Intelligent feeding control method and equipment for seafood temporary rearing tank

Abstract

The invention provides an intelligent feeding control method and device for a seafood temporary rearing jar, and the method comprises the steps: calculating the parameter information of seafood based on an obtained seafood video of the seafood temporary rearing jar, calculating the feeding amount of the seafood based on the parameter information, carrying out the optical flow calculation of selected feature points in two continuous frames of images of the seafood video, and obtaining the feeding amount of the seafood based on the optical flow calculation. The method comprises the following steps: analyzing activity characteristics of seafood in a video based on characteristic point change calculated by optical flow, calculating an activity index of the seafood based on the activity characteristics, analyzing external characteristics of the seafood, performing weighted summation on a health index generated by an analysis result of the external characteristics and the activity index to obtain an adjustment factor, and adjusting the feeding amount by using the adjustment factor. And the feeder is controlled to feed the seafood in the seafood temporary rearing tank according to the adjusted feeding amount, so that the feeding amount of the seafood is comprehensively and accurately controlled, and insufficient feeding or excessive feeding is avoided.

Description

Technical Field

[0001] This application relates to the technical field of seafood temporary cultivation. Specifically, this application relates to an intelligent feeding control method and device for a seafood temporary cultivation tank. Background Art

[0002] A seafood temporary cultivation tank is a facility for temporarily storing and cultivating seafood, which allows seafood to be stored in an environment that simulates its natural habitat after being caught or during transportation. The main purpose of this facility is to maintain the freshness and vitality of seafood, and reduce its stress and mortality before transportation or sale.

[0003] In the traditional seafood temporary cultivation process, feeding often relies on manual experience or automatic feeders, but the feeding amount of seafood lacks scientific basis, which easily leads to underfeeding or overfeeding, affecting the health status and growth environment of seafood.

[0004] For example, in the technical solution with the patent application number 202111363565.2, it collects the characteristic data and water quality data of fish, analyzes the ideal water quality and feeding plan according to the characteristic data. The feeding plan includes feeding time, feeding amount and feeding speed. It adjusts the water quality according to the water quality data and the ideal water quality, and judges whether there is an abnormality in the water quality data at the feeding time. If not, it feeds according to the feeding plan; if there is an abnormality, it adjusts the water quality to normal and then feeds according to the feeding plan. Although this technical solution realizes the scientific control of the feeding amount to a certain extent, it does not deeply analyze the relationship between the characteristic data of fish and the feeding amount, resulting in low control accuracy of the feeding amount.

[0005] In addition, in the technical solution with the patent application number 202111275107.3, it obtains the image information and feeding parameters in the fish tank, identifies the types of fish in the image information, obtains the preset feeding conditions corresponding to the types of fish, and judges whether the feeding parameters meet the preset feeding conditions. If the feeding parameters meet the preset feeding conditions, it determines the quantity and body length of each type of fish, determines the type of fish feed to be fed according to the type of fish, and determines the required feeding amount of fish feed to be fed according to the quantity and body length of the fish, and feeds evenly according to the type of fish feed and the feeding amount. Although this technical solution can determine the feeding amount of fish feed to be fed according to the quantity and body length of fish, it does not consider the health status of fish, and it is also easy to lead to underfeeding or overfeeding.

[0006] Therefore, a more comprehensive solution is needed to accurately control the feeding amount of seafood in a seafood temporary cultivation tank. Summary of the Invention

[0007] The main purpose of this application is to provide an intelligent feeding control method and device for a seafood temporary cultivation tank to achieve intelligent and precise feeding amount control of seafood in the seafood temporary cultivation tank.

[0008] To achieve the above-mentioned invention purpose, this application provides an intelligent feeding control method for a seafood temporary cultivation tank, including:

[0009] Calculating the parameter information of seafood based on the obtained seafood video of the seafood temporary cultivation tank, and calculating the feeding amount of seafood based on the parameter information;

[0010] Performing optical flow calculation on the selected feature points in two consecutive frames of the seafood video, analyzing the activity characteristics of seafood in the video based on the change of feature points calculated by the optical flow, and calculating the activity index of seafood based on the activity characteristics;

[0011] Analyzing the external characteristics of seafood, performing weighted summation on the health index generated from the analysis result of the external characteristics and the activity index to obtain an adjustment factor;

[0012] Adjusting the feeding amount using the adjustment factor, and controlling the feeder to feed the seafood in the seafood temporary cultivation tank according to the adjusted feeding amount.

[0013] Preferably, the calculating the parameter information of seafood based on the obtained seafood video of the seafood temporary cultivation tank includes:

[0014] Identifying the category and quantity of seafood in the seafood video based on the target detection algorithm;

[0015] Calculating the ratio parameter of video pixels to actual length based on the actual size of the seafood temporary cultivation tank and the pixel size of the seafood temporary cultivation tank in the seafood video;

[0016] Detecting the key points of seafood in the seafood video, calculating the pixel distance between the key points of seafood based on the Euclidean distance formula, and multiplying the pixel distance by the ratio parameter to obtain the body size parameter of seafood. The parameter information of the seafood includes the category, quantity, and body size parameter of the seafood.

