An underwater target tracking system based on DBN

Abstract

The application relates to the technical field of ocean exploration, in particular to an underwater target tracking system based on DBN, which comprises a multi-source data processing module, an environment monitoring module, a feature fusion module, a target feature extraction module and a dynamic prediction and identification module. According to the application, the real-time signal-to-noise ratio of sonar and visual data is evaluated, the modal weight is dynamically adjusted, the data reliability under insufficient illumination or complex water quality is ensured, the data fusion strategy is optimized in combination with illumination intensity, turbidity concentration and visible distance, the system has good environmental adaptability, the time sequence features and image edge gradients are utilized to extract key geometric differences, improve feature recognition distinction, screen and optimize significant edge features, highlight target areas and effectively improve recognition accuracy, and based on the offset analysis of DBN node responses, the target tracking conditions can be judged in real time, and the efficiency of underwater exploration and the applicability in complex environments are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ocean exploration, and in particular to an underwater target tracking system based on DBN. BACKGROUND

[0002] The field of ocean exploration involves the use of various scientific methods and tools to study and monitor the marine environment and its phenomena, including physical, chemical, biological and geological exploration, with the aim of understanding the dynamic processes of the ocean, its ecosystems and the impact of the ocean on the global environment, including acoustic positioning, satellite remote sensing, deep-sea submersibles, and various sensors and instruments for measuring ocean temperature, salinity, flow rate, etc. Ocean exploration is not only part of scientific research, but also crucial for maritime safety, ocean resource development, environmental protection and climate change research.

[0003] Among them, the underwater target tracking system based on DBN involves the use of deep belief networks (DBN) to achieve automatic tracking of underwater targets. DBN is a deep learning algorithm that can learn features and patterns from large amounts of data. In the underwater target tracking system, DBN is used to process and analyze data collected from sonar, radar or other sensors to identify and track underwater objects such as submarines, fish schools or other marine organisms. Main uses include oceanographic research, military reconnaissance, fisheries management and search and rescue missions, improving the efficiency and accuracy of underwater detection and surveillance.

[0004] Although there are existing underwater target recognition methods based on sonar or images in the prior art, there are obvious limitations in complex marine environments. Single modal data has significantly decreased recognition accuracy in low light, high turbidity or severe acoustic interference. Existing data fusion methods mostly use static weights or simple weighting, lack adaptive adjustment to environmental changes, resulting in unstable results. Feature extraction relies on edge or spectral features, making it difficult to reflect the dynamic changes of the target in time series and multi-modal environments. In terms of target tracking, there is a lack of deep learning algorithm support, and the discrimination is insufficient, which can easily cause misidentification or loss of targets. Therefore, it is difficult to meet the needs of real-time and accurate detection and tracking of underwater targets. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to propose an underwater target tracking system based on DBN.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: an underwater target tracking system based on DBN, the system comprises:

[0007] The multi-source data processing module obtains the amplitude value and frequency value in the sonar signal and the pixel brightness value and color saturation value in the visual image, performs preliminary signal-to-noise ratio evaluation, compares with the signal-to-noise ratio benchmark in the current marine environment, judges the validity of the modal data, dynamically adjusts the weight of the acoustic and visual data, and generates a modal data weight coefficient;

[0008] The environmental monitoring module monitors the illumination intensity, turbidity particle concentration and visibility distance of the underwater area based on the modal data weight coefficient, analyzes the influence degree of the environmental parameters on the sonar and visual data quality, adjusts the data fusion process, and obtains a data fusion strategy ratio;

[0009] The feature fusion module processes the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the frequency and amplitude change of the short-time data segment, determines the key geometric center distribution difference, and obtains a feature fusion density index;

[0010] The target feature extraction module calls the feature fusion density index, performs spatial and noise distribution analysis on the target region in the sonar image, calculates the interference ratio and reflection feature of the region, screens the region with obvious edge features, optimizes the target feature recognition degree, and obtains an edge recognition intensity value.

[0011] The present application improves that the modal data weight coefficient includes acoustic data weight and visual data weight, the data fusion strategy ratio includes illumination adaptation ratio, turbidity adaptation ratio and visibility adaptation ratio, the feature fusion density index includes frequency change density, amplitude change density and visual edge density, and the edge recognition intensity value includes spatial interference ratio, noise reflection ratio and edge recognition accuracy.

