Aquaculture water quality detection system

Abstract

The invention discloses an aquaculture water quality detection system, and belongs to the technical field of water quality detection. The problems that an existing aquaculture water quality detection system can reduce the detection frequency of the aquaculture water body, and meanwhile it is ensured that the quality of cultured fishes is not prone to decline or die due to the water quality problem of the aquaculture water body in the culture process are solved. Comprising an image acquisition module, a water quality detection module, a secondary data processing module, a control module and an early warning module, and the image acquisition module is used for being arranged at the center of a fishpond to obtain fishpond water surface images in real time. The fish pond water surface and internal images are continuously acquired through the image acquisition module, the cultured fish condition is judged, if abnormity is found, the water quality detection module is used for detecting the water quality, confirming problems and giving an early warning to farmers so as to adjust the water quality in time, the working frequency of the water quality detection module is reduced, the service life of the system is prolonged, and the method is suitable for a high-density culture environment. The detection result is accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality detection, and in particular to an aquaculture water quality detection system. Background Art

[0002] In aquaculture, water quality testing is a key link to ensure a healthy aquaculture environment, increase production and economic benefits. By monitoring water quality parameters such as dissolved oxygen, pH, ammonia nitrogen, nitrite, etc., potential problems can be discovered and solved in a timely manner, providing the best growth environment for aquatic animals. For example, dissolved oxygen is the basis for the survival of aquatic organisms, and its content should be maintained at 5-8 mg / L, at least not less than 4 mg / L. The pH value reflects the acidity and alkalinity of the water body, and the optimal range is 7.5-8.5. Ammonia nitrogen and nitrite are important indicators for water quality monitoring. The ammonia nitrogen content should not exceed 0.2 mg / L, and nitrite should be controlled at a lower level. In addition, modern aquaculture has also introduced the concept of intelligent Internet of Things, which realizes real-time monitoring and early warning of water quality parameters through water quality sensors and data acquisition and transmission equipment. The application of these technologies not only improves the efficiency and accuracy of water quality monitoring, but also provides a scientific basis for aquaculture management, which helps to achieve intelligent and sustainable development of aquaculture.

[0003] However, during the aquaculture process, the water quality of the aquaculture water changes, which will immediately cause a reaction in the farmed fish. If the farmer does not adjust the water quality of the aquaculture water in time, the quality of the farmed fish will decline or even die after a period of time. Therefore, the current aquaculture water quality detection system often uses a collection unit to perform real-time detection of the water quality of the aquaculture water.

[0004] For the aquaculture water quality detection system, the ability to detect in real time will cause the collection unit to detect the aquaculture water too frequently. High-frequency aquaculture water quality detection will not only lead to an increase in the operation and maintenance cost of the entire system, but also bring a large amount of detection data. The constant calculation of a large amount of detection data to estimate the condition of farmed fish will cause the entire aquaculture water quality detection system to be overburdened. However, if the frequency of aquaculture water quality detection system detecting aquaculture water is significantly reduced, detection will not be timely, and then farmed fish may experience quality decline or death due to untimely water quality monitoring and adjustment of the aquaculture water.

[0005] Therefore, the current aquaculture water quality detection system needs to be able to reduce the detection frequency of aquaculture water bodies, while ensuring that farmed fish are not prone to quality degradation or death due to water quality problems in the aquaculture water body during the breeding process.

[0006] Therefore, an aquaculture water quality detection system is proposed to solve or alleviate the above problems. Summary of the invention

[0007] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an aquaculture water quality detection system.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions: An aquaculture water quality detection system comprises an image acquisition module, a water quality detection module, a secondary data processing module, a control module, and an early warning module. The image acquisition module is used to be arranged at the center of a fish pond to obtain a real-time image of the fish pond water surface, to perform a vertical reciprocating movement once every 1 hour and to obtain an image of the inside of the fish pond. The water quality detection module is used to detect water body data in the fish pond. The secondary data processing module is used to receive image data to determine the status of farmed fish and transmit a primary result to the control module, receive water body data to determine the water body status and transmit a secondary result to the control module. The control module is used to control the water quality detection module to start working according to the primary result, and to control the early warning module to alarm according to the secondary result.

