JP7817136B2 - Feeding control device and feeding control method - Google Patents

Description

This disclosure relates to a feeding control device and a feeding control method for controlling feeding of aquatic organisms.

Feeding devices for feeding farmed aquatic animals are known.

Patent Document 1 discloses an aquatic animal farming support system. The aquatic animal farming support system includes an image acquisition unit that acquires still or moving images by using a camera to capture images of a collector lifted from an aquaculture pond. The aquatic animal farming support system also includes an image analysis unit that analyzes the acquired images, and an aquatic animal information acquisition unit that acquires aquatic animal information about the aquatic animals in the aquaculture pond based on the analysis results of the image analysis unit.

Patent Document 2 discloses a feeding system. The feeding system includes a feeding device that feeds farmed fish, a camera, and an image processing device that analyzes the movement of images shown in image data acquired by the camera and acquires information about the movement of the farmed fish. After the feeding device begins feeding, the camera begins acquiring image data.

International Publication No. 2020/250330 International Publication No. 2018/042651

The aquatic animal farming support system of Patent Document 1 requires the collector to be raised and lowered from the farm pond. Furthermore, in the feeding system of Patent Document 2, the camera begins acquiring image data after the feeder begins feeding. Therefore, both the aquatic animal farming support system of Patent Document 1 and the feeding system of Patent Document 2 leave room for more appropriate control of the feeding of aquatic organisms.

One aspect of the present disclosure aims to realize a feeding control device and feeding control method that appropriately controls feeding of aquatic organisms.

To solve the above problem, a feeding control device according to one aspect of the present disclosure includes a first calculation unit that calculates statistics indicating characteristics of the distribution of pixel values within each of multiple divided regions in an image of an aquarium containing aquatic organisms captured by a first imaging device; a determination unit that determines whether the aquatic organisms are feeding based on the statistics calculated by the first calculation unit; and an operation control unit that controls the operation of a feeding device that supplies food to the aquarium based on the determination result of the determination unit.

To solve the above problem, a feeding control method according to one aspect of the present disclosure includes a first calculation step of calculating, for each of a plurality of divided regions in an image captured by a first imaging device of an aquarium containing aquatic organisms, a statistical quantity indicating characteristics of the distribution of pixel values within the divided region; a determination step of determining whether the aquatic organisms are feeding based on the statistical quantity calculated in the first calculation step; and an operation control step of controlling the operation of a feeding device that supplies food to the aquarium based on the determination result in the determination step.

One aspect of the present disclosure provides a feeding control device and a feeding control method that appropriately control feeding of aquatic organisms.

FIG. 1 is a diagram illustrating an example of a schematic configuration of an aquaculture system according to the present disclosure. FIG. 1 is a schematic diagram illustrating an example of an overflow tank according to the present disclosure. 1 is a diagram illustrating an example of a schematic configuration of a feeding control device according to an embodiment of the present disclosure. FIG. 1 is a diagram showing an example of the swimming state of a fish (chub mackerel) captured by a breeding tank camera. FIG. 10 is a diagram for explaining a method for calculating a histogram of variance values in a swimming state while feeding. FIG. 10 is a diagram for explaining a method for calculating a histogram of variance values in a swimming state while feeding. FIG. 10 is a diagram for explaining a method for calculating a first index value and a second index value. FIG. 10 is a diagram for explaining a method for determining whether or not a fish is feeding. FIG. 10 is a flow chart showing an example of operational control of the feeding device by the operation control unit. FIG. 10 is a flow diagram showing an example in which a determination unit determines whether or not a fish is feeding based on both a statistical amount and the amount of remaining food counted by a remaining food amount determination unit. 10 is a flow diagram showing an example in which a determination unit determines whether or not a fish is currently feeding based on the amount of remaining food counted by a remaining food amount determination unit. FIG. 10 is a flow diagram showing an example in which a determination unit determines whether or not a fish is currently feeding based on the amount of remaining food counted by a remaining food amount determination unit. FIG. 10 is a flow diagram showing an example in which a determination unit determines whether or not a fish is currently feeding based on the amount of remaining food counted by a remaining food amount determination unit. FIG. FIG. 10 is a flow chart showing an example of the operation of the adjusting unit to adjust the amount of feed per unit time based on the amount of remaining feed. FIG. 10 is a flow chart for explaining a method for determining whether or not a fish is feeding, taking into consideration the swimming speed of the fish. This is a diagram showing the movement of a fish between two frames expressed as a two-dimensional vector using optical flow. This is a graph comparing the dispersion ratio and the swimming speed of fish at the same time. 1 is a flow chart illustrating an example of a feeding control method according to the present disclosure.

One embodiment of the present disclosure will be described in detail below, with reference to the schematic diagram shown in Figure 1.

First, we will explain the aquaculture system 1 according to the present disclosure. The aquaculture system 1 can be applied to both land-based and marine aquaculture, but the following explanation will be given assuming that it is applied to land-based aquaculture.

(Land-based aquaculture)
Land-based aquaculture is the cultivation of aquatic organisms in an artificially created environment on land. Land-based aquaculture is mainly divided into two types: "free-flowing land-based aquaculture" and "closed-circulation land-based aquaculture." Free-flowing land-based aquaculture is a method in which seawater, etc. is continuously drawn in from the natural environment and used as rearing water. Closed-circulation land-based aquaculture is a method in which rearing water is purified using a filtration system and reused in a circulating manner. Closed-circulation land-based aquaculture is a method with a low environmental impact, as rearing water is not generally discharged, and the market is expected to expand in the future.

In conventional land-based aquaculture, a set amount of feed is supplied to the breeding tank at a set time. Setting the feeding time and amount was based on the breeder's visual inspection and/or experience, placing a heavy burden on the breeder. Furthermore, because the amount of feed varies depending on the fish species and/or the fish's growth process, it was not easy to properly control feeding.

The aquaculture system 1 according to the present disclosure is equipped with the following configuration to appropriately control the feeding of aquatic organisms.

(Aquaculture system according to the present disclosure)
The aquaculture system 1 according to the present disclosure will be described below. The aquaculture system 1 can be applied to both free-flowing land-based aquaculture and closed-circulation land-based aquaculture, but the following description will be applied to closed-circulation land-based aquaculture. The aquaculture system 1 can be applied to the rearing of aquatic organisms, particularly seafood excluding shellfish, and examples of the aquaculture targets include mackerel, pufferfish, flounder, shrimp, octopus, squid, catfish, and eels.

First, to give an overview, the aquaculture system 1 according to the present disclosure performs image analysis to determine whether the farmed fish are in a feeding state or a normal swimming state (not feeding), and controls the operation of the feeding device based on the results to appropriately control the feeding of the fish. The aquaculture system 1 will be described in detail below.

The configuration of the aquaculture system 1 will be explained using Figure 1. Figure 1 is a diagram showing an example of the general configuration of the aquaculture system 1.

The aquaculture system 1 mainly comprises a breeding tank 2, an overflow tank 3, a feeding device 4, a breeding tank camera 5 (first imaging device), a remaining feed camera 6 (second imaging device), and a feeding control device 10.

