Marine fish intelligent culture device and method based on vision and vortex

Abstract

The invention provides an intelligent marine fish culture device and method based on vision and vortex, and relates to the technical field of aquaculture. According to the device, a layered monitoring module is used for collecting environment data, fish body state data and residual bait data, the environment data comprise water temperature data and dissolved oxygen data, and the fish body state data comprise fish body behavior data, barrel bottom capsule distribution data and fish body surface data; the edge calculation unit is used for performing instruction calculation according to threshold values corresponding to the environment data, the fish body state data and the residual bait data to obtain an eddy current regulation and control instruction and a feeding instruction; the intelligent vortex module is used for switching corresponding working modes according to the vortex regulation and control instruction, the working modes comprise a normal breeding mode and a capsule removing mode, and the intelligent feeder is used for feeding the breeding area according to the feeding instruction. The problem of low marine fish disease monitoring and protection efficiency in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aquaculture, in particular to a marine fish intelligent breeding device and method based on vision and vortex. BACKGROUND

[0002] Cryptocryon irritans is a major parasitic ciliate of marine bony fishes, and its life cycle includes four stages: trophont, precyst, cyst and larva. The cyst stage is the key link for disease spread. When the water temperature is 25-27℃, a single cyst can release hundreds of ciliate larvae in 3-7 days, which invade the fish body through the wound or weak mucus layer, leading to large-scale death of fry, causing tens of billions of yuan in losses to the global marine aquaculture industry annually.

[0003] The main problems in disease and pest monitoring and control are as follows:

[0004] The monitoring means is backward: traditional manual observation relies on the experience of breeders, and the cyst misjudgment rate is more than 30%, and the dynamic distribution of the cyst cannot be tracked in real time;

[0005] The control method is limited: the current prevention and control of Cryptocryon irritans mainly includes chemical control, physical control and immune control. The chemicals used in chemical control are harmful to fish, users and the environment, and are prone to drug residues; physical control is time-consuming and labor-intensive and easy to damage the fish; immune control has not been used; the newly proposed biological control relies on the natural environment and lacks a dynamic control mechanism.

[0006] In summary, there is still a technical gap in efficient physical removal and intelligent control of the cyst stage. SUMMARY

[0007] In order to overcome the shortcomings of the prior art, the purpose of the present application is to provide a marine fish intelligent breeding device and method based on vision and vortex, which solves the problem of low efficiency of disease monitoring and protection of marine fish in the prior art.

[0008] To achieve the above purpose, the present application provides the following scheme:

[0009] A marine fish intelligent breeding device based on vision and vortex, comprising:

[0010] an edge computing unit, a breeding barrel main body, a layered monitoring module, an intelligent vortex module and an intelligent feeder provided on the breeding barrel main body at a predetermined position; wherein the barrel bottom of the breeding barrel main body is a conical surface with an inclination angle of 30°, and the inside of the breeding barrel main body includes a breeding area and a vortex area;

[0011] The edge computing unit is connected with the layered monitoring module, the intelligent vortex module and the intelligent feeder respectively;

[0012] The hierarchical monitoring module is configured to collect environmental data, fish body state data, and residual bait data, wherein the environmental data includes water temperature data, dissolved oxygen data, and barrel bottom cyst data, and the fish body state data includes fish body behavior data and fish body surface data; the edge computing unit is configured to perform instruction calculation according to threshold values corresponding to the environmental data, fish body state data, and residual bait data, to obtain vortex control instructions and feeding instructions; the intelligent vortex module is configured to switch a corresponding working mode according to the vortex control instructions, wherein the working mode includes a normal breeding mode and a cyst removal mode, and the intelligent feeder is configured to feed the breeding area according to the feeding instructions.

[0013] Preferably, the hierarchical monitoring module includes:

[0014] a top underwater camera, a side wall underwater camera, and a water quality sensor;

[0015] The top underwater camera is configured to collect fish body behavior data and fish body surface data, the side wall underwater camera is configured to collect barrel bottom cyst distribution data and residual bait data, and the water quality sensor is configured to collect the water temperature data and the dissolved oxygen data.