[0009] Preferably, the image acquisition module includes a transparent tube with an open top and a closed bottom in a columnar structure, an electric push rod vertically fixedly connected to the bottom surface of the transparent tube, a panoramic camera fixedly connected to the movable end of the electric push rod and slidably connected in the transparent tube, and a power supply arranged on the bottom surface of the transparent tube and coupled to the panoramic camera and the electric push rod. When the electric push rod is in a maximum length state, the horizontal height of the panoramic camera is higher than the water surface of the fish pond.

[0010] Preferably, the water body data includes DO, pH value, ammonia nitrogen content, nitrite content, water temperature, chlorophyll content, and sulfide content.

[0011] Preferably, the secondary data processing module is used to receive image data to determine the status of farmed fish and transmit the primary result to the control module, including the following steps: Based on the fish pond water surface image when not feeding, determine whether the farmed fish in the fish pond water surface image have floating head breathing phenomenon. If floating head breathing phenomenon exists, output a result that there is a problem with the water body and floating head breathing phenomenon exists. If floating head breathing phenomenon does not exist, no result needs to be output; According to the image of the fish pond water surface during feeding, determine whether the farmed fish in the image of the fish pond water surface have an indifferent reaction. If the fish have an indifferent reaction, output a result that there is a problem with the water body and the fish has a decreased appetite. If the fish have no indifferent reaction, no result needs to be output; Based on the image inside the fish pond, it is determined whether the farmed fish therein have shortness of breath. If so, a result is output once that there is a problem with the water body and there is shortness of breath. If not, there is no need to output a result.

[0012] Preferably, judging whether farmed fish in the fish pond water surface image have floating head breathing phenomenon according to the fish pond water surface image when not feeding, if the floating head breathing phenomenon exists, outputting a result that there is a problem with the water body and the floating head breathing phenomenon exists, and if the floating head breathing phenomenon does not exist, there is no need to output a result, including the following steps: The water surface image of the fish pond when not feeding is subjected to histogram equalization, and the low-frequency components of illumination and the high-frequency components of reflection are separated to complete homomorphic filtering; The optical flow method is used to calculate the pixel motion vectors between adjacent frames to capture the movement trajectory of the school of fish; The optical flow amplitude is thresholded to complete the binarization of the moving area and distinguish the floating fish school from the environmental interference. Convert the RGB image to HSV space and extract the fish body color features; Fill small holes, smooth the contour, and complete morphological closing operations to separate the precise contour of the floating head fish body from the moving area; The proportion of floating head area to the total water surface area is counted, the spatial distribution of floating head breathing phenomenon is extracted, and the regional density is obtained; Calculate the average optical flow directional entropy of the floating head area, complete the motion pattern analysis, extract the dynamic characteristics of the floating head breathing phenomenon, and characterize the irregularity of the struggling movement; Construct a linear discriminant function for dual feature fusion; Combined with the construction of a linear discriminant function, it is judged that "floating head breathing phenomenon exists".

[0013] Preferably, judging whether the farmed fish in the fish pond water surface image have an indifferent reaction phenomenon according to the fish pond water surface image during feeding, if there is an indifferent reaction phenomenon, outputting a result that there is a problem with the water body and there is a decrease in appetite, and if there is no indifferent reaction phenomenon, there is no need to output a result, including the following steps: In the water surface image of the fish pond during feeding, the diffusion characteristics of the bait after entering the water are used to segment the initial feeding core area through the HSV color threshold to locate the feeding area; Through morphological expansion to simulate the bait diffusion range, define the dynamic analysis area, establish the spatiotemporal analysis benchmark, and complete the dynamic calibration of the feeding area; Dense optical flow is used to calculate the migration speed of fish schools to the feeding area; The average motion vector amplitude of the dynamic analysis area within T seconds after feeding was calculated; Divide the grid in the dynamic analysis area to calculate the motion direction entropy of each unit; The uneaten bait area was extracted by background difference method, and the residual ratio was more than 10%, which was considered as insufficient feeding; If the average motion vector amplitude is less than 0.3px / frame and the residual ratio exceeds 10% at the same time, it is judged that there is a cold response phenomenon.