The breeding tank 2 contains the fish to be cultivated and water suitable for the fish (freshwater, seawater, or brackish water). The breeding tank 2 is set up in an environment appropriate for the type of fish to be cultivated or the growth stage of the fish. The breeding tank 2 may be installed indoors or outdoors. The breeding tank 2 is not limited to a specific size or shape. The breeding tank 2 includes tanks and associated equipment used in conventional land-based aquaculture.

The breeding tank 2 may have a mechanism for overflowing to remove floating matter on the water surface, as well as leftover food and feces from the bottom. This reduces the buildup of dirt on the walls of the breeding tank 2 and keeps them clean. Overflowing water is discharged from the breeding tank 2 into the overflow tank 3.

The overflow tank 3 is installed in the drainage flow path leading from the breeding tank 2 to the downstream side. More specifically, the overflow tank 3 may be installed between the breeding tank 2 and a filter tank (not shown) for filtering the drainage water. Water may be transported by gravity from the breeding tank 2 to the downstream filter tank. If the breeding tank 2 and the filter tank were directly connected without the overflow tank 3, the difference in elevation between the breeding tank 2 and the filter tank could cause all of the water in the breeding tank 2 to leak into the filter tank. The overflow tank 3 is installed between the breeding tank 2 and the filter tank to prevent this from happening. The overflow tank 3 can automatically adjust the operating water level of the breeding tank 2 depending on its installation height.

The overflow tank 3 will be described in more detail with reference to Figure 2. Figure 2 is a schematic diagram showing an example of an overflow tank 3 according to the present disclosure.

Referring to the diagram indicated by reference number 200, the overflow tank 3 is installed in the drainage flow path leading downstream from the breeding tank 2, and collects floating matter on the surface of the water in the breeding tank 2, as well as leftover food and feces at the bottom. The overflow tank 3 is connected to the breeding tank 2 via the first pipe 3a. The overflow tank 3 may be a small tank with a simple structure.

The first pipe 3a sends breeding water containing floating matter on the surface of the breeding tank 2, as well as leftover food and feces at the bottom, to the overflow tank 3. The first pipe 3a may be installed in an appropriate position, such as above or below the overflow tank 3. The first pipe 3a does not have to be directly connected to the overflow tank 3, and the pipe outlet may be positioned above the overflow tank 3, as shown by reference number 200 in Figure 2. There may be one or more first pipes 3a.

The breeding water sent from the breeding tank 2 to the overflow tank 3 is sent to the filter tank via the second pipe 3b.

Referring to Figure 2, which shows an example of the overflow tank 3, a residual food collector 7 is placed on top of the overflow tank 3 and collects residual food contained in the breeding water discharged from the first pipe 3a. The residual food collector 7 has a structure that allows water to pass through and can capture residual food. The residual food collector 7 may be, for example, a net or a colander. The mesh size of the mesh is smaller than the residual food. This allows the residual food collector 7 to capture residual food in the breeding tank 2 that is sent from the breeding tank 2.

The residual food collector 7 only needs to be able to capture residual food in the breeding tank 2, and may be installed not above the overflow tank 3, but in the drainage flow path leading downstream from the breeding tank 2, or in a filter tank located downstream of the overflow tank 3. The residual food collector 7 does not need to collect all residual food; it may only collect a portion of the residual food.

The aquaculture system 1 may have a valve for controlling the flow rate of the breeding water discharged from the breeding tank 2. The valve can adjust the amount of breeding water discharged so that the entire amount of breeding water in the breeding tank 2 does not flow out. In this case, the aquaculture system 1 does not need to have an overflow tank 3, and the residual feed collector 7 only needs to be installed in the drainage flow path downstream of the breeding tank 2.

The overflow tank 3 shown in Figure 2 is an example and is not limited to any particular structure or specifications.

The feeder 4 (feeder) is a device that feeds the fish cultivated in the breeding tank 2. The feed may be live feed (raw fish fillets), moist pellets (semi-raw solid type), or dry pellets (dry solid type). In the following description, the feed is assumed to be in pellet form, which has a higher specific gravity than water. Because the feed has a higher specific gravity than water, feed not eaten by the fish (residual feed) sinks to the bottom of the breeding tank 2 and is sent to the overflow tank 3 via the first pipe 3a. Note that the feed may also have a lower specific gravity than water. In this case, the residual feed is sent to the overflow tank 3 via the first pipe 3a and collected by the residual feed collector 7.

The feeding device 4 is installed above the breeding aquarium 2 and supplies food to the breeding aquarium 2 at feeding times. The feeding device 4 may start and stop feeding at preset times. The feeding device 4 may start and stop feeding and/or adjust the amount of food fed by command from the feeding control device 10 (described below) or manual operation by a zookeeper or the like. The feeding device 4 may be a conventional feeding device and may include, for example, a feed tank, a feed measuring unit, a feeding mechanism, a drive unit, a control unit, and/or a communication unit.

The breeding tank camera 5 captures images of the breeding tank 2 and transmits the captured images (image data) to the feeding control device 10. The captured image shows a planar view of the water surface of the breeding tank 2, and includes an image of area T where food is supplied from the feeding device 4. The breeding tank camera 5 can also capture images of an area in the breeding tank 2 different from area T. Area T may be the entire area of the breeding tank 2. The breeding tank camera 5 and the feeding control device 10 are connected via wire and/or wirelessly. The breeding tank camera 5 can capture both video and still images.

The breeding tank camera 5 may be a device capable of capturing images of objects in both the visible light and infrared light ranges. Some land-based aquaculture facilities are operated with lights turned off at night. Therefore, the breeding tank camera 5 may capture visible images during the day and infrared images at night. Image processing (described below) in the feeding control device 10 may be performed without distinguishing between visible and infrared images.

The remaining feed camera 6 captures images of the remaining feed collector 7, which collects remaining feed. As described above, the remaining feed collector 7 is installed, for example, above the overflow tank 3, in downstream piping, or in a filter tank. The remaining feed camera 6 captures images of the remaining feed collector 7 and transmits the captured images (image data) of the remaining feed to the feeding control device 10. The remaining feed camera 6 and the feeding control device 10 are connected via wire and/or wirelessly. The remaining feed camera 6 can capture both video and still images.

The remaining bait camera 6 may be a device capable of capturing images of objects in both the visible light and infrared light regions. The remaining bait camera 6 may capture visible images during the day and infrared images at night. Image processing (described below) in the remaining bait camera 6 may be performed without distinguishing between visible and infrared images.

(Feeding control device)
The feeding control device 10 determines whether the fish are feeding or not based on (1), (2), and/or (3) below, and controls the operation of the feeding device 4 that feeds the breeding tank 2 based on the determination result.

(1) A statistical quantity that indicates the characteristics of the distribution of pixel values within each of multiple divided regions in the image of the breeding aquarium 2 captured by the breeding tank camera 5.

(2) The amount of remaining food determined based on the image of the remaining food captured by the remaining food camera 6 in the remaining food collector 7.

(3) The amount of fish movement calculated by performing video processing on multiple images of the breeding tank 2 captured by the breeding tank camera 5.

