Fish sampling and counting equipment, statistical method thereof and method for monitoring total number of fishes in net cage

Abstract

The invention provides a fish sampling and counting device, a counting method thereof and a net cage fish total number monitoring method.The fish sampling and counting device comprises a counting grid frame, a suspension supporting device, a counting device and a control system.The suspension supporting device is connected with the counting grid frame and used for keeping the counting grid frame suspended in a water body; a plurality of counting grid spaces which are arranged at intervals in the horizontal direction are arranged in the counting supporting frame; the counting device comprises a plurality of detection units, each counting grid space is correspondingly provided with one detection unit, each detection unit comprises a plurality of detectors arranged on the periphery of the counting grid space, and the detectors generate detection light rays crossing counting grid holes; and the detectors of the counting devices are all in communication connection with the control system.

Description

Technical Field

[0001] This invention relates to the field of aquaculture monitoring technology, and in particular to a fish sampling and counting device and its statistical method, as well as a method for monitoring the total number of fish in net cages. Background Technology

[0002] In aquaculture, particularly in the cage culture of golden pomfret, accurate fish population counting is crucial for optimizing feeding, disease control, and harvest management. Golden pomfret is a high-value marine fish widely farmed in China and Southeast Asia, and precise monitoring of its population directly impacts aquaculture efficiency and resource management.

[0003] In existing aquaculture production, fish population monitoring mainly relies on the following methods, which can be effective under certain conditions, but still have limitations in long-term online monitoring scenarios in large deep-sea cages:

[0004] 1. Visual estimation: This method typically involves estimating fish populations through sampling or direct observation at the edge of the net cage. While suitable for small-scale, nearshore, or indoor aquaculture, it is less effective for deep-sea net cages with diameters of tens of meters and significant depths. Relying solely on visual estimation or sampling for extrapolation often requires multiple people working together, has a limited observation frequency, makes it difficult to obtain continuous data, and results are susceptible to weather conditions, observation location, and the experience of the operators, making accuracy and stability difficult to guarantee.

[0005] 2. Acoustic Detection: Sonar or underwater acoustic equipment has been used to detect fish schools and has been applied to some extent in fisheries resource assessment. However, inside deep-sea cages, the cage structure, counterweight chains, and aeration facilities generate multipath and reflected signals. Furthermore, when fish schools are densely clustered, individual echoes tend to overlap, and the data is easily affected by factors such as water quality and environmental noise, making data interpretation relatively complex. It is usually more suitable for estimating relative biomass or group distribution, and it is quite difficult to conduct daily, detailed monitoring of the number of fish in a single cage.

[0006] 3. Optical Imaging: This method uses underwater cameras to capture images or videos, combined with image recognition algorithms to detect and count fish. It achieves good results in conditions of good water transparency and moderate fish density. However, in actual deep-sea cages, image quality and recognition performance are easily affected by suspended particles, attached organisms, nighttime lighting conditions, and the three-dimensional overlap of fish schools. This makes it difficult to consistently distinguish all individuals and requires high-performance hardware and computational power. It is primarily used for localized behavioral observation or sampling analysis, and is not suitable for long-term online statistics of the entire fish population within the cage.

[0007] 4. Channel Counter: This device guides fish through a counting section sequentially using narrow channels or transfer pipes, and sensors count the fish. It has been used in land-based factory farming and graded transfer systems. However, for larger open-type deep-sea cages, constructing additional forced guidance channels often requires significant modifications to the existing cage structure and water flow environment, resulting in high construction and maintenance costs. Furthermore, it may increase stress on fish under high-density, long-term farming conditions. Therefore, its widespread application in the daily management of deep-sea cages is somewhat limited.

[0008] In deep-sea cage aquaculture, water currents, turbidity, and the random movement of fish further increase the difficulty of fish counting. Current technologies lack an efficient, accurate, and minimally disturbing counting device for fish. Therefore, there is an urgent need for a fish sampling and counting device that can achieve automated and high-precision fish quantity monitoring in complex aquatic environments. Summary of the Invention

[0009] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a fish sampling and counting device and its statistical method, as well as a method for monitoring the total number of fish in net cages, which can realize automated, high-precision and stable fish sampling and counting in complex aquatic environments, and thereby automatically monitor the number of fish in net cages over a long period of time.

[0010] To achieve the above objectives, the present invention provides a fish sampling and counting device, comprising a counting grid frame, a suspension support device, a counting device, and a control system. The suspension support device is connected to the counting grid frame to keep the counting grid frame suspended in the water. The counting support frame has multiple counting grid spaces arranged at intervals in the horizontal direction. The counting device includes multiple detection units, with one detection unit corresponding to each counting grid space. Each detection unit includes multiple detectors arranged around the perimeter of the counting grid space, and the detectors generate detection light rays that pass through the holes of the counting grid. The detectors of the counting device are all communicatively connected to the control system.

