A method and system for evaluating the status of marine aquaculture

Abstract

The present application relates to the technical field of data evaluation, in particular to a marine aquaculture state evaluation method and system, comprising the following steps: collecting images to estimate fish population density, performing three-time difference trend judgment, combining dissolved oxygen change to identify abnormalities, linkage adjustment of risk interval, and output of aquaculture state evaluation result. In the present application, the fish population density is continuously identified and converted by image frames inside the net cage, the density change trend is judged by combining multi-time difference value, the dissolved oxygen change amplitude and statistical threshold are introduced for abnormal identification, and the risk interval boundary is dynamically adjusted when the trend signal and environmental abnormality appear at the same time, so that the risk judgment is kept synchronous with the actual change of water body, linkage analysis of fish population gathering situation and environmental fluctuation is realized, the risk identification sensitivity and interval division flexibility are improved, the real-time and pertinence of marine aquaculture state evaluation are enhanced, and higher matching degree and reliability are ensured for on-site management decision.

Description

Technical Field

[0001] This invention relates to the field of data evaluation technology, and in particular to a method and system for evaluating the status of marine aquaculture. Background Technology

[0002] The field of data assessment technology mainly involves the organization, analysis, comparison, and judgment of collected data, including data acquisition, data cleaning, indicator construction, threshold setting, and result judgment. Typically, it involves establishing an evaluation index system, selecting several quantitative parameters reflecting the state of an object, classifying and summarizing the raw data to calculate ratios or averages, and then classifying or assessing risks according to preset standards. This forms a quantitative description and classification method for the state of the target object, widely used in scenarios such as agricultural production management, environmental monitoring, and resource utilization. Specifically, the marine aquaculture status assessment method refers to the method of judging the state of aquaculture water bodies and the growth status of aquaculture organisms. It mainly focuses on environmental parameters such as water temperature, salinity, dissolved oxygen, pH, ammonia nitrogen concentration, and transparency, as well as production data such as stocking density, feeding amount, growth cycle, and survival rate. Data is collected regularly by deploying water quality monitoring instruments and sampling devices in the aquaculture area. Each parameter is compared item by item according to pre-defined evaluation indicators, and a comprehensive score is calculated. Based on the score range, the data is divided into normal warning or abnormal levels, thereby completing the judgment of the marine aquaculture status.

[0003] Current marine aquaculture status assessment processes focus on periodic data collection and comparison of water quality and production parameters. The evaluation process relies on periodic data aggregation and pre-defined scoring intervals, making it difficult to reflect the spatial distribution and dynamic changes of fish populations within the aquaculture water body. It also lacks the ability to continuously characterize the correlation between density fluctuations and sudden environmental changes. In actual operation, this often leads to delayed status assessments and crude risk identification. When local anomalies occur or fish population aggregation changes, the overall score may still remain at the original level, resulting in untimely early warning triggering and rigid interval divisions, affecting the accuracy of on-site control decisions and the level of aquaculture safety assurance. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for assessing the status of marine aquaculture.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for assessing the status of marine aquaculture, comprising the following steps: S1: Collect image frames from the underwater camera inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the fish density in the cage water by combining the cage volume parameters. S2: Read the fish density records of the net cage water body at three consecutive time points, arrange them in chronological order, perform adjacent time difference calculation to form the first difference and the second difference, perform sign judgment on the first difference and the second difference to determine if the signs are consistent, record the effective and invalid states of the trend trigger flag, and generate the density change trend judgment result; S3: Monitor the dissolved oxygen concentration change records of the dissolved oxygen sensor in the nearshore cages, compare them with the set dissolved oxygen change threshold, and record an abnormality flag as a valid state if the current dissolved oxygen change exceeds the dissolved oxygen change threshold. Otherwise, record an invalid state and generate a dissolved oxygen abnormality judgment result. S4: When the density change trend determination result is recorded as valid and the dissolved oxygen anomaly determination result is recorded as valid, read the upper boundary record of the original normal interval and assign it to the lower boundary position of the warning interval to form a new interval division. In other cases, keep the original interval division unchanged. Determine the interval assignment position of the new interval division and the fish density record of the net cage water body to generate the aquaculture risk interval assignment result. S5: Based on the aquaculture risk zone attribution results, extract the corresponding nearshore cage number, map the zone attribution record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

[0006] As a further aspect of the present invention, the fish density record in the cage water includes density value, collection time identifier, spatial location coordinates, and data confidence level; the density change trend determination result includes trend direction identifier, change intensity level, and trend validity status record; the dissolved oxygen anomaly determination result includes anomaly level classification result, fluctuation amplitude index, environmental stability coefficient, and anomaly status identifier; the aquaculture risk range attribution result includes risk level category, range boundary identifier, risk trigger type, and risk persistence status; and the marine aquaculture status assessment result includes cage status identifier text, risk warning level, spatial distribution layer mark, and comprehensive assessment label.

[0007] As a further aspect of the present invention, the effective and invalid states of the trend trigger flag specifically refer to the following: when the signs are consistent, the absolute values ​​of the first difference and the second difference are processed respectively, and the absolute value of the second difference is compared with the absolute value of the first difference. When the absolute value of the second difference is greater than the absolute value of the first difference, the trend trigger flag is recorded as effective; otherwise, it is recorded as invalid. The set dissolved oxygen change threshold specifically refers to continuously collecting periodic dissolved oxygen change records and calculating the average value and standard deviation, and then adding the average value and standard deviation to obtain the set dissolved oxygen change threshold.

[0008] As a further aspect of the present invention, the step of obtaining the fish density record in the net cage water body specifically includes: S111: Collect image frames from underwater cameras inside nearshore cages, detect the grayscale difference between adjacent image frames and construct a foreground pixel difference matrix, remove connected regions with areas lower than the preset connected region area threshold, retain continuously distributed connected regions and extract the boundary coordinate set to generate a set of fish body outline pixel regions. S112: Based on the set of pixel regions of the fish body outline, count the number of pixels in each connected region, perform a summation operation to obtain the total area of ​​the fish body outline pixels, and at the same time, count the product of the number of rows and columns of the resolution of a single frame image to obtain the total area of ​​pixels in a single frame. Calculate the ratio of the total area of ​​the fish body outline pixels to the total area of ​​pixels in a single frame to generate pixel area ratio data. S113: Extract the nearshore cage volume parameters, calculate the fish distribution per unit volume based on the proportional relationship between the pixel area ratio data and the nearshore cage volume parameters, perform numerical recording encoding operation and associate with timestamp data frames to generate a fish density record for the cage water body.

