental information.

[0036] Multiple types of environmental information acquired for each cage by the environmental sensor 61 and at least one image acquired for each cage by the camera 62 are sent to the judgment device with cage information identifying each cage attached.

[0037] Here, the cage information can be generated based on the reference position where the mobile device 1 starts moving, the moving speed of the mobile device 1, the pitch of the cages in the cage row 90R, and the height at which the camera 62 and the environmental sensor 61 are attached to the mobile main body 10 (a height corresponding to the height of the cage row 90R). Alternatively, a two-dimensional code may be attached to a certain location on the cage row layer 90L, and this may be used as the reference position for movement. In this case, when one of the cameras 62 captures an image of the two-dimensional code, it is determined that the mobile device 1 has reached the reference position, and processing by the movement control means can be performed.

[0038] Alternatively, an identification code may be displayed for each cage within the range photographed by the camera 62, and an image including the identification code may be photographed, whereby cage information can be obtained by image analysis of the identification code.

[0039] The determination device (not shown) is composed of a computer that mainly includes a storage device consisting of a main storage device and an auxiliary storage device, a processor, a communication device that communicates with the control device, an alarm device, an input device such as a keyboard, and an output device including a display. The storage device stores a program that causes the computer to function as environmental abnormality determination means that determines the occurrence of an abnormality based on environmental information, and a program that causes the computer to function as life-or-death determination means that determines the life or death of the birds in the cage that is the subject of the photograph based on images. While the control device is configured to move together with the mobile device 1, the determination device is separate from the mobile device 1 and is installed in a management booth or the like within the facility. The control device and determination device communicate wirelessly.

[0040] Next, the movement of the moving device 1 configured as described above will be described. When the motor 45 is operated to rotate the drive pulley 41 in one direction, the wire 40 moves in one direction while rotating the driven pulley 42, and the slider 11 is pulled by the wire 40 via the connecting plate 40p and the moving body 10. As a result, the slider 11 moves while rolling the roller 11r along the opening edge of the second-highest feeding trough 85, and the moving device 1 moves in one direction. On the other hand, when the drive pulley 41 is rotated in the opposite direction by the motor 45, the same force is transmitted as above, and the moving device 1 moves in the opposite direction. In other words, the forward and reverse rotation of the motor 45 causes the moving device 1 to reciprocate along the front of the cage row layer 90L.

[0041] When the transfer device 1 reciprocates, the first guide member 20 fixed to the transfer body 10 moves while its roller 20r rolls along the opening edge of the third-highest feed trough 85. Similarly, the second guide member 30 moves while its roller 30r rolls along the opening edge of the lowest feed trough 85. In this way, the rollers 20r, 30r of the first and second guide members 20 and 30 roll along the opening edge of the feed trough 85 below the slider 11, stabilizing the movement of the transfer device 1 along the front surface of the cage row layer 90L.

[0042] The above example illustrates a case where the movement of one moving device 1 is driven by one slide drive mechanism. A monitoring system using the monitoring method of this embodiment can be configured, as shown in FIGS. 4 to 6 , in which the movement of two moving devices 2a, 2b is driven by one slide drive mechanism. The two moving devices 2a, 2b each move along two adjacent cage row layers 90L separated by an aisle. As described above, in a typical poultry house, two cage row layers 90L are arranged in pairs with their backs facing each other, and therefore, the two adjacent cage row layers 90L separated by an aisle face each other with their fronts facing each other.

[0043] The moving devices 2a and 2b differ from the moving device 1 in that the moving device 2a is driven by a slide drive mechanism, and the moving device 2b is driven by the slide drive mechanism. Therefore, the upper ends of the moving bodies 10 of the moving devices 2a and 2b are connected by a connecting member 70. Furthermore, air is pressurized and delivered from a single blower 50 to the ducts 52 of the moving devices 2a and 2b. Therefore, the upper ends of the ducts 52 of the moving devices 2a and 2b are connected by a second connecting pipe 53, and the connecting pipe 51 connected to the blower 50 is connected to the middle of the second connecting pipe 53. Except for the above, the configuration of the moving device 2a is the same as that of the moving device 1, and the configurations of the A2, B2, and C2 ranges in FIG. 4 are the same as or symmetrical to the A1, B1, and C1 ranges in FIG. 1, respectively.

[0044] Movement device 2b differs from movement device 2a in that it has slider 12 instead of slider 11, but is otherwise symmetrical to movement device 2a. Slider 12 is a long member extending in the direction of extension of cage row 90R, and has rollers 12r at each end in the longitudinal direction. These rollers 12r are free rollers that can rotate freely around a horizontal axis. Slider 12 is fixed to movement body 10 of movement device 2b by connector 12c, and the lower end of roller 12r abuts from above against the opening edge of the second-from-top feeding trough 85 (see Figure 6).