[0011] Furthermore, the counting grid frame includes multiple concentric circular frames with different radii, and radial connecting rods connecting adjacent circular frames. There are one or more radial connecting rods between adjacent circular frames, and when there are multiple rods, they are distributed along the circular frames at an annular interval, dividing the space between adjacent circular frames into multiple counting grid spaces.

[0012] Furthermore, the multiple detectors of the detection unit are distributed at equal intervals along the circular frame, and the arc length distance between adjacent detectors is 0.5~0.7m.

[0013] Furthermore, the counting grid frame includes a spiral planar spiral frame and multiple partitions disposed in the spiral space of the planar spiral frame. The multiple partitions are distributed at intervals in the spiral space of the planar spiral frame, dividing the spiral space into multiple counting grid spaces.

[0014] Furthermore, the suspension support device includes a suspension body and a support rod. The suspension body is suspended in the water and located below the counting grid frame. The bottom end of the support rod is fixedly connected to the suspension body and to the counting grid frame.

[0015] The present invention also provides a statistical method for the above-mentioned fish sampling and counting device, comprising the following steps:

[0016] A1. Number the detectors in each detection unit of the counting device according to the arrangement order; after the counting device and control system are turned on, the fish passing through the corresponding counting grid space are detected through each detection unit. The control system collects and stores the detection signals of all detectors in the detection units. When the detection light of a detector is blocked, it is recorded as a trigger event.

[0017] A2. The detection data of each detection unit is processed in the following manner:

[0018] A21. The control system records the timestamp t of the trigger event for each detector. 触 The duration of occlusion Δt and the signal amplitude F are denoted as the original triggering event Y0(i, t). 触 , Δt, F), where i is the detector number;

[0019] A22. Noise Reduction and Invalid Trigger Filtering: Set abnormal trigger judgment conditions to judge the original trigger event Y0(i,t). 触 If the event Y(i, t) is a foreign object triggered by something other than a fish, then discard it to obtain the denoised trigger event Y1(i, t). 触 , Δt, F);

[0020] A23. Merge and Deduplication: Set merge and deduplication conditions, and determine if there are several events Y1(i,t) triggered after noise reduction. 触 If the conditions for merging and deduplication (Δt, F) are met, these noisy trigger events are merged into one valid passing event. Individual noisy trigger events that do not meet the conditions for merging and deduplication are considered as a single valid passing event. Finally, the number of valid passing events of the detection unit is obtained.

[0021] A3. Divide the continuous counting time into multiple statistical time windows of length ΔT. Summarize the total number of valid passes for all detection units within each statistical time window as the total number of passes C for that statistical time window, and obtain the counting time series. , where k represents the number of the statistical time window; based on Calculate the mean of all statistical time windows, and denote it as the average count intensity corresponding to that continuous counting time. .

[0022] Furthermore, in step A23, the method for determining whether the merging and deduplication conditions are met is as follows: set an adjacent trigger interval threshold, such as triggering events Y1(i,t) after certain denoising. 触 If the numbers in (Δt, F) are adjacent and the interval between timestamps is less than the threshold of adjacent trigger interval, then the merging and deduplication conditions are met.

[0023] This invention also provides a method for monitoring the total number of fish in a net cage, using the aforementioned fish sampling and counting equipment, comprising the following steps:

[0024] S1. Initial calibration:

[0025] S11. During the fish fry release stage, obtain the initial number of fish released into the net. ;

[0026] S12. After the fish fry are released, a calibration period is set up. During the calibration period, the fish sampling and counting equipment is installed in the net cage and operates continuously. The average counting intensity during the calibration period is obtained through the above statistical methods. And calculate the calibrated unit fish sampling count intensity parameters. ;

[0027] S2. Estimation based on actual measurements:

[0028] S21. Install the fish sampling and counting device in the net cage and operate it continuously. Using the statistical method described above, obtain the average counting intensity corresponding to a continuous counting time segment ending at the current time t, and record it as... ,

[0029] S22. Determine the actual unit fish sampling count intensity parameters. Based on changes in aquaculture time conditions, decide whether to adjust the sampling and counting intensity parameter for the calibrated unit fish. Make corrections; if necessary, set the correction method and perform the correction to obtain the desired result. If no correction is needed, then let ;

[0030] S23. Estimate the number of fish at the current time t. .

[0031] Furthermore, in step S22, the correction method includes: S221, fish age behavior correction: setting a behavior correction factor that changes with fish age. , ,in This is a parameter used to characterize the age of the fish at the current moment.

[0032] Furthermore, in step S22, the correction conditions also include: S222, Environmental and operating condition correction: setting an environmental correction factor. , .

[0033] As described above, the fish sampling and counting device and statistical method thereof, as well as the method for monitoring the total number of fish in net cages, which are involved in this invention, have the following specific beneficial effects:

[0034] 1. The fish sampling and counting device effectively reduces the impact on water areas and fish farming by setting up a counting grid frame with multiple counting grid spaces and detecting the number of fish passing through the counting grid space through multiple detection units. In addition, combined with scientific statistical methods, it can automatically and accurately count the number of fish passing through the counting grid frame, keeping the error of the counting results within a small range.

