WO2017013795A1

WO2017013795A1 - Method and device for monitoring acute toxin using small fish

Info

Publication number: WO2017013795A1
Application number: PCT/JP2015/071000
Authority: WO
Inventors: 陸郎横田; 慎一戸田
Original assignee: 株式会社アニマックス
Priority date: 2015-07-23
Filing date: 2015-07-23
Publication date: 2017-01-26
Also published as: SG11201709839UA; CN208366974U; JPWO2017013795A1; JP6051368B1

Abstract

Provided are a monitoring method and monitoring device whereby a warning of an acute toxin can be issued at an appropriate timing. A configuration is adopted whereby an entire horizontal surface of a water tank 11 is divided in advance into a plurality of sensing blocks B smaller than a horizontal image of a small fish, an image of the entire horizontal surface of the water tank 11 is captured for each predetermined time by a camera device 13, the presence of a small fish is determined for each of the sensing blocks B from the captured image of the water tank 11, the number of presence sensing blocks in which a small fish is determined to be present is accumulated as a monitoring history, and the monitoring history is referenced and a warning of an acute toxin is issued when the number of presence sensing blocks is below a warning standard for a predetermined time period.

Description

Method and apparatus for monitoring acute toxin using small fish

The present invention relates to an acute poison monitoring method and a monitoring device for constantly supplying water from a water source to a water tank in which small fish are released and monitoring the small fish to warn of acute poison.

Conventionally, an inspection method using a chemical reagent or the like is generally used as a method for monitoring the water quality of a water source such as a river, a reservoir, or a dam. However, since harmful substances that can be detected with a reagent are limited and specialized knowledge is required for handling the reagent, there is a problem that it is difficult to monitor water quality on a daily basis by this method.

On the other hand, recently, bioassays using small fish such as medaka have attracted attention as a method that can detect various harmful substances more easily and more easily. Bioassays monitor the status of fish released into a tank containing water collected from a water source. The following patent documents are mentioned as an example of the prior art related to this, for example.
Patent Documents 1 and 2 describe that an alarm signal is output when the number of fish in a monitoring range set in an aquarium is monitored and the number of monitoring changes.

JP 2002-257815 A JP 2003-139964 A

However, in the monitoring methods described in Patent Documents 1 and 2, an alarm signal is output when there is a change in water quality due to the movement behavior of fishes to the downstream side, etc., but how much the water quality changes. Since it is not always clear whether the alarm signal is output, it was necessary for a skilled staff to make a final judgment by checking the water tank on-site.
The present invention has been made paying attention to such a problem, and it is an object of the present invention to provide a monitoring method and a monitoring apparatus in which an alarm is issued at an appropriate timing particularly for acute poison.

The method for monitoring acute toxin according to the present invention is to constantly supply water from a water source to a water tank from which a plurality of small fishes are released, preliminarily divide the entire horizontal surface of the water tank into a plurality of detection blocks smaller than a horizontal image of the small fish, The entire horizontal surface of the aquarium is photographed every predetermined time by the camera device, the presence of small fish is determined for each of the detection blocks from the captured image of the aquarium, and the number of presence detection blocks whose presence is determined is accumulated as a monitoring history Then, referring to the monitoring history, an acute poison is warned when the number of presence detection blocks falls below a warning standard in a predetermined period.

The monitoring device for acute toxin according to the present invention includes a water tank in which water is constantly supplied from a monitored water source and a plurality of small fishes are released, and a plurality of detection blocks smaller than a horizontal image of the small fishes on the entire horizontal surface of the water tank. The presence of a small fish is determined for each of the detection blocks from an operation unit that accepts an operation for classifying and setting, a camera device that photographs the water tank every predetermined time, and a captured image of the water tank. A monitoring recording unit for accumulating the number of presence detection blocks as a monitoring history, and an alarm unit for alarming acute poison when the number of presence detection blocks falls below an alarm standard for a predetermined period with reference to the monitoring history It is characterized by that.

In the present invention, for example, the predetermined period is 96 hours, and when the number of presence detection blocks decreases to half or less during the predetermined period, an acute poison may be warned. In this case, the acute poison alarm is a clear signal indicating that there is an acute poison of about half the lethal concentration in 96 hours in the aquarium water. In addition, since the detection block is smaller than the horizontal image of small fish, the same small fish is judged to exist at the same time by multiple detection blocks, there is no oversight of small fish, and there is a detection error for the number of individuals. It can be suppressed.

It is a block diagram which shows the basic composition of the monitoring apparatus of acute poison made into an example of embodiment. It is a whole perspective view which shows the specific example of the said monitoring apparatus. (A) is a perspective view which shows the specific example of the water tank which comprises the said monitoring apparatus, (b), (c) is a top view explaining all the basic operations of a water supply means. (A) is a top view explaining the process of dividing and setting the horizontal surface of an aquarium into a some detection block, (b), (c) is a top view which shows the specific determination example of presence of the small fish in a detection block. It is. (A) is an example of the photographed image of an aquarium, (b) is a top view which shows an example of the determination result which determined presence of small fish from the captured image. It is a graph which shows an example of the result of having monitored the number of individuals of small fish. It is a flowchart explaining an example of the basic procedure which monitors the number of individuals of small fish. (A) is the other example of the picked-up image of an aquarium, (b) is a top view which shows an example of the determination result which determined the approach to the detection block of a small fish from the picked-up image. It is a graph which shows an example of the result of having monitored the amount of activities of small fish. It is a flowchart explaining an example of the basic procedure which monitors the active mass of small fish. It is a specific example of a display screen of a monitoring history. It is another example of the display screen of a monitoring history. (A) is a graph of the number of individuals normalized by the warning standard of acute poison, (b) is a graph of the number of individuals normalized by the warning standard of acute poison, and (c) is standardized by the warning standard of sudden decrease in activity amount. (D) is a graph of activity amounts normalized by the warning criteria for sudden increase in activity amounts.