[0017] Furthermore, before identifying the category and quantity of seafood in the seafood video based on the target detection algorithm, it further includes:

[0018] Preprocessing the seafood video, and the preprocessing methods include denoising, contrast enhancement, and color correction.

[0019] Preferably, the performing optical flow calculation on the selected feature points in two consecutive frames of the seafood video includes:

[0020] Performing feature point matching on the selected feature points in two consecutive frames of the seafood video;

[0021] Calculate the Pearson correlation coefficient between each feature point in each frame of the image and the corresponding feature point in another frame of the image respectively;

[0022] Organize the calculated Pearson correlation coefficients into a correlation matrix according to the positions of the feature points. Each element in the correlation matrix represents the correlation strength between the corresponding feature points in two consecutive frames of the image;

[0023] Estimate the optical flow vector based on the correlation matrix. The optical flow vector is used to describe the change of the feature points between two consecutive frames of the image.

[0024] Preferably, the analysis of the external features of the seafood includes:

[0025] Extract the video frames of the seafood video, extract the external feature pixels of the seafood in each video frame, and obtain the seafood feature map corresponding to each video frame;

[0026] Multiply the seafood feature maps corresponding to all video frames pixel by pixel, and perform weighted summation to obtain the external feature representation of the seafood;

[0027] Calculate the health index of the seafood in the seafood holding tank based on the external feature representation.

[0028] Further, after calculating the health index of the seafood in the seafood holding tank based on the external feature representation, it further includes:

[0029] Judge whether the health index of the seafood in the seafood holding tank is abnormal;

[0030] When the health index is abnormal, send an alarm prompt message to the user terminal.

[0031] Preferably, the adjustment of the feeding amount by using the adjustment factor includes:

[0032] Calculate the function value with the natural logarithm as the base and the negative of the adjustment factor as the independent variable based on the exponential function;

[0033] Multiply the function value by the feeding amount to obtain the adjusted feeding amount.

[0034] Further, after controlling the feeder to feed the seafood in the seafood holding tank according to the adjusted feeding amount, it further includes:

[0035] After a preset time period of feeding the seafood, extract the water body image of the seafood holding tank from the seafood video obtained again;

[0036] Extract the hue value of each pixel in the water body image, and calculate the average hue value of the water body of the seafood holding tank based on the hue value of each pixel in the water body image;

[0037] When the average hue value is greater than the preset chromaticity value, calculate the gray-level co-occurrence matrix of the water body image, extract the texture features of the water body based on the gray-level co-occurrence matrix, and evaluate the turbidity of the water body based on the texture features of the water body;

[0038] When the turbidity is greater than the turbidity threshold, automatically change the water in the seafood temporary culture tank.

[0039] Further, after controlling the feeder to feed the seafood in the seafood temporary culture tank according to the adjusted feeding amount, it further includes:

[0040] Obtain the feedback on the feeding situation of the seafood;

[0041] Based on the feedback on the feeding situation, adjust the feeding amount again.

[0042] This application also provides an electronic device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method described in any one of the above.

[0043] An intelligent feeding control method and device for a seafood temporary culture tank provided by this application calculate the parameter information of the seafood by analyzing the seafood video, so as to determine a reasonable basic feeding amount, avoid the randomness of the traditional feeding method, and improve the scientificity and rationality of feeding. At the same time, by calculating the optical flow of the feature points in two consecutive frames of images, accurately analyze the activity characteristics of the seafood and calculate the activity index, perform a weighted sum of the health index and the activity index to obtain an adjustment factor, and fine-tune the basic feeding amount through the adjustment factor, so as to comprehensively consider the appearance and activity ability of the seafood, achieve a comprehensive and accurate adjustment of the basic feeding amount, avoid water pollution caused by overfeeding, and at the same time ensure that the seafood obtains sufficient nutrition, which is beneficial to maintaining the stability and health of the temporary culture environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic flow chart of an intelligent feeding control method for a seafood temporary culture tank according to an embodiment of this application;

[0045] Figure 2 It is a schematic flow chart of an intelligent feeding control method for a seafood temporary culture tank according to another embodiment of this application;

[0046] Figure 3 It is a schematic block diagram of the structure of an intelligent feeding control device for a seafood temporary culture tank according to an embodiment of this application;

[0047] Figure 4 It is a schematic block diagram of the structure of an electronic device according to an embodiment of this application.

[0048] The realization, functional features and advantages of the purpose of this application will be further described in conjunction with the embodiments with reference to the accompanying drawings. Detailed implementation manner

[0049] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application 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 only used to explain the present application and are not used to limit the present application.