[0009] As a further aspect of the present invention, the step of obtaining the density change trend determination result specifically includes: S211: Obtain the fish density records of the net cage water body at three consecutive time points, perform sequence rearrangement according to the timestamp ascending order, extract the density values ​​of two adjacent time points and perform subtraction to form the first difference, extract the density values ​​of the next two time points and perform subtraction to form the second difference, establish time index association for the difference, and generate a double difference sequence. S212: Extract the signs of the first difference and the second difference values ​​from the double difference sequence and perform a sign consistency judgment. Record the sign consistency. When the signs are consistent, perform absolute value operations on the first difference and the second difference respectively, compare the absolute value of the second difference with the absolute value of the first difference, and assign the trend trigger flag to a valid state when the absolute value of the second difference is greater than the absolute value of the first difference. Otherwise, assign the invalid state and generate a trend trigger flag record. S213: Based on the trend trigger flag record, read the second difference record result, establish a key-value mapping structure, use the trend trigger flag as the key item and the second difference record result as the value item, perform combined storage and write the time index identifier, and generate the density change trend determination result.

[0010] As a further aspect of the present invention, the step of obtaining the dissolved oxygen anomaly determination result specifically includes: S311: Monitor dissolved oxygen concentration records of nearshore cages, collect dissolved oxygen concentration values ​​at two adjacent time points, perform difference calculation between the value at the later time and the value at the earlier time, associate with timestamp index, and generate dissolved oxygen change records. S312: Collect multiple dissolved oxygen change records based on the dissolved oxygen change records and aggregate them into a periodic sequence. Calculate the mean and standard deviation of the periodic sequence. Perform an addition operation on the mean and standard deviation to generate a set dissolved oxygen change threshold. S313: Extract the current dissolved oxygen change value based on the dissolved oxygen change record. If the current dissolved oxygen change value exceeds the set dissolved oxygen change threshold, record the abnormality flag as valid. Otherwise, record the abnormality flag as invalid. Establish a mapping between the abnormality flag and the timestamp to generate a dissolved oxygen abnormality judgment result.

[0011] As a further aspect of the present invention, the step of obtaining the aquaculture risk range attribution result specifically includes: S411: Obtain the density change trend determination result and the dissolved oxygen anomaly determination result, perform state equivalence comparison operation, when both state values ​​are valid, read the upper boundary record of the original normal interval and write the corresponding value to the lower boundary position of the warning interval, establish a new interval division, when any state value is invalid, keep the original interval division value unchanged, and generate an interval boundary value table. S412: Extract the lower and upper boundary values ​​of the new interval based on the interval boundary value table, obtain the current density value of the fish density record in the net cage water, detect the sampling period duration and the number of sampling points, read the net cage volume parameters and volume benchmark value, and combine the dissolved oxygen concentration record of the nearshore net cage dissolved oxygen sensor, the dissolved oxygen benchmark value and the trend trigger flag to calculate and obtain the risk discrimination index, and compare it with the preset judgment benchmark value. Perform size comparisons and record the determination signs to generate interval assignment results; S413: Based on the interval attribution judgment result, perform interval range matching operation in conjunction with the interval boundary value table, establish a mapping relationship between the judgment symbol and the corresponding interval number and write it into the density time index entry to obtain the aquaculture risk interval attribution result.

[0012] As a further aspect of the present invention, the calculation formula for the risk discrimination index is specifically as follows: ; in, Represents a risk assessment index. Represents the current density value. Represents the lower boundary value of the new interval. Represents the upper boundary value of the new interval. Represents the duration of the sampling period. Represents the number of sampling points. Represents the volume parameters of the net cage. Represents the volume reference value. This represents the dissolved oxygen concentration at the current moment. This represents the dissolved oxygen concentration value at the previous moment. Represents the dissolved oxygen baseline value. This is a dimensionless trend trigger flag value; it is 1 for a valid state and 0 for an invalid state. It represents the time base value and is a dimensionless constant.

[0013] As a further aspect of the present invention, the steps for obtaining the marine aquaculture status assessment results are specifically as follows: S511: Obtain the aquaculture risk zone attribution result, retrieve the zone number and extract the corresponding nearshore cage number, establish a mapping between cage number and zone number, and generate a cage zone mapping matrix; S512: Based on the cage interval mapping matrix, map the interval number to the standard status identifier text encoding value, verify the one-to-one correspondence between the encoding value and the interval number, establish the number and text key value structure, and generate the status identifier encoding set; S513: Based on the status identifier encoding set, extract the nearshore cage location coordinate data, perform cage number and location coordinate matching, write the status identifier text encoding value into the corresponding coordinate record and perform spatial index update operation, establish the location coordinate and status text association structure, and output the marine aquaculture status assessment result.