The configuration of the feeding control device 10 will be explained below with reference to Figure 3. Figure 3 is a diagram showing an example of the general configuration of the feeding control device 10 according to one embodiment of the present disclosure.

As shown in FIG. 3, the feeding control device 10 includes a communication unit 20 and a control unit 30.

The feeding control device 10 may be equipped with a recording unit 40 such as ROM (Read Only Memory), RAM (Random Access Memory), or an external storage device. The recording unit 40 may store predetermined application programs executed by the control unit 30. The recording unit 40 may record a feeding schedule, including the setting of the amount of food to be fed per hour during feeding operations. The recording unit 40 may record images (still images and/or still images) captured by the breeding tank camera 5 and the remaining food camera 6, which will be described later. The recording unit 40 may record the results of judgment by the judgment unit 32, which will be described later, various thresholds used by the judgment unit 32 for judgment, etc. In this way, the recording unit 40 may record any data and/or information related to the operation of the feeding control device 10.

The feeding control device 10 may include an input unit 50 that accepts various user operations, such as a keyboard, mouse, or voice input. The feeding control device 10 may also include a display unit 60 that displays text, images, or abnormality alerts from the abnormality alert unit 37 (described below).

The communication unit 20 is configured to be able to communicate with at least the aquarium camera 5, the remaining bait camera 6, and the feeding device 4. This communication may be wireless or wired. In the case of wireless communication, communication using radio waves or infrared rays may be used. In the case of communication using radio waves, Bluetooth (registered trademark) and Wi-Fi (registered trademark) may be used.

The communications unit 20 can transmit various information related to the operation of the aquaculture system 1 to the manager of the aquaculture system 1 via the internet, the cloud, etc. Examples of such information include abnormality alerts from the abnormality alert unit 37 (described below), images captured by the breeding tank camera 5 and/or the remaining feed camera 6, and the feeding status (feeding schedule, feed amount, etc.) of the feeding device 4.

The control unit 30 controls the execution of the processing of each function provided by the feeding control device 10, and includes a variance value calculation unit 31, a judgment unit 32, an operation control unit 33, a remaining food amount identification unit 34, an adjustment unit 35, an activity amount calculation unit 36, and an abnormality notification unit 37. Each configuration is described below.

(Calculation of statistics)
The variance value calculation unit 31 acquires an image of a breeding tank 2 containing aquatic organisms (fish) captured by the breeding tank camera 5 from the breeding tank camera 5 via the communication unit 20. Then, the variance value calculation unit 31 calculates a statistic indicating the dispersion of pixel values within each of a plurality of divided regions in the captured image. This will be specifically described below using FIG. 4 etc.

Note that, in the following description, the calculated statistical quantity is described as a statistical quantity (variance or standard deviation) indicating the dispersion of pixel values. However, other statistical quantities indicating the characteristics of the distribution of pixel values within each of multiple divided regions in the captured image may also be calculated. Statistical quantities indicating the characteristics of the distribution of pixel values may include, for example, the variance or standard deviation, which indicate the dispersion of pixel values, as well as the mean, median, mode, kurtosis, or skewness of pixel values. Depending on the type of aquatic organism or the type of aquarium camera 5, statistical values such as the mean, median, mode, kurtosis, or skewness of pixel values can be used to determine whether the fish are feeding.

Figure 4 shows an example of the swimming behavior of a fish (chub mackerel) captured by the aquarium camera 5.

The diagram indicated by reference number 400 shows the swimming state of fish under normal circumstances (not feeding). The diagram indicated by reference number 410 shows the swimming state of fish while feeding. Comparing the two, the swimming state of fish under normal circumstances and while feeding is different. When swimming under normal circumstances, the impact of the school of fish on the water surface is small. When swimming while feeding, the fish bite the bait, causing the water surface to become rough.

By focusing on the difference in the swimming behavior of fish during normal times and when they are feeding, the variance calculation unit 31 calculates statistics as follows. This will be explained with reference to Figures 5 and 6. Note that in the following explanation, the statistics will be explained as variances.

Figure 5 is a diagram illustrating a method for calculating a histogram of variance values during normal swimming.

As shown in the diagram indicated by reference number 500, the variance calculation unit 31 divides a captured image showing the normal swimming state of fish into blocks each measuring 10px x 10px. Next, the variance calculation unit 31 calculates the variance of the pixel values within each block for each block. The pixel values are expressed in 256 levels from 0 to 255.

Next, referring to the diagram indicated by reference number 510, the variance value calculation unit 31 creates a histogram. In the histogram, the horizontal axis represents the variance value and the vertical axis represents the frequency. As an example, the variance value calculation unit 31 creates a histogram in which the variance value on the horizontal axis is in increments of 20 (i.e., the class width is 20).

Figure 6 is a diagram illustrating a method for calculating a histogram of variance values during swimming while feeding.

Referring to the diagram indicated by reference number 600, the variance calculation unit 31 divides a captured image showing the swimming state of a feeding fish into blocks each measuring 10px x 10px. Next, the variance calculation unit 31 calculates the variance of the pixel values within each block for each block. The pixel values are expressed in 256 levels from 0 to 255.

Next, referring to the diagram indicated by reference number 610, the variance value calculation unit 31 creates a histogram. The histogram shows variance values on the horizontal axis and frequency on the vertical axis. As an example, the variance value calculation unit 31 creates a histogram in which variance values are divided into increments of 20.

In this way, the variance value calculation unit 31 calculates a variance value indicating the variation in pixel values within each of multiple divided regions in the image of region T of the breeding tank 2 captured by the breeding tank camera 5. The variance value calculation unit 31 then sends data related to the calculated variance value to the determination unit 32.

The variance calculation unit 31 may calculate the standard deviation of pixel values within the divided region as a statistical quantity indicating variation. The divided block size may be other sizes, such as 15px x 15px. The variance values on the horizontal axis of the histogram may be other numerical values, such as in increments of 30. The variance calculation unit 31 may send data associating variance values with frequencies to the determination unit 32, rather than in the form of a histogram.

(Determination by Determination Unit)
The determination unit 32 receives data relating to the variance calculated by the variance calculation unit 31 from the variance calculation unit 31. Then, the determination unit 32 determines whether or not the fish is feeding based on the variance calculated by the variance calculation unit 31. This will be explained in detail below.

First, referring to Figures 5 and 6, comparing the two histograms in Figures 5 and 6, we can see that during normal swimming, the fish swim calmly, so the histogram has a distribution skewed to the left. In contrast, during swimming while feeding, the fish swim actively, so the histogram has a more spread-out distribution.

The determination unit 32 therefore calculates the frequency ratio of the variance values to clearly express the difference between the variance values during normal swimming and during feeding. The determination unit 32 then determines whether the fish is feeding based on this frequency ratio. This will be explained using Figure 7 and other figures.

Figure 7 is a diagram for explaining how the determination unit 32 calculates the first index value and the second index value. The diagram indicated by reference number 700 shows how a predetermined first division and a predetermined second division are set in the histogram indicated by reference number 510 in Figure 5. The diagram indicated by reference number 710 shows how a predetermined first division and a predetermined second division are set in the histogram indicated by reference number 610 in Figure 6.