FIG. 1 shows an acute toxin monitoring apparatus as an example of the embodiment. This monitoring device monitors the quality of water, particularly acute poisons, such as rivers, reservoirs, dams, etc., using a bioassay that uses small fish. ) Is assumed. The medaka grows to about 2.5 centimeters long in the first year of life, but this monitoring device uses about 10 to 12 such medaka, and the replacement of the dead medaka is 96 hours after the death. After that has passed.

First, the basic configuration of this monitoring apparatus will be described.
As an element related to water quality monitoring, the monitoring device 10 divides a water tank 11 that is constantly supplied with water from a water source and releases a plurality of small fish, and a plurality of detection blocks smaller than the horizontal image of the small fish over the entire horizontal surface of the water tank 11. The presence of a small fish is determined for each of the detection blocks from the operation unit 12 that receives the setting operation, the camera device 13 that photographs the water tank 11 every predetermined time, and the captured image of the camera device 13. A monitoring recording unit 14 for accumulating the number of presence detection blocks as a monitoring history, an alarm unit 15 for alarming acute poison when the number of presence detection blocks falls below an alarm standard for a predetermined period with reference to the monitoring history, and monitoring And a display unit 16 for displaying the history. The monitoring device 10 further includes a water supply means 17 for supplying water taken from a water source to the water tank 11 at a constant rate per hour, an automatic feeding device (not shown) for periodically feeding the water tank 11, and an LED for illuminating the water tank 11. A lamp (not shown) is provided.

If the aquarium 11 is thin with a size of 20 x 30 cm square to 25 to 35 cm square and 4 cm to 5 cm deep, it can maintain 7 to 15 small fish, and the entire horizontal plane can be maintained. It is also easy to photograph with the camera device 13 from above.
The operation unit 12 can be configured with, for example, a touch panel, a keyboard, or the like. Note that the detection block classification information and other initial setting information may be read from a USB memory, a memory card, a network, or the like.
The camera device 13 may be a general device using a CMOS sensor or a CCD sensor. The photographed image may be color or monochrome.
What is necessary is just to comprise the monitoring recording part 14 using computer apparatuses, such as a personal computer, for example. Then, it becomes possible to store the monitoring history in the built-in hard disk or the like and write the stored data to a USB memory, a memory card or the like. In addition, if a remote communication terminal is connected to the network, it is possible to receive a remote operation from the communication terminal or the like, or to transfer the detection of acute poison.
The alarm unit 15 is for notifying the staff of the detection of acute poison. For example, the alarm unit 15 may be configured to sound an alarm sound by a buzzer, a siren or the like, or may be configured to display an alarm by a patrol lamp or the like. Good. Or you may comprise so that it may transfer to an external apparatus.
The display unit 16 can be configured by a liquid crystal panel or a CRT. If a touch panel is used, it can be used as the operation unit 12 and the apparatus can be miniaturized.
The water supply means 17 can be constituted by an electric pump or an electromagnetic valve. Moreover, what is necessary is just to use what is generally marketed for an automatic feeding apparatus and an LED lamp.

FIG. 2 is a perspective view showing a specific example of the monitoring device.
The monitoring device 10 has a basic configuration in which a water tank 11 and a monitoring recording unit 14 are accommodated in a dark box 20 that can be opened and closed by

front doors

20a, 20a, etc., and an alarm unit 15 made of a patrol lamp is provided on the upper surface of the dark box 20. More specifically, the monitoring recording unit 14 is accommodated in the upper stage of the dark box 20, the water tank 11 is accommodated in the middle stage, and the sample water tank 21, raw water supply pump, heat exchanger (not shown), etc. are accommodated in the lower stage. ing. The camera device 13, the LED lamp, and the automatic feeding device (not shown) are arranged above the water tank 11. Furthermore, you may provide the fan apparatus etc. (not shown) which spray on the water surface of the water tank 11 when water temperature is high.
The water tank 11 is formed integrally with a settling tank 11a constituting the water supply means 17, and a camera device 13 and an LED lamp are fixed downward on the outer wall surface of the settling tank 11a. The monitoring recording unit 14 includes a numeric keypad constituting the operation unit 12 and a touch panel constituting the display unit 16 on the front surface of the casing. The camera device 13, LED lamp, automatic feeding device, raw water supply pump, heat exchanger and the like are controlled by the monitoring recording unit 14.
If the entire monitoring device 10 is housed in the dark box 20 and internally illuminated by the LED lamp in this way, only the aquarium 11 is illuminated and the other parts are kept dark, so that algae and phytoplankton are generated and propagated. The effect of suppressing is obtained. This effect will be even better if the entire water tank 11 is formed of a low light reflective resin or the like. Since the LED lamp basically irradiates light in one direction, it is possible to illuminate only the aquarium 11, and the water temperature rise in the aquarium 11 due to the irradiation light is slight.