[0014] A marine aquaculture status assessment system includes: The density acquisition module is used to perform S1: acquire image frames from the underwater camera device inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the density of fish in the cage water by combining the cage volume parameters. The trend determination module is used to execute S2: read the fish density records of the net cage water body at three consecutive time points, arrange them in chronological order, perform adjacent time difference calculation to form a first difference and a second difference, perform sign judgment on the first difference and the second difference to determine if the signs are consistent, record the effective and invalid states of the trend trigger flag, and generate a density change trend determination result; The dissolved oxygen identification module is used to perform S3: monitor the dissolved oxygen concentration change records of the dissolved oxygen sensor in the nearshore cage, compare them with the set dissolved oxygen change threshold, and record an abnormality flag as a valid state if the current dissolved oxygen change exceeds the dissolved oxygen change threshold, otherwise record an invalid state, and generate a dissolved oxygen abnormality judgment result. The interval assignment module is used to execute S4: when the density change trend determination result record is valid and the dissolved oxygen anomaly determination result record is valid, read the original normal interval upper boundary record and assign it to the lower boundary position of the warning interval to form a new interval division. In other cases, keep the original interval division unchanged, and determine the interval assignment position by comparing the new interval division record with the fish density record of the net cage water body to generate the aquaculture risk interval assignment result. The status output module is used to execute S5: based on the aquaculture risk zone attribution results, extract the corresponding nearshore cage number, map the zone attribution record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, the density of fish schools is calculated by continuously identifying image frames inside the net cages and combining multiple time-series differences to determine the density change trend. At the same time, the magnitude of dissolved oxygen change and statistical thresholds are introduced for anomaly identification. When trend signals and environmental anomalies occur simultaneously, the risk interval boundaries are dynamically adjusted to keep the risk assessment synchronized with the actual changes in the water body. This enables the linkage analysis of fish school aggregation and environmental fluctuations, improves the sensitivity of risk identification and the flexibility of interval division, enhances the real-time and targeted nature of marine aquaculture status assessment, and ensures that on-site management decisions have a higher degree of matching and reliability. Attached Figure Description

[0016] Figure 1 This is a flowchart of the main steps of the present invention; Figure 2 This is a flowchart illustrating the process of recording and obtaining fish density in the net cage water body according to the present invention. Figure 3 This is a flowchart of the process for obtaining the density change trend determination result in this invention; Figure 4 This is a flowchart of the process for obtaining the dissolved oxygen anomaly determination result in this invention; Figure 5 This is a flowchart illustrating the process of obtaining the aquaculture risk range attribution results in this invention. Figure 6 This is a flowchart illustrating the process of obtaining the marine aquaculture status assessment results according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] Please see Figure 1 A method for assessing the status of marine aquaculture includes the following steps: S1: Collect image frames from the underwater camera inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the fish density in the cage water by combining the cage volume parameters. S2: Read the fish density records of the net cage water at three consecutive time points, arrange them in chronological order, and perform adjacent time difference calculation to form the first difference and the second difference. Perform sign judgment on the first difference and the second difference and record the sign matching. When the signs match, perform absolute value processing on the first difference and the second difference respectively, and compare the absolute value of the second difference with the absolute value of the first difference. When the absolute value of the second difference is greater than the absolute value of the first difference, record the trend trigger flag as valid. Otherwise, record it as invalid. Combine and store the trend trigger flag and the second difference record results to generate the density change trend judgment result. S3: Monitor dissolved oxygen concentration records of nearshore cages, continuously record dissolved oxygen concentration at two time points and perform difference calculation to form dissolved oxygen change records, continuously collect dissolved oxygen change records for multiple periods and calculate the average value and standard deviation, add the average value and standard deviation to construct the set dissolved oxygen change threshold, compare the current dissolved oxygen change with the set dissolved oxygen change threshold, if the current dissolved oxygen change record exceeds the dissolved oxygen change threshold record, the record is marked as valid, otherwise the record is marked as invalid, and a dissolved oxygen anomaly judgment result is generated; S4: When the density change trend determination result is recorded as valid and the dissolved oxygen abnormality determination result is recorded as valid, read the original normal interval upper boundary record and assign it to the lower boundary position of the warning interval to form a new interval division. In other cases, keep the original interval division unchanged, and determine the interval assignment position by comparing the new interval division record with the fish density record of the net cage water body to generate the aquaculture risk interval assignment result. S5: Based on the aquaculture risk zone attribution results, extract the corresponding nearshore cage number, map the zone attribution record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

[0020] Fish density records in net cage water include density values, collection time markers, spatial location coordinates, and data confidence levels. Density change trend determination results include trend direction markers, change intensity levels, and trend validity records. Dissolved oxygen anomaly determination results include anomaly level classification results, fluctuation amplitude indicators, environmental stability coefficients, and anomaly status markers. Aquaculture risk range attribution results include risk level categories, range boundary markers, risk trigger types, and risk persistence status. Marine aquaculture status assessment results include net cage status marker text, risk warning levels, spatial distribution layer markings, and comprehensive assessment labels.

[0021] Please see Figure 2 Step S1 is as follows: S111: Collect image frames from underwater cameras inside nearshore cages, detect the grayscale difference between adjacent image frames and construct a foreground pixel difference matrix, remove connected regions with areas lower than the preset connected region area threshold, retain continuously distributed connected regions and extract the boundary coordinate set to generate a set of fish body outline pixel regions. Continuous image frames were acquired from a camera located 5 meters underwater inside a nearshore fish cage, with the video stream sampling frequency set to 60 frames per second. The original RGB format images were extracted, and the color channels of each pixel were iterated and weighted averages were performed to map them uniformly to an 8-bit single-channel grayscale space. The 100th frame was extracted as the current image frame, and the 99th frame as the adjacent previous image frame. The grayscale values ​​at their corresponding spatial coordinates were extracted and subtracted pixel by pixel. The absolute values ​​of all subtraction differences were processed to obtain a 2D data array of 2,073,600 grayscale differences. A preset dynamic background denoising value of 25 was read. This value was set by extracting the maximum grayscale fluctuation amplitude within the 95% confidence interval of 200 consecutively acquired pure water background images without fish activity. The 2D data array was iterated, assigning 255 to the pixel values ​​corresponding to difference coordinates greater than or equal to 25, and assigning 0 to the pixel values ​​corresponding to coordinates less than 25, thereby constructing a foreground pixel difference matrix. A connected component labeling algorithm is used to scan the matrix, counting the total number of pixels within adjacent, continuous regions with a pixel value of 255, which is used as the area value of each connected component. A connected component area threshold of 150 is set, derived from 50% of the minimum projected pixel area of ​​a standard adult fish at the camera's furthest focal length of 8 meters (300). Connected components with an area value lower than 150 are discarded, and their internal pixel values ​​are reset to 0. Boundary tracing is performed on the remaining continuous connected regions with an area greater than or equal to 150, extracting the coordinate set of the outermost pixels. All boundary coordinates that meet the criteria are concatenated into an array to generate a closed set of pixel regions representing the fish's outline.