First, the determination unit 32 calculates a first index value corresponding to the number of divided areas having variance values that are classified into a predetermined first category among multiple divided areas in an image captured by the breeding aquarium camera 5 of area T of the breeding aquarium 2. In the example shown in Figure 7, the predetermined first category is set to variance values of 0 to 50. The first index value is the number of divided areas having variance values of 0 to 50.

The determination unit 32 also calculates a second index value corresponding to the number of divided regions having variance values that fall into a predetermined second category, which indicates a range of variance values greater than the first category. In the example shown in Figure 7, the predetermined second category is set to variance values of 200 to 250. The second index value is the number of divided regions having variance values of 200 to 250.

The determination unit 32 then calculates a feeding index that indicates the difference in magnitude between the first index value and the second index value, and uses this feeding index to determine whether the fish is feeding. In the example shown in Figure 7, the determination unit 32 calculates a feeding index (variance frequency ratio) that indicates the difference in magnitude between the first index value (the number of divided regions with variance values 0 to 50) and the second index value (the number of divided regions with variance values 200 to 250), and uses this frequency ratio to determine whether the fish is feeding. This is explained using Figure 8.

Figure 8 is a diagram illustrating how the determination unit 32 determines whether a fish is feeding.

In the graph of Figure 8, the horizontal axis indicates the frame number of the captured image, and the vertical axis indicates the dispersion frequency ratio. The frame number can be considered to correspond to time (time-series data). The determination unit 32 calculates the feeding index (dispersion frequency ratio) for each image captured over time by the breeding aquarium camera 5, and creates a graph (or table, etc.) that correlates time with the dispersion frequency ratio.

Referring to the example in Figure 8, there is a noticeable difference in the dispersion frequency ratio between the left and right sides of the horizontal axis, frame numbers 1000 to 1400. This is due to the significant difference in dispersion frequency ratio between normal swimming states and swimming states while feeding. The left side, where the frequency ratio is relatively high, indicates the normal swimming state of the fish, where the difference between the first and second index values is large. The right side, where the frequency ratio is relatively low, indicates the swimming state of the fish while feeding, where the difference between the first and second index values is small.

The determination unit 32 therefore determines that the time period corresponding to the left of frame numbers 1000 to 1400, where the variance ratio is greater than a predetermined threshold value for the variance ratio (e.g., 1.0), is a "normal swimming state." The determination unit 32 can also determine that the time period corresponding to the right of frame numbers 1000 to 1400, where the variance ratio is smaller than the threshold, is a "swimming state while feeding."

In this way, the determination unit 32 can determine whether or not a fish is feeding by taking advantage of the significant difference in the dispersity ratio between normal swimming states and feeding swimming states. In conventional image processing technology, for example, when distinguishing between normal swimming states and feeding swimming states, a method of comparing the brightness of captured images of both states has been considered, but no significant results were obtained. In this regard, the determination unit 32 can determine with high accuracy whether or not a fish is feeding based on the dispersity ratio.

The predetermined first division and the predetermined second division may each be set as appropriate. For example, the predetermined first division may be set to a variance value between 50 and 100, and the predetermined second division may be set to a variance value between 200 and 250.

However, empirically, it is preferable that the lower limit of the variance value in the second category be at least twice the upper limit of the variance value in the first category. This allows the determination unit 32 to determine with greater accuracy whether the fish is feeding.

The first index value and the second index value may each be a value indicating the total area of the divided regions, rather than the number of divided regions. The feeding index may be the ratio or difference between the first index value and the second index value. In either case, the determination unit 32 can determine whether the fish is feeding.

Returning to Figure 3, the determination unit 32 determines whether the fish is feeding and sends information related to the determination result to the operation control unit 33.

The operation control unit 33 controls the operation of the feeding device 4 that supplies food to the breeding tank 2 based on the determination result of the determination unit 32. An example of this is explained using Figure 9. Figure 9 is a flow diagram showing an example of the operation control of the feeding device 4 by the operation control unit 33. Note that in the example of Figure 9, it is assumed that the feeding device 4 is not feeding the fish, and the determination result of the determination unit 32 is "not feeding."

First, in S900, the determination unit 32 determines that the fish are not feeding using the aforementioned statistical method. Next, in S902, the determination unit 32 checks the fish feeding schedule recorded in the recording unit 40. Then, in S904, the determination unit 32 references the fish feeding schedule and determines whether the current time has passed the scheduled feeding start time.

If the answer is NO in S904, the current time has not passed the scheduled feeding start time. Therefore, the judgment unit 32 instructs the operation control unit 33 to continue the stopped state of the feeding device 4. The judgment unit 32 may also execute the "continue stopped feeding" command without issuing any instruction to the operation control unit 33. If the answer is NO in S904, the flow returns to the step between S900 and S902.

If the answer is YES in S904 (S906), the current time has passed the scheduled feeding start time. Therefore, the determination unit 32 instructs the operation control unit 33 to cause the feeding device 4 to start feeding.

The flow then proceeds to S908. Here, the flow from S908 onwards assumes a situation in which the feeding device 4 is not performing a feeding operation despite the operation control unit 33 instructing the feeding device 4 to start feeding. The fact that the feeding device 4 is not performing a feeding operation may be determined, for example, by the determination unit 32 using the aforementioned statistical method to determine that "the fish are not feeding." Alternatively, the determination unit 32 may determine that the feeding device 4 is not performing a feeding operation when it receives a signal indicating that a feeding operation is in progress from the feeding device 4.

In S908, the determination unit 32 refers to the fish feeding schedule and determines whether the current time is a predetermined time after the scheduled feeding start time. The predetermined time is a threshold value used to determine whether an abnormality has occurred in the feeding device 4, and may be set to any appropriate time.

If the answer is YES in S908 (S910), this indicates that the fish have not been fed even though a predetermined amount of time has passed since the scheduled feeding start time. Therefore, the determination unit 32 determines that an abnormality has occurred in the feeding device 4 and causes the abnormality notification unit 37 to issue an abnormality alarm.

If the answer is NO at S908, it is not yet time to determine that an abnormality has occurred in the feeding device 4, so the flow returns to S906, and the determination unit 32 instructs the operation control unit 33 to cause the feeding device 4 to begin feeding.

In this way, the operation control unit 33 can control the operation of the feeding device 4 that supplies food to the breeding aquarium 2 based on the judgment result of the judgment unit 32.

(Determination based on remaining food amount)
Next, the remaining food amount specifying unit 34 will be described. As mentioned above, the remaining food camera 6 is installed in the drainage flow path from the breeding aquarium 2 and captures an image of the remaining food collector 7, which collects remaining food contained in the drainage water. The remaining food amount specifying unit 34 specifies the amount of remaining food based on the remaining food image. The remaining food amount specifying unit 34 can detect the amount of remaining food collected in the remaining food collector 7 using a well-known image detection method, for example, the YOLO (You Only Look Once) algorithm. In this case, the remaining food amount specifying unit 34 (YOLO) detects each piece of remaining food collected in the remaining food collector 7 and counts the number of pieces. The remaining food amount specifying unit 34 (YOLO) may also count the number of remaining food based on the area occupied by the remaining food in the remaining food image.