FIG. 3A is a perspective view showing a specific example of the water tank, and FIGS. 3B and 3C are both plan views for explaining the basic operation of the water supply means.
As shown to Fig.3 (a), the water tank 11 is integrated with the sedimentation tank 11a and the filth tank 11b. The sedimentation tank 11a is a tank that removes precipitates and suspended matters from the water supply to the water tank 11 in advance, and the filth tank 11b is a tank that moves and accumulates excrement of small fish to keep the water tank 11 clean.
The water tank 11 communicates with the sedimentation tank 11a through

water supply ports

11c and 11c provided on one of two opposing side surfaces, and communicates with the filth tank 11b through a water outlet 11d provided on the other side. The bottom surface of the water tank 11 is gently inclined downward from the

water supply ports

11c and 11c toward the water discharge port 11d, and the excrement is sent to the filth tank 11b by the water flow from the

water supply ports

11c and 11c toward the water discharge port 11d. Has been. The water outlet 11d may be provided with a net or the like in order to prevent small fish from escaping.
The settling tank 11a is a four tank type, and a water injection port 11e is provided at the upper part of the first tank, and a high water surface overflow port 11f and a low water surface overflow port 11g are provided at the upper part of the fourth tank. The first to fourth tanks are configured to communicate with adjacent tanks at the lower end or upper end of the partition wall. The

overflow ports

11f and 11g are connected to the

water supply ports

11c and 11c of the water tank 11 by pipes, respectively. A solenoid valve 11h is provided in a pipe connecting the low water surface overflow port 11g and the water supply port 11c.
When the electromagnetic valve 11h is closed, the sedimentation tank 11a discharges the water overflowing from the high water surface overflow port 11f from the water supply port 11c, thereby generating a unidirectional vortex in the water tank 11 as shown in FIG. be able to. On the other hand, when the electromagnetic valve 11h is open, the water overflowed from the low water surface overflow port 11g can be discharged from the water supply port 11c, and a reverse vortex can be generated in the water tank 11 as shown in FIG. . Soil can flow out only from the bottom half of the water tank 11 by only one-way vortex flow, but the sewage flows out from the entire bottom surface of the water tank 11 by appropriately switching the direction of the vortex flow in the water tank 11 by controlling the electromagnetic valve 11h. Is possible.
The four corners of the water tank 11 are curved surfaces, and a vortex stabilizing column 11 i is erected at the center of the water tank 11. Therefore, a stable vortex is obtained and water does not stagnate in the water tank 11. Furthermore, a dead fish collecting net 11j is stretched between the eddy current stabilizing column 11i and the peripheral wall of the water tank 11. The dead fish collection net 11j has enough eyes to allow small fish to swim through. The dead fish is carried by the eddy current and caught by being caught sideways by the dead fish collection net 11j.
The filth tank 11b is divided into a section close to the water tank 11 and a section far from the water tank 11 by a partition wall 11m, and

drainage ports

11k and 11k are provided in the respective sections. The heights of the

drain ports

11k and 11k define the water surface of the water tank 11. The height of the partition wall 11m is set slightly lower than the

drain ports

11k and 11k. A sample water sampling port 11 n communicating with the sample water tank 21 is provided on the bottom surface of the section far from the water tank 11. A solenoid valve or the like is provided on the pipe connected to the sample water tank 21. When the water in the water tank 11 is collected from the sample water collection port 11n, there is no problem that the water surface of the water tank 11 is excessively lowered because of the partition wall 11m.

Next, the basic operation of the monitoring device will be described.
The monitoring method for acute poisoning by the monitoring device 10 is such that a water tank 11 in which a plurality of small fishes are released is constantly supplied from a water source, and the entire horizontal surface of the water tank 11 is divided into a plurality of detection blocks smaller than the horizontal image of the small fishes in advance. Then, the entire horizontal plane of the aquarium 11 is photographed by the camera device 13 every predetermined time, and the presence detection block in which the existence is determined by determining the presence of small fish for each of the detection blocks from the captured image of the aquarium 11. The number is accumulated as a monitoring history, and the monitoring history is referred to, and when the number of presence detection blocks falls below an alarm standard for a predetermined period, an acute poison is alarmed.

For example, in the fish acute toxicity test method stipulated in the OECD test guideline 203, the half-lethal concentration (LC50) is the concentration at which half of the fish die as an effect on the fish when exposed to a chemical substance for 96 hours.
Therefore, as a preferred specific example, if the predetermined period is 96 hours, and the acute poison alarm is configured to alarm when the number of presence detection blocks decreases to half or less during the predetermined period, the acute poison alarm It is a clear signal indicating that there is an acute toxin of about half-lethal concentration in 96 hours. If small fish are 2 to 3 centimeters in length, and each detection block has the same size as two or three or two to four adjacent detection blocks, the size of each small fish can be determined simultaneously. There is no oversight, and detection error is suppressed for the number of individuals.
Whether or not small fish have decreased to half or less in 96 hours is determined by, for example, using 50% of the 96-hour moving average of the number of individuals as an alarm standard for acute poisoning and comparing the current number of individuals with the alarm standard Run with. As a modification, 50% of the number of individuals 96 hours before the present time may be used as an alarm standard for acute poisoning, and the current number of individuals may be compared with the alarm standard.