[0050] An intelligent feeding control method for a seafood temporary holding tank proposed by the present application has an execution entity of an electronic device. The electronic device is electrically connected to a feeder and an underwater camera respectively. The feeder is installed on the tank wall of the seafood temporary holding tank and is used to automatically feed the seafood in the seafood temporary holding tank with feed. The underwater camera is used to collect video information of the seafood or water body in the seafood temporary holding tank. Among them, the electronic device can be any type of mobile or stationary computing device, including a mobile computer or a stationary computing device such as a desktop computer or a PC. The components of the computing device include but are not limited to a memory and a processor. The processor is connected to the memory through a bus, and a database is used to store data. The computing device also includes an access device, and the access device enables the computing device to communicate via one or more networks.

[0051] Reference Figure 1 As shown, in one embodiment, the present application provides an intelligent feeding control method for a seafood temporary holding tank. The method includes:

[0052] S11. Calculate the parameter information of the seafood based on the obtained seafood video of the seafood temporary holding tank, and calculate the feeding amount of the seafood based on the parameter information;

[0053] S12. Perform optical flow calculation on the selected feature points in two consecutive frames of images of the seafood video, analyze the activity characteristics of the seafood in the video based on the change of the feature points calculated by the optical flow, and calculate the activity index of the seafood based on the activity characteristics;

[0054] S13. Analyze the external characteristics of the seafood, and perform weighted summation on the health index generated by the analysis result of the external characteristics and the activity index to obtain an adjustment factor;

[0055] S14. Use the adjustment factor to adjust the feeding amount, and control the feeder to feed the seafood in the seafood temporary holding tank according to the adjusted feeding amount.

[0056] In this embodiment, relevant parameters of the seafood can be extracted from the video of the seafood temporary holding tank through video analysis technology, such as the size (length, width, volume, etc.), quantity, type, etc. of the seafood. For example, image processing algorithms can be used to analyze each frame in the video, identify the contour of the seafood, and then calculate its size parameters, and different types of seafood can be identified and counted through object detection algorithms.

[0057] Based on the calculated seafood parameter information, considering factors such as the growth stage and nutritional requirements of the seafood, calculate an appropriate feeding amount according to a certain feeding standard. For example, for a certain type of seafood, calculate the total feeding amount according to the standard of how many grams of feed are required per kilogram of seafood per day based on its body size and quantity.

[0058] In two consecutive frames of a seafood video, select some points with obvious features, such as the edge points and corner points of the seafood, and then use the optical flow algorithm to calculate the displacement and motion direction of the feature points between the two frames to obtain the optical flow field.

[0059] According to the changes of the feature points in the optical flow field, analyze the activity characteristics of the seafood in the video, such as the movement speed, the change of the movement direction, and the continuity of the movement. For example, the movement speed of the seafood can be reflected by calculating the average displacement magnitude of the feature points, and whether the movement direction of the seafood is regular can be judged by analyzing the distribution of the displacement directions.

[0060] Based on the analyzed activity characteristics, establish an activity index model to quantify the activity characteristics into a numerical value, that is, the activity index. For example, the movement speed, the frequency of the change of the movement direction, etc. can be used as indicators, and the activity index can be calculated by weighted summation. The higher the activity index, the more active the seafood is.

[0061] Analyze the external characteristics of the seafood, such as detecting the color, shape, and whether there are damages on the surface of the seafood, evaluate the health status of the seafood according to the characteristics, and generate a health index. For example, the seafood with normal color, complete shape, and no obvious damage has a higher health index, and vice versa.

[0062] Multiply the health index and the activity index by the corresponding weight coefficients respectively, and then add them to obtain an adjustment factor. The weight coefficients can be determined according to the actual situation. For example, if the health status has a greater impact on the feeding amount, a larger weight can be assigned to the health index; if the activity level has a greater impact on the feeding amount, a larger weight can be assigned to the activity index.

[0063] Adjust the initially calculated feeding amount according to the calculated adjustment factor. For example, if the adjustment factor is greater than 1, it means that the health status and activity level of the seafood are better, and the feeding amount can be appropriately increased; if the adjustment factor is less than 1, it means that the health status or activity level of the seafood is poor, and the feeding amount needs to be reduced. Finally, use the adjusted feeding amount as a control signal to send it to the feeder, so that the feeder feeds the seafood into the seafood temporary culture tank according to the adjusted feeding amount. Thus, by comprehensively considering the parameter information, activity characteristics, and health status of the seafood, the feeding amount can be determined more accurately, avoiding overfeeding or underfeeding, and reducing the probability of diseases caused by improper feeding.