[0016] Preferably, the breeding barrel body includes:

[0017] a top support, a breeding barrel shell, a grid isolation layer, an outer barrel four-corner platform support, and a conical surface;

[0018] The top support is fixedly installed at an upper end of the breeding barrel shell and is configured to carry the top underwater camera;

[0019] The grid isolation layer is horizontally arranged above an intersection of an inner wall of the breeding barrel shell and the conical surface and has a height corresponding to a top end of the central drain valve, and is configured to separate the breeding area from the vortex area;

[0020] The outer barrel four-corner platform support is located at a periphery bottom of the breeding barrel shell and is configured to support and horizontally adjust the breeding barrel body;

[0021] The conical surface is configured to form a spiral sinking flow field in the cyst removal mode to collect barrel bottom cysts to the central drain valve.

[0022] Preferably, the intelligent vortex module includes:

[0023] an edge tangent water inlet pipe, a circulating water drain, and a central drain valve;

[0024] The edge tangent water inlet pipe is uniformly distributed along an inner wall of the breeding barrel shell in a circumferential direction and is configured to form a slow ring circulation at a flow rate of 0.2-0.3 m / s in the normal breeding mode and to increase to 0.5-0.8 m / s to enhance the spiral vortex in the cyst removal mode.

[0025] A water circulation outlet is arranged on the side wall of the culture barrel shell and is higher than the grid isolation layer, which is used to maintain the water level and discharge the surface residual feed and impurities in the normal culture mode;

[0026] The center drain valve is located at the lowest point of the conical surface, which is used to fully open and discharge the cysts and sewage enriched at the bottom of the barrel during the cyst cleaning mode, and is closed to maintain water stability in the normal culture mode.

[0027] Preferably, the center drain valve is an electromagnetic valve with a diameter ranging from 10 to 15 cm.

[0028] Preferably, the time period of the cyst cleaning mode ranges from 30s to 60s.

[0029] Preferably, the mesh diameter of the grid isolation layer ranges from 2 to 5 mm.

[0030] Preferably, the conical surface is composed of smooth polyethylene material.

[0031] An intelligent sea fish culture method based on vision and eddy current, comprising:

[0032] Image and environmental parameter acquisition: the top camera and the inner side wall camera acquire fish group behavior, cyst distribution at the bottom of the barrel, and fish body surface feature data in real time at a preset frequency, and the water quality sensor synchronously acquires water temperature data and dissolved oxygen data;

[0033] Image preprocessing and feature analysis: noise removal and contrast enhancement processing are performed on the collected images, YOLO11 deep learning algorithm is applied to identify cyst small targets and fish behavior, and cyst density and fish abnormal behavior indicators are output in combination with morphological filtering and spatial positioning;

[0034] Dynamic decision and execution: when the cyst density exceeds the threshold value, the water inflow rate is automatically increased and the drain valve is opened, forming a spiral vortex at the bottom of the barrel to concentrate and discharge the cysts, and after single cleaning, the density is reviewed at a set time interval, and if the density is below the threshold value, the process is stopped, and if the density is above the threshold value, the process is repeated, and the cumulative cleaning times do not exceed 3 times;

[0035] Feeding amount regulation: 3 times of feeding are performed at regular intervals every day, and the feeding amount of each time is 4% of the initial weight of the feeding machine. The residual feed data collected by the side wall camera is used to control the feeding strategy. If the residual feed coverage area identified by the camera is >5% after 1 hour of feeding, the next feeding amount is reduced by 15%; if the residual feed coverage area is identified for 2 consecutive times <2%, the feeding amount is increased by 10%.

[0036] Abnormal early warning and remote management: automatic real-time early warning for cyst over-standard, abnormal water temperature data, and abnormal dissolved oxygen data, and alarm is pushed to the administrator terminal, supporting remote intervention and manual operation.