[0022] S112: Based on the set of pixel regions of the fish body outline, count the number of pixels in each connected region, perform a summation operation to obtain the total area of ​​the fish body outline pixels, and at the same time, count the product of the number of rows and columns of the resolution of a single frame image to obtain the total area of ​​pixels in a single frame. Calculate the ratio of the total area of ​​the fish body outline pixels to the total area of ​​pixels in a single frame to generate pixel area ratio data. Read the 75 discrete connected regions marked with globally unique identifiers from the previously generated set of fish outline pixel regions. For each independent connected region, extract the pixel count value contained within it and push it into the summation queue. Perform an accumulation summation operation to calculate the total pixel count of all 75 connected regions, obtaining the total fish outline pixel area of ​​358400. Obtain the horizontal resolution value of 1920 and the vertical resolution value of 1080 of a single frame image, and perform a multiplication operation on these two parameters to obtain the total pixel area of ​​a single frame of 2073600. Use the total fish outline pixel area of ​​358400 as the numerator and the total pixel area of ​​a single frame of 2073600 as the denominator, and perform a division operation to obtain a floating-point ratio of 0.1728. Multiply this ratio by a constant of 100 for format conversion to generate a pixel area ratio of 17.28, and convert this data into a 64-bit floating-point data format and write it into the shared memory pool.

[0023] S113: Extract the volume parameters of nearshore net cages, calculate the fish distribution per unit volume based on the proportional relationship between pixel area ratio data and net cage volume parameters, perform numerical recording encoding operation and associate with timestamp data frames to generate fish density records of net cage water. Physical measurement parameters of nearshore net cages were recorded, and the 3D spatial dimensions of the cage's interior (length 50, width 50, depth 20) were extracted. These three values ​​were multiplied to obtain a cage volume parameter of 50000. The pixel area ratio data of 17.28 was read from the shared memory pool. Based on a pre-set area ratio and volume density conversion coefficient library, a conversion product coefficient of 4.5 was retrieved. This conversion product coefficient of 4.5 was obtained by placing a known number of fish samples in a standard experimental tank and performing linear regression fitting between their actual density and the collected pixel ratio. The pixel area ratio data of 17.28 and the conversion product coefficient of 4.5 were multiplied to calculate the fish distribution per unit volume of 77.76. This distribution value of 77.76 was then subjected to standardized data sequence encoding to obtain a hexadecimal floating-point code. Simultaneously, the network time protocol interface is called to obtain the current precise timestamp of the system, 1710500060. The hexadecimal floating-point code is then bound and mapped to the timestamp data frame using a key-value pair. This generates a structured record of fish density in the net cage water with time attributes, which is then persistently stored in the time-series database.

[0024] Please see Figure 3 Step S2 is as follows: S211: Obtain the fish density records of the net cage water body at three consecutive time points, perform sequence rearrangement according to the timestamp in ascending order, extract the density values ​​of two adjacent time points and perform subtraction to form the first difference, extract the density values ​​of the next two time points and perform subtraction to form the second difference, establish time index association for the difference, and generate a double difference sequence. Retrieve fish density records for net cages at three consecutive time points from the time-series database and extract the timestamp data associated with each record. Obtain timestamps 1710500000, 1710500060, and 1710500120, and rearrange them according to ascending order of value to determine the logical order of the previous, middle, and subsequent timestamps. Extract the corresponding fish density values ​​based on the rearranged timestamp order: 75.20, 76.50, and 79.12. Extract the middle time density value (76.50) and the previous time density value (75.20), and subtract them to obtain the first difference of 1.30. Extract the subsequent time density value (79.12) and the middle time density value (76.50), and subtract them to obtain the second difference of 2.62. The calculated first difference 1.30 and second difference 2.62 are used as keys to establish a time index association structure with their corresponding trigger time points 1710500060 and 1710500120, respectively. This generates a double difference sequence containing the fluctuation gradients of the two time points and pushes it into the memory queue.

[0025] S212: Extract the signs of the first and second differences from the double difference sequence and perform a sign consistency judgment. Record the sign consistency. When the signs are consistent, perform absolute value operations on the first and second differences respectively, compare the absolute values ​​of the second and first differences, and assign a trend trigger flag as valid when the absolute value of the second difference is greater than the absolute value of the first difference. Otherwise, assign an invalid state and generate a trend trigger flag record. For the double-difference sequence in the memory queue, the sign of the first difference 1.30 and the second difference 2.62 are extracted as positive. An equality matching operation is performed on the signs of both, and if the signs are consistent, the result is recorded as logically true in the status register. For this logically true state, absolute value operations are performed on the first difference 1.30 and the second difference 2.62 respectively, obtaining the absolute value of the first difference (1.30) and the absolute value of the second difference (2.62). The size comparison operator in the instruction set is read, and the absolute value of the second difference (2.62) is compared with the absolute value of the first difference (1.30). Since 2.62 is greater than 1.30, the fish gathering and fluctuation in the aquatic environment is determined to be accelerating. In this case, the trend trigger flag is assigned a valid state, represented by a Boolean value of 1; if the absolute value of the second difference is not greater than the absolute value of the first difference, it is assigned an invalid state, represented by a Boolean value of 0. The current Boolean value 1 is encapsulated and output to generate a trend trigger flag record.

[0026] S213: Based on the trend trigger flag record, read the second difference record result, establish a key-value mapping structure, use the trend trigger flag as the key item, use the second difference record result as the value item, perform combined storage and write it to the time index identifier, and generate the density change trend determination result. Read the previously obtained trend trigger flag record, extract its state Boolean value of 1, and simultaneously retrieve the corresponding second difference record result 2.62 from the double difference sequence. Create a new key-value mapping structure in the cache dictionary, using the trend trigger flag's Boolean value of 1 as the key and the second difference record result 2.62 as the value, and perform a dictionary combination storage operation. Extract the corresponding later time index identifier 1710500120 from the time series database and append it as metadata to the header of the key-value mapping structure. After completing all data concatenation, solidify and output the dictionary structure to generate the density change trend determination result for subsequent analysis.