[0012] The present application improves that the multi-source data processing module comprises:

[0013] The sonar data extraction submodule obtains the amplitude value and frequency value in the sonar signal, calculates the signal-to-noise ratio under each frequency according to the environmental influence factors and the amplitude and frequency data change range, and obtains a sonar signal-to-noise ratio result;

[0014] The visual image extraction submodule obtains the pixel brightness value and color saturation value in the visual image, calculates the color stability of each image region according to the different illumination conditions and color distribution, and obtains an image stability result;

[0015] The data weight calculation submodule assigns weights to the sonar and visual image data based on the sonar signal-to-noise ratio result and the image stability result, and uses the formula: ;

[0016] The modal data weight coefficient WS is obtained, wherein SNR i represents the sonar signal-to-noise ratio result under the i-th frequency, and Wsnr The STB represents the weighting factor of the sonar data. j W represents the image stability result within the j-th region. stb The weighting coefficients represent the visual image data, where n represents the frequency of the sonar signal and m represents the number of regions in the visual image.

[0017] The present invention is improved in that the environmental monitoring module includes:

[0018] The illumination monitoring submodule monitors the illumination intensity of the underwater area based on the modal data weighting coefficients. According to the data collected by the illumination sensor, it analyzes the illumination intensity fluctuations in the time series, identifies the average illumination intensity and its standard deviation, and obtains the illumination intensity analysis results.

[0019] The particle concentration analysis submodule analyzes the turbidity particle data recorded by the turbidity sensor based on the light intensity analysis results, using the following formula: ,

[0020] Obtain the adjusted concentration of turbid particles Optimize murky data that is greatly affected by lighting conditions, among which... It is the average light intensity. This represents the concentration of turbid particles at the b-th measurement point. It is the total number of measurement points. It is the maximum concentration of turbid particles during the measurement period;

[0021] The data fusion adjustment submodule, based on the adjusted turbidity particle concentration, integrates water depth and water temperature, analyzes the influence of environmental parameters, optimizes the fusion process of sonar and visual data, matches the current underwater environment, and obtains the data fusion strategy ratio.

[0022] The present invention is improved in that the feature fusion module includes:

[0023] The sonar feature extraction submodule monitors the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the short-term changes in frequency and amplitude within each data segment, calculates the characteristic fluctuation index of the sonar signal, and obtains the sonar characteristic fluctuation value.

[0024] The visual feature analysis submodule collects edge gradient changes in visual data, analyzes the gradient value of each image unit, calculates the average gradient value, and correlates it with the geometric center position of each image unit to determine the geometric center distribution of the image units and obtain the visual geometric distribution differences.

[0025] The fusion density calculation submodule uses the formula based on the differences between the sonar feature fluctuation values ​​and the visual geometric distribution: ,

[0026] The feature fusion density is calculated to obtain a feature fusion density index DR, wherein FR u GR represents the u-th sonar feature fluctuation value u n represents the u-th visual geometric distribution difference value dr N represents the total sample size of data.

[0027] The target feature extraction module comprises:

[0028] The feature density analysis submodule calls the feature fusion density index, extracts the reflection intensity and position information of the pixel points according to the target region data in the sonar image, analyzes the spatial distribution, marks the difference density partition, and generates the density distribution feature;

[0029] The noise distribution calculation submodule calls the density distribution feature, analyzes the noise signal intensity, identifies the key area of noise interference by comparing the ratio of the reflection signal and the noise signal, calculates the correlation between the noise and the density, and obtains the noise interference ratio;

[0030] The edge feature optimization submodule filters the area with obvious edge features according to the noise interference ratio, analyzes the area edge, and adopts the formula: ;

[0031] The target feature recognition degree is optimized to obtain an edge recognition intensity value SQ, wherein GQ p is the gradient intensity value of the p-th pixel point, CQ p is the edge curvature, EQ p is the reflection intensity value, RQ p is the reflection feature of the area, NQ p is the noise signal intensity, n sq is the total number of analyzed pixel points.

[0032] The system further comprises:

[0033] The dynamic prediction recognition module performs difference analysis on the node response value output by the DBN model based on the edge recognition intensity value, calculates the offset amplitude of the node response, performs dynamic prediction according to the offset amplitude, judges whether the current response meets the target tracking condition, and obtains a target tracking analysis result;

[0034] The target tracking analysis result comprises a node response offset, tracking prediction accuracy, and condition compliance.

[0035] The dynamic prediction recognition module comprises:

[0036] The edge recognition analysis submodule analyzes the node response values ​​output by the DBN model based on the edge recognition strength value, detects the edge recognition strength of the nodes, identifies the intensity differences of the node response values, and obtains edge recognition deviation data.