The step of dividing and setting the horizontal plane of the water tank 11 into a plurality of detection blocks is executed by the monitoring recording unit 14 when a predetermined operation is received by the operation unit 12. For example, the shooting direction and magnification of the camera device 13 are manually adjusted so that the shooting range of the camera device 13 coincides with the entire horizontal surface of the water tank 11, and a predetermined operation is performed with the operation unit 12 while viewing the captured image of the water tank 11. Thus, an exclusion block NB to be excluded in the detection process is arbitrarily selected and set.

FIG. 4 (a) is a plan view for explaining a process of setting the horizontal plane of the aquarium into a plurality of detection blocks. As the detection block B, the shooting range of the camera device 13 is divided into 24 × 32 blocks. If the shooting direction and magnification of the camera device 13 are adjusted so that the shooting range of the camera device 13 coincides with the entire horizontal surface of the water tank 11, the entire horizontal surface of the water tank 11 is divided into a plurality of detection blocks B as shown in the figure. The By selecting the four corners of the water tank 11, the part of the eddy current stabilizing column 11i, and the part of the dead fish collection net 11j as the exclusion block NB, it becomes possible to make only the living small fishes the target of existence determination. That is, there is no possibility of erroneously determining the eddy current stabilizing column 11i, dead fish, etc. as small fish. If the water tank 11 is 24 × 32 cm wide, the detection block B is approximately 1 cm square when the shooting range of the camera device 13 is adjusted to coincide with the entire horizontal surface of the water tank 11. On the other hand, since small fish is assumed to be 2.5 cm or more in length, one small fish will be detected simultaneously by 2 to 4 detection blocks B, and detection of small fish will be suppressed. It is done.

The step of photographing the horizontal plane of the water tank 11 with the camera device 13 every predetermined time may be executed, for example, every 0.2 to 2 seconds.

The process of determining the presence of small fish for each of the detection blocks B from the captured image of the aquarium 11 and accumulating the number of presence detection blocks determined to exist as a monitoring history is executed by the monitoring recording unit 14. Here, it is desirable to prevent a small suspended matter or the like from being erroneously determined as a small fish by regarding a fish of a certain size or more as a small fish.
Specifically, for example, each of the detection blocks B is composed of 5 × 5 detection points S, and when any object is photographed at 13 or more detection points S among them, the object is a small fish. It is good to judge its existence by considering it. FIGS. 4B and 4C show specific determination examples.
In the case of FIG. 4B, a part of the small fish F is photographed at 14 detection points S out of 5 × 5 detection points S constituting the detection block B. Since this satisfies the condition of the small fish, the presence of the small fish is determined for this detection block B.
In the case of FIG. 4C, the floating substance U is photographed at 8 detection points S out of 5 × 5 detection points S constituting the detection block B. Since this does not satisfy the conditions for the small fish, the presence of the small fish is not determined for this detection block B. It should be noted that the number of detection points S constituting the detection block B described here and the conditions for estimating small fishes are merely examples, and are not limited thereto.
FIGS. 5A and 5B show an example of a captured image of an aquarium and an example of a determination result obtained by determining the presence of small fish from the captured image. In FIG. 5B, the exclusion block NB is indicated by a white frame, and the small fish presence detection block EB is indicated by a hatched frame.
The dead fish DF caught on the dead fish collection net 11j is automatically excluded from the detection target by the exclusion block NB. Since there are 11 small fish F living and the number of presence detection blocks is 37, the number of presence detection blocks per animal is about 3.4.
Even if the number of small fish F is the same, the number of presence detection blocks should vary within a certain range for each captured image, and therefore a short-time moving average value is calculated and used as the number of presence detection blocks. It is desirable. Specifically, for example, a horizontal plane of the aquarium 11 is photographed every 0.5 seconds, the presence of small fish is determined for each detection block B from the photographed image, and a one-minute moving average value of the presence detection block EB is calculated. Then, the value is adopted as the current number of presence detection blocks. The number of presence detection blocks is not limited to the 1-minute moving average value of the presence detection block EB, and may be a 30-second moving average value or a 3-minute moving average value, for example.
What is necessary is just to accumulate | store the present number of presence detection blocks calculated | required in this way as a monitoring history.

With reference to the monitoring history, the monitoring recording unit 14 and the alarm unit 15 execute a process of alarming acute poisoning when the number of presence detection blocks falls below the alarm standard in a predetermined period.
Specifically, the present number of presence detection blocks is calculated as a one-minute moving average, and at the same time, a 96-hour moving average value of the number of presence detection blocks is calculated, and 50% of the 96-hour moving average value is acute. As a poison alarm standard, an acute poison alarm may be issued when the number of presence detection blocks falls below the alarm standard. The 96-hour moving average value changes slowly even in a situation where small fish F dies one after another. Therefore, even if 50% of the 96-hour moving average value is used as an alarm standard for acute poisoning, the acute poison There will be no inconveniences such as delayed alarms. In addition, when the number of presence detection blocks falls below the alarm standard, timing by the monitoring timer is started, and then the state where the number of presence detection blocks falls below the alarm standard continues until a predetermined time (for example, 3 minutes) elapses. An acute poison alarm may be issued as a condition. Of course, if the number of presence detection blocks recovers the alarm standard before the predetermined time has elapsed, the acute poison alarm is stopped. In this way, false alarms can be suppressed.