[0027] Please see Figure 4 Step S3 is as follows: S311: Monitor dissolved oxygen concentration records of nearshore cages, collect dissolved oxygen concentration values ​​at two adjacent time points, perform difference calculation between the value at the later time and the value at the earlier time, associate with timestamp index, and generate dissolved oxygen change records. Dissolved oxygen concentration records were periodically reported by a sensor deployed 8 meters underwater in nearshore net cages, with a sampling frequency of once every 60 seconds. Data from two adjacent time points were extracted: 6.80 for the previous time and 6.50 for the subsequent time. The difference between the subsequent value of 6.50 and the previous value of 6.80 was calculated, yielding a difference of -0.30. The timestamp 1710500120 corresponding to the subsequent time was extracted as the primary key index. The calculated difference of -0.30 was then bound to this timestamp as a foreign key, generating an independent dissolved oxygen change record, which was then appended to a sliding window buffer queue of length 120.

[0028] S312: Collect multi-period dissolved oxygen change records based on dissolved oxygen change records and aggregate them into a periodic sequence. Calculate the mean and standard deviation of the periodic sequence. Perform an addition operation on the mean and standard deviation to generate the set dissolved oxygen change threshold. Dissolved oxygen variation records for the past 120 sampling periods are read from the sliding window buffer queue. These discretely distributed data are serialized and assembled in ascending order of timestamp, forming a one-dimensional periodic sequence of length 120. This one-dimensional periodic sequence is iterated through, and the absolute values ​​of the 120 variation differences are summed. The sum is divided by the total number 120 to calculate the periodic sequence mean, which is 0.15. Further, the difference between each value in the periodic sequence and the mean of 0.15 is extracted and squared. All 120 squared values ​​are summed, divided by 120, and the square root is taken to calculate the periodic sequence standard deviation, which is 0.04. The periodic sequence mean of 0.15 and the periodic sequence standard deviation of 0.04 are added together to obtain the set dissolved oxygen variation threshold, which is 0.19.

[0029] Table 1. Dynamic characteristics of dissolved oxygen periodic sequence; S313: Extract the current dissolved oxygen change value based on the dissolved oxygen change record. If the current dissolved oxygen change value exceeds the set dissolved oxygen change threshold, record the abnormality flag as valid. Otherwise, record the abnormality flag as invalid. Establish a mapping between the abnormality flag and the timestamp to generate a dissolved oxygen abnormality judgment result. As shown in Table 1, three sets of recent periodic dissolved oxygen statistics were extracted. Based on the latest generated dissolved oxygen change record, the absolute value of the current dissolved oxygen change (0.30) was extracted. The set dissolved oxygen change threshold (0.19) for periodic sequence number 1 in Table 1 was read, and the current dissolved oxygen change value (0.30) was compared with the set threshold (0.19). Since 0.30 is significantly greater than the threshold standard of 0.19, an anomaly flag was recorded in the system memory as valid, represented by the value 1. If the change value at the current time is 0.11, the trigger condition is not met, and an anomaly flag was recorded as invalid, represented by the value 0. The timestamp of the current data trigger (1710500120) was obtained, and a hash table mapping was established between the anomaly flag (value 1) and this timestamp. The resulting dissolved oxygen anomaly judgment was encapsulated and sent to the data input end of the early warning rule engine.

[0030] Please see Figure 5 Step S4 is as follows: S411: Obtain the density change trend judgment result and the dissolved oxygen anomaly judgment result, perform state equivalence comparison operation, when both state values ​​are valid, read the upper boundary record of the original normal interval and write the corresponding value to the lower boundary position of the warning interval, establish a new interval division, when either state value is invalid, keep the original interval division value unchanged, and generate an interval boundary value table. Retrieve the density change trend judgment result and dissolved oxygen anomaly judgment result residing in the early warning rule engine. Extract the effective state value 1 for the density change trend and the effective state value 1 for the dissolved oxygen anomaly. Perform a state equivalence comparison operation on these two state values, determining that both state values ​​are effective states. Based on the triggering condition of dual effective states, read the data structure of the original normal interval in the configuration file and extract the upper boundary record value 80.00 of the original normal interval. Take 80.00 from the upper boundary variable and directly overwrite it to the memory address at the lower boundary of the early warning interval. Simultaneously expand the upper boundary of the early warning interval to the preset safety maximum value of 120.00, thereby establishing a new interval division for high-risk environments. If any state value in the aforementioned equivalence comparison operation is 0, skip the overwrite instruction and keep the original interval boundary value unchanged. Integrate all updated upper and lower boundary values ​​into a two-dimensional list to generate the interval boundary value table as shown in Table 2.

[0031] Table 2. Values ​​of the boundary of the cage environment zone; S412: Extract the lower and upper boundary values ​​of the new interval from the interval boundary value table, obtain the current density value of fish population in the cage water, detect the sampling period and number of sampling points, read the cage volume parameters and volume baseline value, monitor the dissolved oxygen concentration record of the nearshore cage dissolved oxygen sensor, and combine the dissolved oxygen baseline value and trend trigger flag, using the formula: ; The risk discrimination index is calculated and compared with the preset judgment benchmark value. Perform size comparisons and record the determination signs to generate interval assignment results; where, Represents a risk assessment index. Represents the current density value. Represents the lower boundary value of the new interval. Represents the upper boundary value of the new interval. Represents the duration of the sampling period. Represents the number of sampling points. Represents the volume parameters of the net cage. Represents the volume reference value. This represents the dissolved oxygen concentration at the current moment. This represents the dissolved oxygen concentration value at the previous moment. Represents the dissolved oxygen baseline value. This is a dimensionless trend trigger value, set to 1 for valid states and 0 for invalid states. It is used to incrementally compensate the risk assessment index when the density change trend accelerates. The time reference value is a dimensionless constant whose value can be set according to the sampling period, such as 60 seconds or 1 second, and is used to normalize the sampling time term into a dimensionless ratio.