[0037] The present application discloses the following technical effects:

[0038] The present application provides a kind of based on vision and eddy sea fish intelligent breeding device, comprising: edge computing unit, breeding barrel main part and its layered monitoring module being arranged on the breeding barrel main part preset position, intelligent eddy module and intelligent feeder;Wherein, the barrel bottom of the breeding barrel main part is the circular conical surface of inclination 30 °, and breeding barrel main part inside includes breeding area and eddy area;The edge computing unit is connected with the layered monitoring module, intelligent eddy module and intelligent feeder respectively;The layered monitoring module is used to collect environmental data, fish body state data and residual bait data, wherein the environmental data includes: water temperature data and dissolved oxygen data, and the fish body state data includes: fish body behavior data, barrel bottom cyst distribution data and fish body surface data;The edge computing unit is used to carry out instruction calculation according to the threshold value corresponding to the environmental data, fish body state data and residual bait data, to obtain eddy control instruction and feeding instruction;The intelligent eddy module is used to switch corresponding working mode according to the eddy control instruction, and the working mode includes: normal breeding mode and cyst removal mode, and the intelligent feeder is used to feed the breeding area according to the feeding instruction.The present application monitors cyst distribution and fish behavior in real time by multiple cameras, accurately identifies targets using YOLO11 algorithm, forms natural spiral eddy by 30 ° inclination of barrel bottom and edge tangent water inlet, realizes efficient removal of barrel bottom cyst and intelligent feeding.Constructs "monitoring-eddy-feeding" closed loop system, significantly improves the efficiency of stimulating cryptosporidium cyst removal, reduces energy consumption and labor cost, is suitable for industrialized indoor aquaculture scene, cuts off the key link of disease transmission, and guarantees the healthy growth of grouper. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described in the following only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0040] Figure 1 A three-dimensional schematic view of the sea fish intelligent breeding device based on vision and eddy provided by the embodiment of the present application is provided.

[0041] Figure 2 A three-dimensional schematic view of the sea fish intelligent breeding device based on vision and eddy provided by the embodiment of the present application is provided.

[0042] Figure 3 A bottom valve schematic view of the sea fish intelligent breeding device based on vision and eddy provided by the embodiment of the present application is provided.

[0043] Figure 4 The control logic schematic diagram provided for the embodiments of the present application.

[0044] Reference signs:

[0045] 1 - intelligent feeder; 2 - top support; 3 - water quality sensor; 4 - top underwater camera; 5 - culture barrel shell; 6 - grid isolation layer; 7 - outer barrel four corner platform support; 8 - edge tangent water inlet pipe; 9 - side wall underwater camera; 10 - conical surface; 11 - circulating water drain; 12 - central drain valve. DETAILED DESCRIPTION

[0046] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0047] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0048] As shown in Figures 1-3 The present application provides an intelligent marine fish culture device based on vision and vortex, comprising:

[0049] An edge computing unit, a culture barrel main body, a layered monitoring module, an intelligent vortex module and an intelligent feeder 1 provided on a preset position of the culture barrel main body; wherein the barrel bottom of the culture barrel main body is a conical surface 10 with an inclination angle of 30°, and the inside of the culture barrel main body comprises a culture area and a vortex area;

[0050] The edge computing unit is connected with the layered monitoring module, the intelligent vortex module and the intelligent feeder 1 respectively;

[0051] The layered monitoring module is used for collecting environmental data, fish body state data and residual bait data, wherein the environmental data includes water temperature data and dissolved oxygen data, the fish body state data includes fish body behavior data, barrel bottom cyst distribution data and fish body surface data; the edge computing unit is used for performing instruction calculation according to the threshold values corresponding to the environmental data, fish body state data and residual bait data, to obtain vortex control instructions and feeding instructions; the intelligent vortex module is used for switching the corresponding working mode according to the vortex control instructions, wherein the working mode includes a normal culture mode and a cyst removal mode; and the intelligent feeder 1 is used for feeding the culture area according to the feeding instructions.