[0035] 2. The method for monitoring the total number of fish in net cages is based on fish sampling and counting equipment and its statistical methods. By using counting, it can accurately estimate the number of fish in net cages, keeping the error in fish number estimation within a small range. Compared with traditional methods such as manual visual inspection and simple acoustic / optical counting, it significantly reduces error fluctuations. The method for monitoring the total number of fish in net cages can achieve long-term continuous monitoring of fish number estimation, continuously outputting the fish number, which can be used as a basis for optimizing and adjusting aquaculture conditions. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the fish sampling and counting device of the present invention.

[0037] Figure 2 This is a schematic diagram of the counting grid frame and detector in this invention.

[0038] Figure 3 This is a schematic diagram of the planar spiral-shaped counting grid frame in this invention.

[0039] Explanation of icon numbers:

[0040] 1-Net cage, 11-Float frame, 12-Net cover, 2-Counting grid frame, 21-Counting grid space, 22-Circular frame, 23-Radial connecting rod, 24-Planar spiral frame, 25-Separator, 3-Detector, 4-Suspension body, 5-Support rod, 6-Pull rope, 7-Equipment box, 8-Solar panel. Detailed Implementation

[0041] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification.

[0042] It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are merely for illustrative purposes to aid those skilled in the art and to facilitate understanding. They are not intended to limit the scope of the invention and therefore have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to size, without affecting the effectiveness and objectives of the invention, should still fall within the scope of the technical content disclosed herein. Furthermore, the terms "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity and not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention's implementation.

[0043] It should also be noted that when a component is referred to as being "fixed to" or "set on" another component, it can be directly on the other component or may be connected to an intermediary component. When a component is referred to as being "connected to" another component, it can be directly connected to the other component or indirectly connected to the other component through an intermediary component.

[0044] Furthermore, the use of terms such as "first" and "second" in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed in this application.

[0045] See Figure 1 and Figure 3 This invention provides a fish sampling and counting device, including a counting grid frame 2, a suspension support device, a counting device, and a control system. The suspension support device is connected to the counting grid frame 2 to keep the counting grid frame 2 suspended in the water. The counting support frame has multiple counting grid spaces 21 arranged at intervals in the horizontal direction, and each counting grid space 21 is independent of the others. The counting device includes multiple detection units, and each counting grid space 21 is correspondingly provided with a detection unit. The detection unit includes multiple detectors 3 arranged around the counting grid space 21. The detectors 3 generate detection light rays that cross the counting grid holes. The detectors 3 of the counting device are all communicatively connected to the control system.

[0046] The fish sampling and counting device of this invention can be set up in the water of aquaculture cage 1 to collect samples and count fish activity in the aquaculture area, and can be used to calculate the number of fish. The counting grid frame 2 is suspended in the water within the cage 1 by a suspension support device, specifically located within the netting 12 of the cage 1. The size of the counting grid frame 2 is determined based on the horizontal area of ​​the netting 12, and the counting grid frame 2 is set at an appropriate depth from the water surface, which is determined according to the habits of the farmed fish. The counting grid spaces 21 of the counting grid frame 2 are used for fish to pass through. By setting multiple independent counting grid spaces 21, each counting grid space 21 is equipped with a detection unit to detect the fish passing through it. This results in relatively small counting grid spaces 21, which can effectively detect the number of fish passing through them, providing accurate counting and minimal impact from water turbidity. The entire counting grid frame 2 can occupy most of the area above the water in the cage 1 as needed, thus effectively and extensively collecting the number of fish passing through without affecting the main areas where the fish live. This sample quantity can then be used to accurately estimate the number of fish in the cage 1. The fish sampling and counting device can continuously and in real-time sample and count fish in the net cage 1 to obtain the number of fish passing through the counting grid frame 2. For each counting grid space 21, the detection light density of the detector 3 of its corresponding detection unit can be set according to actual needs. When a fish passes through, the detection light of the corresponding detector 3 is blocked, and the detector 3 sends a corresponding trigger signal to the control system. The control system determines the fish passing through the counting grid space 21 within a certain period of time based on the trigger signal. By summarizing the fish passing through all counting grid spaces 21 within a certain period of time, the number of fish passing through the counting grid frame 2 can be obtained as the counting result.

[0047] See Figures 1 to 3 The present invention will be further described below with reference to specific embodiments:

[0048] In this embodiment, see Figure 1 and Figure 2 As a preferred design, the counting grid frame 2 is a concentric circular grid frame structure, including multiple circular frames 22 with different radii and concentric arrangement, and radial connecting rods 23 connecting adjacent circular frames 22. There are multiple radial connecting rods 23 between adjacent circular frames 22 and they are distributed at intervals along the circular frames 22, dividing the space between adjacent circular frames 22 into multiple counting grid spaces 21. For example, there can be four radial connecting rods 23 between adjacent circular frames 22, and they are distributed in a circular array. In this way, the large circular space between two adjacent circular frames 22 is divided into four arc-shaped counting grid spaces 21.