[0064] An intelligent feeding control method for a seafood temporary cultivation tank provided by this application calculates the parameter information of seafood by analyzing seafood videos, thereby determining a reasonable basic feeding amount, avoiding the randomness of traditional feeding methods, and improving the scientificity and rationality of feeding. At the same time, by calculating the optical flow of feature points in two consecutive frames of images, the activity characteristics of seafood are accurately analyzed and the activity index is calculated. The health index and the activity index are weighted and summed to obtain an adjustment factor, and the basic feeding amount is fine-tuned through the adjustment factor, so as to comprehensively consider the appearance and activity ability of seafood, achieve comprehensive and accurate adjustment of the basic feeding amount, avoid water pollution caused by overfeeding, and at the same time ensure that seafood obtains sufficient nutrition, which is beneficial to maintaining the stability and health of the temporary cultivation environment.

[0065] Reference Figure 2 As shown, in one embodiment, calculating the parameter information of seafood based on the obtained seafood video of the seafood temporary cultivation tank may specifically include:

[0066] S111. Identify the category and quantity of seafood in the seafood video based on the target detection algorithm;

[0067] S112. Calculate the ratio parameter of video pixels to actual length based on the actual size of the seafood temporary cultivation tank and the pixel size of the seafood temporary cultivation tank in the seafood video;

[0068] S113. Detect the key points of seafood in the seafood video, calculate the pixel distance between the key points of seafood based on the Euclidean distance formula, multiply the pixel distance by the ratio parameter to obtain the body size parameter of the seafood, and the parameter information of the seafood includes the category, quantity and body size parameter of the seafood.

[0069] In this embodiment, a target detection algorithm suitable for seafood videos, such as YOLO, SSD, etc., can be selected. This algorithm can quickly and accurately identify different categories of seafood and their quantities in the video and give their position information. For example, use a labeled seafood data set to train the target detection model so that it can identify common seafood species, such as shrimp, crab, fish, etc. Then apply the trained model to the seafood video, detect frame by frame, identify the category of seafood in each frame, and count the quantity of various seafood.

[0070] Obtain the actual size of the seafood temporary cultivation tank, including length, width, height, etc. In the seafood video, calculate the pixel size of the seafood temporary cultivation tank in the video, such as the pixel values of length and width. Compare the actual size with the pixel size and calculate the ratio parameter of video pixels to actual length. For example, if the actual length of the seafood temporary cultivation tank is 1 meter and the pixel length in the video is 1000 pixels, then the ratio parameter is 1 meter / 1000 pixels = 0.001 meter / pixel.

[0071] Detect the key points of seafood in the seafood video, such as the head, tail, and various joints of the shrimp. Use the Euclidean distance formula to calculate the pixel distances between the key points of the seafood. For example, for two key points of the shrimp's head and tail, calculate the Euclidean distance between their pixel coordinates in the video. Multiply the pixel distance by the scale parameter to obtain the actual body size parameters of the seafood, such as the length and width of the shrimp. These body size parameters can be used to describe the size and shape of the seafood, and combined with the seafood category and quantity, form complete seafood parameter information.

[0072] This embodiment is based on an object detection algorithm, which can accurately identify the category and quantity of seafood, providing basic data for initially calculating the feeding amount of seafood. Through key point detection and scale parameter calculation, the body size parameters of seafood can be accurately calculated.

[0073] In one embodiment, before using the object detection algorithm to identify the category and quantity of seafood in the seafood video, it further includes:

[0074] Preprocess the seafood video, and the preprocessing methods include denoising, contrast enhancement, and color correction.

[0075] This embodiment can use methods such as Gaussian filtering and median filtering to remove the noise in the video and improve the video quality. Gaussian filtering is suitable for Gaussian noise, and median filtering is suitable for salt-and-pepper noise. In addition, methods such as histogram equalization and adaptive histogram equalization are used to enhance the contrast of the video, making the contours and details of the seafood clearer. Finally, operations such as white balance correction and color space conversion are performed on the seafood video to make the colors of the video more realistic and consistent, avoiding color deviations caused by light changes.

[0076] This embodiment can significantly improve the quality of the video through preprocessing steps such as denoising, contrast enhancement, and color correction, making subsequent analysis more accurate and reliable.

[0077] In one embodiment, the optical flow calculation for the selected feature points in two consecutive frames of the seafood video includes:

[0078] Match the feature points of the selected feature points in two consecutive frames of the seafood video;

[0079] Calculate the Pearson correlation coefficient of each feature point in each frame with the corresponding feature point in the other frame respectively;

[0080] Organize the calculated Pearson correlation coefficients into a correlation matrix according to the positions of the feature points. Each element in the correlation matrix represents the correlation strength between the corresponding feature points in two consecutive frames of the image;

[0081] Estimate the optical flow vector based on the correlation matrix, where the optical flow vector is used to describe the change of feature points between two consecutive frames of images.