[0052] Specifically, the camera equipment is composed of a top underwater camera (resolution ≥ 1080P, waterproof level IP68) and an inner side wall camera (180° wide-angle lens, waterproof level IP68), which are fixed on the barrel by corrosion-resistant supports to realize 360° image collection without dead angle in the breeding barrel, covering the water surface, middle layer of water and barrel bottom area. The top camera is used for monitoring the swimming behavior of grouper and the state of fish body surface, and the inner side wall camera focuses on the distribution data of encapsulation and residual feed at the barrel bottom. The sensor unit integrates a water temperature sensor (accuracy ±0.5℃) and a dissolved oxygen sensor (measurement range 0-20 mg / L, accuracy ±0.2 mg / L) to collect real-time environmental parameters of the breeding barrel and provide data support for dynamic adjustment of the encapsulation detection threshold and the vortex control strategy.

[0053] Barrel bottom structure design:

[0054] The barrel bottom is a conical surface 10 with a 30° inclination angle, and a smart drain valve (solenoid valve) with a diameter of 10-15 cm is arranged at the center to connect an external circulating water system. The water inlet pipes are uniformly distributed at the edge of the barrel bottom (50-80 cm away from the center), and the tangential direction water inlet design (45° angle with the barrel wall) is adopted to ensure the spiral flow of water along the inclination angle of the barrel bottom.

[0055] Working mode:

[0056] Normal breeding mode (without encapsulation):

[0057] The edge water inlet pipe continuously supplies water at a flow rate of 0.2 m / s, and the center valve of the barrel bottom is closed to maintain slow circulation of the water body to avoid residual feed deposition.

[0058] Encapsulation removal mode (encapsulation detected):

[0059] When the inner side wall camera detects that the water temperature is 25℃ and the encapsulation density at the barrel bottom is >5 / ㎡, the flow rate of the edge water inlet pipe is increased to 0.5 m / s, and the center valve is fully opened (opening degree 100%). The 30° inclination angle of the barrel bottom and the tangential water inlet form a clockwise spiral vortex (flow rate 0.6-0.8 m / s) to make the encapsulation gather towards the center with the water flow and be discharged. The single removal time is 30-60 s, and after the end, it is automatically switched back to the normal mode.

[0060] Control logic: combined with the inclination angle characteristics of the barrel bottom, the encapsulation detection threshold is lowered by 1 / ㎡ for every 1℃ increase in water temperature, and the removal is triggered in advance in high temperature environment.

[0061] The 30-minute re-inspection is performed after single removal, and if the density is still >5 / ㎡, the removal process is repeated, and the cumulative removal times are ≤3 to avoid excessive drainage affecting the water quality.

[0062] Self-adaptive feeding system:

[0063] Feeding device: including fish monitoring camera (resolution 6 million pixels) and quantitative feeding machine (accuracy ± 2%), the former monitors the body surface state and swimming behavior of Epinephelus, and the latter feeds 3 times a day, each time is 4% of the initial weight

[0064] Feeding control strategy: the residual bait data collected by the side wall camera controls the feeding strategy, if the residual bait coverage area identified by the camera is > 5% after 1 hour of feeding, the next feeding amount is reduced by 15%; if the residual bait coverage area is identified for 2 times in succession < 2%, the feeding amount is increased by 10%.

[0065] Edge computing unit:

[0066] Built-in YOLO11 target detection algorithm, trained for capsule morphological features (diameter 50-300μm, surface ridge protrusion), recognition accuracy ≥95%, false detection rate <5%; trained for residual bait morphological features, recognition accuracy ≥95%, false detection rate <5%

[0067] Integrated multi-modal data fusion algorithm, fusing capsule density, fish abnormal behavior, water temperature and dissolved oxygen data, outputting valve opening and closing, feeding amount adjustment and other control instructions, forming a "monitoring-analysis-execution" closed loop.