[0014] Preferably, judging whether the farmed fish have shortness of breath according to the internal image of the fish pond, if there is shortness of breath, outputting a result that there is a problem with the water body and there is shortness of breath, and if there is no shortness of breath, there is no need to output a result, comprises the following steps: Retinex theory is used to enhance the visibility of low-light areas in the image inside the fish pond and complete color correction; The YOLOv8 target detection model is used to detect the farmed fish bounding box, output the candidate area with a confidence level > 0.8, and locate the key points of the gill cover edge based on HRNet; Lucas-Kanade optical flow velocity was calculated within a 10×10 pixel window around the operculum keypoint; The operculum spacing sequence was Fourier transformed to extract the main frequency component. When breathing was rapid, the main frequency was >1.5 Hz; The main frequency distribution histogram of all farmed fish tested was counted, and the abnormal proportion was calculated. If the abnormal proportion was greater than 5%, it was judged as tachypnea; Expand the range of 20 pixels outside the operculum area to verify the consistency of the optical flow direction. If the consistency angle is less than 15°, it is determined to be a valid operculum movement. Use Kalman filtering to predict the operculum motion state of the occluded fish; If the main frequency is >1.5hz and the abnormal ratio is >5%, it is judged that shortness of breath exists.

[0015] Preferably, the receiving of water body data to determine the water body condition and transmitting the secondary result to the control module comprises the following steps: After receiving the water body data and the primary result, the water body data is matched with the water body threshold, wherein the water body threshold includes a DO threshold, a pH value threshold, an ammonia nitrogen content threshold, a nitrite content threshold, a water temperature threshold, a chlorophyll amount threshold, and a sulfide content threshold; The match result is compared with the first result. If they are the same, the second result is output with the same content as the first result. If they are not the same, the first result is re-judged and the above steps are repeated.

[0016] Preferably, the secondary data processing module includes a processor, and the control module includes a controller.

[0017] The present invention has the following beneficial effects: The present invention can continuously acquire images through the image acquisition module, and adjust the position of the image acquisition module at intervals, and then acquire the internal image of the fish pond in addition to the image of the fish pond water surface, and judge the farmed fish conditions through the fish pond water surface image and the fish pond internal image. If the farmed fish are in an abnormal condition, the water quality of the farmed water body is detected through the water quality detection module to further confirm the water quality problems and the problems encountered by the farmed fish, so as to give early warning to the breeders, so that the breeders can timely adjust the water quality of the farmed water body in the farmed fish pond, and avoid the situation that the farmed fish are degraded or die due to the water quality problems of the farmed water body during the breeding process. At the same time, only the image acquisition module is used to perform image detection, and there is no need to perform real-time detection on the water quality detection module, which significantly reduces the working frequency of the water quality detection module in the system and ensures the service life of the whole system. In addition, this system is suitable for fish ponds with high-density fish farming. The difference from conventional fish pond farming is that in the face of a high-density farming environment of, for example, 50 kg / cubic meter, this system can still judge the condition of farmed fish based on the fish pond water surface image and the fish pond internal image fed back by the image acquisition module. For the conventional sparse fish farming environment, conventional methods cannot be applied in this high-density farming environment, and cannot accurately output the condition of farmed fish, nor can they accurately output the water quality test results of the farmed fish pond. The method in this system can accurately output the results in the face of high-density farming, ensuring the accuracy of the test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 It is a structural block diagram of the present invention.

[0020] 1. Image acquisition module; 2. Water quality detection module; 3. Secondary data processing module; 4. Control module; 5. Early warning module. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0024] In the description of the present invention, it should be understood that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, or are the orientations or positional relationships in which the product of the invention is conventionally placed when in use, or are the orientations or positional relationships conventionally understood by those skilled in the art. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0025] Furthermore, the terms “first”, “second”, “third”, etc. are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0026] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0027] A water quality detection system for aquaculture, such as Figure 1As shown, it includes an image acquisition module 1, a water quality detection module 2, a secondary data processing module 3, a control module 4, and an early warning module 5. The image acquisition module 1 is used to be set at the center of the fish pond to obtain the fish pond water surface image in real time, and to perform vertical reciprocating movement once every 1 hour to obtain the internal image of the fish pond. The water quality detection module 2 is used to detect the water body data in the fish pond. The water body data includes DO, pH value, ammonia nitrogen content, nitrite content, water temperature, chlorophyll content, and sulfide content. The secondary data processing module 3 is used to receive image data to determine the status of farmed fish and transmit the primary result to the control module 4, receive water body data to determine the water body status and transmit the secondary result to the control module 4, and the control module 4 is used to control the water quality detection module 2 to start working according to the primary result, and control the early warning module 5 to alarm according to the secondary result.