[0082] In this embodiment, a feature point detection algorithm (such as SIFT, ORB, etc.) can be used to detect feature points in two consecutive frames of images. These feature points are usually points with obvious features in the image, such as corner points, edge points, etc., which can better represent the local features of the image. At the same time, a descriptor is extracted for each detected feature point. The descriptor is a vector that quantitatively represents the features of the area around the feature point and can reflect the local information of the feature point, such as gradient direction, magnitude, etc. Using the descriptors of the feature points, match them between the two frames of images to find the corresponding feature point pairs in the two frames. For example, methods such as brute-force matching or FLANN matching can be used to determine the matching point pairs according to the similarity of the descriptors.

[0083] For each pair of matched feature points, calculate the Pearson correlation coefficient between their pixel values in the two frames of images. The Pearson correlation coefficient is used to measure the linear correlation degree between two variables, and its value range is between -1 and 1. When the value is close to 1, it indicates a strong positive correlation between the two variables; when the value is close to -1, it indicates a strong negative correlation; when the value is close to 0, it indicates a weak correlation. In the two frames of images, obtain the pixel values of each feature point and its neighborhood respectively to form two sequences of pixel values, and then calculate the Pearson correlation coefficient of these two sequences.

[0084] Organize the calculated Pearson correlation coefficients into a matrix according to the positions of the feature points in the image. The rows and columns of the matrix correspond to the positions of the feature points in the two frames of images respectively, and each element in the matrix represents the correlation strength between the corresponding feature points. Using the correlation strength information in the correlation matrix, estimate the optical flow vector of each feature point between two consecutive frames of images. The optical flow vector represents the displacement and motion direction of the feature point between the two frames and can be solved by the least squares method so that the optical flow vector can best explain the correlation strength and displacement relationship between the feature points.

[0085] In addition, the average magnitude of all optical flow vectors in the optical flow field can be calculated, and then a deviation range can be set, such as 2 times the standard deviation of the average magnitude. For optical flow vectors greater than the deviation range, it can be considered that they do not conform to the physical laws and are excluded. For example, if the magnitudes of most vectors in the optical flow field are within a certain range, and there are individual vectors that deviate significantly from this range, they can be excluded.

[0086] This embodiment can accurately find the corresponding feature points in two consecutive frames of images, and through the Pearson correlation coefficient, it can quantify the correlation strength between feature points, providing a more accurate data basis for the estimation of optical flow vectors. In addition, the optical flow vectors estimated based on the correlation matrix can more accurately describe the motion of feature points, which helps to analyze the activity characteristics of seafood.

[0087] In one embodiment, the analysis of the external features of the seafood includes:

[0088] Extract the video frames of the seafood video, extract the external feature pixels of the seafood in each video frame, and obtain the seafood feature map corresponding to each video frame;

[0089] Multiply the seafood feature maps corresponding to all video frames pixel by pixel, and perform weighted summation to obtain the external feature representation of the seafood;

[0090] Calculate the health index of the seafood in the seafood holding tank based on the external feature representation.

[0091] This embodiment can use the computer vision library OpenCV to read the seafood video, extract each frame image in the seafood video frame by frame, and then use edge detection and color histogram to extract the features of the seafood in each video frame to obtain the external feature pixels of the seafood, and organize the extracted feature pixels into a seafood feature map. Each pixel value in the seafood feature map represents a certain external feature of the seafood (such as color, shape, texture, etc.).

[0092] Perform pixel-by-pixel multiplication on the seafood feature maps corresponding to all video frames to obtain a comprehensive feature map to highlight the features that are consistent in multiple frames. Perform weighted summation on the comprehensive feature map, and the weights can be assigned according to the importance of the feature map or the time series. For example, an adaptive weight assignment method can be used to make the weights of important features higher.

[0093] The comprehensive feature map obtained through weighted summation is the external feature representation of the seafood and can be used for health index calculation. For example, the distribution and changes of the feature values in the external feature representation can be analyzed, such as color uniformity, edge sharpness, texture complexity, etc. According to the feature analysis results, a health index model is established to map the feature values to the health index. For example, seafood with high color uniformity and clear edges has a higher health index, and vice versa.

[0094] This embodiment can extract features frame by frame from video frames, can capture the external features of seafood more comprehensively, and through pixel-by-pixel multiplication and weighted summation, can comprehensively consider the feature information of multiple video frames to obtain a more stable external feature representation. In addition, calculating the health index based on the external feature representation can objectively and quantitatively evaluate the health status of seafood.

[0095] In one embodiment, after calculating the health index of the seafood in the temporary storage tank based on the external feature representation, the following steps are further included:

[0096] Determine whether the health index of the seafood in the temporary storage tank is abnormal;

[0097] When the health index is abnormal, send an alarm prompt message to the user terminal.

[0098] In this embodiment, a reasonable threshold range of the health index can be set according to factors such as the type of seafood, growth stage, and temporary storage environment. This threshold range can be determined based on historical data and breeding experience.

[0099] Obtain the health index of the seafood in real time and compare it with the set threshold range. If the health index exceeds the set threshold range, it is determined as abnormal, and an alarm prompt message is sent to the user terminal. The alarm prompt message should include key data such as the time of alarm occurrence, alarm level, monitored object generating the alarm, data triggering the alarm, etc., so that the user can quickly understand the abnormal situation and take corresponding measures.