[0049] Preferably, the arc length and radial width of the arc-shaped counting grid space 21 are within an appropriate range, specifically set according to factors such as the size and habits of the fish, neither too large nor too small, to better detect the number of fish passing through and avoid the problem of fish overlapping and affecting the counting judgment. The radial connecting rod 23 between adjacent circular frames 22 can also be a single rod, so that the larger circular space between two adjacent circular frames 22 constitutes a counting grid space 21. The outermost diameter and number of layers of the circular frames 22 can be configured according to the size of the net cage 1 and the target fish species' activity water layer. Preferably, the outermost diameter is approximately 25 meters and the spacing between adjacent layers is approximately 0.6 meters to cover the main activity water layer of the golden pomfret.

[0050] In this embodiment, see Figure 2 Furthermore, the multiple detectors 3 of the detection unit are distributed at equal intervals along the circular frame 22, and the arc length distance between adjacent detectors 3 is 0.5~0.7m. The detection rays of each detector 3 are radial and distributed at equal intervals, so that it can be determined whether there are fish passing through each part of the arc-shaped counting grid space 21 without affecting each other.

[0051] See Figure 3 In another embodiment, the counting grid frame 2 can also adopt a spiral frame structure. The counting grid frame 2 includes a spiral planar spiral frame 24 and multiple separators 25 disposed in the spiral space of the planar spiral frame 24. The planar spiral frame 24 is shaped like a mosquito coil, and the multiple separators 25 are distributed at intervals in the spiral space of the planar spiral frame 24, dividing the spiral space into multiple counting grid spaces 21. The multiple detectors 3 of the detection unit are indirectly mounted on the planar spiral frame 24. In other embodiments, the counting grid frame 2 can also be configured with other shapes, such as rectangles, ellipses, etc., with a long strip shape being preferred.

[0052] In this embodiment, see Figure 2 Furthermore, the counting grid frame 2 is made of water-resistant and corrosion-resistant aluminum alloy or high-strength plastic, and its surface is coated with an anti-biofouling coating to adapt to long-term seawater environments.

[0053] In this embodiment, as a preferred design, the detector 3 in the counting device is an active infrared sensor, including a transmitting end and a receiving end, which are arranged opposite each other on both sides of the counting grid space 21. Specifically, see... Figure 2 The generating end and the receiving end are respectively set on two adjacent circular frames 22. When the fish enters the generating end and the receiving end, the infrared light (detection light) emitted by the generating end is blocked, and the intensity of the infrared light received by the receiving end changes, thereby outputting different signals and sending them to the control system. The control system can record information such as the timestamp corresponding to the signal change.

[0054] In this embodiment, as a preferred design, the suspension support device includes a suspending body 4 and a support rod 5. The suspending body 4 is suspended in the water inside the net cage 1 and located below the counting grid frame 2. The bottom end of the support rod 5 is fixedly connected to the suspending body 4 and to the counting grid frame 2. The suspending body 4 can be a buoyancy box, generating buoyancy and supporting the counting grid frame 2 above it through the support rod 5, thereby effectively stabilizing the counting grid frame 2. Furthermore, the support rod 5 is connected to the counting grid frame 2 in the middle, which reduces the required number and volume of suspending bodies 4, thus minimizing the impact on the aquaculture in the net cage 1. Preferably, the upper end of the support rod 5 extends above the water surface, and the section of the support rod 5 above the water surface is also connected to the float frame 11 of the net cage 1 through multiple ropes 6, thereby stabilizing the horizontal position of the support rod 5 and, consequently, stabilizing the horizontal position of the counting grid frame 2 within the net cage 1.

[0055] In this embodiment, as a preferred design, a lighting device is also included. This device comprises underwater LED lights mounted on the counting grid frame 2, and these lights can be mounted on some or all of the circular frames 22. The lighting parameters, including brightness, flashing frequency, and duty cycle, can be adjusted by the control system to simulate the light stimulation preferred by the farmed fish, thereby increasing the probability of fish activity near the counting grid frame. Different operating modes can also be set according to diurnal variations and aquaculture management needs, such as a weak induction mode during the day and a stronger induction mode at night. By adjusting the lighting and sensor parameters, the system can adapt to different water quality and lighting conditions.

[0056] In this embodiment, as a preferred design, the control system includes an embedded microprocessor capable of preprocessing, event recognition, and statistical summarization of multi-sensor data, executing various processing algorithms, and outputting corresponding processing results. The control system also includes a data storage module and a transmission module for transmitting data signals to the detector 3. It supports local storage and remote data uploading, preferably transmitting the detection data wirelessly to locations such as shore or ship for remote monitoring, facilitating long-term historical data analysis and host computer monitoring.