[0028] The process of receiving water body data to judge the water body condition and transmitting the secondary result to the control module 4 includes the following steps: After receiving the water body data and the primary result, the water body data is matched with the water body thresholds, which include DO threshold, pH value threshold, ammonia nitrogen content threshold, nitrite content threshold, water temperature threshold, chlorophyll content threshold, and sulfide content threshold; The match result is compared with the first result. If they are the same, the second result is output with the same content as the first result. If they are not the same, the first result is re-judged and the above steps are repeated.

[0029] The image acquisition module 1 includes a transparent tube with an open top and a closed bottom in a columnar structure, an electric push rod vertically fixedly connected to the bottom surface of the transparent tube, a panoramic camera fixedly connected to the movable end of the electric push rod and slidably connected in the transparent tube, and a power supply arranged on the bottom surface of the transparent tube and coupled to the panoramic camera and the electric push rod. When the electric push rod is in a maximum length state, the horizontal height of the panoramic camera is higher than the water surface of the fish pond. The secondary data processing module 3 includes a processor, and the control module 4 includes a controller.

[0030] The aquaculture water quality detection system of the present invention can continuously obtain images of the water surface of the fish pond and images of the interior of the fish pond through the image acquisition module 1, thereby realizing real-time monitoring of the status of farmed fish. When it is necessary to obtain images of the interior of the fish pond, the panoramic camera is driven by an electric push rod to complete a lifting and reciprocating movement, and the panoramic camera can obtain images of the interior of the fish pond through the transparent tube. In this way, by adjusting the position of the image acquisition module 1 at intervals, images of different depths can be obtained, further enriching the monitoring data, and making the judgment of the status of farmed fish more accurate and comprehensive.

[0031] Secondly, when the system finds that farmed fish are in an abnormal condition, it can promptly start the water quality detection module 2 to detect the water quality of the farming water body. This linkage mechanism effectively reduces the working frequency of the water quality detection module 2, avoiding the high operation and maintenance costs and system burdens brought about by high-frequency or even real-time water quality detection in traditional systems. The condition of farmed fish is preliminarily judged through image data, and water quality detection is only performed when abnormalities are confirmed. This not only ensures the timeliness of the detection, but also improves the overall efficiency and service life of the system.

[0032] Furthermore, the system is particularly suitable for fish pond environments with high-density fish farming. Under high-density farming conditions, such as a farming density of 50 kg / cubic meter, traditional water quality testing and farmed fish status monitoring methods are often difficult to accurately output results. The system, through the image information fed back by the image acquisition module 1, can accurately judge the status of farmed fish and ensure that the test results can be accurately output in a high-density farming environment, which helps to ensure the healthy growth of farmed fish and reduce quality degradation or death due to water quality problems.

[0033] Finally, the secondary data processing module 3 of the system can comprehensively process and analyze the received image data and water body data, further confirm water quality problems and specific problems encountered by farmed fish, and transmit the results to the control module 4. The control module 4 controls the early warning module 5 to alarm according to these results, and promptly notifies the farmers to take corresponding water quality adjustment measures. This intelligent data processing and early warning mechanism enables farmers to respond quickly and adjust the water quality of the aquaculture water in time, avoiding aquaculture risks caused by water quality problems and improving the overall stability and reliability of aquaculture.

[0034] Preferably, the secondary data processing module 3 is used to receive image data to determine the status of farmed fish and transmit the primary result to the control module 4, including the following steps: Based on the fish pond water surface image when not feeding, determine whether the farmed fish in the fish pond water surface image have floating head breathing phenomenon. If floating head breathing phenomenon exists, output a result that there is a problem with the water body and floating head breathing phenomenon exists. If floating head breathing phenomenon does not exist, no result needs to be output; According to the image of the fish pond water surface during feeding, determine whether the farmed fish in the image of the fish pond water surface have an indifferent reaction. If the fish have an indifferent reaction, output a result that there is a problem with the water body and the fish has a decreased appetite. If the fish have no indifferent reaction, no result needs to be output; Based on the image inside the fish pond, it is determined whether the farmed fish therein have shortness of breath. If so, a result is output once that there is a problem with the water body and there is shortness of breath. If not, there is no need to output a result.