[0068] Further, the hierarchical monitoring module includes:

[0069] Top underwater camera 4, side wall underwater camera 9 and water quality sensor 3;

[0070] The top underwater camera 4 is used to collect fish behavior data and fish body surface data, the side wall underwater camera 9 is used to collect capsule distribution data and residual bait data at the bottom of the barrel, and the water quality sensor 3 is used to collect water temperature data and dissolved oxygen data.

[0071] Further, the culture barrel body includes:

[0072] Top bracket 2, culture barrel shell 5, grid isolation layer 6, outer barrel four corner platform bracket 7, conical surface 10;

[0073] The top bracket 2 is fixedly installed at the upper end of the culture barrel shell 5, and is used to bear the top underwater camera 4;

[0074] The grid isolation layer 6 is horizontally arranged above the intersection of the inner wall of the culture barrel shell 5 and the conical surface 10, and the height corresponds to the top end of the central drain valve 12, and is used to separate the culture area and the vortex area;

[0075] The outer barrel four corner platform bracket 7 is located at the periphery of the bottom of the culture barrel shell 5, and is used to support and horizontally adjust the culture barrel body;

[0076] The conical surface 10 is used to form a spiral sinking flow field in the capsule cleaning mode, and to converge the capsule at the bottom of the barrel to the central drain valve 12.

[0077] Further, the intelligent vortex module comprises:

[0078] The edge tangent water inlet pipe 8, the circulating water drain 11 and the central drain valve 12;

[0079] The edge tangent water inlet pipe 8 is evenly distributed along the inner wall of the culture barrel shell 5, and is used to form a slow circular flow at a flow rate of 0.2-0.3 m / s in the normal culture mode, and to increase to 0.5-0.8 m / s to enhance the spiral vortex in the capsule cleaning mode;

[0080] The circulating water drain 11 is arranged on the side wall of the culture barrel shell 5 and is higher than the grid isolation layer 6, and is used to maintain the water level and discharge the surface residual feed and plankton impurities in the normal culture mode;

[0081] The central drain valve 12 is located at the lowest point of the conical surface 10, and is used to fully open to discharge the enriched capsule and sewage at the bottom of the barrel during the capsule cleaning mode, and is closed to maintain the stability of the water body in the normal culture mode.

[0082] Further, the central drain valve 12 is an electromagnetic valve, and the diameter of the electromagnetic valve ranges from 10 cm to 15 cm.

[0083] Further, the time period of the capsule cleaning mode ranges from 30 s to 60 s.

[0084] Further, the mesh diameter of the grid isolation layer 6 ranges from 2 mm to 5 mm.

[0085] Further, the conical surface 10 is composed of smooth polyethylene material.

[0086] As shown in Figure 4 Corresponding to the above device, the embodiment also discloses specific control steps:

[0087] Feature extraction and analysis:

[0088] In this link, through the multi-modal data fusion algorithm and the hierarchical feature extraction framework, the precise quantification of grouper abnormal behavior and capsule quantity is realized, and the specific process is as follows:

[0089] Image acquisition and preprocessing:

[0090] The top camera (1080P, 5 frames / second) acquires the fish body overhead and surface images in real time, removes the water body fluctuation noise after Gaussian filtering, and inputs the YOLO11 model for fish body detection and tracking.

[0091] Core index calculation:

[0092] Average swimming speed monitoring: Track the single-tail grouper through video stream, calculate the total distance of movement within 300 frames (1 minute), and obtain the average swimming speed V (unit: cm / s). If V≥10 cm / s is detected for 10 consecutive minutes (normal state V<5 cm / s), it is determined as "abnormal behavior".

[0093] Abnormal trigger rules:

[0094] When the above abnormalities occur in ≥30% of the fish body in a single frame, and repeat ≥3 times within 10 consecutive minutes, the edge computing unit generates an abnormal behavior alarm, and the encapsulation detection module starts urgent scanning.