[0057] In this embodiment, see Figure 1 Furthermore, an equipment box 7 is also provided, which is located at the top of the support rod 5, or it can be located in the float frame 11 of the net cage 1. The microprocessor, data storage module, and transmission module of the control system can all be housed in the equipment box 7. The equipment box 7 also contains a battery power supply device to power the counting device, control system, and lighting device, specifically using a waterproof lithium battery pack as the main power source. Furthermore, a solar panel 8 is also included, installed on the top of the equipment box 7, to charge the battery power supply device, thereby ensuring the long-term use of the fish sampling and counting equipment. Alternatively, the battery power supply device can be connected to an external power supply system.

[0058] The present invention also provides a statistical method for the above-mentioned fish sampling and counting method, comprising the following steps:

[0059] A1. Number the detectors 3 in each detection unit of the counting device according to the arrangement order, or number each detection unit as well; turn on the counting device and control system, and detect the fish passing through the corresponding counting grid space 21 through each detection unit. The control system collects and stores the detection signals of the detectors 3 of all detection units. When the detection light of a detector 3 is blocked, it is recorded as a trigger event.

[0060] A2. The detection data of each detection unit is processed in the following manner:

[0061] A21. The control system records the timestamp t of the trigger event for each detector 3. 触 The occlusion duration Δt and signal amplitude F are denoted as the original trigger event, denoted by Y0(i, t). 触 Let F(t, Δt) represent the signal amplitude, where i is the number of detector 3. This records each original trigger event and its corresponding parameter information. The signal amplitude F reflects the degree to which the detection light is blocked. The less the detection light is blocked, the larger the signal amplitude F. For example, when the detection light is partially blocked by a bubble or a small object, the signal amplitude F output by detector 3 changes, but the signal amplitude F is still relatively large. However, when the light is completely blocked by a fish, the output signal amplitude F changes and becomes 0.

[0062] A22. Noise Reduction and Invalid Trigger Filtering: Set abnormal trigger judgment conditions to judge the original trigger event Y0(i,t). 触 If Δt,F is a foreign object triggering event not caused by a fish, it is discarded to obtain the denoised trigger event, denoted as Y1(i,t). 触 ,Δt,F).

[0063] Preferably, a corresponding threshold can be set based on the duration of occlusion, the signal amplitude (reflecting the degree to which the detection light is blocked), and the interval between adjacent triggers. For a given original trigger event Y0(i,t) 触 The original triggering event Y0(i, t) is judged based on the following sub-condition: Δt, F). For example, if one of the following sub-conditions is met, the original triggering event Y0(i, t) can be considered as such. 触 The following conditions must be met for abnormal triggering: 1) The duration of obstruction Δt is less than the set threshold, indicating that the obstruction is too small and not caused by fish obstruction; 2) The signal amplitude F is greater than the set threshold, indicating that the triggering is caused by foam or small debris; 3) There are objects with the same number i and timestamp t. 触Two original triggers with an interval less than the set threshold indicate that the same detector 3 was blocked for a very short period of time, possibly caused by debris floating nearby. This method can eliminate abnormal triggers caused by floating debris, bubbles, etc., achieving noise reduction and filtering of invalid triggers.

[0064] A23. Merge and Deduplication: Set merge and deduplication conditions, and determine if there are several events Y1(i,t) triggered after noise reduction. 触 If the conditions for merging and deduplication are met, these noisy trigger events are merged into a single valid passing event. Individual noisy trigger events that do not meet the conditions for merging and deduplication are treated as a single valid passing event C. Finally, the valid passing events of the detection unit are obtained.

[0065] Preferably, the method for determining whether the merging and deduplication conditions are met is as follows: set an adjacent trigger interval threshold H, where H is preferably 0.5 to 1 second. If there are several noisy trigger events Y1(i,t) 触 If the numbers of detectors 3 (Δt, F) are adjacent and the interval between their timestamps is less than the threshold H between adjacent triggers, then these detectors 3 are considered to meet the merging and deduplication conditions. This means that the fish is swimming horizontally and these detectors 3 have triggered several times, which can be merged into one valid passing event. However, if a certain noisy trigger event cannot find other noisy trigger events with adjacent numbers and timestamp intervals less than H, then it means that the merging conditions are not met. This means that the fish only passed through one detector 3 at that time, causing only one trigger, which can be considered as one valid passing event. Each valid passing event represents one fish passing through.

[0066] A3. Divide the continuous counting time into multiple statistical time windows of length ΔT. Summarize the total number of valid passes for all detection units within each statistical time window as the total number of passes C for that statistical time window, and obtain the counting time series. , where k represents the number of the statistical time window; based on Calculate the mean of all statistical time windows, and denote it as the average count intensity corresponding to that continuous counting time. Specifically, let the number of statistical time windows be... ,Right now Based on sets This allows us to obtain the total number of fish that pass through during the entire continuous counting time and calculate the average counting intensity. .

[0067] The fish sampling device of the present invention can be used not only in large net cages 1, but also in natural waters (such as coastal areas) to monitor the population dynamics of golden pomfret.