FIG. 6 is a graph showing the results of monitoring the number of small fish individuals for 7 days (168 hours) by the above method. The vertical axis of the graph G1 represents the number of presence detection blocks (number of individuals). The initial number of small fish is 12. Moreover, the period of 7 days is an example, and water quality monitoring is generally performed continuously for a long period of time.
In the graph G1, the solid line indicates the number of presence detection blocks. The number of presence detection blocks gradually decreases with the death of small fish.
The broken line indicates the 96-hour moving average value of the number of presence detection blocks, which is gently decreased after the number of presence detection blocks.
The alternate long and short dash line indicates 50% of the 96-hour moving average, which is used as a criterion for determining the half death due to acute poisoning at each time point, that is, as an acute poison warning criterion. When the number of presence detection blocks falls below this alarm standard, an acute poison alarm is issued. At this point, a portion of the water in the aquarium may be transferred to the sample water tank for storage.
The two-dot chain line indicates 75% of the 96-hour moving average value, which is used as a criterion for determining the death due to acute poisoning at each time point, that is, the acute poison warning standard. When the number of presence detection blocks falls below this warning standard, an acute poisoning warning is issued. Thus, if it is configured to issue an acute poisoning warning at an earlier point than the acute poisoning warning, preparations such as countermeasures for acute poisoning can be performed with a margin.
In this example, one animal died at 26 hours, and an acute poison warning was issued when a total of 4 animals died at 111 hours. In addition, an acute poison warning was issued when a total of 8 animals died at 140 hours.

FIG. 7 is a flowchart for explaining an example of the basic procedure of the method. Here, only the actual procedure of monitoring will be described assuming that the process of dividing the entire horizontal surface of the water tank into a plurality of detection blocks has already been performed.
Steps 100 and 101 are steps in which the entire horizontal surface of the water tank is photographed every predetermined time (for example, 0.5 seconds) by the camera device.
Steps 102 and 103 are steps of determining the presence of small fish for each detection block from the captured image of the aquarium and accumulating the number of presence detection blocks as a monitoring history.
Steps 104 to 108 are steps for alarming acute poisoning with reference to the monitoring history when the number of presence detection blocks falls below the alarm criterion during a predetermined period.
Specifically, in step 104, a one-minute moving average value of the number of presence detection blocks is calculated. By treating this average value as the current number of presence detection blocks, the influence of detection variations is suppressed.
In step 105, a 96-hour moving average value of the number of presence detection blocks is calculated. 50% of this average value is used as a warning standard for acute poisons, and 75% is used as a warning standard for acute poisons.
In steps 106 and 107, the current number of presence detection blocks is compared with the alarm standard for acute poisons, and if the number of presence detection blocks is below this alarm standard, an acute poison alarm is issued. This basic procedure ends when the acute poison alarm is issued, but may return to step 100 without ending.
In steps 108 and 109, the current number of presence detection blocks is compared with an acute poison warning standard, and if the number of presence detection blocks is below this warning standard, an acute poison warning is issued. After this, the process returns to step 100.

Next, the monitoring of the amount of activity that is useful when executed together with the monitoring of the number of individuals will be explained. In short, small fishes exhibit the property that their behavior changes in response to acute poisons, etc., so the activity amount of small fishes should be monitored to help predict acute poisons. For example, the activity of small fish tends to increase rapidly at the stage when acute poison begins to mix in water (frenzy behavior). In addition, as the concentration of acute toxin increases, the activity of small fish tends to decrease rapidly (slow behavior). Warning of such behavioral changes can be very helpful in predicting acute toxins. Also, if the date and time when such behavior change occurred is recorded, when an acute poison that causes more than half of small fish to die later is detected, the date and time when the acute poison began to occur, and the concentration of the acute poison is high. This is a useful clue to analyze the date and time.

The activity amount monitoring method is similar to the individual number monitoring method.
That is, the monitoring method of the activity amount by the monitoring device 10 is that the entire horizontal surface of the aquarium 11 is photographed by the camera device 13 every predetermined time, and the small fish enters the block for each detection block B from the photographed image of the aquarium 11. It consists of the basic procedure of accumulating the number of intrusion detection blocks that have been determined to be entered as a monitoring history, and warning the activity abnormality when the number of intrusion detection blocks reaches the warning criteria by referring to the monitoring history. . The detection block B can be used as it is for monitoring the number of individuals.

The process of determining the entry of small fishes into the block for each of the detection blocks B from the captured image of the aquarium 12 and accumulating the number of entry detection blocks determined to enter as a monitoring history is executed by the monitoring recording unit 14. The
Specifically, the presence of small fish is determined for each of the detection blocks B from the two most recent consecutive captured images, and the determination results are compared, and the small fish enters the detection block B. Judging. That is, the presence of small fish is not determined last time, and the entry of small fish may be determined for the detection block B in which the presence of small fish is determined this time.

