JP2000201568A - Automatic feeding system for aquatic organism - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an automatic feeding system for aquatic organisms designed to put feeding levels closer to the actual feed intake demand for aquatic, organisms to raise feeding efficiency and contribute to preventing environmental water from contamination as well. SOLUTION: This automatic feeding system for aquatic organisms works as follows: a database 60 prepares in advance a standard feeding pattern varying stepwise according to the feed intake process for aquatic organism colonies for each aquatic organismic species and raising environmental condition; an ITV camera 40 photographs the activities of aquatic organismic colonies raised, and the resultant pictures thus taken are subjected to image processing by a relevant image processing means 502 to effect numeric conversion; a pattern read-out means reads out the standard feeding pattern meeting the raising environmental conditions from the database; a correction means 70 corrects the standard feeding pattern based on the image-processed data with further subdivided time unit compared to the respective steps of the standard feeding pattern, and a feeding control means 501 executes a feeding control based on the resultant feeding pattern thus corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic feeding apparatus for aquatic organisms, and in particular, measures and analyzes the feeding behavior (feeding activity) of a group of aquatic organisms bred (cultured) by image processing. The present invention relates to an automatic feeding device for aquatic organisms, which controls an appropriate feeding amount according to a feeding requirement of an aquatic organism.

[0002]

2. Description of the Related Art Generally, an appropriate amount of food for aquatic organisms is determined by a ratio to the weight of a target aquatic organism group.
If this proper amount of feed is given per unit time, the target aquatic organisms will grow smoothly.

[0003] By the way, feeding of aquatic organisms has conventionally been generally performed by humans, and the feeder judges the appropriate number of times of feeding per day and the appropriate amount of feeding per one according to the growth state of the target aquatic organism. . In this case, the largest factors that feeders use as judgment materials are the underwater behavior of the target aquatic organism,
That is, the activity amount (feeding amount).

[0004] For example, in the case of red sea bream, which is a main cultured fish, when the target fish demands a high level of food at the beginning of feeding, very active and intense waves stand near the water surface. From the middle to the end of feeding, as the degree of demand for food decreases, the animal moves below the water surface and, even when fed, does not eat and the amount of activity decreases.

The feeder feeds the fish while observing the growth status and feeding behavior of the school of fish. The basics of feeding are to feed the individuals in the target fish school evenly, and to limit the feeding rate to 10
That is, approaching 0%. However, in reality, it is very difficult to feed all the individuals in a school of fish evenly and to achieve a food intake close to 100%.

[0006] Causes include the skill of the feeder and time constraints. Feeders need to consider the season, temperature, water temperature and weather. In recent years, the scale of aquaculture has been gradually increased with the rise of the aquaculture industry, and the time available for feeders to devote to the target aquatic organisms has been considerably restricted.

In order to solve this problem, automatic feeders are commercially available, but most of the automatic feeders currently on the market are of a type in which a fixed amount of food is fed at a fixed interval by a timer (Japanese Patent Laid-Open Publication No. H11-163873). No. 63-98334), and the feed requirements of the target aquatic organisms were not sufficiently considered. Therefore, it is said that the feeding rate when using it is considerably lower than that for the skilled person.

[0008] Another cause is that the behavior of the target aquatic organism in the water during feeding cannot be sufficiently grasped. The behavior of the target aquatic organism can be accurately observed only near the water surface, and the behavior in the depth direction is extremely difficult to observe.

In order to solve these problems, Japanese Patent Application Laid-Open No.
As described in JP-A-262040, the behavior of the aquatic organisms near the water surface during feeding is imaged, or the distribution of the aquatic organisms in the depth direction and the feeding behavior (the amount of activity or the amount of activity) is imaged. A method of digitizing by image processing and controlling the amount of feed in stages based on the digitized image processing data is being studied. Also, in this automatic feeding device, the feeding environment is corrected by the feeding environment because the feeding environment such as the season, the atmospheric pressure, and the dissolved oxygen concentration affects the feeding behavior.