Instead of YOLO, the remaining food amount identification unit 34 may use SSD (Single Shot MultiBox Detector), R-CNN, or an object detection method derived therefrom to detect the amount of remaining food collected in the remaining food collector 7.

For example, consider the case where the remaining bait collector 7 that collects remaining bait is a net. If the net is always placed on top of the overflow tank 3, the remaining bait amount determination unit 34 will detect the total amount of remaining bait collected by the net. If the net is not always placed on top of the overflow tank 3 and is periodically replaced with a new net, the remaining bait amount determination unit 34 may start counting the amount of remaining bait from 0 when the net is replaced with a new one.

The determination unit 32 can also determine whether the fish is feeding based on both the aforementioned statistical quantity (variance or standard deviation) and the amount of remaining food counted by the remaining food amount determination unit 34. An example of this is explained with reference to Figure 10. Figure 10 is a flow diagram showing an example in which the determination unit determines whether the fish is feeding based on both the statistical quantity and the amount of remaining food counted by the remaining food amount determination unit.

In S1000, the determination unit 32 determines whether the remaining food amount N counted by the remaining food amount determination unit 34 satisfies "N > N1." N1 is a threshold value used to determine whether the pet is currently feeding, and may be set to an appropriate value.

If the answer to S1000 is YES (S1002), this indicates that the amount of remaining food is large. Therefore, the judgment unit 32 judges that the animal is "not feeding" without taking the aforementioned statistical quantities into consideration.

If the answer is NO in S1000 (S1004), the determination unit 32 determines whether the fish is feeding based on the aforementioned statistics.

If the answer is YES in S1004 (S1006), the judgment unit 32 judges that the animal is "feeding."

If the answer is NO in S1004 (S1008), the judgment unit 32 judges that the animal is not currently feeding.

The determination method described above is one example, but in this way the determination unit 32 can determine whether or not the fish are feeding based on both the aforementioned statistical amount and the amount of remaining food counted by the remaining food amount determination unit 34. The threshold value itself may be set appropriately based on the amount of food fed, the number of fish, the season, the frequency of replacing the remaining food collector 7, and/or the growth of the fish. The threshold value may also be set based on the rate at which the amount of remaining food increases, rather than the amount of remaining food. The set threshold value may be recorded in the recording unit 40.

The determination unit 32 can also determine whether the fish is feeding based solely on the amount of remaining food counted by the remaining food amount determination unit 34. This will be explained below with reference to Figures 11 to 13. Each of Figures 11 to 13 is a flow chart showing an example in which the determination unit determines whether the fish is feeding based on the amount of remaining food counted by the remaining food amount determination unit.

Referring to Figure 11, in S1100, after feeding by the feeding device 4 begins, the determination unit 32 determines whether the remaining food amount N counted by the remaining food amount determination unit 34 satisfies "N > N2." N2 is a threshold value used to determine whether feeding is in progress, and may be set to an appropriate value.

If the answer is YES at S1100 (S1102), this indicates that there is a large amount of remaining food. Therefore, the determination unit 32 determines that the animal is "not feeding." Next, at S1104, the determination unit 32 instructs the operation control unit 33 to cause the feeding device 4 to stop feeding.

If the answer is NO at S1100 (S1106), this indicates that the amount of remaining food is low. Therefore, the judgment unit 32 judges that the animal is "feeding." Then, at S1108, the judgment unit 32 instructs the operation control unit 33 to continue feeding by the feeding device 4. Because S1108 corresponds to "continue stopping feeding," the judgment unit 32 may execute "continue stopping feeding" without issuing any instruction to the operation control unit 33.

Next, referring to FIG. 12, in S1200, the determination unit 32 determines whether a predetermined time has elapsed since the feeding device 4 started feeding. The predetermined time may be set as appropriate.

If S1200 returns NO (S1202), this indicates that the feeding was terminated prematurely. Therefore, the judgment unit 32 instructs the operation control unit 33 to maintain the feeding amount (kg/h). Since S1202 corresponds to "maintaining the feeding amount," the judgment unit 32 may execute "maintaining the feeding amount" without issuing any instruction to the operation control unit 33.

If the answer is YES in S1200 (S1204), the determination unit 32 then determines whether the remaining food amount N counted by the remaining food amount determination unit 34 satisfies "N < N3". N3 is a threshold value used to determine whether to increase the amount of food fed. N3 may be set to an appropriate value.

If S1204 returns NO (S1206), this indicates that the amount of remaining feed is not small. Therefore, the judgment unit 32 instructs the operation control unit 33 to maintain the feed amount (kg/h). Since S1206 corresponds to "maintain feed amount," the judgment unit 32 may execute "maintain feed amount" without issuing any instruction to the operation control unit 33.

If S1204 returns YES (S1208), this indicates that the amount of remaining feed is low. Therefore, the judgment unit 32 instructs the adjustment unit 35 (described below) to increase the feed amount (kg/h).

Referring to FIG. 13, in S1300, the determination unit 32 determines whether a predetermined time has elapsed since the feeding device 4 started feeding. The predetermined time may be set as appropriate.

If the result at S1302 is NO (S1302), this indicates that the feeding was terminated prematurely. Therefore, the judgment unit 32 instructs the operation control unit 33 to maintain the feeding amount (kg/h). Since S1302 corresponds to "maintaining the feeding amount," the judgment unit 32 may execute "maintaining the feeding amount" without issuing any instruction to the operation control unit 33.

If the answer is YES in S1300 (S1304), it is determined whether the remaining food amount N counted by the remaining food amount determination unit 34 satisfies "N≧N4". N4 is a threshold value used to determine whether to reduce the amount of food fed. N4 is a number equal to or greater than 1 and may be determined as appropriate.

If the result of S1304 is NO (S1306), this indicates that there is no need to reduce the feed amount. Therefore, the judgment unit 32 instructs the operation control unit 33 to maintain the feed amount (kg/h). Since S1304 corresponds to "maintain feed amount," the judgment unit 32 may execute "maintain feed amount" without issuing any instruction to the operation control unit 33.

If S1304 returns YES (S1308), this indicates that the feed amount needs to be reduced. Therefore, the determination unit 32 instructs the adjustment unit 35 (described below) to reduce the feed amount (kg/h).

In this way, the determination unit 32 can determine whether the fish are feeding based on the amount of remaining food counted by the remaining food amount determination unit 34, and can also adjust the amount of food provided.

The adjustment unit 35 adjusts the amount of food to be fed per hour based on the amount of remaining food for the feeding device 4, which feeds the breeding aquarium 2 based on a feeding schedule that includes settings for the amount of food to be fed per hour in the feeding operation. The adjustment unit 35 may obtain data related to the feeding schedule from the recording unit 40. The adjustment unit 35 may obtain data related to the amount of remaining food from the remaining food amount determination unit 34.