[0100] This embodiment can monitor the health status of the seafood in real time, reduce the workload of manual monitoring, and improve the monitoring efficiency. When the health index is abnormal, an alarm prompt message can be sent to the user in a timely manner, enabling the user to take measures quickly to avoid further expansion of losses. By timely discovering and handling abnormal situations of the health index, the death or quality decline of seafood caused by health problems can be effectively reduced, and economic losses can be lowered.

[0101] In one embodiment, the adjustment of the feeding amount using the adjustment factor includes:

[0102] Calculate the function value with the negative of the adjustment factor as the independent variable based on the exponential function with the natural logarithm as the base;

[0103] Multiply the function value by the feeding amount to obtain the adjusted feeding amount.

[0104] In this embodiment, an exponential function with the natural logarithm as the base (i.e., the base is e), in the form of f(x) = e -x , where x is the adjustment factor, can be selected. Substitute the adjustment factor into the exponential function to calculate the corresponding function value. This function value reflects the influence degree of the adjustment factor on the feeding amount. The larger the adjustment factor, the smaller the function value.

[0105] Multiply the calculated function value by the initial feeding amount to obtain the adjusted feeding amount. The adjusted feeding amount can be dynamically adjusted according to the change of the adjustment factor to adapt to the actual needs of the seafood.

[0106] In this embodiment, through the exponential function, the feeding amount can be dynamically adjusted according to the change of the adjustment factor, enabling the feeding amount to more precisely adapt to the actual needs of the seafood. At the same time, the non-linear characteristic of the exponential function makes the feeding amount more sensitive to the change of the adjustment factor, and can better reflect the influence of the health status and activity level of the seafood on the feeding amount.

[0107] In one embodiment, after controlling the feeder to feed the seafood in the seafood temporary culture tank according to the adjusted feeding amount, the following steps are further included:

[0108] After a preset time period for feeding the seafood, extract the water body image of the seafood temporary culture tank from the seafood video obtained again;

[0109] Extract the hue value of each pixel in the water body image, and calculate the average hue value of the water body in the seafood temporary culture tank based on the hue values of each pixel in the water body image;

[0110] When the average hue value is greater than the preset chromaticity value, calculate the gray-level co-occurrence matrix of the water body image, extract the texture features of the water body based on the gray-level co-occurrence matrix, and evaluate the turbidity of the water body based on the texture features of the water body;

[0111] When the turbidity is greater than the turbidity threshold, automatically change the water in the seafood temporary culture tank.

[0112] In this embodiment, a reasonable time period can be set, for example, every 2 hours or 4 hours, to extract the water body image from the seafood video. Convert the water body image to the HSV color space and extract the hue value of each pixel. Calculate the average value of the hue values of all pixels to obtain the average hue value of the water body.

[0113] When the average hue value is greater than the preset chromaticity value, calculate the gray-level co-occurrence matrix of the water body image, extract the texture features of the water body based on the gray-level co-occurrence matrix, and evaluate the turbidity of the water body based on the texture features of the water body. For example, convert the water body image to a grayscale image and calculate the gray-level co-occurrence matrix. This gray-level co-occurrence matrix is used to capture the spatial structure and texture information in the image.

[0114] In addition, extract texture features from the gray-level co-occurrence matrix, such as contrast, entropy, energy, etc. These features can reflect the texture complexity and uniformity of the water body. Evaluate the turbidity of the water body according to the extracted texture features. For example, a water body with high contrast and large entropy usually indicates that the water body is more turbid.

[0115] Finally, a reasonable turbidity threshold can be set according to historical data and experience. When the detected turbidity exceeds the turbidity threshold, trigger the automatic water change mechanism, open the water change valve, and perform water body replacement.

[0116] For example, assume that in a seafood holding tank, water body images are extracted from the seafood videos obtained again every 2 hours. The average hue value calculated from the hue values of each pixel in the extracted water body image is 120 (range 0 - 255). The preset chromaticity value is 100. When the average hue value of 120 is greater than 100, the gray-level co-occurrence matrix of the water body image is calculated, and texture features are extracted. Assume that the extracted contrast is 0.8, the entropy is 0.9, and the energy is 0.6. Based on the said features, the turbidity of the water body is evaluated as high. When the turbidity exceeds the preset turbidity threshold of 0.7, the device automatically triggers a water change mechanism to change the water in the seafood holding tank.

[0117] This embodiment can monitor the hue and turbidity of the water body in real time, reduce the workload of manual monitoring, and improve the monitoring efficiency; when the turbidity of the water body is abnormal, it can timely trigger an automatic water change mechanism to keep the water body clean and avoid seafood health problems caused by water body pollution. In addition, by keeping the water body clean, it can effectively reduce the death or quality decline of seafood caused by water body pollution, reduce economic losses, and improve the breeding efficiency.