[0032] Extract the lower boundary value of the new interval with the interval level label 2 (80.00) and the upper boundary value of the new interval (120.00) from Table 2. Obtain the fish density record of the net cage water at the current sampling time from the time series database, and read the current density value (85.50). Call the system clock to detect and extract the current sampling period duration as 60. Query the sampling queue pool to detect and extract the number of sampling points within this period as 10. Read the 3D physical parameter calculation results of the net cage and extract the net cage volume parameter (50000). Read the system baseline configuration file and extract the volume baseline value (10000). Access the sensor real-time data stream, extract the current dissolved oxygen concentration value (6.50) and the previous dissolved oxygen concentration value (6.80), and obtain the absolute difference of the corresponding dissolved oxygen change record (0.30). Read the dissolved oxygen baseline value (5.00) set in the hydrological standard library. Extract the effective status of the trend trigger flag generated by the preceding judgment and convert it to obtain the trend trigger flag value (1). Assign L the new lower boundary value of 80.00, U the new upper boundary value of 120.00, D the current density value of 85.50, T the sampling period duration of 60, N the number of sampling points of 10, V the cage volume parameter of 50000, and 10000 the volume baseline value. The current dissolved oxygen concentration is 6.50. The dissolved oxygen concentration value at the previous moment was 6.80. A difference of 0.30 in the recorded dissolved oxygen change is assigned ΔDO, and a dissolved oxygen baseline value of 5.00 is assigned ΔDO. The trend trigger flag value of 1 is assigned. Substitute all the above parameters into the formula: ; Based on the input values, mathematical logic operations are performed to calculate the final value of the risk discrimination index R, which is 31.4225. The advantage of this formula lies in its ability to correlate high-frequency sampling characteristics (T to N ratio) with spatial capacity load (V to N ratio). The ratio is multiplicatively combined and passed through the tail. The Boolean compensation allows the formula to release a significantly amplified risk index in the initial stage when fish exhibit abnormal aggregation tendencies. Since 31.4225 far exceeds the benchmark value of 1, a positive sign indicating a severe deviation from the standard is recorded. This result indicates that the water body is currently in a state of dual danger—a dramatic increase in density and depletion of dissolved oxygen—providing a rigid qualitative condition for subsequent interval rating. The positive sign is output as the interval assignment result.

[0033] S413: Based on the interval attribution judgment result, perform interval range matching operation in combination with the interval boundary value table, establish a mapping relationship between the judgment symbol and the corresponding interval number and write it into the density time index entry to obtain the aquaculture risk interval attribution result; Extract the positive sign of the generated interval attribution judgment result and extract the interval boundary value table from Table 2. Extract the current density value 85.50 and perform a numerical comparison operation with the lower boundary value and upper boundary value corresponding to each interval level label in Table 2. It is found that 85.50 is greater than 80.00 and less than 120.00, and successfully matches the interval level label 2. Establish a numeric key-value pair mapping relationship between the judgment symbol containing high-risk semantics and the obtained interval level label 2. Call the time series database API interface, locate the extended field in the density time index entry with index number 1710500120, write the constructed numeric key-value pair into this field, execute the transaction commit command, and save the generated complete aquaculture risk interval attribution result.

[0034] Please see Figure 6 The S5 steps are as follows: S511: Obtain the aquaculture risk zone attribution results, retrieve the zone number and extract the corresponding nearshore cage number, establish a mapping between cage number and zone number, and generate a cage zone mapping matrix; Access the specified data table in the time-series database to retrieve the aquaculture risk range attribution result just written. Perform a JSON format parsing operation to extract the range level label value 2. Send a query request to the network topology management service to perform a correlation search in the cage registry based on range level label 2, extracting the corresponding nearshore cage number value 1005 that experienced environmental fluctuations. Allocate a set of 2D array memory blocks in system memory, write the cage number value 1005 into the first column index position, and write the range level label value 2 into the second column index position. By establishing a two-way binding record for this row of data, generate a cage range mapping matrix for the next step of risk text translation.

[0035] Table 3. Dictionary of marine aquaculture facility status mapping; S512: Based on the cage interval mapping matrix, map the interval number to the standard status identifier text encoding value, verify the one-to-one correspondence between the encoding value and the interval number, establish the number and text key value structure, and generate the status identifier encoding set; The generated cage interval mapping matrix in memory is read, and the interval level label value 2 is extracted. The system's preset state translation configuration dictionary is accessed, which predefines hard conversion rules for interval level labels to standard state identifier text encoding values. Based on these rules, the interval level label value 2 is mapped to the text encoding value 4 representing a secondary high-risk warning state. A consistency check function is invoked, and the mapped encoding value 4 is compared bidirectionally with the source interval level label value 2 to confirm that the correspondence is unique in the configuration dictionary and that no out-of-bounds conflict has occurred. After successful verification, a key-value mapping structure between the cage number value 1005 and the state identifier encoding value 4 is established in the cache pool. This structure is then encapsulated in the state identifier encoding set and distributed externally.

[0036] S513: Based on the status identifier encoding set, call the nearshore cage location coordinate data, perform cage number and location coordinate matching, write the status identifier text encoding value into the corresponding coordinate record and perform spatial index update operation, establish the location coordinate and status text association structure, and output the marine aquaculture status assessment result. Referring to Table 3, obtain the externally distributed status identifier code set and extract the cage number value 1005 and status identifier code value 4. Initiate a location query request command based on cage number value 1005 to the external geographic information system server. Obtain the JSON response packet returned by the server and parse it to extract the latitude coordinates (22.542) and longitude coordinates (114.053) of the nearshore cage entity in the WGS84 coordinate system. In the local spatial database, perform a record positioning operation based on the cage number, matching the latitude and longitude coordinates with the cage number for spatial attributes. Subsequently, perform string serialization conversion on the status identifier code value 4 and write it to the business status field of the current spatial record. Trigger the built-in update mechanism of the spatial database engine to reconstruct the R-tree spatial index structure for this cage node, enabling a strong association between the location coordinate data and the status text data in the underlying storage block. Integrate all the data entities containing latitude, longitude, and coded status, and output the final marine aquaculture status assessment result to the front-end monitoring screen.