[0035] Among them, according to the fish pond water surface image when not feeding, it is judged whether the farmed fish in the fish pond water surface image has floating head breathing phenomenon. If the floating head breathing phenomenon exists, a result is output once that there is a problem with the water body and the floating head breathing phenomenon exists. If the floating head breathing phenomenon does not exist, there is no need to output a result once, including the following steps: Perform histogram equalization on the water surface image of the fish pond when not feeding , separate the low-frequency components of illumination and the high-frequency components of reflection to complete homomorphic filtering ,in, is the number of pixels at gray level i, N is the total number of pixels, is a high pass filter; The optical flow method is used to calculate the pixel motion vectors between adjacent frames to capture the movement trajectory of the fish school. ,in, and is the spatial gradient, is the time gradient, u and v are the optical flow velocity components; Threshold segmentation is performed on the optical flow amplitude to complete the binarization of the moving area and distinguish between floating fish and environmental interference ,in, is the binary motion area; Convert RGB images to HSV space to extract fish color features , set the threshold range Perform masking operations; Fill small holes, smooth the contours, and complete morphological closing operations to separate the precise contours of the floating head fish from the moving area. ,in, For expansion, is erosion, B is a circular structural element, and A is a binary motion area; Count the proportion of floating head area to the total water surface area, extract the spatial distribution of floating head breathing phenomenon, and obtain the regional density ,in, is the total water surface area; Calculate the average optical flow directional entropy of the floating head area, complete the motion pattern analysis, extract the dynamic characteristics of the floating head breathing phenomenon, and characterize the irregularity of the struggling movement. ,in, is the optical flow direction angle, is the direction histogram probability; Constructing the linear discriminant function Perform dual feature fusion, where , ; Combined with the construction of linear discriminant function judgment, if any of the following conditions is met, it is judged as "there is floating head breathing phenomenon", .

[0036] According to the water surface image of the fish pond when feeding, it is judged whether the farmed fish in the water surface image of the fish pond have a cold reaction phenomenon. If there is a cold reaction phenomenon, a result is outputted once, that is, there is a problem with the water body and there is a decrease in appetite. If there is no cold reaction phenomenon, there is no need to output a result once, which includes the following steps: In the fish pond water surface image during feeding, the diffusion characteristics of the bait after entering the water are used to segment the initial feeding core area through the HSV color threshold. , locate the feeding area; By simulating the bait diffusion range through morphological expansion, the dynamic analysis area is defined. , establish a temporal and spatial analysis benchmark and complete the dynamic calibration of the feeding area; Using dense optical flow to calculate the migration speed of fish schools to the feeding area , where W is the local window, , , is the spatiotemporal gradient; Statistical analysis of the average motion vector amplitude in the dynamic analysis area within T seconds after feeding , during normal feeding ; Divide the grid in the dynamic analysis area to calculate the motion direction entropy of each unit. ,in, is the probability distribution of the 8-direction histogram, and the eating behavior is characterized by a high entropy value, that is, ; Extract the uneaten bait area by background subtraction method , if the residual ratio exceeds 10%, it is judged as insufficient feeding; If the average motion vector amplitude is less than 0.3px / frame and the residual ratio exceeds 10% at the same time, it is judged that there is a cold response phenomenon.

[0037] Among them, judging whether the farmed fish therein have shortness of breath according to the internal image of the fish pond, if there is shortness of breath, outputting a result that there is a problem with the water body and there is shortness of breath, and if there is no shortness of breath, there is no need to output a result, including the following steps: Retinex theory is used to enhance the visibility of low-light areas in the image inside the fish pond and complete color correction. ,in, is the Gaussian filter kernel; The YOLOv8 target detection model is used to detect the farmed fish bounding box, output the candidate area with a confidence level > 0.8, and locate the key points of the gill cover edge based on HRNet. and , ,in, For pre-training model parameters, output the gill cover opening and closing distance ; Calculate the Lucas-Kanade optical flow velocity in a 10×10 pixel window around the operculum keypoint ; Opercular spacing sequence Perform Fourier transform to extract the main frequency component , the main frequency is >1.5hz when breathing is rapid; The main frequency distribution histogram of all farmed fish detected is statistically analyzed , and calculate its abnormal proportion , if the abnormal ratio , it is judged as shortness of breath; Expand the range of 20 pixels outside the gill cover area to verify the consistency of the optical flow direction. , if the consistency angle is <15°, it is judged as effective operculum movement; Using Kalman filtering to predict the operculum motion state of the occluded fish , where A and B are state transfer matrices, is the process noise; If the main frequency is greater than 1.5hz and the abnormal ratio , it is judged that there is shortness of breath.