[0010]

SUMMARY OF THE INVENTION The above-mentioned Japanese Patent Application Laid-open No. Hei 9-2
In the automatic feeding technique described in Japanese Patent No. 62040, since the feeding amount to be controlled is corrected according to the breeding environment conditions, the feeding amount is made closer to the required feeding amount of aquatic organisms than the conventional automatic feeding device. However, there were the following points to be improved.

That is, according to the above-mentioned prior art, the amount of feed is controlled stepwise (eg, at the beginning of feeding, at the middle of feeding, at the end of feeding) over time, and the feeding environment conditions (eg, season, The feed amount to be controlled is uniformly corrected according to the atmospheric pressure, water temperature, dissolved oxygen, etc.). In practice, however, aquatic organisms can behave exceptionally during each stage of the feeding process. For example, in the early stage of feeding, motivation to eat is active, but even in the initial stage, the fish temporarily go back to the upper layer for alert and return to the middle layer, and the amount of food consumption decreases. In some cases, the actual food demand and feeding amount of aquatic organisms did not match due to variations in daily feeding activities that could not be explained solely by breeding environmental conditions.

The present invention has been made in view of the above points, and an object of the present invention is to increase the feeding efficiency by making the feeding amount of an automatic feeding device closer to the actual feeding demand of aquatic organisms, and to improve the feeding efficiency of environmental water. It is an object of the present invention to provide an automatic feeding device that can contribute to the prevention of food pollution.

[0013]

In order to achieve the above object, the present invention basically proposes the following means for solving the problems.

[0014] That is, a breeding means for breeding the aquatic organism group, a database prepared in advance at least for each breeding environment condition, a standard feeding pattern that changes stepwise according to the feeding process of the aquatic organism group, Imaging means for photographing the activity state of the group, image processing means for image-processing this image and digitizing it, means for reading a standard feeding pattern suitable for breeding environment conditions from the database, and reading the read standard feeding pattern. Correction means for correcting based on the quantified image processing data in units of time subdivided from each stage of the standard feeding pattern, and feeding control means for executing feeding control based on the corrected feeding pattern, It is characterized by having.

With this configuration, the standard feeding pattern for each stage is finely corrected so as to match the actual feeding process that changes every moment, so that it is adapted to the actual feeding activity of the aquatic organism. Enables precise feeding control.

[0016]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is an overall configuration diagram showing an automatic feeding apparatus for aquatic organisms according to an example of the present embodiment. Here, the case where the aquatic organism is a cultured fish such as red sea bream or yellowtail is shown.

[0018] The fish 10 can be raised indoors or outdoors.
(Cultured aquarium or marine fish). In the underwater 30 in the breeding means 20, the activity state (feeding activity) of the fish is photographed by the ITV camera (industrial television camera) 40, and an image signal 40S is output to the image processing means 502 of the control device 50.

The control device 50 is constituted by a small computer and includes a feed control means 501, an image processing means 502, a feed pattern reading means 503, and a feed pattern correction means (feed amount / feed speed correction means) 70.

The image processing means 502 includes the ITV camera 40
The image (imaging signal 40S) obtained from is subjected to image processing, and the activity state (behavior) of the aquatic organisms is image-processed and digitized. Here, the water depth position distribution state of the fish school 10 is quantified.

The database 60 stores a standard feeding pattern that changes stepwise (for example, an early stage of feeding, a middle stage of feeding, and a last stage of feeding) in accordance with the feeding process of the aquatic organism group. Prepare in advance for each breeding environment condition (memory)
are doing. Here, the breeding environment conditions are, for example, season,
Pressure, water temperature, dissolved oxygen, and the like are input to the control device 50 by a detection sensor or input via an operator according to the type.

The feeding pattern reading means 503 reads a standard feeding pattern from the database 60 that is suitable for the aquatic organism group and the breeding environment conditions.

