Infectious spleen and kidney necrosis virus detection method and system based on tangential flow technology

Abstract

The invention discloses an infectious spleen and kidney necrosis virus detection method and system based on a tangential flow technology. Fish school videos and water body data are obtained by arranging monitoring equipment. A fish school monitoring image set is formed through key frame extraction, and fish school color and contour features are extracted by using an image recognition model. And in the two culture cycles, analyzing fish school feature data in combination with preset standard features, generating disease indexes, and screening out predicted infection and non-infection areas. Comparing the prediction infection areas of the two periods, marking the same area as an infection analysis area, and calculating a diffusion index. The infection condition is judged according to the diffusion index and an early warning threshold value, if infection is predicted to exist, culture water in an infection analysis area is sampled, viruses are filtered and detected through the tangential flow technology, and virus infection data are generated. According to the invention, the disease of the cultured fish school can be accurately analyzed, so that subsequent targeted virus detection is facilitated, and a scientific decision basis is provided for farmers.

Description

Technical Field

[0001] The present invention relates to the field of data analysis and virus detection, and more specifically, to a method and system for detecting infectious spleen and kidney necrosis virus based on tangential flow technology. Background Art

[0002] Infectious spleen and kidney necrosis is a disease caused by the infectious spleen and kidney necrosis virus that seriously harms the aquaculture industry. Traditional detection methods are complex and time-consuming, and cannot meet the needs of large-scale rapid screening. Existing technologies lack prediction and detection analysis of the virus, making it difficult to improve aquaculture efficiency. Summary of the invention

[0003] The present invention overcomes the defects of the prior art and proposes an infectious spleen and kidney necrosis virus detection method and system based on tangential flow technology.

[0004] The first aspect of the present invention provides a method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology, comprising:

[0005] Based on the target water aquaculture area, multiple aquaculture monitoring areas are divided;

[0006] Monitoring equipment is installed in each aquaculture monitoring area to obtain video data and water data of farmed fish through the monitoring equipment;

[0007] Extract key frames from the video data to form a fish monitoring image set. Based on a breeding area, extract image data from the fish monitoring image set and import it into the image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features, and obtain fish color feature data and contour feature data.

[0008] In the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the color feature data and contour feature data of the fish school are analyzed for ISKNV virus symptoms. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the disease index; in the second breeding cycle, through the above screening process, the second predicted infected area and the second predicted non-infected area are screened out based on the disease index;

[0009] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, mark the same area based on the two cycle analysis to obtain the infection analysis area. In the two cycles, calculate the ratio of the number of predicted infection areas to predicted non-infected areas in the adjacent area corresponding to the infection analysis area, and obtain two ratios corresponding to the two breeding cycles. The difference between the two ratios is used as the diffusion index.

[0010] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses using tangential flow technology, and virus infection data is generated.

[0011] In this scheme, multiple aquaculture monitoring areas are divided based on the target water aquaculture area, specifically:

[0012] Obtain basic information on the target water aquaculture area;

[0013] Basic information includes the scale of farming, the size of the water body farming area, and the types of farmed fish;

[0014] The breeding monitoring needs are analyzed through basic information, and corresponding monitoring points are set, and multiple breeding monitoring areas are divided through the monitoring points.

[0015] In this solution, monitoring equipment is set up in each aquaculture monitoring area, and video data and water body data of farmed fish are obtained through the monitoring equipment, specifically:

[0016] Monitoring equipment is installed in each aquaculture monitoring area, including a camera and a water monitoring device;

[0017] Video data and water data of farmed fish are obtained through monitoring equipment.

[0018] In this solution, the video data is subjected to key frame extraction to form a fish monitoring image set. Based on a breeding area, image data is extracted from the fish monitoring image set and imported into an image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features. The color feature data and contour feature data of the fish are obtained as follows:

[0019] Extract key frames from video data to form a fish school monitoring image set;

[0020] Perform image smoothing, noise reduction, and enhancement preprocessing on fish monitoring image sets;

[0021] Based on a farming area, image data is extracted from the fish monitoring image set and imported into the image recognition model;

[0022] Fish schools are identified and their features are extracted through an image recognition model. The feature extraction includes two dimensions: color features and contour features of fish body parts, and the color feature data and contour features of the fish schools are obtained.

[0023] In this scheme, in the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the fish color feature data and contour feature data are analyzed for ISKNV virus symptoms. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result. The first predicted infected area and the first predicted non-infected area are screened out based on the disease index, specifically:

[0024] In the first breeding cycle, the color characteristic data and the contour characteristic data of the corresponding fish school are obtained;

[0025] Obtain color contrast features and contour contrast features based on ISKNV virus symptoms from the system database;

[0026] The color feature data and contour feature data of the corresponding fish school are compared with the color contrast feature and contour contrast feature respectively. The difference comparison is calculated based on the cosine similarity of the feature data, and two difference values ​​are obtained. The mean is calculated based on the two difference values, and the virus infection prediction and evaluation are performed to obtain the disease index;

[0027] The breeding monitoring areas with disease index higher than the preset index are marked as the first predicted infected areas, and the rest are marked as the first predicted non-infected areas.