An example of the operation of the adjustment unit 35 to adjust the amount of feed fed per hour based on the amount of remaining food is described with reference to Figure 14. Figure 14 is a flow chart showing an example of the operation of the adjustment unit 35 to adjust the amount of feed fed per hour based on the amount of remaining food.

At S1400, the determination unit 32 determines whether the remaining food amount N counted by the remaining food amount determination unit 34 satisfies "N > N5." N5 is a threshold value used to determine whether the amount of food fed has increased or decreased, and may be set to an appropriate value.

If the result of S1400 is NO (S1402), this indicates that the amount of remaining food is less than or equal to N5. The judgment unit 32 then determines whether the "N≦N5" state has continued for a predetermined time. The predetermined time may be set as appropriate.

If S1402 returns NO (S1404), this indicates that the remaining feed amount has not remained below N5 for a predetermined period of time, meaning that there is no need to increase the amount of feed given to the fish. Therefore, the determination unit 32 instructs the operation control unit 33 to maintain the feed amount (kg/h). Since S1404 corresponds to "maintain feed amount," the determination unit 32 may execute "maintain feed amount" without issuing any instruction to the operation control unit 33.

If S1402 returns YES (S1406), this indicates that the amount of remaining food has remained below N5 for a predetermined period of time, meaning that the amount of food given to the fish needs to be increased. Therefore, the determination unit 32 instructs the adjustment unit 35 to increase the amount of food given (kg/h).

If S1400 returns YES (S1408), this indicates that the amount of remaining food is greater than N5. Therefore, the judgment unit 32 determines whether there is less than a predetermined time until the feeding end time based on the feeding schedule.

If the answer is NO in S1408 (S1410), this indicates that there is more than a predetermined amount of time remaining until the feeding end time, meaning that the amount of feed given to the fish needs to be reduced. Therefore, the judgment unit 32 instructs the adjustment unit 35 to reduce the amount of feed (kg/h).

If S1408 returns YES (S1412), this indicates that there is less than a predetermined amount of time remaining until the feeding end time, meaning that there is no need to reduce the amount of feed given to the fish. Therefore, the determination unit 32 instructs the operation control unit 33 to maintain the feed amount (kg/h). Because S1412 corresponds to "maintain feed amount," the determination unit 32 may execute "maintain feed amount" without issuing any instruction to the operation control unit 33.

The method for adjusting the feed amount described above is one example, but in this way, the adjustment unit 35 can adjust the feed amount per hour based on the amount of remaining food for the feeding device 4, which feeds the breeding aquarium 2 based on a feeding schedule that includes setting the feed amount per hour in the feeding operation.

(Determination based on swimming speed)
Next, a method for determining whether or not a fish is feeding, taking into consideration the swimming speed of the fish, will be described with reference to Fig. 15. Fig. 15 is a flow chart for explaining a method for determining whether or not a fish is feeding, taking into consideration the swimming speed of the fish.

First, in S1500, the behavioral amount calculation unit 36 acquires an image of area T of the breeding aquarium 2 captured by the breeding aquarium camera 5 from the breeding aquarium camera 5 via the communication unit 20.

Next, in S1502, the activity amount calculation unit 36 calculates the activity amount (swimming speed) of the fish by performing video image processing on the multiple captured images. The activity amount calculation unit 36 may calculate the swimming speed of the fish using a well-known image detection method, for example, an optical flow algorithm. The optical flow may be running continuously.

Optical flow indicates the distribution of the apparent speed of an object in an image. By estimating the optical flow between captured images, the speed of an object in a video can be measured. Figure 16 is a diagram using optical flow to represent the movement of a fish between two frames as a two-dimensional vector. Using this method, the behavioral amount calculation unit 36 can calculate the swimming speed by performing video processing on multiple captured images.

Returning to FIG. 15, in S1504, the determination unit 32 determines whether the activity level (swimming speed) of the fish is below a predetermined value. At this time, it is assumed that the feeding device 4 is feeding the fish.

If the answer is NO in S1504 (S1506), this indicates that the fish is feeding and is actively moving. Therefore, the determination unit 32 determines that the fish is "feeding."

If the answer is NO in S1504 (S1508), this indicates that the fish has fed sufficiently and is not actively moving. Therefore, the determination unit 32 determines that the fish is "not feeding."

Steps S1504, S1506, and S1508 will be explained in more detail using Figure 17. Figure 17 is a graph comparing the dispersion frequency ratio and the swimming speed of fish at the same time.

In the diagram indicated by reference number 1700, the horizontal axis indicates time and the vertical axis indicates the dispersion frequency ratio. The time period indicated by F in the graph indicates feeding. In the diagram indicated by reference number 1710, the horizontal axis indicates time and the vertical axis indicates the swimming speed (pixels) of the fish. The time period indicated by F in the graph indicates feeding. The times in reference numbers 1700 and 1710 are the same.

As can be seen from Figure 17, the swimming speed increases initially during feeding F, but decreases after a while. Taking advantage of this characteristic of fish, the determination unit 32 determines that the fish are not feeding if the swimming speed of the fish falls below a predetermined value while feeding is being performed by the feeding device 4. As an example, as time passes after feeding begins, the fish feel full and their swimming speed decreases. Therefore, if the swimming speed of the fish drops to 50% of the speed immediately after feeding began, the determination unit 32 determines that the fish are not feeding. The predetermined value may be set appropriately depending on the amount of feed, the number of fish, the season, and/or the growth state of the fish. The predetermined value may be recorded in the recording unit 40.

Note that the method for measuring the swimming speed of fish is not limited to optical flow, and other methods such as image analysis techniques based on frame difference or individual tracking may also be used.

As described above, the determination unit 32 can determine whether a fish is feeding based on (1), (2), and/or (3) below.

(1) A statistical quantity that indicates the characteristics of the distribution of pixel values within each of multiple divided regions in the image of the breeding aquarium 2 captured by the breeding tank camera 5.

(2) The amount of remaining food determined based on the image of the remaining food captured by the remaining food camera 6 in the remaining food collector 7.

(3) The amount of fish movement calculated by performing video processing on multiple images of the breeding tank 2 captured by the breeding tank camera 5.

The determination unit 32 can determine whether the fish are feeding based on the individual results of (1) to (3). The determination unit 32 can also determine whether the fish are feeding based on at least two or more results extracted from (1) to (3). For example, the determination unit 32 can determine whether the fish are feeding based on the results of (1) and (2), (2) and (3), or (1) and (3). Furthermore, the determination unit 32 can determine whether the fish are feeding based on all of the results of (1) to (3). Which of the results of (1) to (3) the determination unit 32 uses to determine whether the fish are feeding may be changed as appropriate depending on the type of fish, the state of growth of the fish, and/or the season. However, by determining whether the fish are feeding based on at least two or more results extracted from (1) to (3), the aquaculture system 1 and the feeding control device 10 can achieve the effect of more appropriately controlling the feeding of the fish.

As a result, the feeding control device 10 and the aquaculture system 1 equipped with the feeding control device 10 according to the present disclosure can control the feeding of fish more appropriately than before. The biggest challenge facing the aquaculture industry is said to be the cost of fish feed. One factor behind this is that the cost of fish meal used in fish feed has tripled over the past 15 years. Feed costs account for half of the total cost, and can reach 80% depending on the fish species.