The feeding control means 501 calculates a feeding amount and a feeding time defined by the standard feeding pattern based on the image processing data (this calculation is accompanied by a correction from the correcting means 70), and the calculated control signal Feeding control (feeding amount / feeding time control) is executed via the feeding device 80 based on 50S.

The correction means 70 performs various image analysis based on the image processing data from the image processing means 502, for example, determination of a feeding stage (for example, an initial stage of feeding, a middle period of feeding, and an end stage of feeding),
Analysis of the water depth distribution state of the school of fish 10 (analysis of the activity of the aquatic organisms), and based on the analysis, a predetermined correction formula [formula for calculating Δk (t) of the following formula 1] is used to determine the feeding correction amount and the feeding stage The amount of correction of the switching timing is obtained, and this correction signal 7
0S is sent to the feed control means 501. Feeding control means 501
Inputs the correction signal 70S and executes the feeding control in consideration of the correction amount to the feeding amount and the feeding time defined by the standard feeding pattern described above. In other words, the feed control unit 501 executes the feed control based on the corrected feeding pattern.

The correcting means 70 corrects the standard feeding pattern read from the database 60 based on the image processing data. The important point here is that the correction means 70 is more subdivided than each stage of the standard feeding pattern (for example, the initial feeding stage, the middle feeding stage, and the final feeding stage) when correcting the standard feeding pattern. That is, the correction is performed based on the numerically processed image processing data in a minute time unit.

The feeding device 80 is provided with, for example, a feed supply port on the bottom surface of the feed tank 90, and further provided with a rotating disk with a feed supply hole rotating along the bottom surface. This feeder feeds the feed 11 from the feed tank 90 so that the feed feed hole of the feed tank 90 matches the feed feed port on the bottom of the feed tank 90. By changing the rotating disk speed, the feed speed (feed count per unit time) Thus, the amount of feed can be changed to control the amount of feed. In that sense, here, the feed amount and the feed speed are in a two-sided relationship.

In the feeding device 80, the feed 11 is supplied from the feed tank 90 to the fish 10 in the breeding means 20 in accordance with the feed amount (feed speed) determined by the feed control means 501.

Next, each component of the above automatic feeding device will be described in detail.

[0029] The breeding means 20 is a breeding aquarium or a cage, whether indoors or outdoors. The shape of the breeding aquarium, for example,
It is rectangular or cylindrical, made of acrylic or concrete, and varies in size. The shape of the cage is rectangular and is made of mesh or steel with a mesh smaller than that of the fish 10 and has a volume of 300.
m ^3, and the like.

Here, the functions of the automatic feeding control device 50 (feeding control means 501, image processing means 502, feeding pattern reading means 503, correction means 70) will be described in detail with reference to FIGS.

First, the image processing means 501 in the control device 50 determines whether the image signal 40S obtained by the ITV camera 40 is
Then, a water depth position defining step 51 and an image processing step 52 are executed as shown in the flowchart of FIG.

In the water depth position definition step 51, the screen image of the behavior of the school of fish is divided into a plurality of layers in the water depth direction, and these layers are defined as water depth positions. Specifically, the imaging area on the captured screen is divided into a plurality in the depth direction, for example,
As shown in FIG. 3, the water depth position is divided into three, and the first and second from the top are defined as the upper layer, the third to fifth are the middle layers, and the sixth to tenth are defined as the low layers. The definition of the water depth position may be set arbitrarily.

In the image processing step 52, the behavior of the school of fish at each water depth position (upper, middle, low) defined in the water depth definition step 51 is image-processed and quantified. For example, fish 10
When the underwater is photographed by illuminating the underwater with lighting means, the background is bright and the fish is photographed dark, so the dark part (representing the fish) is set to “1” and the bright part (background) is set to “0”. 2
The activity amount (distribution state) of the upper, middle, and lower (lower) fish schools can be calculated based on the binarized number of pixels. The details of the activity amount quantification process itself are described in Japanese Patent Application Laid-Open No. 9-262040.

In the feeding pattern reading step 53 executed by the feeding pattern reading means 503, a standard feeding pattern satisfying the conditions is read from the feeding pattern database 60 prepared in advance for each fish species and breeding condition.