[0028] In this solution, in the second breeding cycle, the second predicted infected area and the second predicted non-infected area are screened out based on the symptom index through the above screening process, and the first breeding cycle and the second breeding cycle are continuous cycles.

[0029] In this scheme, it is determined whether the first predicted infection area and the second predicted infection area have the same area. If so, the same area based on the analysis of the two cycles is marked to obtain the infection analysis area. In the two cycles, the ratio of the number of predicted infection areas and predicted non-infected areas in the adjacent areas corresponding to the infection analysis area is calculated respectively, and two ratios corresponding to the two breeding cycles are obtained. The difference between the two ratios is used as the diffusion index, which is specifically:

[0030] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, select the same predicted infection area based on the analysis and screening of the two breeding cycles and mark it as the infection analysis area;

[0031] In the first breeding cycle, the ratio of the number of predicted infected areas to the number of predicted non-infected areas in the neighboring areas of the infected analysis area is calculated, and the average ratio is calculated based on the number of infected analysis areas;

[0032] The second breeding cycle was analyzed and the corresponding average ratio was obtained;

[0033] The diffusion index was obtained by subtracting the average ratio of the first breeding cycle from that of the second breeding cycle.

[0034] In this scheme, the diffusion index and the warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses through tangential flow technology, and virus infection data is generated, specifically:

[0035] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infected analysis area is used as the sample area for water sampling;

[0036] The water samples are filtered and the viruses are enriched in the hollow fiber membrane through the tangential flow technology. The enrichment efficiency and purity of the virus particles are controlled by adjusting the fluid pressure and flow rate. The virus particles are then detected to generate virus infection data based on the infection analysis area.

[0037] The second aspect of the present invention also provides an infectious spleen and kidney necrosis virus detection system based on tangential flow technology, the system comprising: a memory, a processor, the memory comprising an infectious spleen and kidney necrosis virus detection program based on tangential flow technology, the infectious spleen and kidney necrosis virus detection program based on tangential flow technology when executed by the processor implements the following steps:

[0038] Based on the target water aquaculture area, multiple aquaculture monitoring areas are divided;

[0039] Monitoring equipment is installed in each aquaculture monitoring area to obtain video data and water data of farmed fish through the monitoring equipment;

[0040] Extract key frames from the video data to form a fish monitoring image set. Based on a breeding area, extract image data from the fish monitoring image set and import it into the image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features, and obtain fish color feature data and contour feature data.

[0041] In the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the color feature data and contour feature data of the fish school are analyzed for ISKNV virus symptoms. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the disease index; in the second breeding cycle, through the above screening process, the second predicted infected area and the second predicted non-infected area are screened out based on the disease index;

[0042] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, mark the same area based on the two cycle analysis to obtain the infection analysis area. In the two cycles, calculate the ratio of the number of predicted infection areas to predicted non-infected areas in the adjacent area corresponding to the infection analysis area, and obtain two ratios corresponding to the two breeding cycles. The difference between the two ratios is used as the diffusion index.

[0043] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses using tangential flow technology, and virus infection data is generated.

[0044] The third aspect of the present invention also provides a computer-readable storage medium, which includes an infectious spleen and kidney necrosis virus detection program based on tangential flow technology. When the infectious spleen and kidney necrosis virus detection program based on tangential flow technology is executed by a processor, the steps of the infectious spleen and kidney necrosis virus detection method based on tangential flow technology as described in any of the above items are implemented.

[0045] The present invention discloses an infectious spleen and kidney necrosis virus detection method and system based on tangential flow technology, and obtains fish video and water body data by setting monitoring equipment. A fish monitoring image set is formed by key frame extraction, and the color and contour features of the fish are extracted by using an image recognition model. In two breeding cycles, the fish characteristic data is analyzed in combination with preset standard features, a symptom index is generated, and predicted infected and non-infected areas are screened out. The predicted infected areas of the two cycles are compared, the same areas are marked as infection analysis areas, and the diffusion index is calculated. The infection situation is judged according to the diffusion index and the early warning threshold. If the infection is predicted, the breeding water in the infection analysis area is sampled, and the virus is filtered and detected using tangential flow technology to generate virus infection data. Through the present invention, accurate analysis of the symptoms of farmed fish can be achieved, so that targeted virus detection can be carried out later, providing a scientific decision-making basis for farmers. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A flow chart showing a method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology of the present invention is shown;

[0047] Figure 2 A block diagram of an infectious spleen and kidney necrosis virus detection system based on tangential flow technology of the present invention is shown. DETAILED DESCRIPTION

[0048] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0050] Figure 1 The flowchart of the infectious spleen and kidney necrosis virus detection method based on tangential flow technology of the present invention is shown.