[0037] The node response offset calculation submodule uses the edge recognition deviation data and the following formula: ;

[0038] Calculate the offset magnitude PF of the z-th node. z, Obtain the offset magnitude data, where R z,y E represents the y-th response value of the z-th node. y n represents the preset edge value of the y-th response. pf WF represents the total number of response events. k The weight coefficient of the k-th node, m pf Represents the total number of nodes;

[0039] The target tracking judgment submodule determines whether a node meets the target tracking conditions based on the offset amplitude data, identifies nodes that meet the conditions, and obtains the target tracking analysis results.

[0040] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0041] In this invention, by evaluating the real-time signal-to-noise ratio of sonar and visual data and dynamically adjusting modal weights, the reliability of data is ensured under conditions of insufficient light or complex water quality. By combining light intensity, turbidity concentration, and visibility distance, the data fusion strategy is optimized, enabling the system to have good environmental adaptability. By utilizing time series features and image edge gradients, key geometric differences are extracted to improve feature recognition discrimination. Significant edge features are screened and optimized to highlight the target area, effectively improving recognition accuracy. Based on the offset analysis of DBN node response, target tracking conditions can be judged in real time, improving system response speed and tracking accuracy, and enhancing the efficiency of underwater detection and its applicability in complex environments. Attached Figure Description

[0042] Figure 1 This is a system flowchart of the present invention;

[0043] Figure 2 This is a flowchart of the multi-source data processing module in this invention;

[0044] Figure 3 This is a flowchart of the environmental monitoring module in this invention;

[0045] Figure 4 This is a flowchart of the feature fusion module in this invention;

[0046] Figure 5A flow chart of the target feature extraction module in the present application;

[0047] Figure 6 A flow chart of the dynamic prediction identification module in the present application. DETAILED DESCRIPTION

[0048] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0049] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited. EMBODIMENT

[0050] Please refer to Figure 1 The present application provides a technical solution: a DBN-based underwater target tracking system comprising:

[0051] The multi-source data processing module acquires the amplitude value and frequency value in the sonar signal and the pixel brightness value and color saturation value in the visual image, performs preliminary signal-to-noise ratio evaluation according to the variation range of the data and the environmental influence factors, compares with the signal-to-noise ratio benchmark in the current marine environment, judges the effectiveness of the modal data, dynamically adjusts the weight of the acoustic and visual data, and generates a modal data weight coefficient;

[0052] The environmental monitoring module monitors the illumination intensity, turbidity particle concentration and visibility distance of the underwater area based on the modal data weight coefficient, analyzes the influence degree of the environmental parameters on the quality of the sonar and visual data, adjusts the data fusion process, matches the current underwater environment, and obtains a data fusion strategy ratio;

[0053] The feature fusion module processes the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the frequency and amplitude variation of the short-time data segment, combines the edge gradient variation of the visual data, determines the key geometric center distribution difference, and obtains a feature fusion density index;

[0054] The target feature extraction module calls the feature fusion density index to analyze the spatial and noise distribution of the target region in the sonar image, calculate the interference ratio and reflection characteristics of the region, screen the region with obvious edge features, and optimize the target feature recognition degree to obtain an edge recognition intensity value.

[0055] The dynamic prediction recognition module performs difference analysis on the node response value output by the DBN model based on the edge recognition intensity value, calculates the offset amplitude of the node response, performs dynamic prediction according to the offset amplitude, judges whether the current response meets the target tracking condition, and obtains a target tracking analysis result.

[0056] The modal data weight coefficient includes acoustic data weight and visual data weight, the data fusion strategy ratio includes illumination adaptation ratio, turbidity adaptation ratio, and visibility adaptation ratio, the feature fusion density index includes frequency change density, amplitude change density, and visual edge density, the edge recognition intensity value includes spatial interference ratio, noise reflection ratio, and edge recognition accuracy, and the target tracking analysis result includes node response offset, tracking prediction accuracy, and condition compliance.

[0057] The signal-to-noise ratio benchmark in the current marine environment refers to the "typical environmental noise level" or "empirical signal-to-noise ratio standard" of the sonar or underwater acoustic signal in the actual application area or monitored sea area. Specifically, it is a reference value or threshold for judging and comparing the quality of collected sonar signals, and is set according to historical monitoring data of the target sea area and marine acoustic standards. It is the "reference signal-to-noise ratio level" of the sonar signal at a specific frequency under the condition of no strong interference and no obvious target in the target water area, and provides a benchmark for multi-source data effectiveness evaluation and weight adaptive allocation.