The feeding control means 501 includes an image processing step 52.
Predict the feeding activity of the school of fish based on the number of pixels indicating the fish distribution amount of each layer determined in the above and the standard feeding pattern determined in the feeding pattern reading step, and also receive a correction signal 70S from the feeding amount correction means 70, Determine the amount of feeding. The number of pixels indicating the fish distribution amount obtained in the image processing step 52 is sent to the feed amount correction means 70, and the correction means 70 analyzes the image to output a correction signal 70S.

In the feeding speed step 54 executed by the feeding control means 501, the image processing value obtained in the image processing step 52
A signal 70S regarding the feed pattern determined in the feed pattern reading step 53 and the feed correction amount from the correction means 70
The appropriate feeding amount (feeding speed) and time are determined from the above, and a feeding control signal 50S is output to the feeding device 80.

FIG. 4 shows an example of the standard feeding pattern. In FIG. 4, the horizontal axis indicates time, and the vertical axis indicates feed amount. For example, a large amount of feed is fed in the early stage when fish feeding activity is active, and the bait demand-following type in which the amount of feed is reduced in the middle and late stages (FIG. 4a), or the amount of feed is suppressed in the initial stage. There is a motivation-feeding type in which a large amount of food is fed during the middle period and a small amount is fed at the end period after the desire to eat is increased (FIG. 4b). If the breeding environment conditions are appropriate for the fish 10, feeding is performed in a bait request behavior tracking type. If the water temperature is low and the willingness to eat is inactive, feed is provided in a motivating manner, and if the feeding is not to be stopped halfway, the feeding is stopped. These standard feeding patterns are stored in the feeding pattern database 60.

Next, a flowchart of the steps executed by the correction means 70 will be described with reference to FIG.

The correction means 70 is first provided with the feed control means 50
When receiving the correction command signal 70S 'from Step 1, a timing correction step 71, a feed correction step 72, and a timing step 73 described below are executed.

In the timing correction step 71, the current feeding stage is determined based on the image processing data (water depth distribution data of the school of fish during the feeding activity) from the image processing means 501. The determination of the feeding stage will be described with reference to FIG.

In FIG. 6, the horizontal axis represents the time axis during feeding activity, and the vertical axis represents the number of pixels indicating the amount of fish school (for example,
Of the binary pixels), and the data on the number of pixels is displayed separately for the number of pixels in the upper layer, the number of pixels in the middle layer, and the number of pixels in the lower layer.

In the early stage of feeding, the school of fish is active in the feeding behavior, so that it is present in the upper layer. As the feeding progresses, the willingness to eat begins to decrease, and eventually the swim starts to spread.

That is, the initial stage of feeding: the number of pixels in the upper layer ≧ the number of pixels in the middle layer, and the number of pixels in the upper layer ≧ the number of pixels in the lower layer.The middle stage of feeding: the number of pixels in the middle layer ≧ the number of pixels in the upper layer ≧ the number of pixels in the lower layer. The number of pixels in the lower layer ≧ the number of pixels in the upper layer sets the feeding stage (feeding stage), and the feeding stage is determined based on the magnitude relationship of the number of pixels. Note that the number of pixels in the upper layer, the number of pixels in the middle layer, and the number of pixels in the lower layer as described above are reversed due to the exceptional behavioral pattern of the fish, even if the feeding is temporarily in the initial stage. The same reversal phenomenon may occur in the middle stage of feeding and the end of feeding, so in order to avoid erroneous judgment at such times,
It is desirable to sample the water depth distribution data several times and apply the above-described determination condition based on the accumulated value, average value, or frequency. In the present embodiment, this is done.

Based on the determination of the feeding stage, the timing of switching the feeding stage (ie, feeding time) is corrected.

Next, a feed correction step 72 is executed. FIG. 7 shows the details of the feed correction step 72.