[0051] like Figure 1 As shown, the first aspect of the present invention provides a method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology, comprising:

[0052] S102, dividing a plurality of aquaculture monitoring areas based on the target water body aquaculture area;

[0053] S104, setting up monitoring equipment in each aquaculture monitoring area, and obtaining video data and water body data of aquaculture fish through the monitoring equipment;

[0054] S106, extract key frames from the video data to form a fish monitoring image set, extract image data from the fish monitoring image set based on a breeding area and import the image recognition model to extract fish features, the feature extraction includes two dimensions: color feature and fish body part contour feature, and obtain fish color feature data and contour feature data;

[0055] S108, in the first breeding cycle, taking the breeding monitoring area as the analysis unit, performing ISKNV virus symptom analysis on the color feature data and the contour feature data of the fish school, performing comparison in the analysis process in combination with the preset standard features, generating a symptom index of the breeding area based on the comparison result, and screening out a first predicted infected area and a first predicted non-infected area based on the symptom index; in the second breeding cycle, through the above screening process, screening out a second predicted infected area and a second predicted non-infected area based on the symptom index;

[0056] S110, determining whether the first predicted infection area and the second predicted infection area have the same area, if so, marking the same area based on the two cycle analysis to obtain the infection analysis area, and calculating the number ratio of the predicted infection area to the predicted non-infected area in the adjacent area corresponding to the infection analysis area in the two cycles, and obtaining two ratios corresponding to the two breeding cycles, and taking the difference between the two ratios as the diffusion index;

[0057] S112, determine whether there is an infection at present through the diffusion index and the warning threshold. If so, use the infection analysis area as the sample area, sample the aquaculture water based on the sample area, filter and detect the virus on the water sample through the tangential flow technology, and generate virus infection data.

[0058] It should be noted that ISKNV virus is infectious spleen and kidney necrosis virus.

[0059] According to an embodiment of the present invention, the aquaculture area of ​​the target water body is divided into multiple aquaculture monitoring areas, specifically:

[0060] Obtain basic information on the target water aquaculture area;

[0061] Basic information includes the scale of farming, the size of the water body farming area, and the types of farmed fish;

[0062] The breeding monitoring needs are analyzed through basic information, and corresponding monitoring points are set, and multiple breeding monitoring areas are divided through the monitoring points.

[0063] It should be noted that each breeding monitoring area corresponds to a monitoring point.

[0064] According to an embodiment of the present invention, a monitoring device is set in each aquaculture monitoring area, and video data and water body data of aquaculture fish are obtained through the monitoring device, specifically:

[0065] Monitoring equipment is installed in each aquaculture monitoring area, including a camera and a water monitoring device;

[0066] Video data and water data of farmed fish are obtained through monitoring equipment.

[0067] It should be noted that the water body monitoring device can measure relevant water body environmental data, such as water temperature, pH, dissolved oxygen content, etc.

[0068] According to an embodiment of the present invention, the video data is subjected to key frame extraction to form a fish school monitoring image set. Based on a breeding area, image data is extracted from the fish school monitoring image set and imported into an image recognition model to extract fish school features. The feature extraction includes two dimensions: color features and fish body part contour features. The color feature data and contour feature data of the fish school are obtained as follows:

[0069] Extract key frames from video data to form a fish school monitoring image set;

[0070] Perform image smoothing, noise reduction, and enhancement preprocessing on fish monitoring image sets;

[0071] Based on a farming area, image data is extracted from the fish monitoring image set and imported into the image recognition model;

[0072] Fish schools are identified and their features are extracted through an image recognition model. The feature extraction includes two dimensions: color features and contour features of fish body parts, and the color feature data and contour features of the fish schools are obtained.

[0073] It should be noted that the image recognition model can use target recognition algorithm models such as CNN and YOLO. Since ISKNV infection can easily cause fish to have wide mouths, protruding eyes, enlarged spleens, and black fish bodies, the contour features of the fish body are mainly extracted to analyze the biological characteristics of the fish, such as the fish mouth, fish eyes, and fish body.

[0074] According to an embodiment of the present invention, in the first breeding cycle, the breeding monitoring area is used as the analysis unit, the fish color feature data and the contour feature data are subjected to ISKNV virus symptom analysis, the analysis process is compared with the preset standard features, and the symptom index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the symptom index, specifically:

[0075] In the first breeding cycle, the color characteristic data and the contour characteristic data of the corresponding fish school are obtained;

[0076] Obtain color contrast features and contour contrast features based on ISKNV virus symptoms from the system database;

[0077] The color feature data and contour feature data of the corresponding fish school are compared with the color contrast feature and contour contrast feature respectively. The difference comparison is calculated based on the cosine similarity of the feature data, and two difference values ​​are obtained. The mean is calculated based on the two difference values, and the virus infection prediction and evaluation are performed to obtain the disease index;

[0078] The breeding monitoring areas with disease index higher than the preset index are marked as the first predicted infected areas, and the rest are marked as the first predicted non-infected areas.

[0079] It should be noted that the difference value is the distance value calculated based on the cosine similarity. The larger the distance value, the greater the difference, which means the lower the probability of ISKNV symptoms, and the lower the corresponding symptom index. The symptom index is inversely proportional to the mean of the two difference values.

[0080] According to an embodiment of the present invention, in the second breeding cycle, the second predicted infected area and the second predicted non-infected area are screened out based on the symptom index through the above screening process, and the first breeding cycle and the second breeding cycle are continuous cycles.

[0081] It should be noted that the analysis process of the second breeding cycle is based on the fish school monitoring image set of the second breeding cycle.