[0045] With this configuration, when the moving body 10 of the moving device 2a is moved by the driving of the motor 45, the moving body 10 of the moving device 2b moves in response to this. Accordingly, the slider 12, the first guide member 20, and the second guide member 30 roll with their respective rollers 12r, 20r, and 30r in contact with the opening edge of the feeding trough 85.

[0046] Next, processing based on the environmental information acquired by the environmental sensor 61 will be described. As described above, the poultry house is elongated in the direction of the cage row 90R. An air inlet is provided on one of a pair of walls perpendicular to the longitudinal direction, and an exhaust fan is provided on the other wall. Air inside the poultry house is exhausted to the outside by the exhaust fan, and air is drawn into the poultry house from the outside through the air inlet, forcing ventilation. Because air flows longitudinally within the poultry house, differences in the rearing environment for the birds occur between upstream and downstream. The larger the poultry house and the longer the cage row 90R, the greater the environmental differences. In typical poultry facilities, birds born on the same day are reared by the poultry house or by the cage row layer 90L. Therefore, environmental differences along the length of the poultry house are undesirable and must be adjusted to be uniform. Accurately understanding the environment is important for properly adjusting the environment within the poultry house.

[0047] In this embodiment, the environmental sensors 61 are attached to the moving devices 1, 2a, and 2b, which move in the direction (horizontal direction) along the length of the cage row 90R. Therefore, a single environmental sensor 61 can obtain environmental information for all cages belonging to that height of the cage row 90R. Therefore, compared to installing environmental sensors 61 at fixed locations, a smaller number of environmental sensors 61 can obtain a larger amount of environmental information, allowing for an accurate understanding of the bird rearing environment. The environmental sensors 61 send the acquired environmental information to the determination device along with cage information. Because the moving devices 1, 2a, and 2b are equipped with multiple types of environmental sensors 61 at each height of the cage row 90R, the environmental information is sent to the determination device along with cage information and sensor information identifying the type of environmental sensor 61.

[0048] The determination device stores the environmental information in association with the cage information and the sensor information in the storage device. The environmental abnormality determination means determines whether an abnormality has occurred based on the environmental information. The environmental abnormality determination means determines whether an abnormality has occurred, for example, by comparing the numerical values ​​constituting the environmental information with a predetermined numerical range.

[0049] When it is determined that an abnormality has occurred, the environmental abnormality determination means activates an alarm device to issue an audible and / or optical warning, and also adds abnormality detection information indicating that an abnormality has occurred to the environmental information at that time.

[0050] <Temperature Sensor> As the air flows longitudinally through the poultry house, it is heated by the body heat of the birds, so in cages belonging to a certain cage row 90R, the more downstream the cage, the higher the environmental temperature. Furthermore, because the heated air rises, the higher the cage in a certain cage row layer 90L, the higher the environmental temperature. Therefore, in large poultry houses, temperature differences between locations are very large, and making the temperature more uniform is a major challenge. In the past (Patent Document 1), temperature sensors were placed near auxiliary air inlets at the top of the side walls of the poultry house, and the number of measurement points was limited.

[0051] In contrast, in this embodiment, by using a temperature sensor as the environmental sensor 61 attached to the horizontally moving moving devices 1, 2a, and 2b, it is possible to continuously measure the temperature in the horizontal direction with a single temperature sensor. In addition, because multiple temperature sensors are attached to the moving devices 1, 2a, and 2b according to the height of each of the multiple cage rows 90R, the number of measurement points in the vertical direction is sufficient. Therefore, it is possible to accurately measure the temperature of each cage in which birds are raised, and to obtain temperature distribution within the poultry house that was not possible with conventional methods. This makes it possible to appropriately adjust the forced ventilation conditions and adjust the temperature throughout the poultry house to a temperature suitable for raising birds.

[0052] <Humidity Sensor> As the air flows longitudinally through the poultry house, the humidity increases due to the birds' breathing. Furthermore, during periods of high outside temperatures, such as summer, a cooling pad is placed at the air inlet of the poultry house. The cooling pad is a breathable pad, and by dripping water onto the pad while air passes through, the air is cooled by the heat of evaporation, and the cooled air is supplied to the poultry house. When such a cooling pad is used, the humidity environment within the poultry house becomes more complex. By using a humidity sensor as the environmental sensor 61 attached to the moving devices 1, 2a, and 2b, it is possible to obtain sufficient environmental information related to humidity both horizontally and vertically.