In the feeding correction step 72, first, a current feeding step is determined in a feeding step determination step 72-1. Here, if the image processing data of each layer divided into a plurality of layers (for example, upper layer, middle layer, and lower layer) is defined as “type of image processing data”, in the next step 72-2, the above-mentioned data used for correcting the feed amount is described. The type of the image processing data (the number of pixels indicating the fish school water depth position distribution) is switched for each stage of the feeding process, and the increase or decrease of the number of pixels is calculated, and the correction amount determination step 72-3 is executed.

That is, if the current feeding process is determined to be the initial stage, the increase or decrease of the upper image processing data (the number of pixels indicating the upper fish school; the so-called upper pixel) is determined in step 72-2. In the correction amount determination step 72-3, a feed correction amount corresponding to the increase / decrease of pixels is obtained. In this case, the feed amount and the feed speed may be separately controlled, and the correction amount of the feed speed may be obtained in addition to the feed amount. The same applies to the middle stage of feeding and the end of feeding. In the middle period, the image processing data of the middle layer is used as correction data to determine the increase or decrease in the number of pixels in the middle layer (the number of pixels indicating the middle fish group). By using the processing data as correction data, the increase / decrease in the number of pixels in the low layer (the number of pixels indicating the low-level fish school) is determined, and the correction amount Δk (t) of the feed amount (feed rate) is determined from the increase / decrease in the number of pixels.

The feed control means 501 calculates the correction amount Δk
The corrected feeding amount (corrected feeding pattern value) KH (t) is calculated from (t) and the feeding amount (feeding pattern value) KP (t) defined by the standard feeding pattern. The equation for calculating KP (t) and the equation for calculating the correction amount Δk (t) are shown in Equation 1. Note that the obtained correction amount is stored in the feeding pattern database 60.

[0049]

KH (t) = KP (t) + Δk (t) Therefore, Δk (t) = E (t) × I (t) KH (t): Feeding amount after correction KP (t): Feeding pattern Value Δk (t): Correction amount E (t): Feeding stage I (t): Pixel increase / decrease number In the limiter step 73, it functions as a limiter for the time until the switching timing of the feeding stage. If each feeding stage exceeds a predetermined time, the feeding speed is controlled and the process moves to the next feeding stage. At the end, feeding is terminated. Also,
Feeding is also terminated when the entire feeding time reaches a predetermined time. By using restrictions on the feeding speed and the time until the timing of switching the feeding stage, wasteful feeding and time can be saved.

As is apparent from the equation (1), the correction factor includes the feeding stage E (t), which is an important factor in the present invention, and in particular, each stage of the read standard feeding pattern. It is important to execute the feed correction based on the above equation 1 in more detailed time units.

FIG. 8 shows the change over time in the feed amount corrected according to the present embodiment. The upper part of FIG. 8 shows image processing data (the number of pixels indicating the distribution state of the upper, middle, and lower part fish schools) at each stage of the feeding process, and the lower part shows the control state of the amount of feed.

As is clear from FIG. 8, in this embodiment, the standard feeding pattern for each stage is finely corrected in each stage so as to match the actual feeding process which changes every moment. Enables precise feeding control adapted to the actual feeding activity of aquatic organisms. For example, since the feeding activity is very active at the initial stage, the feeding is larger than the predetermined standard feeding pattern. As the feeding behavior gradually decreases, the amount of feeding is also corrected in accordance with the feeding pattern. Thus, by ensuring a feeding pattern similar to the increase or decrease in the activity state (the number of pixels) of the school of fish, it is possible to finally achieve lean feeding.

According to the present embodiment, in addition to the breeding environment conditions, the automatic feeding is enabled by taking into account the daily unpredictable activities of the fish (the activities which cannot be measured under the breeding environment conditions), so that the feeding is more accurate than ever. Feeding guarantee can be realized.