[0082] According to an embodiment of the present invention, it is determined whether the first predicted infection area and the second predicted infection area have the same area. If so, the same area based on the analysis of the two cycles is marked to obtain the infection analysis area. In the two cycles, the ratio of the number of predicted infection areas to the predicted non-infected areas in the adjacent areas corresponding to the infection analysis area is calculated respectively, and two ratios corresponding to the two breeding cycles are obtained. The difference between the two ratios is used as the diffusion index, which is specifically:

[0083] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, select the same predicted infection area based on the analysis and screening of the two breeding cycles and mark it as the infection analysis area;

[0084] In the first breeding cycle, the ratio of the number of predicted infected areas to the number of predicted non-infected areas in the neighboring areas of the infected analysis area is calculated, and the average ratio is calculated based on the number of infected analysis areas;

[0085] The second breeding cycle was analyzed and the corresponding average ratio was obtained;

[0086] The diffusion index was obtained by subtracting the average ratio of the first breeding cycle from that of the second breeding cycle.

[0087] It should be noted that the adjacent areas may be adjacent areas that are in contact with each other or adjacent areas within a certain preset range, which is specifically set by the user.

[0088] In the prediction area analysis of the first breeding cycle and the second breeding cycle, due to the different fish identification results, the corresponding prediction area conditions are also different. Therefore, based on one breeding monitoring area, if both cycles are marked as predicted infection areas, this area is the infection analysis area.

[0089] The diffusion index calculation formula is as follows:

[0090]

[0091] Among them, P is the diffusion index, K is the correction coefficient, N is the total number of infection analysis areas, and D i , C represent the predicted number of infected areas and the predicted number of non-infected areas in the neighboring area corresponding to the i-th infection analysis area in the first breeding cycle, respectively, and F i 、E i They respectively represent the predicted number of infected areas and the predicted number of non-infected areas in the neighboring areas corresponding to the i-th infection analysis area in the second breeding cycle. This represents a ratio calculation.

[0092] It is worth mentioning that the infectious spleen and kidney necrosis virus is generally a sudden viral infection in aquaculture, accompanied by explosive death of aquaculture organisms in a short period of time. Therefore, its incidence is difficult to predict. If the virus is frequently tested, it will greatly increase the consumption of manpower and material resources and increase the cost of aquaculture, making it difficult to achieve an increase in aquaculture benefits. In addition, ISKNV has a high degree of infectivity, making it difficult to monitor, predict and detect aquatic virus infections.

[0093] In aquaculture, if infected by ISKNV, sick fish will have symptoms such as wide mouths and blackened bodies, sometimes accompanied by cramp-like tremors, and may also have enlarged spleens, erosions and congestion, and enlarged kidneys. In general underwater image analysis, since fish also die in general and underwater images are relatively fuzzy, the analysis of fish symptoms through underwater fish image monitoring often has large errors. Therefore, it is necessary to conduct an overall fish analysis of the breeding area and a more practical and precise analysis plan to achieve accurate analysis of farmed fish symptoms, so as to conduct targeted virus detection in the future and provide scientific decision-making basis for farmers.

[0094] Therefore, the present invention analyzes the characteristics of fish schools in different breeding areas, calculates the corresponding disease index, and screens out the predicted infected areas and predicted non-infected areas under different cycles. The predicted infected areas are areas with a higher probability of virus existence, and the predicted non-infected areas are areas with a lower probability of virus existence. Through the analysis of two cycles, the common predicted infected areas are screened out. If there are common predicted infected areas, it means that based on the time dimension, there are areas with a fixed prediction result of high probability of infection. This area has high representativeness and is the key area for virus analysis. Further, the present invention calculates the diffusion index of two consecutive cycles to further comprehensively judge whether the key area is contagious. At the same time, because the underwater image is relatively fuzzy and there are also recognition errors in judging the changes in the fish body, there are certain misjudgments. Based on this, the present invention sets multi-cycle data for virus diffusion prediction and analysis, and judges the probability of virus existence from the overall perspective, thereby reducing misjudgments, and then reducing the consumption of unnecessary manpower and material resources. At the same time, it realizes real-time early warning of viruses and disease conditions.

[0095] In addition, users can adjust the cycle length to improve the sensitivity of monitoring to adapt to different breeding situations.

[0096] The larger the diffusion index is and the positive it is, the higher the probability of infection in the corresponding area is, that is, the higher the probability of virus infection is.

[0097] According to an embodiment of the present invention, the diffusion index and the warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, aquaculture water is sampled based on the sample area, and the water sample is filtered and virus detected by tangential flow technology, and virus infection data is generated, specifically:

[0098] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infected analysis area is used as the sample area for water sampling;

[0099] The water samples are filtered and the viruses are enriched in the hollow fiber membrane through the tangential flow technology. The enrichment efficiency and purity of the virus particles are controlled by adjusting the fluid pressure and flow rate. The virus particles are then detected to generate virus infection data based on the infection analysis area.

[0100] The present invention improves the accuracy of aquaculture prediction and virus detection through two aspects: predictive data and detection mode (detected through tangential flow technology), thereby reducing aquaculture risks and costs and improving aquaculture benefits.