[0037] A marine aquaculture status assessment system includes: The density acquisition module is used to perform S1: acquire image frames from the underwater camera device inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the density of fish in the cage water by combining the cage volume parameters. The trend determination module is used to execute S2: read the fish density records of the net cage water at three consecutive time points, arrange them in chronological order, perform adjacent time difference calculation to form the first difference and the second difference, perform sign judgment on the first difference and the second difference to determine if the signs are consistent, record the valid and invalid states of the trend trigger flag, and generate the density change trend determination result; The dissolved oxygen identification module is used to perform S3: monitor the dissolved oxygen concentration change records of the dissolved oxygen sensor in the nearshore cage, compare them with the set dissolved oxygen change threshold, and record an abnormality flag as a valid state if the current dissolved oxygen change exceeds the dissolved oxygen change threshold, otherwise record an invalid state, and generate a dissolved oxygen abnormality judgment result. The interval assignment module is used to execute S4: When the density change trend judgment result record is valid and the dissolved oxygen abnormality judgment result record is valid, the upper boundary record of the original normal interval is read and assigned to the lower boundary position of the warning interval to form a new interval division. In other cases, the original interval division remains unchanged. The new interval division record is compared with the fish density record of the net cage water body to determine the interval assignment position and generate the aquaculture risk interval assignment result. The status output module is used to execute S5: based on the aquaculture risk zone assignment results, extract the corresponding nearshore cage number, map the zone assignment record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

[0038] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for assessing the status of marine aquaculture, characterized in that, Includes the following steps: S1: Collect image frames from the underwater camera inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the fish density in the cage water by combining the cage volume parameters. S2: Read the fish density records of the net cage water body at three consecutive time points, arrange them in chronological order, perform adjacent time difference calculation to form the first difference and the second difference, perform sign judgment on the first difference and the second difference to determine if the signs are consistent, record the effective and invalid states of the trend trigger flag, and generate the density change trend judgment result; S3: Monitor the dissolved oxygen concentration change records of the dissolved oxygen sensor in the nearshore cages, compare them with the set dissolved oxygen change threshold, and record an abnormality flag as a valid state if the current dissolved oxygen change exceeds the dissolved oxygen change threshold. Otherwise, record an invalid state and generate a dissolved oxygen abnormality judgment result. S4: When the density change trend determination result is recorded as valid and the dissolved oxygen anomaly determination result is recorded as valid, read the upper boundary record of the original normal interval and assign it to the lower boundary position of the warning interval to form a new interval division. In other cases, keep the original interval division unchanged. Determine the interval assignment position of the new interval division and the fish density record of the net cage water body to generate the aquaculture risk interval assignment result. S5: Based on the aquaculture risk zone attribution results, extract the corresponding nearshore cage number, map the zone attribution record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

2. The method for assessing the status of marine aquaculture according to claim 1, characterized in that: The fish density record in the cage water includes density value, collection time identifier, spatial location coordinates, and data confidence level. The density change trend determination result includes trend direction identifier, change intensity level, and trend validity status record. The dissolved oxygen anomaly determination result includes anomaly level classification result, fluctuation amplitude index, environmental stability coefficient, and anomaly status identifier. The aquaculture risk range attribution result includes risk level category, range boundary identifier, risk trigger type, and risk persistence status. The marine aquaculture status assessment result includes cage status identifier text, risk warning level, spatial distribution layer mark, and comprehensive assessment label.

3. The method for assessing the status of marine aquaculture according to claim 1, characterized in that: The effective and invalid states of the trend trigger flag specifically refer to the following: when the signs are consistent, the absolute values ​​of the first difference and the second difference are processed respectively, and the absolute value of the second difference is compared with the absolute value of the first difference. When the absolute value of the second difference is greater than the absolute value of the first difference, the trend trigger flag is recorded as effective; otherwise, it is recorded as invalid. The set dissolved oxygen change threshold specifically refers to continuously collecting periodic dissolved oxygen change records and calculating the average value and standard deviation, and then adding the average value and standard deviation to obtain the set dissolved oxygen change threshold.

4. The method for assessing the status of marine aquaculture according to claim 1, characterized in that, The specific steps for obtaining the fish density record in the net cage water body are as follows: S111: Collect image frames from underwater cameras inside nearshore cages, detect the grayscale difference between adjacent image frames and construct a foreground pixel difference matrix, remove connected regions with areas lower than the preset connected region area threshold, retain continuously distributed connected regions and extract the boundary coordinate set to generate a set of fish body outline pixel regions. S112: Based on the set of pixel regions of the fish body outline, count the number of pixels in each connected region, perform a summation operation to obtain the total area of ​​the fish body outline pixels, and at the same time, count the product of the number of rows and columns of the resolution of a single frame image to obtain the total area of ​​pixels in a single frame. Calculate the ratio of the total area of ​​the fish body outline pixels to the total area of ​​pixels in a single frame to generate pixel area ratio data. S113: Extract the nearshore cage volume parameters, calculate the fish distribution per unit volume based on the proportional relationship between the pixel area ratio data and the nearshore cage volume parameters, perform numerical recording encoding operation and associate with timestamp data frames to generate a fish density record for the cage water body.

5. The method for assessing the status of marine aquaculture according to claim 1, characterized in that, The specific steps for obtaining the density change trend determination result are as follows: S211: Obtain the fish density records of the net cage water body at three consecutive time points, perform sequence rearrangement according to the timestamp ascending order, extract the density values ​​of two adjacent time points and perform subtraction to form the first difference, extract the density values ​​of the next two time points and perform subtraction to form the second difference, establish time index association for the difference, and generate a double difference sequence. S212: Extract the signs of the first difference and the second difference values ​​from the double difference sequence and perform a sign consistency judgment. Record the sign consistency. When the signs are consistent, perform absolute value operations on the first difference and the second difference respectively, compare the absolute value of the second difference with the absolute value of the first difference, and assign the trend trigger flag to a valid state when the absolute value of the second difference is greater than the absolute value of the first difference. Otherwise, assign the invalid state and generate a trend trigger flag record. S213: Based on the trend trigger flag record, read the second difference record result, establish a key-value mapping structure, use the trend trigger flag as the key item and the second difference record result as the value item, perform combined storage and write the time index identifier, and generate the density change trend determination result.