[0053] <Carbon dioxide gas sensor> As the air flows longitudinally through the poultry house, the concentration of carbon dioxide gas contained in the air increases due to the birds' breathing. Because carbon dioxide is heavier than air, the lower the cages in a certain cage row layer 90L, the higher the carbon dioxide gas concentration in the environment. The carbon dioxide gas concentration in the air is an indicator of whether ventilation and air mixing in the poultry house are being performed appropriately. By using a carbon dioxide gas sensor as the environmental sensor 61 attached to the movement devices 1, 2a, and 2b, environmental information regarding carbon dioxide gas concentration can be obtained sufficiently both horizontally and vertically.

[0054] <Ammonia Gas Sensor> In a typical poultry house, a conveyor is placed below each cage row 90R to receive the droppings excreted by the birds, and the conveyor is operated periodically to transport the droppings outside the poultry house. If droppings remain inside the poultry house, it can cause foul odors and the proliferation of pathogens. Because ammonia gas is generated from the droppings, the ammonia gas concentration in the air serves as an indicator of whether the droppings are being properly disposed of. By using an ammonia gas sensor as the environmental sensor 61 attached to the moving devices 1, 2a, and 2b, environmental information regarding the ammonia gas concentration can be obtained sufficiently in both the horizontal and vertical directions.

[0055] <Wind Speed ​​Sensor, Wind Direction Sensor> Because the forced ventilation described above generates a longitudinal air flow within the poultry house, information on wind speed and wind direction serves as an indicator of whether ventilation is being performed appropriately. Furthermore, a high wind speed has a significant effect of stirring the air within the poultry house, which can make the environment uniform, but can also cause stress to the birds. By using a wind speed sensor and a wind direction sensor as the environmental sensor 61 attached to the movement devices 1, 2a, and 2b, environmental information related to wind speed and wind direction can be obtained sufficiently both horizontally and vertically.

[0056] The wind speed and direction are also affected by the movement of the mobile devices 1, 2a, and 2b themselves, to which the environmental sensor 61 is attached. The wind speed and direction are also affected by the compressed air sent from the blower 50. To address this issue, the mobile devices 1, 2a, and 2b can be moved and the blower 50 operated in advance in a state where there is no air movement due to ventilation, and the wind speed and direction at that time are measured. This allows the wind speed and direction measured in a state where ventilation is occurring to be corrected. Alternatively, a method can be used in which the mobile devices 1, 2a, and 2b and the blower 50 are stopped for a short period of time when environmental information is acquired using the wind speed sensor and wind direction sensor.

[0057] <Illuminance Sensor> There is an optimum illuminance for raising chickens, which varies depending on the stage of development of the chickens. Illuminance also affects the egg productivity of the chickens. Therefore, adjusting the illuminance is important in the chicken rearing environment, but in large chicken coops, the number of lighting fixtures is enormous, making their management a challenge. By using an illuminance sensor as the environmental sensor 61 attached to the moving devices 1, 2a, and 2b, environmental information related to illuminance can be obtained sufficiently in both the horizontal and vertical directions.

[0058] Next, we will explain the judgments based on the images acquired by the camera 62. These judgments can include (1) determining whether the birds are alive or dead, (2) determining abnormalities in the appearance of living birds, (3) determining whether there is an excess or deficiency of feed, (4) determining whether eggs are remaining, and (5) determining whether the air is contaminated by dust.

[0059] <(1) Determining whether a bird is dead or alive> To determine whether an image contains a dead bird, only still images may be used, only moving images may be used, or both may be used. Methods for determining whether a bird is dead or alive include: (1-1) a method of image analysis of still images (using feature points), (1-2) a method of image analysis of still images (comparison of multiple still images), (1-3) a method of using a trained model generated by machine learning of still images, (1-4) a method of simultaneously capturing the same area with an infrared camera and a visible light camera, (1-5) a method of using a trained model generated by machine learning of moving images, (1-6) a method of extracting and comparing an image at a certain point in time and an image after a predetermined time has elapsed from images constituting a moving image, and (1-7) a method of making a determination based on moving images when the death of a bird is determined based on still images.

[0060] (1-1) Method of Image Analysis of Still Images (Use of Feature Points) Image analysis is used to extract feature points specific to dead birds or feature points specific to live birds from a still image, making it possible to determine whether a still image contains a dead bird. For example, during active periods such as when the birds are being fed, live birds are standing. Therefore, if an image is taken so that the bottom of the cage, i.e., the space between the feeding trough 85 and the egg collection conveyor, is included in the image, vertical lines that represent the outlines of the legs of live birds can be detected by edge extraction.