[0101] According to an embodiment of the present invention, judging whether there is currently an infection by using the diffusion index and the warning threshold value further includes:

[0102] In the second cycle, water body data and meteorological data of each infection analysis area were obtained;

[0103] The water body data and meteorological data are converted into multi-dimensional characteristic data to form environmental characteristic data, which includes characteristic data of three dimensions: water temperature, pH, and meteorology;

[0104] The environmental feature data and the preset feature data are vectorized and the difference is calculated based on the Euclidean distance, and the weight value of each infection analysis area is set based on the size of the difference;

[0105] According to the weight value, based on the diffusion index of all infection analysis areas, the corresponding weighted diffusion index is calculated;

[0106] The weighted diffusion index is used to predict and evaluate viral infection.

[0107] It should be noted that due to the characteristics of viral infection and transmission, there is a certain correlation with environmental changes. For areas with large changes in environmental characteristics, environmental characteristics can be introduced to set weights for each infection analysis area, and weighted average can be performed when the diffusion index is subsequently calculated to improve the accuracy of diffusion analysis and provide more accurate virus predictions for breeding. The preset feature data is the environmental characteristics suitable for viral infection and transmission set based on ISKNV, which is used to compare and refer to the suitability of the current environmental state for ISKNV. The diffusion index and the early warning threshold are used to determine whether there is currently an infection, and an evaluation can also be made based on the weighted diffusion index.

[0108] Figure 2 A block diagram of an infectious spleen and kidney necrosis virus detection system based on tangential flow technology of the present invention is shown.

[0109] The second aspect of the present invention further provides an infectious spleen and kidney necrosis virus detection system 2 based on tangential flow technology, the system comprising: a memory 21, a processor 22, the memory 21 comprising an infectious spleen and kidney necrosis virus detection program based on tangential flow technology, the infectious spleen and kidney necrosis virus detection program based on tangential flow technology when executed by the processor 22 implements the following steps:

[0110] Based on the target water aquaculture area, multiple aquaculture monitoring areas are divided;

[0111] Monitoring equipment is installed in each aquaculture monitoring area to obtain video data and water data of farmed fish through the monitoring equipment;

[0112] Extract key frames from the video data to form a fish monitoring image set. Based on a breeding area, extract image data from the fish monitoring image set and import it into the image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features, and obtain fish color feature data and contour feature data.

[0113] In the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the color feature data and contour feature data of the fish school are analyzed for ISKNV virus symptoms. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the disease index; in the second breeding cycle, through the above screening process, the second predicted infected area and the second predicted non-infected area are screened out based on the disease index;

[0114] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, mark the same area based on the two cycle analysis to obtain the infection analysis area. In the two cycles, calculate the ratio of the number of predicted infection areas to predicted non-infected areas in the adjacent area corresponding to the infection analysis area, and obtain two ratios corresponding to the two breeding cycles. The difference between the two ratios is used as the diffusion index.

[0115] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses using tangential flow technology, and virus infection data is generated.

[0116] It should be noted that ISKNV virus is infectious spleen and kidney necrosis virus.

[0117] According to an embodiment of the present invention, the aquaculture area of ​​the target water body is divided into multiple aquaculture monitoring areas, specifically:

[0118] Obtain basic information on the target water aquaculture area;

[0119] Basic information includes the scale of farming, the size of the water body farming area, and the types of farmed fish;

[0120] The breeding monitoring needs are analyzed through basic information, and corresponding monitoring points are set, and multiple breeding monitoring areas are divided through the monitoring points.

[0121] It should be noted that each breeding monitoring area corresponds to a monitoring point.

[0122] According to an embodiment of the present invention, a monitoring device is set in each aquaculture monitoring area, and video data and water body data of aquaculture fish are obtained through the monitoring device, specifically:

[0123] Monitoring equipment is installed in each aquaculture monitoring area, including a camera and a water monitoring device;

[0124] Video data and water data of farmed fish are obtained through monitoring equipment.

[0125] It should be noted that the water body monitoring device can measure relevant water body environmental data, such as water temperature, pH, dissolved oxygen content, etc.

[0126] According to an embodiment of the present invention, the video data is subjected to key frame extraction to form a fish school monitoring image set. Based on a breeding area, image data is extracted from the fish school monitoring image set and imported into an image recognition model to extract fish school features. The feature extraction includes two dimensions: color features and fish body part contour features. The color feature data and contour feature data of the fish school are obtained as follows:

[0127] Extract key frames from video data to form a fish school monitoring image set;

[0128] Perform image smoothing, noise reduction, and enhancement preprocessing on fish monitoring image sets;

[0129] Based on a farming area, image data is extracted from the fish monitoring image set and imported into the image recognition model;

[0130] Fish schools are identified and their features are extracted through an image recognition model. The feature extraction includes two dimensions: color features and contour features of fish body parts, and the color feature data and contour features of the fish schools are obtained.

[0131] It should be noted that the image recognition model can use target recognition algorithm models such as CNN and YOLO. Since ISKNV infection can easily cause fish to have wide mouths, protruding eyes, enlarged spleens, and black fish bodies, the contour features of the fish body are mainly extracted to analyze the biological characteristics of the fish, such as the fish mouth, fish eyes, and fish body.