[0118] In one embodiment, after controlling the feeder to feed the seafood in the seafood holding tank according to the adjusted feeding amount, it further includes:

[0119] Obtaining the feedback on the feeding situation of the seafood;

[0120] Based on the feedback on the feeding situation, adjusting the feeding amount again.

[0121] This embodiment can use an underwater camera to monitor the feeding situation in the seafood holding tank in real time, record the reactions of the seafood to the feed, such as the feeding speed, the feeding amount, etc. In addition, data on the seafood feeding, including the feeding time, the feeding amount, the remaining feed amount, etc., can also be collected through sensors or manual records.

[0122] Meanwhile, computer vision technology is used to analyze the feeding behavior of the seafood, such as the feeding frequency, the duration, the uniformity of feeding, etc. The collected feeding data is analyzed to evaluate whether the feeding condition of the seafood is good, and whether there is a situation of insufficient feeding or overfeeding.

[0123] In addition, a strategy for adjusting the feeding amount can also be formulated according to the analysis results. If the feeding situation of the seafood is good, the feeding amount can be appropriately increased; if the feeding situation is not good, the feeding amount needs to be reduced to dynamically adjust the feeding amount in real time or regularly according to the feeding feedback of the seafood to ensure that the feeding amount matches the actual needs of the seafood.

[0124] For example, assume that in a seafood temporary holding tank, it is found through video monitoring that the feeding speed of the shrimp is slow and there is a large amount of feed remaining. After analysis, it is considered that the feeding situation of the shrimp is not good, probably because the feeding amount is too much. Then, the device automatically reduces the next feeding amount from the original 100 grams to 80 grams to better match the actual feeding needs of the shrimp.

[0125] This embodiment can more accurately adjust the feeding amount by real-time monitoring and analyzing the feeding situation of seafood, avoid overfeeding or underfeeding, reduce feed waste, improve the utilization rate of feed, and reduce breeding costs. At the same time, it ensures that seafood can obtain an appropriate amount of nutrition, avoids health problems caused by overfeeding or underfeeding, and promotes the healthy growth of seafood.

[0126] Reference Figure 3 As shown, an intelligent feeding control device for a seafood temporary holding tank is also provided in an embodiment of the present application. The device includes:

[0127] A calculation module 31, configured to calculate parameter information of seafood based on the obtained seafood video of the seafood temporary holding tank, and calculate the feeding amount of the seafood based on the parameter information;

[0128] An analysis module 32, configured to perform optical flow calculation on selected feature points in two consecutive frames of the seafood video, analyze the activity characteristics of the seafood in the video based on the change of the feature points calculated by the optical flow, and calculate the activity index of the seafood based on the activity characteristics;

[0129] A weighted summation module 33, configured to analyze the external characteristics of the seafood, perform weighted summation on the health index generated from the analysis result of the external characteristics and the activity index to obtain an adjustment factor;

[0130] A control module 34, configured to use the adjustment factor to adjust the feeding amount and control the feeder to feed the seafood in the seafood temporary holding tank according to the adjusted feeding amount.

[0131] The intelligent feeding control device for a seafood temporary holding tank provided by the present application calculates the parameter information of the seafood by analyzing the seafood video, thereby determining a reasonable basic feeding amount, avoiding the randomness of the traditional feeding method, and improving the scientificity and rationality of feeding. At the same time, through the optical flow calculation of the feature points in two consecutive frames, the activity characteristics of the seafood are accurately analyzed and the activity index is calculated. The health index and the activity index are weighted and summed to obtain an adjustment factor, and the basic feeding amount is fine-tuned through the adjustment factor, thereby comprehensively considering the appearance and activity ability of the seafood, realizing a comprehensive and accurate adjustment of the basic feeding amount, avoiding water pollution caused by overfeeding, and at the same time ensuring that the seafood obtains sufficient nutrition, which is beneficial to maintaining the stability and health of the temporary holding environment.

[0132] As described above, it can be understood that each component of the intelligent feeding control device for a seafood temporary holding tank proposed in this application can achieve the functions of any one of the intelligent feeding control methods for a seafood temporary holding tank as described above, and the specific structure will not be elaborated.

[0133] Referring Figure 4 as shown, an electronic device is further provided in an embodiment of the present application, and its internal structure can be as Figure 4 shown. The electronic device includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor designed for the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a storage medium and an internal memory. The storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the storage medium. The database of the electronic device is used to store relevant data of the intelligent feeding control method for a seafood temporary holding tank. The network interface of the electronic device is used to communicate with external devices through a network connection. When the computer program is executed by the processor, it implements an intelligent feeding control method for a seafood temporary holding tank.

[0134] In an embodiment of the present application, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, it implements an intelligent feeding control method for a seafood temporary holding tank.