[0095] Bucket bottom capsule number detection (small target accurate identification):

[0096] Image acquisition and target recognition:

[0097] The grid inside the wall camera (waterproof level IP68) collects 10 bucket bottom images of the conical surface at a frequency of 2 frames / second, after contrast enhancement preprocessing, input YOLO11 model to detect capsules (training data label capsule diameter 50-300 μm, surface ridge features).

[0098] Bucket bottom residual bait quantity detection (small target accurate identification):

[0099] Image acquisition and target recognition:

[0100] The grid inside the wall camera (waterproof level IP68) collects 10 bucket bottom images of the conical surface at a frequency of 2 frames / second, after contrast enhancement preprocessing, input YOLO11 model to detect residual bait.

[0101] Dynamic threshold adjustment:

[0102] The edge computing unit automatically corrects the detection threshold according to the water temperature sensor data (accuracy ±0.5℃):

[0103] Reference threshold: 5 / ㎡ (water temperature 25℃);

[0104] For every 1℃ increase in temperature, the threshold is lowered by 1 / ㎡ (e.g. threshold adjusted to 3 / ㎡ at 27℃), to improve the sensitivity of capsule identification in high temperature environment.

[0105] Density quantification method:

[0106] The pixel coordinates of the detected capsules and residual bait are converted to physical positions through camera calibration parameters, with the center of the bucket bottom as the origin, dividing 1m×1m grid, calculating the number of capsules in each grid, and outputting the area density (pieces / ㎡).

[0107] Multi-modal data fusion and decision:

[0108] Edge computing unit integrates two types of data through timestamp alignment technology (error <100 ms) and executes hierarchical control logic: abnormal behavior linkage:

[0109] If abnormal behavior alarm and capsule density >5 / m2 are triggered at the same time, increase water inflow by 50% and start the bucket bottom drain valve (10s on / 30s off), with an initial duration of 2 minutes, and then dynamically extend according to the capsule density.

[0110] Environmental parameter warning:

[0111] If the water temperature >27℃ and the capsule density >3 / m2, send a warning SMS to the administrator through the 4G module, prompting manual intervention.

[0112] Capsule flushing intelligent control:

[0113] Triggering conditions:

[0114] When the underwater camera under the grid detects that the 25℃ bucket bottom capsule density >5 / m2, immediately trigger the 50% increase in water inflow and start the bucket bottom drain valve (10s on / 30s off), with an initial duration of 2 minutes.

[0115] Dynamic regulation rules:

[0116] If the capsule density >5 / m2 at 25℃, automatically extend the water inflow increase and drain valve opening time to 4 minutes. After one round of flushing, the system automatically rechecks the capsule density, and if it is still >5 / m2, start the second round after 10 minutes, until the density <5 / m2.

[0117] Environmental parameter linkage:

[0118] When the water temperature >27℃, the capsule detection threshold is automatically lowered to 3 / m2, and the flushing process is started at the same time. The time is extended to 3 minutes to improve the efficiency of killing in high temperature period.

[0119] Multi-dimensional abnormal data alarm:

[0120] Grouper abnormal behavior lasts ≥10 minutes (such as average swimming speed ≥10cm / s);

[0121] Water temperature >30℃ or <20℃;

[0122] Dissolved oxygen <5mg / L or >12mg / L, and lasts ≥5 minutes

[0123] Warning execution logic:

[0124] Abnormal data is sent in real time through the 4G communication module to the administrator terminal, including:

[0125] Abnormal type (such as "excessive encapsulation density", "water temperature data anomaly", "dissolved oxygen data anomaly");

[0126] Specific parameter value (such as "encapsulation density: 6 / ㎡", "water temperature: 31℃");

[0127] Suggested operation (such as "please check if the water quality sensor is working properly").