[0132] According to an embodiment of the present invention, in the first breeding cycle, the breeding monitoring area is used as the analysis unit, the fish color feature data and the contour feature data are subjected to ISKNV virus symptom analysis, the analysis process is compared with the preset standard features, and the symptom index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the symptom index, specifically:

[0133] In the first breeding cycle, the color characteristic data and the contour characteristic data of the corresponding fish school are obtained;

[0134] Obtain color contrast features and contour contrast features based on ISKNV virus symptoms from the system database;

[0135] The color feature data and contour feature data of the corresponding fish school are compared with the color contrast feature and contour contrast feature respectively. The difference comparison is calculated based on the cosine similarity of the feature data, and two difference values ​​are obtained. The mean is calculated based on the two difference values, and the virus infection prediction and evaluation are performed to obtain the disease index;

[0136] The breeding monitoring areas with disease index higher than the preset index are marked as the first predicted infected areas, and the rest are marked as the first predicted non-infected areas.

[0137] It should be noted that the difference value is the distance value calculated based on the cosine similarity. The larger the distance value, the greater the difference, which means the lower the probability of ISKNV symptoms, and the lower the corresponding symptom index. The symptom index is inversely proportional to the mean of the two difference values.

[0138] According to an embodiment of the present invention, in the second breeding cycle, the second predicted infected area and the second predicted non-infected area are screened out based on the symptom index through the above screening process, and the first breeding cycle and the second breeding cycle are continuous cycles.

[0139] It should be noted that the analysis process of the second breeding cycle is based on the fish school monitoring image set of the second breeding cycle.

[0140] According to an embodiment of the present invention, it is determined whether the first predicted infection area and the second predicted infection area have the same area. If so, the same area based on the analysis of the two cycles is marked to obtain the infection analysis area. In the two cycles, the ratio of the number of predicted infection areas to the predicted non-infected areas in the adjacent areas corresponding to the infection analysis area is calculated respectively, and two ratios corresponding to the two breeding cycles are obtained. The difference between the two ratios is used as the diffusion index, which is specifically:

[0141] Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, select the same predicted infection area based on the analysis and screening of the two breeding cycles and mark it as the infection analysis area;

[0142] In the first breeding cycle, the ratio of the number of predicted infected areas to the number of predicted non-infected areas in the neighboring areas of the infected analysis area is calculated, and the average ratio is calculated based on the number of infected analysis areas;

[0143] The second breeding cycle was analyzed and the corresponding average ratio was obtained;

[0144] The diffusion index was obtained by subtracting the average ratio of the first breeding cycle from that of the second breeding cycle.

[0145] It should be noted that the adjacent areas may be adjacent areas that are in contact with each other or adjacent areas within a certain preset range, which is specifically set by the user.

[0146] In the prediction area analysis of the first breeding cycle and the second breeding cycle, due to the different fish identification results, the corresponding prediction area conditions are also different. Therefore, based on one breeding monitoring area, if both cycles are marked as predicted infection areas, this area is the infection analysis area.

[0147] The diffusion index calculation formula is as follows:

[0148]

[0149] Among them, P is the diffusion index, K is the correction coefficient, N is the total number of infection analysis areas, and D i , C represent the predicted number of infected areas and the predicted number of non-infected areas in the neighboring area corresponding to the i-th infection analysis area in the first breeding cycle, respectively, and F i、E i They respectively represent the predicted number of infected areas and the predicted number of non-infected areas in the neighboring areas corresponding to the i-th infection analysis area in the second breeding cycle. This represents a ratio calculation.

[0150] It is worth mentioning that the infectious spleen and kidney necrosis virus is generally a sudden viral infection in aquaculture, accompanied by explosive death of aquaculture organisms in a short period of time. Therefore, its incidence is difficult to predict. If the virus is frequently tested, it will greatly increase the consumption of manpower and material resources and increase the cost of aquaculture, making it difficult to achieve an increase in aquaculture benefits. In addition, ISKNV has a high degree of infectivity, making it difficult to monitor, predict and detect aquatic virus infections.

[0151] In aquaculture, if infected by ISKNV, sick fish will have symptoms such as wide mouths and blackened bodies, sometimes accompanied by cramp-like tremors, and may also have enlarged spleens, erosions and congestion, and enlarged kidneys. In general underwater image analysis, since fish also die in general and underwater images are relatively fuzzy, the analysis of fish symptoms through underwater fish image monitoring often has large errors. Therefore, it is necessary to conduct an overall fish analysis of the breeding area and a more practical and precise analysis plan to achieve accurate analysis of farmed fish symptoms, so as to conduct targeted virus detection in the future and provide scientific decision-making basis for farmers.