[0061] On the other hand, in the case of dead birds, the body is located at the bottom of the cage, so even if the image is analyzed, the vertical lines that form the outlines of the legs will not be detected. Therefore, by extracting the vertical outlines as feature points from the still image, determining the amount of these feature points, and comparing them with a predetermined threshold, it is possible to determine whether the still image contains a dead bird or not.

[0062] Furthermore, in images of only live birds, the bird's body is barely visible in the space below the feeding trough 85, whereas in images including dead birds, the bird's body is visible in the space below the feeding trough 85. The bird's body and legs are different colors. Therefore, by using the color of the bird's body or the color of its legs as a feature point and using the proportion of that color in the area below the feeding trough 85 in the still image as a feature amount, it is possible to determine whether the still image is alive or dead.

[0063] (1-2) Method for Image Analysis of Still Images (Comparison of Multiple Still Images) Because the moving devices 1, 2a, and 2b reciprocate in the direction of the cage row 90R, still images can be captured by the camera 62 during both forward and backward movements. The difference in pixel values ​​is calculated for each coordinate between a still image taken during forward movement and a still image taken during backward movement of the same cage. The difference in pixel values ​​for stationary parts in both images is zero, while differences in pixel values ​​occur for moving parts. Stationary parts include dead birds, as well as the wires, bars, frame, and feeding trough 85 that make up the cage. Because the outlines of the wires, bars, frame, and feeding trough 85 are linear, they can be distinguished by image analysis from dead birds, which have non-linear outlines. On the other hand, parts with differences in pixel values ​​can be determined to be live birds.

[0064] (1-3) Method using a trained model generated by machine learning of still images A trained model is generated by machine learning using a large number of still images containing only live birds (hereinafter referred to as "trainer still images of live birds") and a large number of still images containing dead birds (hereinafter referred to as "trainer still images containing dead birds") as training data. The generated trained model is stored in the storage device of the determination device. By inputting still images actually captured by camera 62 into the trained model, the probability that the image contains only live birds or the probability that the image contains dead birds is output. In the case of machine learning using a neural network, a trained model is generated using trainer images (trainer still images of live birds, trainer still images containing dead birds) with information on whether the birds are alive or dead as training data, and then trainer images without information on whether the birds are alive or dead are input into the trained model, and the weight coefficients between neurons are adjusted so that a probability appropriate for determining whether the birds are alive or dead is output.

[0065] It is desirable that the teacher images be captured under the same conditions as when the camera 62 actually captures images as the mobile devices 1, 2a, and 2b move. The numerous live bird teacher still images desirably consist of a variety of still images showing birds facing different directions and birds in different positions in the depth direction and left and right direction of the cage. It is desirable that the numerous dead bird-containing teacher still images desirably consist of a variety of still images showing birds that have fallen over and only their bodies are visible but not their legs, birds that have fallen over and their bodies and legs facing sideways are visible, birds with missing body parts, and birds with discolored feathers. It is desirable that each teacher image include images with eggs on the egg conveyor, images with different numbers of eggs, and images with no eggs.

[0066] In cases where the imaging areas of the two images are different, such as when the teacher image is an image of the inside of the cage, while the still image actually captured by camera 62 (hereinafter sometimes referred to as the "input still image") captures the outside of the cage, or when the teacher image is an image of only the lower part of the cage focusing on the bird's legs, while the input still image captures the entire cage, it is desirable to match the imaging area of ​​the teacher image to that of the input still image, make them the same size, and then input the input still image into the trained model. The process of matching the imaging areas can be performed, for example, by extracting the frame or feeding trough 85 through image analysis and using these as references.

[0067] By inputting a still image into the trained model and comparing the output value (probability) with a predetermined threshold, it is possible to determine whether the still image contains a dead bird or not.

[0068] (1-4) Method of simultaneously photographing the same photographing area with an infrared camera and a visible light camera In this method, a pair of cameras 62, a visible light camera and an infrared camera, are used to simultaneously photograph the same photographing object. The infrared camera detects infrared rays emitted from the object, converts them into temperature, and visualizes the temperature of the subject by imaging different temperatures as different colors. The live / dead determination means acquires the simultaneously photographed visible light image and infrared image, along with cage information for the photographing object. The live / dead determination means extracts the portion of the image containing the bird by image analysis of the visible light image and identifies that portion by coordinates. The portion containing the bird can be extracted based on a color specific to the bird, such as the color of its feathers.