6. The method for assessing the status of marine aquaculture according to claim 1, characterized in that, The specific steps for obtaining the dissolved oxygen anomaly determination result are as follows: S311: Monitor dissolved oxygen concentration records of nearshore cages, collect dissolved oxygen concentration values ​​at two adjacent time points, perform difference calculation between the value at the later time and the value at the earlier time, associate with timestamp index, and generate dissolved oxygen change records. S312: Collect multiple dissolved oxygen change records based on the dissolved oxygen change records and aggregate them into a periodic sequence. Calculate the mean and standard deviation of the periodic sequence. Perform an addition operation on the mean and standard deviation to generate a set dissolved oxygen change threshold. S313: Extract the current dissolved oxygen change value based on the dissolved oxygen change record. If the current dissolved oxygen change value exceeds the set dissolved oxygen change threshold, record the abnormality flag as valid. Otherwise, record the abnormality flag as invalid. Establish a mapping between the abnormality flag and the timestamp to generate a dissolved oxygen abnormality judgment result.

7. The method for assessing the status of marine aquaculture according to claim 1, characterized in that, The specific steps for obtaining the aquaculture risk range attribution results are as follows: S411: Obtain the density change trend determination result and the dissolved oxygen anomaly determination result, perform state equivalence comparison operation, when both state values ​​are valid, read the upper boundary record of the original normal interval and write the corresponding value to the lower boundary position of the warning interval, establish a new interval division, when any state value is invalid, keep the original interval division value unchanged, and generate an interval boundary value table. S412: Extract the lower and upper boundary values ​​of the new interval based on the interval boundary value table, obtain the current density value of the fish population in the cage water, detect the sampling period duration and the number of sampling points, read the cage volume parameters and volume baseline value, and combine the dissolved oxygen concentration record of the nearshore cage dissolved oxygen sensor, the dissolved oxygen baseline value and the trend trigger flag to calculate and obtain the risk discrimination index, and compare it with the preset judgment baseline value. Perform size comparisons and record the determination signs to generate interval assignment results; S413: Based on the interval attribution judgment result, perform interval range matching operation in conjunction with the interval boundary value table, establish a mapping relationship between the judgment symbol and the corresponding interval number and write it into the density time index entry to obtain the aquaculture risk interval attribution result.

8. The method for assessing the status of marine aquaculture according to claim 7, characterized in that, The specific formula for calculating the risk discrimination index is as follows: ； in, Represents a risk assessment index. Represents the current density value. Represents the lower boundary value of the new interval. Represents the upper boundary value of the new interval. Represents the duration of the sampling period. Represents the number of sampling points. Represents the volume parameters of the net cage. Represents the volume reference value. This represents the dissolved oxygen concentration at the current moment. This represents the dissolved oxygen concentration value at the previous moment. Represents the dissolved oxygen baseline value. This is a dimensionless trend trigger flag value; it is 1 for a valid state and 0 for an invalid state. It represents the time base value and is a dimensionless constant.

9. The method for assessing the status of marine aquaculture according to claim 1, characterized in that, The specific steps for obtaining the marine aquaculture status assessment results are as follows: S511: Obtain the aquaculture risk zone attribution result, retrieve the zone number and extract the corresponding nearshore cage number, establish a mapping between cage number and zone number, and generate a cage zone mapping matrix; S512: Based on the cage interval mapping matrix, map the interval number to the standard status identifier text encoding value, verify the one-to-one correspondence between the encoding value and the interval number, establish the number and text key value structure, and generate the status identifier encoding set; S513: Based on the status identifier encoding set, extract the nearshore cage location coordinate data, perform cage number and location coordinate matching, write the status identifier text encoding value into the corresponding coordinate record and perform spatial index update operation, establish the location coordinate and status text association structure, and output the marine aquaculture status assessment result.

10. A marine aquaculture status assessment system, characterized in that, The system is used to implement the marine aquaculture status assessment method according to any one of claims 1-9, comprising: The density acquisition module is used to perform S1: acquire image frames from the underwater camera device inside the nearshore cage, identify continuous fish outline pixel areas and count the total pixel area, convert the total fish outline pixel area to the total pixel area of ​​a single frame, and estimate the density of fish in the cage water by combining the cage volume parameters. The trend determination module is used to execute S2: read the fish density records of the net cage water body at three consecutive time points, arrange them in chronological order, perform adjacent time difference calculation to form a first difference and a second difference, perform sign judgment on the first difference and the second difference to determine if the signs are consistent, record the effective and invalid states of the trend trigger flag, and generate a density change trend determination result; The dissolved oxygen identification module is used to perform S3: monitor the dissolved oxygen concentration change records of the dissolved oxygen sensor in the nearshore cage, compare them with the set dissolved oxygen change threshold, and record an abnormality flag as a valid state if the current dissolved oxygen change exceeds the dissolved oxygen change threshold, otherwise record an invalid state, and generate a dissolved oxygen abnormality judgment result. The interval assignment module is used to execute S4: when the density change trend determination result record is valid and the dissolved oxygen anomaly determination result record is valid, read the original normal interval upper boundary record and assign it to the lower boundary position of the warning interval to form a new interval division. In other cases, keep the original interval division unchanged, and determine the interval assignment position by comparing the new interval division record with the fish density record of the net cage water body to generate the aquaculture risk interval assignment result. The status output module is used to execute S5: based on the aquaculture risk interval attribution results, extract the corresponding nearshore cage number, map the interval attribution record to standard status identifier text, and display the status of the corresponding nearshore cage location to form a marine aquaculture status assessment result.

Images