[0152] Therefore, the present invention analyzes the characteristics of fish schools in different breeding areas, calculates the corresponding disease index, and screens out the predicted infected areas and predicted non-infected areas under different cycles. The predicted infected areas are areas with a higher probability of virus existence, and the predicted non-infected areas are areas with a lower probability of virus existence. Through the analysis of two cycles, the common predicted infected areas are screened out. If there are common predicted infected areas, it means that based on the time dimension, there are areas with a fixed prediction result of high probability of infection. This area has high representativeness and is the key area for virus analysis. Further, the present invention calculates the diffusion index of two consecutive cycles to further comprehensively judge whether the key area is contagious. At the same time, because the underwater image is relatively fuzzy and there are also recognition errors in judging the changes in the fish body, there are certain misjudgments. Based on this, the present invention sets multi-cycle data for virus diffusion prediction and analysis, and judges the probability of virus existence from the overall perspective, thereby reducing misjudgments, and then reducing the consumption of unnecessary manpower and material resources. At the same time, it realizes real-time early warning of viruses and disease conditions.

[0153] In addition, users can adjust the cycle length to improve the sensitivity of monitoring to adapt to different breeding situations.

[0154] The larger the diffusion index is and the positive it is, the higher the probability of infection in the corresponding area is, that is, the higher the probability of virus infection is.

[0155] According to an embodiment of the present invention, the diffusion index and the warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, aquaculture water is sampled based on the sample area, and the water sample is filtered and virus detected by tangential flow technology, and virus infection data is generated, specifically:

[0156] The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infected analysis area is used as the sample area for water sampling;

[0157] The water samples are filtered and the viruses are enriched in the hollow fiber membrane through the tangential flow technology. The enrichment efficiency and purity of the virus particles are controlled by adjusting the fluid pressure and flow rate. The virus particles are then detected to generate virus infection data based on the infection analysis area.

[0158] The present invention improves the accuracy of aquaculture prediction and virus detection through two aspects: predictive data and detection mode (detected through tangential flow technology), thereby reducing aquaculture risks and costs and improving aquaculture benefits.

[0159] The third aspect of the present invention also provides a computer-readable storage medium, which includes an infectious spleen and kidney necrosis virus detection program based on tangential flow technology. When the infectious spleen and kidney necrosis virus detection program based on tangential flow technology is executed by a processor, the steps of the infectious spleen and kidney necrosis virus detection method based on tangential flow technology as described in any of the above items are implemented.

[0160] The present invention discloses an infectious spleen and kidney necrosis virus detection method and system based on tangential flow technology, and obtains fish video and water body data by setting monitoring equipment. A fish monitoring image set is formed by key frame extraction, and the color and contour features of the fish are extracted by using an image recognition model. In two breeding cycles, the fish characteristic data is analyzed in combination with preset standard features, a symptom index is generated, and predicted infected and non-infected areas are screened out. The predicted infected areas of the two cycles are compared, the same areas are marked as infection analysis areas, and the diffusion index is calculated. The infection situation is judged according to the diffusion index and the early warning threshold. If the infection is predicted, the breeding water in the infection analysis area is sampled, and the virus is filtered and detected using tangential flow technology to generate virus infection data. Through the present invention, accurate analysis of the symptoms of farmed fish can be achieved, so that targeted virus detection can be carried out later, providing a scientific decision-making basis for farmers.

[0161] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0162] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0163] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0164] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, and other media that can store program codes.

[0165] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0166] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology, characterized in that: include: Based on the target water aquaculture area, multiple aquaculture monitoring areas are divided; Monitoring equipment is installed in each aquaculture monitoring area to obtain video data and water data of farmed fish through the monitoring equipment; Extract key frames from the video data to form a fish monitoring image set. Based on a breeding area, extract image data from the fish monitoring image set and import it into the image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features, and obtain fish color feature data and contour feature data. In the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the color feature data and contour feature data of the fish school are subjected to SKNV virus symptom analysis. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the disease index; in the second breeding cycle, through the above screening process, the second predicted infected area and the second predicted non-infected area are screened out based on the disease index; Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, mark the same area based on the two cycle analysis to obtain the infection analysis area. In the two cycles, calculate the ratio of the number of predicted infection areas to predicted non-infected areas in the adjacent area corresponding to the infection analysis area, and obtain two ratios corresponding to the two breeding cycles. The difference between the two ratios is used as the diffusion index. The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses using tangential flow technology, and virus infection data is generated.

2. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 1, characterized in that: Based on the target water body aquaculture area, multiple aquaculture monitoring areas are divided, specifically: Obtain basic information on the target water aquaculture area; Basic information includes the scale of farming, the size of the water body farming area, and the types of farmed fish; The breeding monitoring needs are analyzed through basic information, and corresponding monitoring points are set, and multiple breeding monitoring areas are divided through the monitoring points.

3. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 1, characterized in that: The monitoring equipment is set in each aquaculture monitoring area, and the video data and water body data of the aquaculture fish are obtained through the monitoring equipment, specifically: Monitoring equipment is installed in each aquaculture monitoring area, including a camera and a water monitoring device; Video data and water data of farmed fish are obtained through monitoring equipment.

4. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 1, characterized in that: The video data is subjected to key frame extraction to form a fish school monitoring image set. Based on a breeding area, image data is extracted from the fish school monitoring image set and imported into an image recognition model to extract fish school features. The feature extraction includes two dimensions: color features and fish body part contour features. The color feature data and contour feature data of the fish school are obtained as follows: Extract key frames from video data to form a fish school monitoring image set; Perform image smoothing, noise reduction, and enhancement preprocessing on fish monitoring image sets; Based on a farming area, image data is extracted from the fish monitoring image set and imported into the image recognition model; Fish schools are identified and their features are extracted through an image recognition model. The feature extraction includes two dimensions: color features and contour features of fish body parts, and the color feature data and contour features of the fish schools are obtained.

5. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 4, characterized in that: In the first breeding cycle, the breeding monitoring area is used as the analysis unit, the color feature data and the contour feature data of the fish school are subjected to SKNV virus symptom analysis, the analysis process is compared with the preset standard features, and the symptom index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the symptom index, specifically: In the first breeding cycle, the color characteristic data and the contour characteristic data of the corresponding fish school are obtained; Obtain color contrast features and contour contrast features based on SKNV virus symptoms from the system database; The color feature data and contour feature data of the corresponding fish school are compared with the color contrast feature and contour contrast feature respectively. The difference comparison is calculated based on the cosine similarity of the feature data, and two difference values ​​are obtained. The mean is calculated based on the two difference values, and the virus infection prediction and evaluation are performed to obtain the disease index; The breeding monitoring areas with disease index higher than the preset index are marked as the first predicted infected areas, and the rest are marked as the first predicted non-infected areas.

6. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 5, characterized in that: In the second breeding cycle, the second predicted infected area and the second predicted non-infected area are screened out based on the symptom index through the above screening process, and the first breeding cycle and the second breeding cycle are continuous cycles.

7. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 6, characterized in that: The method determines whether the first predicted infection area and the second predicted infection area have the same area. If so, the same area based on the analysis of the two cycles is marked to obtain the infection analysis area. In the two cycles, the ratio of the number of predicted infection areas to the number of predicted non-infected areas in the adjacent areas corresponding to the infection analysis area is calculated respectively, and two ratios corresponding to the two breeding cycles are obtained. The difference between the two ratios is used as the diffusion index, which is specifically: Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, select the same predicted infection area based on the analysis and screening of the two breeding cycles and mark it as the infection analysis area; In the first breeding cycle, the ratio of the number of predicted infected areas to the number of predicted non-infected areas in the neighboring areas of the infected analysis area is calculated, and the average ratio is calculated based on the number of infected analysis areas; The second breeding cycle was analyzed and the corresponding average ratio was obtained; The diffusion index was obtained by subtracting the average ratio of the first breeding cycle from that of the second breeding cycle.

8. The method for detecting infectious spleen and kidney necrosis virus based on tangential flow technology according to claim 7, characterized in that: The diffusion index and the warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses through tangential flow technology, and virus infection data is generated, specifically: The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infected analysis area is used as the sample area for water sampling; The water samples are filtered and the viruses are enriched in the hollow fiber membrane through the tangential flow technology. The enrichment efficiency and purity of the virus particles are controlled by adjusting the fluid pressure and flow rate. The virus particles are then detected to generate virus infection data based on the infection analysis area.

9. An infectious spleen and kidney necrosis virus detection system based on tangential flow technology, characterized in that: The system includes: a memory and a processor, wherein the memory includes an infectious spleen and kidney necrosis virus detection program based on tangential flow technology, and when the infectious spleen and kidney necrosis virus detection program based on tangential flow technology is executed by the processor, the following steps are implemented: Based on the target water aquaculture area, multiple aquaculture monitoring areas are divided; Monitoring equipment is installed in each aquaculture monitoring area to obtain video data and water data of farmed fish through the monitoring equipment; Extract key frames from the video data to form a fish monitoring image set. Based on a breeding area, extract image data from the fish monitoring image set and import it into the image recognition model to extract fish features. The feature extraction includes two dimensions: color features and fish body contour features, and obtain fish color feature data and contour feature data. In the first breeding cycle, the breeding monitoring area is used as the analysis unit, and the color feature data and contour feature data of the fish school are subjected to SKNV virus symptom analysis. The analysis process is compared with the preset standard features, and the disease index of the breeding area is generated based on the comparison result, and the first predicted infected area and the first predicted non-infected area are screened out based on the disease index; in the second breeding cycle, through the above screening process, the second predicted infected area and the second predicted non-infected area are screened out based on the disease index; Determine whether the first predicted infection area and the second predicted infection area have the same area. If so, mark the same area based on the two cycle analysis to obtain the infection analysis area. In the two cycles, calculate the ratio of the number of predicted infection areas to predicted non-infected areas in the adjacent area corresponding to the infection analysis area, and obtain two ratios corresponding to the two breeding cycles. The difference between the two ratios is used as the diffusion index. The diffusion index and warning threshold are used to determine whether there is an infection at present. If so, the infection analysis area is used as the sample area, and aquaculture water is sampled based on the sample area. The water samples are filtered and tested for viruses using tangential flow technology, and virus infection data is generated.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes an infectious spleen and kidney necrosis virus detection program based on tangential flow technology. When the infectious spleen and kidney necrosis virus detection program based on tangential flow technology is executed by a processor, the steps of the infectious spleen and kidney necrosis virus detection method based on tangential flow technology as described in any one of claims 1 to 8 are implemented.

Images