[0054]

According to the present invention, the aquatic organism group and the degree to which each individual in the group requires food (feeding requirement), the feeding behavior induced thereby, and the feeding behavior due to a decrease in food requirement. Although the change in the amount of food is quantified by image processing, the feeding pattern for each predetermined feeding stage is subdivided (refined) in each stage according to the amount of action of the aquatic organisms, so that an appropriate amount of food is always Feeding, improving the feeding rate, reducing the cost of feeding, streamlining the feeding operation, and preventing environmental water pollution.

Further, since the supply amount and the duration are automatically controlled in accordance with the change in the feeding activity of the aquatic organisms, the human burden of the feeding operation is reduced, and the feeding operation is made more efficient.

[Brief description of the drawings]

FIG. 1 is a diagram showing an overall configuration according to an embodiment of the present invention.

FIG. 2 is a flowchart showing an operation of feeding control.

FIG. 3 is an explanatory view showing a water depth position defining step.

FIG. 4 is a diagram showing an example of a feeding pattern.

FIG. 5 is a flowchart showing an operation of feeding correction in a feeding stage.

FIG. 6 is an explanatory diagram showing a principle of determining a feeding stage.

FIG. 7 is a flowchart showing a feed amount correction step.

FIG. 8 is a time chart showing a state after a feeding pattern is corrected.

[Explanation of symbols]

10 fish, 11 food, 20 breeding means, 30 underwater, 4
0: imaging means (ITV camera), 50: automatic feeding control device, 501: feeding control means, 502: image processing means, 5
03: feeding pattern reading means, 60: database, 70: feeding amount control means, 80: feeding means, 90: feeding tank.

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Naoki Haramai 7-1-1, Omika-cho, Hitachi City, Ibaraki Prefecture Inside Hitachi, Ltd.Hitachi Research Laboratory Co., Ltd. 7-1-1, Hitachi Research Laboratory, Hitachi, Ltd. (72) Inventor Kenichi Soma 7-1-1, Omika-cho, Hitachi City, Ibaraki Pref. Hitachi, Ltd. Hitachi Research Laboratory, Ltd. (72) Inventor Bunji Yoshitomi Tokyo 559-6 Kitano-cho, Hachioji-shi, Japan F-term in Central Research Institute of Fisheries Co., Ltd. F-term (reference) 2B104 CF01 CF17 5B057 AA15 DA07 DB02 DB08 DC04

Claims (6)

[Claims] 1. A breeding means for breeding a group of aquatic organisms, a database in which a standard feeding pattern that changes stepwise according to a feeding process of the group of aquatic organisms is prepared in advance at least for each breeding environment condition, Imaging means for photographing the activity state of the group; image processing means for image-processing this video to digitize it; means for reading a standard feeding pattern suitable for breeding environment conditions from the database; and reading the read standard feeding pattern. Correction means for correcting based on the quantified image processing data in units of time subdivided from each stage of the standard feeding pattern, and feeding control means for executing feeding control based on the corrected feeding pattern. An automatic feeding device for aquatic organisms, comprising: 2. The automatic feeding apparatus for aquatic organisms according to claim 1, wherein the database prepares a standard feeding pattern for each type of aquatic organisms and breeding environmental conditions in advance. 3. The automatic feeding apparatus for aquatic organisms according to claim 1, wherein the correction means corrects a feeding amount or a feeding speed defined by the standard feeding pattern. 4. The apparatus according to claim 1, wherein the correction unit is configured to correct a timing of switching a feeding stage defined by the standard feeding pattern based on image processing data of a water depth position distribution of the aquatic organisms. The automatic feeding device for aquatic organisms according to any one of claims 1 to 3. 5. A screen for imaging the behavior of the aquatic organisms is divided into a plurality of layers in the depth direction, and the behavior of the aquatic organisms for each layer is image-processed and quantified, and each layer is defined as a type of image processing data. The aquatic according to any one of claims 1 to 4, wherein the correction unit is configured to switch the type of the digitized image processing data used for the correction for each stage of the feeding process. Automatic feeding device for living things. 6. The automatic feeding apparatus for aquatic organisms according to claim 1, further comprising limiter means for determining a feeding amount, a feeding speed, or a time until a switching timing of a feeding stage.