[0069] The live / dead determination means detects the color of pixels in the infrared image at the same coordinates as those identified as the area in the visible light image where the bird is located. Since dead birds have a lower body temperature than live birds, if there are pixels with a color indicating a low temperature within a predetermined range in the area in the infrared image where the bird should be located, the bird can be determined to be dead. Therefore, by setting a threshold value that is lower than the general body temperature of a live bird, it is possible to determine whether a dead bird is included in a still image.

[0070] (1-5) Method using a trained model generated by machine learning of video images A large number of videos containing only surviving birds (hereinafter referred to as "trainer videos of surviving birds") and a large number of videos containing dead birds (hereinafter referred to as "trainer videos containing dead birds") are used as training data to generate a trained model for the video images by machine learning. The generated trained model is stored in the storage device of the determination device.

[0071] By inputting the video captured by the camera 62 into the trained model, the probability that the video contains only live birds or the probability that the video contains dead birds is output. For machine learning, for example, a convolutional neural network, which excels in image recognition, can be combined with a recurrent neural network, which is an expanded neural network that can handle time-series data.

[0072] In the case of machine learning using these neural networks, a trained model is generated using training data consisting of teacher images with information about whether the bird is alive or dead (trainer images of live birds, teacher images containing dead birds), and then teacher images without information about whether the bird is alive or dead are input into the trained model, while adjusting the weighting coefficients between neurons so that a probability appropriate for determining whether the bird is alive or dead is output.

[0073] It is desirable that the teacher moving images be captured under the same conditions as those under which the camera 62 actually captures moving images within the facility. As described above, it is desirable that the teacher moving images consist of a variety of moving images, just as in the case where the teacher images are still images.

[0074] By inputting video images actually captured by the camera 62 into the trained model and comparing the output value (probability) with a predetermined threshold, it is possible to determine whether or not the video images contain dead birds.

[0075] (1-6) A method for extracting and comparing an image at a certain time point and an image after a predetermined time has elapsed from images constituting a moving image. From the consecutive images constituting the moving image, a still image at a certain time point T and a still image at a certain time point (T + ΔT) after a predetermined time has elapsed are extracted, and a still image at a time point (T + 2ΔT) after the same amount of time has elapsed is also extracted, thus extracting multiple still images at time intervals of ΔT. The pixel value difference is then calculated for each identical coordinate between the previous and next images in the time series. Stationary portions of the moving image have zero pixel value differences, while moving portions have pixel value differences. Stationary portions include dead birds, as well as the wires, bars, frame, and feeding trough 85 that make up the cage. These have linear contours, so they can be distinguished by image analysis from dead birds, which have non-linear contours. On the other hand, portions with pixel value differences can be determined to be live birds.

[0076] (1-7) A method for making a judgment based on a moving image when the death of a bird has been judged based on a still image Judgment based on a still image may not correctly judge whether a bird is alive or dead. For example, in the following cases, a sitting (crouching) bird may be judged to be dead. - In judgment (1-1) using image analysis of a still image, when the bird's legs are used as feature points. - In judgment (1-3) using a trained model generated by machine learning of a still image, when machine learning is performed using training data as training data a teacher image that focuses on whether the image contains a bird's legs.

[0077] Furthermore, the rate at which body temperature drops after death is not large. Therefore, if the threshold is set low in the judgment using an infrared camera (1-4), only birds that have been dead for a long time can be detected. If the threshold is set high in an attempt to detect birds that have died recently, there is a risk that surviving birds will be judged as dead, due to individual differences in body temperature among birds.

[0078] Therefore, when it is determined by any of methods (1-1) to (1-4) that a dead bird is included in a still image, a determination based on a moving image (1-5) or (1-6) can be made. In this case, when it is determined that a dead bird is included in a still image, a signal is transmitted from the determination device to the imaging control means of the control device. This signal includes cage information from which the still image determined to include a dead bird was acquired. The imaging control means identifies the cage to be photographed based on this cage information and controls the camera 62 to capture a moving image for a predetermined period of time (e.g., 5 to 10 seconds). The moving image, with the cage information attached, is transmitted to the determination device, where the above determination is made. In this way, by combining determination based on a still image and determination based on a moving image, it is possible to more accurately determine whether a dead bird is included in an image.

[0079] When it is determined by any of methods (1-1) to (1-7) that the image contains a dead bird, the live / dead determination means stores dead bird detection information in association with the image at that time, and activates an alarm device to issue an audible and / or optical alarm. Since the image associated with the dead bird detection information is associated with cage information, a worker who receives the alarm can quickly take appropriate action, such as removing the dead bird from the cage.