FIGS. 8A and 8B show an example of a captured image of an aquarium and an example of a determination result for determining the entry of a small fish from the captured image.
In FIG. 8A, among two consecutive captured images, the small fish F in the current captured image is indicated by a solid line, and the small fish F in the previous captured image is indicated by a dotted line. That is, the small fish that was at the position of the dotted line in the previous captured image has moved to the position of the solid line in the current captured image.
Since the entry detection block MB is a block in which the presence of small fish has not been determined in the previous time and the presence of small fish has been determined in this time, the position and dotted line of the small fish indicated by the solid line in FIG. As can be seen by comparing the position of the small fish shown, the entry detection block MB is shown by a hatched frame in FIG. 8B.
In this example, there are a total of 11 small fish F that survive, but the total number of entry detection blocks is 14. The number of intrusion detection blocks is proportional to the amount of activity of the entire small fish F if the time difference between the previous shooting and the current shooting is small, that is, if the moving distance of the small fish F is shorter (less than the total length of the small fish). It will be a thing. For example, if 0.5 second is adopted as the time difference, a good detection result can be obtained. For example, the one-minute moving average value of the number of entering blocks obtained in this way may be treated as the value of the number of entering blocks at the current time.

With reference to the monitoring history, when the number of approach detection blocks reaches the warning standard, the process of warning the activity abnormality is executed by the monitoring recording unit 14 and the warning unit 15.
Although there are no particular restrictions on the warning criteria here, the moving average value of the activity amount for the predetermined period is obtained from the monitoring history, and the warning reference value for the sudden increase in activity amount and the warning reference value for the sudden decrease in activity amount are based on that value. It may be determined.
Specifically, as described above, the moving average value for 1 minute is calculated as the current number of intrusion detection blocks, and at the same time, the 96 hour moving average value of the number of intrusion detection blocks is also calculated. Warning criteria for sudden increase in volume, 50% of the value as warning criteria for sudden decrease in activity amount, when the current number of detected blocks exceeds the warning criterion for sudden increase in activity amount, or falls below the warning criterion for sudden decrease in activity amount In addition, a warning is given for abnormal activity. When the number of intrusion detection blocks exceeds the warning criteria for sudden increase in activity amount or falls below the warning criteria for sudden decrease in activity, time monitoring by the monitoring timer is started, and then the entry detection is made until a predetermined time (for example, 3 minutes) has passed. An activity abnormality warning may be issued on the condition that the number of blocks has exceeded the warning criterion for sudden increase in activity amount or has remained below the warning criterion for sudden decrease in activity. Of course, if the number of intrusion detection blocks recovers from the alarm standard before the predetermined time has elapsed, the alert for the activity abnormality warning is stopped. In this way, false alarms can be suppressed.
Such behavioral changes occur when the acute poison begins to occur and when the concentration of the acute toxin becomes high. It is also possible to determine the strength. This is because the time difference between the warning of acute poison and the warning of abnormal activity (rapid increase in activity) is roughly the same as the time required for half of the small fish to die. It is based on the consideration that can be estimated. For example, when an acute poison is warned, if an alarm of abnormal activity (rapid increase in activity amount) has been warned 48 hours before that, it can be presumed that an acute poison that is about twice as strong as the lethal concentration for 96 hours is mixed.
Therefore, at least when the acute poison is alarmed, the monitoring device 10 refers to the monitoring history, estimates the strength of the acute poison based on the transition of the number of presence detection blocks and the number of detection blocks, and displays on the display unit 16. It is good to make it the structure to display.
Such a behavior change is not only abnormal in water quality but also occurs normally when a small fish is surprised by something, for example, when it makes a loud sound or the water tank shakes. It also occurs when water supply stops or the water temperature fluctuates. Therefore, a microphone that detects noise, an accelerometer that detects vibration, etc. may be provided, and when these sensors detect an abnormality or when water supply stops, an activity abnormality warning may not be issued for a certain period of time. .

FIG. 9 is a graph showing the results of monitoring the activity amount of small fish for 7 days (168 hours) by the activity amount monitoring method. The vertical axis of the graph G2 is the number of entry detection blocks.
In the graph G2, a solid line indicates the number of entrance detection blocks (activity amount). The number of intrusion detection blocks suddenly increases from the 15th hour after the start of monitoring, once returns to the original level, rapidly decreases at the 84th hour, and then returns to the original level. Although the number of intrusion detection blocks actually varies greatly, it is assumed here that it has changed in stages for simplicity.
The broken line indicates the 96-hour moving average value of the number of the entrance detection blocks, and gently decreases behind the number of the entrance detection blocks.
The alternate long and short dash line indicates 50% of the 96-hour moving average, which is used as a warning standard for sudden decrease in activity at each time. When the number of intrusion detection blocks falls below this warning standard, an activity amount sudden decrease warning is issued.
The two-dot chain line shows 150% of the 96-hour moving average value, which is used as a warning standard for sudden increase in activity at each time. When the number of intrusion detection blocks exceeds this warning threshold value, an activity amount sudden increase warning is issued.
In this example, an activity amount sudden increase warning is issued at the 15th hour. This warning suggests the occurrence of water quality abnormalities such as acute poisoning. Also, a warning about sudden decrease in activity is issued at 84 hours. This warning suggests the seriousness of water quality abnormalities.