[0128] The embodiment also provides a sea fish intelligent breeding method based on vision and eddy current, which comprises the following steps:

[0129] Image and environmental parameter acquisition: the top camera and the inner side wall camera acquire fish group behavior data, barrel bottom encapsulation distribution data, barrel bottom residual bait distribution data and fish body surface feature data in real time at a preset frequency, and the water quality sensor synchronously acquires water temperature data and dissolved oxygen data;

[0130] Image preprocessing and feature analysis: noise removal and contrast enhancement processing are performed on the collected images, YOLO11 deep learning algorithm is applied to identify encapsulation small targets, residual bait small targets and fish behavior, and encapsulation density, residual bait density and fish abnormal behavior index are output in combination with morphological filtering and spatial positioning;

[0131] Dynamic decision and execution: when the encapsulation density exceeds a threshold value, the water inflow speed is automatically increased and the drain valve is opened to form a barrel bottom spiral eddy current, the encapsulation is concentrated and discharged, the density is reviewed after single removal for a set time, the process is stopped when the density is below the threshold value, the process is repeated when the density is above the threshold value, and the cumulative removal times are not more than 3 times;

[0132] Feeding amount regulation: 3 times of feeding are performed at regular time intervals every day, and the feeding amount of each time is 4% of the initial weight of the feeding machine. The residual bait data collected by the side wall camera is used to control the feeding strategy. If the residual bait coverage area identified by the camera is greater than 5% after 1 hour of feeding, the feeding amount of the next time is reduced by 15%; if the residual bait coverage area is identified for 2 times in succession and is less than 2%, the feeding amount is increased by 10%.

[0133] Abnormal early warning and remote management: real-time early warning is automatically performed on excessive encapsulation, water temperature data anomaly and dissolved oxygen data anomaly, and the alarm is pushed to the administrator terminal, and remote intervention and manual operation are supported.

[0134] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other.

[0135] The principles and implementation manners of the present application are described by using specific examples in the present application, and the above examples are only used to help understand the method of the present application and its core idea; meanwhile, for the general technical personnel in the art, the specific implementation manners and application ranges will be changed according to the idea of the present application. In conclusion, the content of the present specification should not be understood as the limitation of the present application.

Claims

1. A visual and eddy current-based intelligent marine fish farming device, characterized in that, The application relates to an edge computing unit, a breeding barrel main body and a layered monitoring module, an intelligent vortex module and an intelligent feeder arranged on the breeding barrel main body at a preset position. The edge computing unit is connected with the layered monitoring module, the intelligent vortex module and the intelligent feeder. The layered monitoring module is used for collecting environmental data, fish body state data and residual bait data, wherein the environmental data comprises water temperature data and dissolved oxygen data, the fish body state data comprises fish body behavior data, bottom capsule distribution data and fish body surface data; the edge computing unit is used for performing instruction calculation according to corresponding threshold values of the environmental data, the fish body state data and the residual bait data, so as to obtain vortex control instructions and feeding instructions; the intelligent vortex module is used for switching corresponding working modes according to the vortex control instructions, wherein the working modes comprise a normal breeding mode and a capsule cleaning mode; and the intelligent feeder is used for feeding the breeding area according to the feeding instructions. The layered monitoring module comprises a top underwater camera, a side wall underwater camera and a water quality sensor.

2. The intelligent marine fish farming device based on vision and eddy current according to claim 1, characterized in that, The top underwater camera is used for collecting fish body behavior data and fish body surface data; the side wall underwater camera is used for collecting bottom capsule distribution data and residual bait data; and the water quality sensor is used for collecting water temperature data and dissolved oxygen data. The breeding barrel main body comprises a top support, a breeding barrel shell, a grid isolation layer, an outer barrel four-corner platform support and a conical surface. The top support is fixedly installed at the upper end of the breeding barrel shell and is used for bearing the top underwater camera.