[0080] <(2) Determining Abnormalities in the Appearance of Surviving Birds> Birds suffering from the heat have their mouths open. The color and distribution of the color around the beak differs when the mouth is open and when it is closed. Sick birds may have a different color comb than healthy birds. When birds in the same cage are injured by pecking at each other, bleeding can be seen in the feathers. The occurrence of such abnormalities in birds that appear in appearance is determined based on images. The determination device stores a program that causes a computer to function as an appearance abnormality determination means. Methods for determination include image analysis based on color and its distribution position, and a method that uses a large number of images of healthy birds and birds with abnormal appearances as training images and uses a trained model generated by machine learning.

[0081] When the visual abnormality determining means determines that an abnormality has occurred in the bird, it stores visual abnormality information in association with the image at that time and activates an alarm device.

[0082] (3) Determining whether or not there is sufficient or insufficient food The drawings illustrate an example in which the camera 62 is positioned so that it can photograph the bottom of the cage (see FIGS. 1 and 4). In addition to the camera 62 positioned in this way, another camera 62 capable of photographing the interior of the feeding trough 85 is attached to the moving device 1, 2a, or 2b. Such a camera 62 is positioned so that it photographs the interior of the feeding trough 85 from above at each height of the multiple cage rows 90R. The storage device of the determination device stores a program that causes the computer to function as a feed amount determination unit. The color of the feed is different from the color of the inner surface of the feeding trough 85, and the feeding trough 85 is shaped to widen upward. Therefore, the amount of feed per unit length of the feeding trough 85 can be determined by analyzing the image captured from above.

[0083] The feed amount determination means quantifies the grasped feed amount and compares it with a predetermined numerical range, which is an allowable range. If the comparison results in a determination that the feed amount is either too much or too little, the feed amount determination means stores abnormal feed amount information in association with the image at that time and activates an alarm device.

[0084] <(4) Determining whether eggs are retained> In chicken coops, eggs are collected by operating an egg collection conveyor at predetermined intervals. If eggs are retained on the egg collection conveyor, they will not be able to receive newly released eggs from the cage floor 91 and will fall. Also, eggs may be unable to be released due to some obstacle, such as the presence of a dead bird, and will end up retained in the cage. To determine whether eggs are retained in this way, a program that causes a computer to function as egg retention determination means is stored in the storage device of the determination device. Because the outline of an egg has no straight lines and is clearly different from the outline of a bird, eggs can be extracted from an image by image analysis.

[0085] The egg retention determination means determines whether the eggs extracted from the image are on a conveyor or in a cage based on their coordinates, and quantifies the proportion of eggs occupying a predetermined area in the image. When the means determines that eggs are retained as a result of comparing the quantification with a predetermined allowable range, the egg retention determination means stores the egg retention information in association with the image at that time, and activates an alarm device.

[0086] (5) Determining Air Contamination by Dust In the forced ventilation described above, the air flows along the longitudinal direction of the poultry house, and the air is fresher the further upstream, while the more dust there is in the air the further downstream. Naturally, dust floating in the air has a negative impact on the health of the birds. Therefore, the level of air contamination by dust is detected based on images acquired by the camera 62. The storage device of the determination device stores a program that causes the computer to function as dust contamination determination means. Methods for detecting the level of contamination include an image analysis method that compares images acquired at different levels of air contamination by dust with images actually acquired by the camera 62 at the site, and a method that uses a trained model generated by machine learning using a large number of images acquired at different levels of air contamination by dust as training images.

[0087] The dust contamination determination means quantifies the detected level of contamination and compares it with a predetermined numerical tolerance range. When the level of contamination exceeds the tolerance range, the dust contamination determination means determines that dust contamination has occurred, associates dust contamination detection information with the image at that time, stores it, and activates an alarm device.

[0088] Next, we will explain the process of combining the environmental information acquired by the environmental sensor 61 with the images acquired by the camera 62. This process can include (21) examining the correlation between bird deaths and the environmental information, and (22) associating the environmental information with the images when it is determined that an abnormality has occurred based on the environmental information.

[0089] <(21) Examination of the correlation between bird deaths and environmental information> By combining information obtained for a certain type of environmental information in a certain cage row layer 90L, a distribution map of environmental information such as temperature distribution, humidity distribution, and carbon dioxide concentration distribution can be obtained. Therefore, a program that causes a computer to function as a means for generating such a distribution map is stored in the storage device of the determination device. When it is determined that an image contains a dead bird, the location of the cage in which the dead bird is located can be determined from the cage information associated with the image.