[0038] Through the above method steps, the image acquisition module 1 is used to analyze the conditions of high-density farmed fish in real time, including floating head breathing, decreased appetite, and rapid breathing. The water quality detection module 2 is activated only when abnormal behavior is detected. Compared with the fixed frequency detection and high frequency detection of the traditional system, the number of water quality detection times can be reduced significantly, and the energy consumption and data volume of the sensor can be significantly reduced. If the behavior of the fish school is normal, the system may trigger a water quality detection once every few hours or even dozens of hours. Once an abnormal condition is detected, the water quality detection is immediately activated and a linkage warning is issued.

[0039] In addition, this system can accurately identify abnormal behavior of fish in high-density farming conditions through the use of algorithms such as optical flow, color space conversion, and morphological closing operations. There are certain standards for the response time of abnormal detection to avoid false detections, thereby significantly improving the accuracy. In addition, in high-density farming conditions, it can eliminate the effects of water turbidity and reflection on image quality, and distinguish between effective fish movement and environmental interference, such as gill cover opening and closing and water flow fluctuations, to ensure that the rapid breathing of farmed fish can be output in high-density fish farming environments. Changes in bait on the water surface and the reactions of farmed fish can be used to determine whether farmed fish have a decreased appetite when farmed fish are farmed at high density, thereby enabling this system to accurately identify abnormal conditions of farmed fish.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An aquaculture water quality detection system, characterized in that: The invention comprises an image acquisition module (1), a water quality detection module (2), a secondary data processing module (3), a control module (4), and an early warning module (5); the image acquisition module (1) is used to be arranged at the center of the fish pond to obtain a real-time image of the water surface of the fish pond, to perform a vertical reciprocating movement once every 1 hour and to obtain an image of the inside of the fish pond; the water quality detection module (2) is used to detect water body data in the fish pond; the secondary data processing module (3) is used to receive image data to determine the status of farmed fish and transmit a primary result to the control module (4); receive water body data to determine the status of the water body and transmit a secondary result to the control module (4); the control module (4) is used to control the water quality detection module (2) to start working according to the primary result, and to control the early warning module (5) to alarm according to the secondary result.

2. The aquaculture water quality detection system according to claim 1, characterized in that: The image acquisition module (1) comprises a transparent cylinder with an open top and a closed bottom in a columnar structure, an electric push rod vertically fixedly connected to the bottom surface of the transparent cylinder, a panoramic camera fixedly connected to the movable end of the electric push rod and slidably connected to the transparent cylinder, and a power supply arranged on the bottom surface of the transparent cylinder and coupled to the panoramic camera and the electric push rod. When the electric push rod is in a maximum length state, the horizontal height of the panoramic camera is higher than the water surface of the fish pond.

3. The aquaculture water quality detection system according to claim 1, characterized in that: The water body data include DO, pH value, ammonia nitrogen content, nitrite content, water temperature, chlorophyll content, and sulfide content.

4. The aquaculture water quality detection system according to claim 1, characterized in that: The secondary data processing module (3) is used to receive image data to determine the status of farmed fish and transmit the primary result to the control module (4), comprising the following steps: Based on the fish pond water surface image when not feeding, determine whether the farmed fish in the fish pond water surface image have floating head breathing phenomenon. If floating head breathing phenomenon exists, output a result that there is a problem with the water body and floating head breathing phenomenon exists. If floating head breathing phenomenon does not exist, no result needs to be output; According to the image of the fish pond water surface during feeding, determine whether the farmed fish in the image of the fish pond water surface have an indifferent reaction. If the fish have an indifferent reaction, output a result that there is a problem with the water body and the fish has a decreased appetite. If the fish have no indifferent reaction, no result needs to be output; Based on the image inside the fish pond, it is determined whether the farmed fish therein have shortness of breath. If so, a result is output once that there is a problem with the water body and there is shortness of breath. If not, there is no need to output a result.