3. The intelligent marine fish farming device based on vision and eddy current according to claim 1, characterized in that, The grid isolation layer is horizontally arranged above the intersection of the inner wall of the breeding barrel shell and the conical surface and has a height corresponding to the top end of the central drainage valve, and is used for separating the breeding area from the vortex area. The outer barrel four-corner platform support is located at the bottom periphery of the breeding barrel shell and is used for supporting and horizontally adjusting the breeding barrel main body. The conical surface is used for forming a spiral sinking flow field in the capsule cleaning mode to collect the bottom capsules to the central drainage valve. The intelligent vortex module comprises an edge tangent water inlet pipe, a circulating water drainage port and a central drainage valve. The edge tangent water inlet pipe is uniformly distributed along the inner wall of the breeding barrel shell and is used for forming a ring-shaped slow circulation at a flow rate of 0.2-0.3 m / s in the normal breeding mode and increasing to 0.5-0.8 m / s to enhance the spiral vortex in the capsule cleaning mode. The circulating water drainage port is arranged on the side wall of the breeding barrel shell and is higher than the grid isolation layer, and is used for maintaining the water level and discharging the surface residual bait and floating impurities in the normal breeding mode.

4. The intelligent marine fish farming device based on vision and eddy current according to claim 1, characterized in that, The central drainage valve is located at the lowest point of the conical surface and is used for fully opening to discharge the bottom enriched capsules and sewage in the capsule cleaning mode and closing to maintain the water body stable in the normal breeding mode. The central drainage valve is an electromagnetic valve, and the diameter of the electromagnetic valve ranges from 10 cm to 15 cm. The time period of the capsule cleaning mode ranges from 30 s to 60 s. The mesh diameter of the grid isolation layer ranges from 2 mm to 5 mm. The conical surface is composed of smooth polyethylene material.

5. The intelligent marine fish farming device based on vision and eddy current according to claim 4, characterized in that, ​ 6. The intelligent marine fish farming device based on vision and eddy current according to claim 1, characterized in that, ​ 7. The intelligent marine fish farming device based on vision and eddy current according to claim 3, characterized in that, ​ 8. The intelligent marine fish farming device based on vision and eddy current according to claim 1, characterized in that, ​ 9. A visual and eddy current-based intelligent marine fish farming method applied to the device of any one of claims 1-8, characterized in that, The method comprises: Image and environmental parameter acquisition: the top camera and the inner side wall camera collect fish group behavior data, capsule distribution data on the bottom of the barrel, residual bait distribution data on the bottom of the barrel, fish body surface feature data in real time at a preset frequency, and the water quality sensor synchronously collects water temperature data and dissolved oxygen data; Image preprocessing and feature analysis: noise removal and contrast enhancement processing are performed on the collected images, YOLO11 deep learning algorithm is applied to identify capsule small targets and fish behavior, and morphological filtering and spatial positioning are combined to output capsule density and fish abnormal behavior indicators; Dynamic decision and execution: when the capsule density exceeds a threshold value, the water inflow speed is automatically increased and the drain valve is opened to form a spiral vortex at the bottom of the barrel, the capsules are concentrated and discharged, the density is reviewed after single removal at a set time interval, and the process is repeated when the density is lower than the threshold value or higher than the threshold value; the cumulative removal times do not exceed 3 times; Feeding amount regulation: 3 times of feeding are performed at regular intervals every day, and the feeding amount of each time is 4% of the initial weight of the feeding machine; the residual bait data collected by the side wall camera is used to control the feeding strategy; if the residual bait coverage area identified by the camera is greater than 5% after 1 hour of feeding, the feeding amount of the next time is reduced by 15%; if the residual bait coverage area is less than 2% for 2 times in succession, the feeding amount is increased by 10%; Abnormal early warning and remote management: real-time early warning is automatically performed on capsule over-standard, abnormal water temperature data and abnormal dissolved oxygen data, and the alarm is pushed to the administrator terminal, and remote intervention and manual operation are supported.

Images