[0068] The present invention also provides a method for monitoring the total number of fish in a net cage, which uses the fish sampling and counting device described above and includes the following steps:

[0069] S1. Initial calibration:

[0070] S11. During the fish fry stocking stage, obtain the initial number of fish stocked in net cage 1. ;

[0071] S12. After the fish fry are released, a calibration period is set up. During the calibration period, fish sampling and counting equipment is installed in the net cages and operates continuously. The average counting intensity during the calibration period is obtained through the above statistical methods. And calculate the calibrated unit fish sampling count intensity parameters. Preferably, the calibration period can be set 1 to 3 days after the fry are introduced, during which the mortality rate is extremely low, and it is assumed that the fish population in net cage 1 remains relatively stable. At the same time, the fish population is close to the dispersion level of normal aquaculture. The length of the calibration period can be set according to actual needs, and one or more can be set.

[0072] During the calibration period, under the assumption of relatively stable fish behavior and environmental conditions, the sampling count intensity parameter per unit fish was then calibrated. ,in The physical meaning is: in the initial stage of stocking, the expected number of effective sampling counts triggered by each fish per unit time; the combined effects of duplicate counting and missed counting are included in this parameter. The control system will... The basic calibration coefficients of cage 1 are saved and used as the basis for subsequent actual monitoring and estimation based on fish age correction and real-time estimation.

[0073] S2. Estimation based on actual measurements:

[0074] S21. Install the fish sampling and counting equipment in the net cage and operate it continuously. Using the statistical method described above, obtain the average counting intensity corresponding to a continuous counting time segment ending at the current time t, and denote it as... Specifically, a continuous count segment with a duration of W ending at the current time t is called a sliding monitoring time window. During continuous monitoring, the sliding monitoring time window slides with time, thereby obtaining the average count intensity that changes over time. .

[0075] S22. Determine the actual unit fish sampling count intensity parameters. Based on changes in aquaculture time conditions, decide whether to adjust the sampling and counting intensity parameter for the calibrated unit fish. Make corrections; if necessary, set the correction method and perform the correction to obtain the desired result. If no correction is needed, then let .

[0076] Specifically, as the rearing time progresses, the fish's body length, weight, and swimming behavior change, altering the water layer where the fish are active, their swimming speed, and their aggregation level. This results in the fish count intensity per unit no longer being constant. Therefore, to ensure the accuracy of the total number calculation, it is necessary to consider the average count intensity as a function of the rearing time t. The corresponding actual unit fish sampling count intensity parameter When changes in aquaculture conditions, such as fish body length, weight, swimming behavior, activity level, swimming speed, and aggregation degree, are small and have little impact on the intensity of fish counts, or when the accuracy of the estimation results is not critical, it may be considered unnecessary to adjust the fish count. Make corrections, that is, at this time .

[0077] When needed Corrections are made to obtain the average count intensity. Matching In this case, the correction methods include:

[0078] S221. Age-related behavior modification: Setting behavior modification factors that change with fish age. , ,in This is a parameter used to characterize the age of the fish at the current moment.

[0079] Specifically, when the intensity of fish count per unit changes over time, to avoid using only the count from the stocking stage... To introduce systematic bias, this embodiment introduces a behavior correction factor that varies with fish age. This will allow the actual unit fish sampling count intensity parameter at any given time to be obtained. Represented as ,in It can be represented by the fish's weight after release or the average weight of the fish. This is used to reflect the changes in the probability and number of repeated passes of fish passing through the counting grid frame 2 of the fish sampling and counting device after the fish transition from the fry stage to the growth stage and fattening stage.

[0080] In practical monitoring work, a staged constant approximation method can be used to determine the constant. Specifically, it includes:

[0081] 1) Divide a complete aquaculture cycle into several growth stages, for example: growth stage 1: fry stage (0-D1) days; growth stage 2: rapid growth stage (D1-D2) days; growth stage 3: fattening stage (D2-D3 days), where D1, D2, and D3 can be set according to the aquaculture experience of the fish species or the company's production procedures.

[0082] 2) Statistical analysis of historical batch data: Using the fish sampling and counting device and statistical method described above, the counting time series and corresponding fish population estimates (e.g., obtained through fish weighing, sampling surveys, or production models) are collected at different growth stages. The average unit fish sampling count intensity parameter for the three growth stages is calculated. , , .

[0083] 3) Define the age correction factor for each growth stage: , Indicates the growth stage.

[0084] In this way, the fish age correction factor for the growth stage is obtained through historical data. Then, the average count intensity at a certain time t and the corresponding actual unit fish sampling count intensity parameters Once it is determined which stage of the fish's growth the monitoring is taking place at, it can be determined... = Furthermore, by using continuous historical data, the complete breeding cycle is divided into more growth stages; that is, the number of growth stages (m) is taken as a larger and more appropriate value to obtain the corresponding... Then, by fitting the function in the form of a continuous function... For example, polynomials or exponential functions can be used to achieve more detailed corrections to the age of the fish.