5. The aquaculture water quality detection system according to claim 4, characterized in that: The method comprises the following steps: judging whether the farmed fish in the fish pond water surface image have floating head breathing phenomenon according to the fish pond water surface image when not feeding; if the floating head breathing phenomenon exists, outputting a result that there is a problem with the water body and the floating head breathing phenomenon exists; and if the floating head breathing phenomenon does not exist, there is no need to output a result. The water surface image of the fish pond when not feeding is subjected to histogram equalization, and the low-frequency components of illumination and the high-frequency components of reflection are separated to complete homomorphic filtering; The optical flow method is used to calculate the pixel motion vectors between adjacent frames to capture the movement trajectory of the school of fish; The optical flow amplitude is thresholded to complete the binarization of the moving area and distinguish the floating fish school from the environmental interference. Convert the RGB image to HSV space and extract the fish body color features; Fill small holes, smooth the contour, and complete morphological closing operations to separate the precise contour of the floating head fish body from the moving area; The proportion of floating head area to the total water surface area is counted, the spatial distribution of floating head breathing phenomenon is extracted, and the regional density is obtained; Calculate the average optical flow directional entropy of the floating head area, complete the motion pattern analysis, extract the dynamic characteristics of the floating head breathing phenomenon, and characterize the irregularity of the struggling movement; Construct a linear discriminant function for dual feature fusion; Combined with the construction of a linear discriminant function, it is judged that "floating head breathing phenomenon exists".

6. The aquaculture water quality detection system according to claim 4, characterized in that: The method of judging whether the farmed fish in the fish pond water surface image have an indifferent reaction phenomenon according to the fish pond water surface image during feeding, and if the indifferent reaction phenomenon exists, outputting a result that there is a problem with the water body and there is a decrease in appetite, and if the indifferent reaction phenomenon does not exist, there is no need to output a result, including the following steps: In the water surface image of the fish pond during feeding, the diffusion characteristics of the bait after entering the water are used to segment the initial feeding core area through the HSV color threshold to locate the feeding area; Through morphological expansion to simulate the bait diffusion range, define the dynamic analysis area, establish the spatiotemporal analysis benchmark, and complete the dynamic calibration of the feeding area; Dense optical flow is used to calculate the migration speed of fish schools to the feeding area; The average motion vector amplitude of the dynamic analysis area within T seconds after feeding was calculated; Divide the grid in the dynamic analysis area to calculate the motion direction entropy of each unit; The uneaten bait area was extracted by background difference method, and the residual ratio was more than 10%, which was considered as insufficient feeding; If the average motion vector amplitude is less than 0.3px / frame and the residual ratio exceeds 10% at the same time, it is judged that there is a cold response phenomenon.

7. The aquaculture water quality detection system according to claim 4, characterized in that: The method of judging whether the farmed fish have shortness of breath according to the internal image of the fish pond, if there is shortness of breath, outputting a result that there is a problem with the water body and there is shortness of breath, and if there is no shortness of breath, there is no need to output a result, includes the following steps: Retinex theory is used to enhance the visibility of low-light areas in the image inside the fish pond and complete color correction; The YOLOv8 target detection model is used to detect the farmed fish bounding box, output the candidate area with a confidence level > 0.8, and locate the key points of the gill cover edge based on HRNet; Lucas-Kanade optical flow velocity was calculated within a 10×10 pixel window around the operculum keypoint; The operculum spacing sequence was Fourier transformed to extract the main frequency component. When breathing was rapid, the main frequency was >1.5 Hz; The main frequency distribution histogram of all farmed fish tested was counted, and the abnormal proportion was calculated. If the abnormal proportion was greater than 5%, it was judged as tachypnea; Expand the range of 20 pixels outside the operculum area to verify the consistency of the optical flow direction. If the consistency angle is less than 15°, it is determined to be a valid operculum movement. Use Kalman filtering to predict the operculum motion state of the occluded fish; If the main frequency is >1.5hz and the abnormal ratio is >5%, it is judged that shortness of breath exists.

8. The aquaculture water quality detection system according to claim 1, characterized in that: The receiving of water body data, determining the water body condition and transmitting the secondary result to the control module (4) comprises the following steps: After receiving the water body data and the primary result, the water body data is matched with the water body threshold, wherein the water body threshold includes a DO threshold, a pH value threshold, an ammonia nitrogen content threshold, a nitrite content threshold, a water temperature threshold, a chlorophyll amount threshold, and a sulfide content threshold; The match result is compared with the first result. If they are the same, the second result is output with the same content as the first result. If they are not the same, the first result is re-judged and the above steps are repeated.

9. The aquaculture water quality detection system according to claim 1, characterized in that: The secondary data processing module (3) comprises a processor, and the control module (4) comprises a controller.

Images