[0090] This makes it possible to examine whether there is a correlation between the distribution of environmental information and the location of the cage in which the bird died. For example, if the cage in which the bird died is included in an area where the temperature is locally high, but other environmental information in that area is normal, it can be inferred that high temperature was the cause of the bird's death. Alternatively, if the cage in which the bird died is included in an area where the carbon dioxide concentration is locally high, but other environmental information in that area is normal, it can be inferred that carbon dioxide concentration was the cause of the bird's death.

[0091] <(22) Processing for Associating Related Information with Images When It is Determined That an Anomaly Has Occurred Based on Environmental Information> As described above, when it is determined that an abnormality has occurred based on environmental information, the environmental abnormality determination means adds abnormality detection information to the environmental information at that time. The environmental information is associated with cage information that identifies the cage from which the information was obtained. Therefore, by extracting images associated with the same cage information, the environmental information at the time the abnormality occurred is associated with the image of the cage in which the environmental abnormality occurred, and a new database is generated.

[0092] Therefore, a program that causes a computer to function as a means for generating a database that associates environmental information with images is stored in the storage means of the determination device. In this association, environmental information and images can be associated one-to-one. Alternatively, images of multiple cages that are adjacent in photographing order to the cage in which the environmental abnormality occurred can be combined with the image of the cage in which the environmental abnormality occurred, and associated with the environmental information at the time the abnormality occurred. In this case, images from the time when the environmental abnormality occurred, including the time before and after that time, are combined with the environmental information.

[0093] The data in this database makes it possible to check, through images, the state of the birds when an environmental abnormality occurs. For example, if an environmental abnormality occurs and the bird appears to be in pain, even if it does not die, it is clear that the type of environmental information (temperature, humidity, carbon dioxide gas, etc.) is affecting the bird's health. Alternatively, if an environmental abnormality is determined to have occurred but the bird remains in a normal state, it is possible that the numerical ranges previously set as criteria for determining an abnormality are inappropriate.

[0094] As described above, according to the monitoring method of this embodiment, because the environmental sensors 61 and cameras 62 are attached to the moving device, it is possible to acquire both environmental information and images by simply moving one moving device along the cage row layer 90L. Furthermore, because the environmental sensors 61 and cameras 62 are both attached at each height of the multiple cage rows 90R, it is possible to acquire environmental information and images for all of the cages belonging to the cage row layer 90L by moving the moving device in the horizontal direction. In other words, it is possible to acquire a large amount of environmental information and images with a small number of environmental sensors 61 and cameras 62.

[0095] Because the slide drive mechanism including the motor 45 is supported by the cage row layer 90L, which is a fixed structure within the poultry house, power can be supplied to the moving devices 1, 2a, and 2b via a wire. The environmental sensor 61 and camera 62 are attached to the moving devices 1, 2a, and 2b, and the control device is also attached to the moving devices 1, 2a, and 2b. Therefore, power can be supplied to the environmental sensor 61, camera 62, and control device via a wire via the moving devices 1, 2a, and 2b. Therefore, unlike when a battery is used as the power source, the moving devices 1, 2a, and 2b can be reciprocated at a high frequency, allowing the environmental sensor 61 to acquire a large amount of environmental information and the camera 62 to acquire a large number of images.

[0096] The environmental information and images are stored in association with the cage information, so that when an abnormality occurs in the environmental information or when an abnormality is detected from the images, the cage in which the abnormality occurred can be identified and a prompt response can be taken.

[0097] Since multiple types of environmental sensors 61 are attached to a single mobile device, multiple types of environmental information can be acquired at once as the mobile device moves. In addition, in this embodiment, the judgments made based on images acquired by camera 62 are not limited to determining whether the birds are alive or dead, but also include a variety of judgments such as determining abnormalities in the appearance of surviving birds, determining whether there is an excess or deficiency of feed, determining whether eggs are remaining, and determining whether the air is polluted by dust, and these judgments are made simultaneously as the mobile devices 1, 2a, and 2b move.

[0098] Furthermore, since environmental information and images are acquired simultaneously for the same cage, it is possible to examine whether there is a correlation between abnormalities in the environmental information and abnormalities identified from the images.

[0099] The moving devices 1, 2a, 2b are equipped with sliders 11, 12 that are driven by a slide drive mechanism and move using the opening edge of the feed trough 85 as a rail. Therefore, there is no need to install new rails in the cage row layer 90L to move the moving devices 1, 2a, 2b, and the moving devices 1, 2a, 2b can be moved using the existing equipment, the feed trough 85, which is always included in the cage row layer 90L.