[0085] S222, Environmental and Operating Condition Correction: Setting Environmental Correction Factors The actual unit fish sampling count intensity parameter is further expanded to Because fish activity levels are influenced not only by age but also by environmental factors such as feeding methods, water temperature, and diurnal rhythms, in order to obtain more accurate data... Environmental correction factors can be introduced. This correction method can be introduced or not, depending on the actual situation.

[0086] In this embodiment, This includes, but is not limited to, whether it is during feeding time (indicator variable), water temperature, light intensity, and diurnal variation. Environmental correction factors can be obtained by fitting historical data using methods such as regression analysis, for example, using a log-linear form. ,in, This indicates whether it is feeding time; if yes, it is set to 1, otherwise it is set to 0. Indicates water temperature. This indicates whether it is nighttime; a value of 1 indicates nighttime, and a value of 0 indicates otherwise. These are parameters estimated from historical data. In most applications, only fish age correction can be enabled as needed, while environmental correction can be omitted, or simplified to a constant correction factor for both feeding and non-feeding modes. and Zero, in non-feeding mode , and All are zero. In other embodiments, the environmental correction factor... It can also be determined through other existing methods, such as setting it based on experience.

[0087] Behavior Modification Factor and environmental correction factors All of these can be set in the control system before the formal monitoring work begins, and the control system can directly call them during the calculation.

[0088] S23. Estimate the number of fish at the current time t. Specifically, without considering modifications, it is: If behavior modification factors are enabled simultaneously and environmental correction factors Then there is If environment modification is not enabled, it can be simplified to .

[0089] Fish population monitoring methods can be performed continuously and in real time, allowing the control system to update the estimated results at preset time intervals (e.g., every 10 minutes or every hour). And output, while also showing the estimation results The data is uploaded to a host computer or cloud platform via a wireless module to achieve long-term automatic monitoring of the number of fish in the net cage.

[0090] To reduce the impact of systematic errors and random fluctuations, the fish population monitoring method in this embodiment can also take the following measures:

[0091] 1) Averaging multiple calibrations: Within the same cage 1's rearing cycle, a short period can be selected again as a secondary calibration period during the middle stage or before harvesting the fish. The corresponding calibration results obtained at this time are then averaged. To fix and update , will the new By combining the data with count data and updating the correction factors at each stage, the accuracy of long-term estimates can be improved.

[0092] 2) Abnormal time window removal: In the count time series The system automatically identifies statistical time windows corresponding to special events such as strong winds and waves, maintenance of cage 1, or abnormal feeding, and marks or removes count data that significantly deviates from normal activity patterns to avoid distorting the average count intensity. The estimation of correction factors causes interference.

[0093] 3) Joint analysis of multiple cages 1: Joint regression analysis is performed on the calibration and operation data of multiple cages 1 in the same aquaculture farm to optimize the correction factors at each stage and improve the robustness of the model.

[0094] The fish sampling and counting device and its statistical method, as well as the method for monitoring the total number of fish in net cages of the present invention, have the following specific beneficial effects:

[0095] 1. The fish sampling and counting device sets up a counting grid frame 2 with multiple counting grid spaces 21, and uses multiple detection units to detect the number of fish passing through the counting grid spaces 21. This can effectively reduce the impact on the water area and fish farming. In addition, combined with scientific statistical methods, it can automatically and accurately count the number of fish passing through the counting grid frame 2, and control the error of the counting results within a small range.

[0096] 2. The method for monitoring the total number of fish in net cages is based on fish sampling and counting equipment and its statistical methods. By using counting, it can accurately estimate the number of fish in net cage 1, and control the error of the fish number estimation within a small range. Compared with traditional methods such as manual visual inspection and simple acoustic / optical counting, it significantly reduces error fluctuations. The method for monitoring the total number of fish in net cages can achieve long-term continuous monitoring of the fish number and continuously output the fish number, which can be used as a basis for optimizing and adjusting the aquaculture conditions.

[0097] In summary, this invention effectively overcomes the various shortcomings of the prior art and has high industrial application value.

[0098] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A fish sampling and counting device, characterized in that: The system includes a counting grid frame (2), a suspension support device, a counting device, and a control system. The suspension support device is connected to the counting grid frame (2) and is used to keep the counting grid frame (2) suspended in the water. The counting support frame has multiple counting grid spaces (21) arranged at intervals in the horizontal direction. The counting device includes multiple detection units, and each counting grid space (21) is provided with a corresponding detection unit. The detection unit includes multiple detectors (3) arranged around the counting grid space (21). The detectors (3) generate detection light rays that cross the counting grid holes. The detectors (3) of the counting device are all communicatively connected to the control system.