[0135] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium provided in the present application and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0136] It should be noted that in this text, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, apparatus, article or method comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article or method. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, apparatus, article or method comprising such element.

[0137] The above are only the preferred embodiments of the present application, and do not limit the patent scope of the present application accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. An intelligent feeding control method for a seafood temporary holding tank, characterized in that, including: Calculating parameter information of the seafood based on the obtained seafood video of the seafood temporary storage tank, and calculating the feeding amount of the seafood based on the parameter information; Performing optical flow calculation on selected feature points in two consecutive frames of the seafood video, analyzing the activity characteristics of the seafood in the video based on the changes in the feature points calculated by the optical flow, and calculating the activity index of the seafood based on the activity characteristics; Analyzing the external characteristics of the seafood, and performing weighted summation on the health index generated from the analysis result of the external characteristics and the activity index to obtain an adjustment factor; Adjusting the feeding amount using the adjustment factor, and controlling the feeder to feed the seafood in the seafood temporary storage tank according to the adjusted feeding amount.

2. The method according to claim 1, wherein The calculating the parameter information of the seafood based on the obtained seafood video of the seafood temporary storage tank includes: Identifying the category and quantity of the seafood in the seafood video based on the target detection algorithm; Calculating the ratio parameter of the video pixel to the actual length based on the actual size of the seafood temporary storage tank and the pixel size of the seafood temporary storage tank in the seafood video; Detecting the key points of the seafood in the seafood video, calculating the pixel distance between the key points of the seafood based on the Euclidean distance formula, and multiplying the pixel distance by the ratio parameter to obtain the body size parameter of the seafood. The parameter information of the seafood includes the category, quantity and body size parameter of the seafood.

3. The method according to claim 2, wherein Before the identifying the category and quantity of the seafood in the seafood video based on the target detection algorithm, it further includes: Performing preprocessing on the seafood video, and the preprocessing methods include denoising, contrast enhancement and color correction.

4. The method according to claim 1, characterized in that, The performing optical flow calculation on selected feature points in two consecutive frames of the seafood video includes: Performing feature point matching on selected feature points in two consecutive frames of the seafood video; Calculating the Pearson correlation coefficient between each feature point in each frame of the image and the corresponding feature point in the other frame of the image respectively; Organizing the calculated Pearson correlation coefficients into a correlation matrix according to the positions of the feature points. Each element in the correlation matrix represents the correlation strength between the corresponding feature points in two consecutive frames of the image; Estimating the optical flow vector based on the correlation matrix, and the optical flow vector is used to describe the change of the feature points between two consecutive frames of the image.

5. The method according to claim 1, wherein The analyzing the external characteristics of the seafood includes: Extracting the video frames of the seafood video, performing external feature pixel extraction on the seafood in each video frame to obtain the seafood feature map corresponding to each video frame; Performing pixel-by-pixel multiplication on all the seafood feature maps corresponding to the video frames, and performing weighted summation to obtain the external feature representation of the seafood; Calculating the health index of the seafood in the seafood temporary storage tank based on the external feature representation.

6. The method according to claim 5, characterized in that, After the calculating the health index of the seafood in the seafood temporary storage tank based on the external feature representation, it further includes: Judging whether the health index of the seafood in the seafood temporary storage tank is abnormal; When the health index is abnormal, sending an alarm prompt message to the user terminal.

7. The method according to claim 1, characterized in that The adjusting the feeding amount using the adjustment factor includes: Calculating the function value with the negative of the adjustment factor as the independent variable based on the exponential function with the natural logarithm as the base; Multiplying the function value by the feeding amount to obtain the adjusted feeding amount.

8. The method according to claim 1, characterized in that After controlling the feeder to feed the seafood in the seafood temporary holding tank according to the adjusted feeding amount, the following steps are further included: After a preset time period for feeding the seafood, extract the water body image of the seafood temporary holding tank from the seafood video obtained again; Extract the hue value of each pixel in the water body image, and calculate the average hue value of the water body in the seafood temporary holding tank based on the hue value of each pixel in the water body image; When the average hue value is greater than the preset chromaticity value, calculate the gray-level co-occurrence matrix of the water body image, extract the texture features of the water body based on the gray-level co-occurrence matrix, and evaluate the turbidity of the water body based on the texture features of the water body; When the turbidity is greater than the turbidity threshold, automatically change the water in the seafood temporary holding tank.

9. The method according to claim 1, wherein After controlling the feeder to feed the seafood in the seafood temporary holding tank according to the adjusted feeding amount, the following steps are further included: Obtain the feedback on the feeding situation of the seafood; Readjust the feeding amount based on the feedback on the feeding situation.

10. An electronic device, characterized in that, It includes: A processor; A memory; Wherein, the memory stores a computer program, and when the processor executes the computer program, it implements the intelligent feeding control method for a seafood temporary holding tank according to any one of claims 1 to 9.

Images