[0100] In the above, the case where the environmental sensor 61 and the camera 62 are attached to the moving devices 1, 2a, 2b has been described, but a configuration in which a microphone (not shown) is also attached may be adopted. A highly directional microphone may be attached at each height of the cage row 90R, or one microphone may be attached to each moving device 1, 2a, 2b. A program that causes a computer to function as abnormal sound detection means is stored in the storage device of the determination device.

[0101] The sound information collected by the microphone is sent to a determination device associated with the cage information. When a microphone is attached at each height of the cage row 90R, the cage information and the sound information are associated in a one-to-one relationship. When there is one microphone per moving device 1, 2a, 2b, the cage information of multiple cages (cages stacked vertically) that are located at the same horizontal position is associated with the sound information.

[0102] Birds that are in poor health may make strange noises. Therefore, the abnormal sound detection means determines whether the sound information is abnormal by comparing the frequency (pitch) and amplitude (loudness) of the sound information with a predetermined numerical range. When the sound information is determined to be abnormal, the abnormal sound detection means extracts an image associated with the same cage information as the cage information associated with the sound information. The image may be a still image or a moving image. This associates the sound information at the time of the abnormality with an image of the cage in which the sound abnormality occurred (and a cage located horizontally). At this time, images of multiple cages adjacent in the photographing order to the cage in which the sound abnormality occurred (and a cage located horizontally) may be combined with the sound information. In this case, images from the time when the sound abnormality occurred, including the time before and after, are combined with the sound information.

[0103] From the image associated with the sound information, the bird that made the strange noise can be identified and its condition can be confirmed.

[0104] The present invention has been described above by citing preferred embodiments, but the present invention is not limited to the above embodiments, and various improvements and design changes are possible within the scope of the gist of the present invention.

[0105] For example, in the above example, the number of stages of the cage rows 90R in the cage row layer 90L is four, but the number of stages is not limited to this.

[0106] In the above example, the determination device is configured as a computer separate from the control device, but this is not limiting. A control device attached to the mobile device may also function as the determination device. Alternatively, a determination device separated from the mobile device may also function as the control device and communicate with the mobile device wirelessly.

[0107] The moving device can also be configured to function as a feeding device. In this case, multiple hoppers (not shown) for supplying feed to the feeding troughs 85 are attached to the moving body 10. The number of hoppers is equal to the number of cage rows 90R in the cage row layer 90L, i.e., the number of feeding troughs 85. The height of each hopper is adjusted so that feed can be supplied to the feeding troughs 85 from above. The multiple hoppers are connected by ducts 52 extending vertically. When feed is introduced from the feed supply device into the opening of the topmost hopper, the feed is supplied to the feeding troughs 85 via each hopper.

[0108] This configuration has the advantage that a single device can both acquire environmental information and images of the cages and supply food to the feed trough 85. However, providing multiple hoppers to the moving device increases the weight of the moving device. Therefore, moving the moving device along the opening edge of the feed trough 85 may place a heavy burden on the feed trough 85. Therefore, in this case, it is desirable to provide a rail on the upper part of the cage row layer 90L that engages with the slider of the moving device.

Claims

1. A method for monitoring a poultry house in which chickens are raised using cage row layers in which cage rows, each of which has a plurality of cages arranged side by side horizontally, are stacked in multiple tiers, the method comprising: using a moving device to which an environmental sensor and a camera are attached, the moving device reciprocating along the front of the cage row layer in the direction in which the cage rows extend; one or more types of environmental sensors are attached to the moving device according to the height of each of the multiple tiers of cage rows; the cameras capture still or moving images as images, and one or more cameras are attached to the moving device according to the height of each of the multiple tiers of cage rows; as the moving device moves, at least one piece of environmental information is obtained for each of the cages by the environmental sensor, and at least one image is obtained for each of the cages by the camera; and the environmental information and the images are each associated with cage information that identifies the cage and stored in a storage device.

2. The chicken coop monitoring method described in claim 1, characterized in that a determination device is used that determines the occurrence of an abnormality based on at least one of the environmental information and the image, and when it is determined that an abnormality has occurred, the environmental information and the image associated with the same cage information are combined and stored in a storage device.

3. The method for monitoring a poultry house described in claim 1, characterized in that the cameras used are cameras that capture images of the inside of the cages belonging to each of the cage rows, as well as cameras that capture images of the inside of feeding troughs arranged along the front of each of the cage rows.

4. A method for monitoring a poultry house as described in claim 1, characterized in that a microphone is attached to the moving device, and the sounds made by the birds are collected by the microphone.

5. The method for monitoring a poultry house described in claim 1, characterized in that the moving device comprises a slider that moves along the opening edge of at least one of the feeding troughs arranged along the front of each of the cage rows.

Images