2. The fish sampling and counting device according to claim 1, characterized in that: The counting grid frame (2) includes multiple concentric circular frames (22) with different radii and a radial connecting rod (23) connecting adjacent circular frames (22). There are one or more radial connecting rods (23) between adjacent circular frames (22), and when there are multiple rods, they are distributed along the circular frames (22) at an annular interval and divide the space between adjacent circular frames (22) into multiple counting grid spaces (21).

3. The fish sampling and counting device according to claim 2, characterized in that: The multiple detectors (3) of the detection unit are distributed at equal intervals along the circular frame (22), and the arc length distance between adjacent detectors (3) is 0.5~0.7m.

4. The fish sampling and counting device according to claim 1, characterized in that: The counting grid frame (2) includes a spiral planar spiral frame (24) and a plurality of partitions (25) disposed in the spiral space of the planar spiral frame (24). The plurality of partitions (25) are distributed at intervals in the spiral space of the planar spiral frame (24) to divide the spiral space into a plurality of counting grid spaces (21).

5. The fish sampling and counting device according to claim 1, characterized in that: The suspension support device includes a suspension body (4) and a support rod (5). The suspension body (4) is suspended in the water and located below the counting grid frame (2). The bottom end of the support rod (5) is fixedly connected to the suspension body (4) and fixedly connected to the counting grid frame (2).

6. A statistical method for a fish sampling and counting device as described in any one of claims 1 to 5, characterized in that: Includes the following steps: A1. Number the detectors (3) in each detection unit of the counting device according to the arrangement order; after the counting device and control system are turned on, the fish passing through the corresponding counting grid space (21) are detected by each detection unit. The control system collects and stores the detection signals of the detectors (3) of all detection units. When the detection light of a detector (3) is blocked, it is recorded as a trigger event. A2. The detection data of each detection unit is processed in the following manner: A21. The control system records the timestamp t of the trigger event of each detector (3). 触 The duration of occlusion Δt and the signal amplitude F are denoted as the original triggering event Y0(i, t). 触 , Δt, F), where i is the number of detector (3); A22. Noise Reduction and Invalid Trigger Filtering: Set abnormal trigger judgment conditions to judge the original trigger event Y0(i,t). 触 If the event Y(i, t) is a foreign object triggered by something other than a fish, then discard it to obtain the denoised trigger event Y1(i, t). 触 , Δt, F); A23. Merge and Deduplication: Set merge and deduplication conditions, and determine if there are several events Y1(i,t) triggered after noise reduction. 触 If the conditions for merging and deduplication (Δt, F) are met, these noisy trigger events are merged into one valid passing event. Individual noisy trigger events that do not meet the conditions for merging and deduplication are considered as a single valid passing event. Finally, the number of valid passing events of the detection unit is obtained. A3. Divide the continuous counting time into multiple statistical time windows of length ΔT. Summarize the total number of valid passes for all detection units within each statistical time window as the total number of passes C for that statistical time window, and obtain the counting time series. , where k represents the number of the statistical time window; based on Calculate the mean of all statistical time windows, and denote it as the average count intensity corresponding to that continuous counting time. .

7. The statistical method according to claim 6, characterized in that: In step A23, the method for determining whether the merging and deduplication conditions are met is as follows: set an adjacent trigger interval threshold, such as triggering events Y1(i,t) after certain denoising. 触 If the numbers in (Δt, F) are adjacent and the interval between timestamps is less than the threshold of adjacent trigger interval, then the merging and deduplication conditions are met.

8. A method for monitoring the total number of fish in a net cage, characterized in that: The method employs the fish sampling and counting device as described in any one of claims 1 to 5, and includes the following steps: S1. Initial calibration: S11. During the fry stocking stage, obtain the initial number of fish stocked in net cage (1). ; S12. A calibration period is set up for a period of time after the fish fry are released. During the calibration period, the fish sampling and counting device is installed in the net cage and operates continuously. The average counting intensity during the calibration period is obtained by using the statistical method described in claim 6 or 7. And calculate the calibrated unit fish sampling count intensity parameters. ; S2. Estimation based on actual measurements: S21. Install the fish sampling and counting device in the net cage and operate it continuously. Using the statistical method described in claim 6 or 7, obtain the average counting intensity corresponding to a continuous counting time segment ending at the current time t, and record it as... , S22. Determine the actual unit fish sampling count intensity parameters. Based on changes in aquaculture time conditions, decide whether to adjust the sampling and counting intensity parameter for the calibrated unit fish. Make corrections; if necessary, set the correction method and perform the correction to obtain the desired result. If no correction is needed, then let ; S23. Estimate the number of fish at the current time t. .

9. The method for monitoring the total number of fish in a net cage according to claim 8, characterized in that: In step S22, the correction method includes: S221. Age-related behavior modification: Setting behavior modification factors that change with fish age. , ,in This is a parameter used to characterize the age of the fish at the current moment.

10. The method for monitoring the total number of fish in a net cage according to claim 9, characterized in that: In step S22, the correction conditions also include: S222, Environmental and Operating Condition Correction: Setting Environmental Correction Factors , .

Images