FIG. 10 is a flowchart for explaining an example of a basic procedure of the activity amount monitoring method. Here, only the actual procedure of monitoring will be described assuming that the process of dividing the entire horizontal surface of the aquarium into a plurality of detection blocks has been executed.
Steps 200 and 201 are steps in which the entire horizontal surface of the water tank 11 is photographed every predetermined time (for example, 0.5 seconds) by the camera device.
Steps 202 and 203 are steps in which the entry of small fish is determined for each detection block from the captured image of the aquarium, and the number of entry detection blocks for which entry has been determined is accumulated as a monitoring history.
Steps 204 and 205 are steps for alarming acute poisoning when the number of presence detection blocks falls below an alarm criterion during a predetermined period with reference to the monitoring history.
Specifically, in step 204, a one-minute moving average value of the number of entering detection blocks is calculated. By treating this average value as the current number of intrusion detection blocks, it is possible to suppress the influence of detection variations.
In step 205, a 96-hour moving average value of the number of entry detection blocks is calculated. 50% of the average value is used as a warning criterion for sudden decrease in activity amount, and 150% is used as a warning criterion for sudden increase in activity amount.
In steps 206 and 207, the current number of detected entry blocks is compared with a warning criterion for sudden increase in activity amount, and if the number of detected approach blocks exceeds this warning criterion, an alert for sudden increase in activity amount is issued. After this, the process returns to step 200.
In steps 208 and 209, the current number of approaching detection blocks is compared with a warning criterion for a sudden decrease in activity amount. If the number of approaching detection blocks is below this warning criterion, an alert for sudden decrease in activity amount is issued. After this, the process returns to step 200.

As described above, the monitoring device 10 warns of acute poisoning or warns of abnormal activity of small fish, but during monitoring, the monitoring history may be displayed on the display unit in real time. A specific example of the display of the monitoring history will be described below.

FIG. 11 is a specific example of a monitor history display screen.
This display screen W1 includes a monitoring data display field S1 for the number of small fish individuals, activity amount, etc., a graph display field S2 for monitoring data such as the number of small fish individuals, activity amount, etc., and a captured image display field S3 for the aquarium 11. And are provided.
In the monitoring data display column S1, the number of individuals, the amount of activity, etc. for each time may be read from the monitoring history and numerically displayed. The time unit is not particularly limited, and it is preferable that every 2 hours, 1 hour, 10 minutes, 1 minute, or the like can be freely selected. Here, the number of individuals per hour, the amount of activity, the alarms generated at that time, the alarms issued, etc. are grouped in slots, and 12 hours are displayed in a table format. The display range of the table T1 may be moved by a scroll operation or the like.
The graph display field S2 may be a graph obtained by reading the number of individuals, the amount of activity, etc. for each time from the monitoring history. It is preferable that the period for displaying the graph can be freely selected from 7 days, 3 days, 1 day, and the like. The display range of the graph may be freely moved by a scroll operation or the like. Here, the number of individuals and the amount of activity are displayed as independent graphs G3 and G4, respectively.
In graph G3, for 7 days from 6/4 to 6/10, the number of individuals, 96-hour moving average, warning standard for acute poison (50% of 96-hour moving average), warning standard for acute poison (96-hour moving average) 75%).
In graph G4, for 7 days from 6/4 to 6/10, the activity amount, the 96-hour moving average, the warning criterion for sudden decrease in activity amount (50% of the 96-hour moving average), and the warning criterion for sudden increase in activity amount (96 150% of the time moving average).
In the captured image display column, the captured image of the date and time selected by the predetermined operation in the monitoring data display column or the captured image P1 of the date and time selected by the predetermined operation in the graph display column is switched and displayed.
If it is set as a display mode like graphs G3 and G4, the transition of the number of individuals and activity amount of small fish can be grasped only by looking at graphs G3 and G4.
It is convenient to accept a date designation operation or the like so that the daily report data of the corresponding date, that is, the monitoring history for that day can be called up in the monitoring data display column S1 and the graph display column S2.

FIG. 12 shows another example of the monitor history display screen. This display screen W2 is different from the display screen W shown in FIG. 11 in the form of the graph displayed in the graph display field S2. Other common elements are denoted by the same reference numerals, and description thereof is omitted.
The graph G5 here shows the number of individuals against the warning standard for acute poison, the number of individuals against the warning standard for acute poison, the amount of activity against the warning standard for sudden decrease in activity amount, and the warning standard for sudden increase in activity amount. As a separate line for the amount of activity against each other, the alarm standard for acute poison, the warning standard for acute poison, the warning standard for sudden decrease in activity amount, and the warning standard for sudden increase in activity amount are placed in the same position. It is displayed superimposed on one screen.
If the display mode is as shown in the graph G5, the corresponding alarm and warning are issued when each line falls below the standard (when the sudden increase in activity amount exceeds the standard), so the graph G5 was seen. It will be possible to predict when alarms and warnings will be issued.