In this regard, the aquaculture system 1 and feeding control device 10, by being equipped with the above-described configuration, can control the feeding of fish more appropriately than before, thereby reducing the cost of fish feed.

In addition, the aquaculture system 1 and feeding control device 10 can reduce feed waste, enabling more environmentally friendly land-based and marine aquaculture.

Furthermore, the aquaculture system 1 and feeding control device 10 can appropriately set feeding times and amounts, thereby reducing the burden on keepers.

(Abnormality notification)
Next, returning to FIG. 3, the abnormality notification unit 37 will be described.

The abnormality notification unit 37 can notify of an abnormality (such as a malfunction) in the feeding device 4. For example, if the determination unit 32 determines that the fish are not currently feeding even though a predetermined amount of time has passed since the scheduled feeding start time for the feeding device 4, the abnormality notification unit 37 may notify that an abnormality has occurred in the feeding device 4 (for example, S912 in FIG. 9 ). The abnormality notification unit 37 may notify the keeper and/or operator of the aquaculture system 1, etc., of the abnormality in the feeding device 4 by electronic means such as email and/or by voice, etc. The abnormality notification unit 37 can also display the abnormality on the display unit 60.

A malfunction in the feeding device can directly affect the life or death of the fish. In this regard, the feeding control device 10 is equipped with an abnormality notification unit, which allows the determination result by the determination unit 32 to be associated with the presence or absence of an abnormality in the feeding device 4. This allows the feeding control device 10 to operate the aquaculture system 1 more safely.

(Feeding control method)
Next, an example of a feeding control method according to the present disclosure will be described with reference to Fig. 18. Fig. 18 is a flow chart for explaining an example of a feeding control method according to the present disclosure.

First, in S1800, operation of the aquaculture system 1 is started.

Next, in S1802, the breeding tank camera 5 captures an image of area T in the breeding tank 2 and transmits the captured image (image data) to the feeding control device 10. In addition, the remaining feed camera 6 captures an image of the remaining feed collector 7, which collects remaining feed contained in the wastewater, and transmits the remaining feed image (image data) to the feeding control device 10. The breeding tank camera 5 and remaining feed camera 6 can capture images at any time after the aquaculture system 1 starts operating.

Next, in S1804, the determination unit 32 calculates a feeding index that indicates the difference in magnitude between the first index value and the second index value.

Next, in S1806, the determination unit 32 determines whether the feeding device 4 is malfunctioning. As an example, if the determination unit 32 determines that the fish are not feeding even though a predetermined amount of time has passed since the scheduled feeding start time for the feeding device 4, the determination unit 32 determines that the feeding device 4 is malfunctioning. The abnormality notification unit 37 then notifies the user that an abnormality has occurred in the feeding device 4. In S30, the amount of remaining food in S40, described below, may also be included in the determination criteria.

In S1808, the remaining food amount determination unit 34 calculates the remaining food amount based on the remaining food image.

In S1810, the activity amount calculation unit 36 calculates the activity amount (swimming speed) of the fish by performing video image processing on multiple captured images.

S1806, S1808, and S1810 may be processed in parallel or may be executed in a predetermined order.

In S1812, based on the results of S1804 to S1810, the judgment unit 32 determines whether to continue operating the feeding device 4. The judgment unit 32 compares the threshold values for the feeding index, remaining food amount, and/or swimming speed to determine whether to continue operating the feeding device 4 or stop operating it. The judgment unit 32 can also determine whether to continue operating the feeding device 4 or stop operating it based on the threshold values for the feeding index, remaining food amount, or swimming speed.

In this way, the aquaculture system 1 according to the present disclosure performs image analysis to determine whether the farmed fish are swimming while feeding or normally (not feeding), and controls the operation of the feeding device based on the results to appropriately control the feeding of the fish.

(Software implementation example)
The functions of the feeding control device 10 (hereinafter referred to as the "device") are realized by a program that causes a computer to function as the device, and by a program that causes a computer to function as each control block of the device (particularly each part included in the control unit 30).

In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program. The functions described in each of the above embodiments are realized by executing the program using this control device and storage device.

The above program may be stored non-transitory on one or more computer-readable storage media. These storage media may or may not be included in the device. In the latter case, the program may be supplied to the device via any wired or wireless transmission medium.

Furthermore, some or all of the functions of each of the above control blocks can be realized by logic circuits. For example, integrated circuits incorporating logic circuits that function as each of the above control blocks are also included in the scope of the present invention. In addition, the functions of each of the above control blocks can also be realized by, for example, a quantum computer.

Furthermore, each process described in each of the above embodiments may be executed by AI (Artificial Intelligence). In this case, the AI may run on the control device, or on another device (for example, an edge computer or a cloud server).

〔summary〕
A feeding control device according to aspect 1 of the present disclosure includes a first calculation unit that calculates statistics indicating characteristics of the distribution of pixel values within each of a plurality of divided areas in an image of an aquarium containing aquatic organisms captured by a first imaging device, a determination unit that determines whether the aquatic organisms are feeding based on the statistics calculated by the first calculation unit, and an operation control unit that controls the operation of a feeding device that feeds the aquarium based on the determination result of the determination unit.

With the above configuration, the feeding control device according to aspect 1 of the present disclosure can appropriately control feeding of aquatic organisms.

A feeding control device according to aspect 2 of the present disclosure is the same as in aspect 1, in which the determination unit determines whether the aquatic organism is feeding using a feeding index that indicates the difference in magnitude between a first index value corresponding to the number of divided areas among the plurality of divided areas that have the statistical quantity classified into a predetermined first division, and a second index value corresponding to the number of divided areas that have the statistical quantity classified into a predetermined second division that indicates a range of statistical quantities larger than the first division.

With this configuration, the determination unit can determine with high accuracy whether an aquatic organism is feeding based on the feeding index.

A feeding control device according to aspect 3 of the present disclosure is the same as in aspect 2, except that the lower limit value of the statistical quantity in the second category is at least twice the upper limit value of the statistical quantity in the first category.

With this configuration, the determination unit can determine with greater accuracy whether an aquatic organism is feeding.

A feeding control device according to Aspect 4 of the present disclosure is the same as Aspect 1, except that the captured image is an image showing a planar view of the water surface of the aquarium, and includes the area to which food is supplied from the feeding device.

With this configuration, the captured image can more reliably capture the fish biting the bait. As a result, the feeding control device according to aspect 4 of the present disclosure can more appropriately control the feeding of aquatic organisms.

The feeding control device according to aspect 5 of the present disclosure, in any of aspects 1 to 4 above, further includes a remaining food amount determination unit that determines the amount of remaining food based on a remaining food image taken of a remaining food collector that is installed in the drainage flow path from the aquarium and collects remaining food contained in the drainage water, and the determination unit determines whether the aquatic organism is feeding based on the statistics and the remaining food amount.

With this configuration, the determination unit determines whether the aquatic organism is feeding based on two indicators: statistics and the amount of remaining food, allowing for more accurate determination.

A feeding control device according to aspect 6 of the present disclosure is the same as in aspect 5 above, and further includes an adjustment unit that adjusts the amount of food to be fed per hour based on the amount of remaining food, for the feeding device that feeds the aquarium based on a feeding schedule that includes setting the amount of food to be fed per hour in the feeding operation.

With the above configuration, the feeding control device according to aspect 6 of the present disclosure can reduce feed waste, saving on feed costs and realizing environmentally friendly land-based aquaculture.

A feeding control device according to aspect 7 of the present disclosure is any one of aspects 1 to 6, further comprising a second calculation unit that calculates the activity level of the aquatic organism by performing video image processing on the captured images, and the determination unit determines that the aquatic organism is not feeding if the activity level is below a predetermined value while the feeding device is feeding the organism.

With this configuration, the determination unit can make a more accurate determination than for aquatic organisms, also based on the amount of activity of the aquatic organisms.

A feeding control device according to aspect 8 of the present disclosure, in any of aspects 1 to 7 described above, further includes an abnormality notification unit that notifies the user that an abnormality has occurred in the feeding device when the determination unit determines that the aquatic organism is not currently feeding even though a predetermined amount of time has passed since the scheduled feeding start time set by the feeding device.

A malfunction in the feeding device can directly affect the life or death of aquatic organisms. In this regard, the feeding control device according to aspect 8 of the present disclosure, by being equipped with the above-described configuration, can correlate the judgment result by the judgment unit with the presence or absence of an abnormality in the feeding device, thereby enabling safer operation of the aquaculture system to which the feeding control device is applied.

A feeding control method according to aspect 9 of the present disclosure includes a first calculation step of calculating, for each of a plurality of divided regions in an image captured by a first imaging device of an aquarium containing aquatic organisms, a statistical quantity indicating characteristics of the distribution of pixel values within the divided region; a determination step of determining whether the aquatic organisms are feeding based on the statistical quantity calculated in the first calculation step; and an operation control step of controlling the operation of a feeding device that supplies food to the aquarium based on the determination result in the determination step.

With the above configuration, the feeding control method according to aspect 9 of the present disclosure can appropriately control feeding of aquatic organisms.

This disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the disclosed technical means are also included in the technical scope of this disclosure. Furthermore, new technical features can be created by combining the disclosed technical means.

1 Aquaculture system 2 Breeding tank 3 Overflow tank 3a First pipe 3b Second pipe 4 Feeding device 5 Breeding tank camera (first imaging device)
6. Remaining bait camera (second imaging device)
7: remaining feed collector 10: feeding control device 20: communication unit 30: control unit 31: variance value calculation unit (first calculation unit)
32 Determination unit 33 Operation control unit 34 Remaining food amount identification unit 35 Adjustment unit 36 Activity amount calculation unit (second calculation unit)
37 Abnormality notification unit 40 Recording unit 50 Input unit 60 Display unit

Claims (9)

a first calculation unit that calculates, for each of a plurality of divided regions in a captured image of an aquarium containing aquatic organisms captured by a first imaging device, a statistical quantity that indicates a characteristic of a distribution of pixel values within the divided region;
a determination unit that determines whether the aquatic organism is feeding based on a frequency distribution of the statistical quantities for each of the plurality of divided regions in the captured image, the frequency distribution being based on the statistical quantities calculated by the first calculation unit; and
A feeding control device comprising: an operation control unit that controls the operation of the feeding device that feeds the aquarium based on the judgment result of the judgment unit. a first calculation unit that calculates, for each of a plurality of divided regions in a captured image of an aquarium containing aquatic organisms captured by a first imaging device, a statistical quantity that indicates a characteristic of a distribution of pixel values within the divided region;
a determination unit that determines whether the aquatic organism is in a feeding state based on the statistics calculated by the first calculation unit;
an operation control unit that controls the operation of the feeding device that supplies food to the aquarium based on the determination result of the determination unit;
The determination unit determines whether the aquatic organism is feeding using a feeding index that indicates the difference in magnitude between a first index value corresponding to the number of divided areas among the multiple divided areas that have the statistical quantity classified into a predetermined first division, and a second index value corresponding to the number of divided areas that have the statistical quantity classified into a predetermined second division that indicates a range of statistical quantities larger than the first division. The feeding control device of claim 2, wherein the lower limit value of the statistical quantity in the second category is at least twice the upper limit value of the statistical quantity in the first category. The feeding control device according to claim 1 , wherein the captured image is an image showing a plan view of the water surface of the aquarium, and includes an area to which food is supplied from the feeding device.
a first calculation unit that calculates, for each of a plurality of divided regions in a captured image of an aquarium containing aquatic organisms captured by a first imaging device, a statistical quantity that indicates a characteristic of a distribution of pixel values within the divided region;
a determination unit that determines whether the aquatic organism is in a feeding state based on the statistics calculated by the first calculation unit;
an operation control unit that controls the operation of the feeding device that supplies food to the aquarium based on the determination result of the determination unit;
A remaining food amount specifying unit is further provided which specifies the amount of remaining food based on a remaining food image obtained by capturing an image of a remaining food collector which is installed in a drainage flow path from the aquarium and collects remaining food contained in the drainage,
The feeding control device, wherein the determination unit determines whether the aquatic organism is currently feeding based on the statistics and the amount of remaining food. The feeding control device of claim 5 further comprises an adjustment unit that adjusts the amount of food to be fed per hour based on the amount of remaining food, for the feeding device that feeds the aquarium based on a feeding schedule that includes setting the amount of food to be fed per hour in the feeding operation. a second calculation unit that calculates the amount of activity of the aquatic organism by performing video processing on the plurality of captured images;
The feeding control device according to claim 1 , wherein the determination unit determines that the aquatic organism is not in a feeding state when the amount of activity falls below a predetermined value while the feeding device is feeding the aquatic organism. a first calculation unit that calculates, for each of a plurality of divided regions in a captured image of an aquarium containing aquatic organisms captured by a first imaging device, a statistical quantity that indicates a characteristic of a distribution of pixel values within the divided region;
a determination unit that determines whether the aquatic organism is in a feeding state based on the statistics calculated by the first calculation unit;
an operation control unit that controls the operation of the feeding device that supplies food to the aquarium based on the determination result of the determination unit;
A feeding control device further comprising an abnormality notification unit that notifies the user that an abnormality has occurred in the feeding device when the determination unit determines that the aquatic organism is not currently feeding even though a predetermined time has passed since the scheduled start time of feeding set by the feeding device. a first calculation step of calculating, for each of a plurality of divided regions in a captured image of an aquarium containing aquatic organisms captured by a first imaging device, a statistic indicating a characteristic of a distribution of pixel values within the divided region;
a determination step of determining whether the aquatic organism is feeding based on a frequency distribution of the statistical quantities for each of the plurality of divided regions in the captured image, the frequency distribution being based on the statistical quantities calculated in the first calculation step;
A feeding control method including an operation control step of controlling the operation of a feeding device that feeds the aquarium based on the judgment result in the judgment step.