The graph G1 will be described in more detail.
FIGS. 13 (a) to (d) show the number of individuals with respect to the warning standard for acute poison, the number of individuals with respect to the warning standard for acute poison, the amount of activity with respect to the warning standard for sudden decrease in activity, and the amount of activity. This shows the amount of activity against the rapidly increasing warning standard as an independent graph.
For example, in the graph G3 of FIG. 11, when the number of individuals is normalized by the acute toxic alarm criterion, that is, when the number of individuals is divided by the acute toxic alarm criterion at each time, the divided number of individuals becomes the acute toxic alarm criterion. Relative value to. Therefore, if the relative value of the alarm standard for acute poisoning is plotted on a horizontal line, a graph G6 shown in FIG. 12A is obtained.
Similarly to the graph G3 of FIG. 11, the number of individuals with respect to the warning standard for acute poisoning, the amount of activity with respect to the warning standard for sudden decrease in activity amount, and the amount of activity with respect to the criterion for rapid increase in activity amount are shown in FIG. ), (C), and graphs G7, G8, and G9 as shown in (d) are obtained.
In graphs G6 to G9 shown in FIGS. 13 (a) to 13 (d), the warning standard for acute poison, the warning standard for acute poison, the warning standard for sudden decrease in activity amount, and the warning standard for sudden increase in activity amount are in the same position. When superimposed, a graph G5 shown in FIG. 11 is obtained.

In the method of monitoring the number of small fish and the amount of activity by analyzing the captured image of the aquarium as described above, a normal monitoring result cannot be obtained if the aquarium water is more than a certain level. Therefore, during monitoring of small fish, the turbidity of water may be monitored, and the alarm and warning may be prohibited while the turbidity exceeds a predetermined standard.
Turbidity can be easily quantified from the average brightness of the entire captured image of the aquarium. For example, when the average brightness of the entire captured image of the aquarium is measured every predetermined time and accumulated in the monitoring history, and the current average brightness falls below the brightness standard calculated from the predetermined period with reference to the monitoring history In addition, it is possible to determine the high turbidity and prohibit the alarm and warning. This greatly reduces false alarms.
The turbidity data may be displayed numerically or graphically in the same manner as the monitoring data such as the number of small fish individuals and the amount of activity on the monitoring history display screen.

DESCRIPTION OF SYMBOLS 10 Monitoring apparatus 11 Water tank 13 Camera apparatus 12 Operation part 14 Monitoring recording part 15 Alarm part B Detection block EB Presence detection block F Small fish MB Intrusion detection block

Claims

Always supply water from a water source to a tank where multiple small fishes were released,
Dividing the entire horizontal surface of the aquarium into a plurality of detection blocks smaller than a horizontal image of small fish,
The entire horizontal surface of the aquarium is photographed every predetermined time by a camera device,
Determine the presence of small fish for each of the detection blocks from the captured image of the aquarium, accumulate the number of presence detection blocks determined to exist as a monitoring history,
An acute poison monitoring method using small fish, characterized in that an acute poison is warned when the number of presence detection blocks falls below a warning standard within a predetermined period with reference to the monitoring history.
In the monitoring method of acute poison using the small fish of Claim 1,
The method for monitoring acute poisoning using small fish, wherein the alarm criterion is calculated from an average value of the number of presence detection blocks over the predetermined period.
In the monitoring method of acute poison using the small fish of Claim 1 or 2,
Determining the entry of small fish into the block for each of the detection blocks from the captured image of the aquarium, further accumulate the number of entry detection blocks determined to enter, as a monitoring history,
An acute poison monitoring method using a small fish, wherein an abnormal activity is warned when the number of intrusion detection blocks reaches a warning standard with reference to the monitoring history.
In the monitoring method of acute poison using the small fish of Claim 3,
At least when an acute poison is alarmed, a small fish characterized in that the strength of acute poison is estimated and displayed based on the transition of the number of presence detection blocks and the number of entry detection blocks with reference to the monitoring history. A method for monitoring acute poisons used.
In the monitoring method of acute poison using the small fish as described in any one of Claims 1 thru | or 4,
Each of the detection blocks has a predetermined number of detection points arranged,
The method for monitoring acute poisons using small fish, wherein the determination of the presence of small fish in each of the detection blocks is performed based on the number of detection points at which the small fish in the block is detected.
In the monitoring method of acute poison using the small fish as described in any one of Claims 1 thru | or 5,
The predetermined period is 96 hours;
An acute poison monitoring method using small fish, characterized in that an acute poison is warned when the number of presence detection blocks decreases to half or less during the predetermined period.
A tank where water is constantly supplied from a monitored water source and a plurality of small fishes are released;
An operation unit that receives an operation of setting a plurality of detection blocks smaller than a horizontal image of small fish over the entire horizontal surface of the aquarium;
A camera device for photographing the water tank every predetermined time;
A monitoring recording unit that determines the presence of small fish for each of the detection blocks from the captured image of the aquarium, and accumulates the number of presence detection blocks determined to exist as a monitoring history;
An acute poison monitoring apparatus using small fish, comprising an alarm unit for alarming acute poison when the number of presence detection blocks falls below an alarm standard for a predetermined period with reference to the monitoring history .