[0058] As a judgment benchmark for "whether the newly collected data is effective and can be used for target identification", for example, if the real-time signal-to-noise ratio is lower than the benchmark, it means that the current data quality is poor and the weight needs to be reduced. Compared with the real-time signal-to-noise ratio, it is used to adjust the weight distribution of sonar data and visual data, and improve the adaptive ability of the system to complex environment. It can cope with the fluctuation of background noise caused by seasonal changes of marine environment, weather changes (wind, rain and snow), human interference, etc.

[0059] For example, after long-term monitoring of a certain sea area, the "background noise signal-to-noise ratio benchmark" of this region is 20 dB. If the sonar signal-to-noise ratio of the system currently monitored at a certain frequency is 15 dB, it is determined to be lower than the benchmark, and the data weight of this frequency is reduced. If it is higher than 20 dB, it means that the current signal quality is excellent, and the weight is increased.

[0060] Please refer to Figure 2 , the multi-source data processing module comprises:

[0061] The sonar data extraction submodule obtains the amplitude value and the frequency value in the sonar signal, calculates the signal-to-noise ratio at each frequency according to the environmental influence factor and the amplitude and frequency data change range, and obtains the sonar signal-to-noise ratio result;

[0062] The sonar signal is monitored in the underwater sensor or sonar detection device, and the sonar signal is converted into an electrical signal form through a numerical conversion module. The amplitude value can be extracted by a wave amplitude measuring instrument of the electrical signal. The instrument records the fluctuation amplitude of the sonar signal by using a sensor, obtains the difference between the peak value and the valley value of each waveform, and records the difference as the amplitude value. The frequency value can be extracted by a frequency analysis module. The period waveform of the sonar signal is subjected to fast Fourier transform (FFT), and the main frequency of the sonar signal is extracted as the frequency value according to the frequency value corresponding to the position with the largest amplitude in the FFT result, so as to realize the extraction of the frequency value. In the identification of the environmental influence factor, the water temperature, salinity, and sea current rate need to be monitored. The water temperature data is collected by a temperature sensor, the salinity data is measured by a salinometer, and the sea current rate data is monitored by a flowmeter. The parameters are converted into numerical values and introduced into a calculation model. According to the amplitude value, the frequency value, and the environmental influence factor at different frequencies, the signal-to-noise ratio at each frequency is calculated. For the calculation of the signal-to-noise ratio, the amplitude value at each frequency is divided by the corresponding environmental noise amplitude N i, The logarithm is taken again, and the calculation formula is: In the formula, SNR i is the signal-to-noise ratio at the i-th frequency, A i is the sonar amplitude value at the i-th frequency, N i is the environmental noise amplitude at the i-th frequency. The signal-to-noise ratio at each frequency point is taken as a single data point, the signal-to-noise ratio results of all frequency points are calculated and recorded, and the sonar signal-to-noise ratio result is obtained.

[0063] The visual image extraction submodule obtains the pixel brightness value and the color saturation value in the visual image, calculates the color stability of each image region according to the different light conditions and color distribution, and obtains the image stability result.

[0064] For the acquisition process of visual image, underwater camera is needed to collect visual data in water in real time, and image data is converted into RGB format. In the process, YUV color space conversion method can be used to extract Y channel data in the image as the brightness value. By traversing each image pixel point, the Y value of each pixel point is recorded to realize the extraction of pixel brightness value. For the extraction of color saturation value, RGB image can be converted into HSV color space. In HSV space, the S channel of the image represents the color saturation value. Image data traversal algorithm is used to extract the S channel value of each pixel point to realize the extraction of pixel color saturation value. After the above data extraction is completed, in order to calculate the color stability of each image region, the image needs to be divided into several regions. According to the horizontal and vertical coordinate axes of the image, the image region is divided into fixed width and height. The standard deviation of pixel brightness value and color saturation value in each region is calculated as the color stability of the region. The calculation formula is: In the formula, STB j is the color stability of the jth region, is the brightness value or color saturation value of the kth pixel point, is the average value of the brightness value or color saturation value of all pixel points in the region, and p is the total number of pixel points in the region. The color stability of each image region is calculated to obtain the image stability result.

[0065] The data weight calculation sub-module assigns weights to the data of sonar and visual image based on the sonar signal-to-noise ratio result and the image stability result. The formula is:

[0066] The modal data weight coefficient WS is obtained, wherein SNR i represents the sonar signal-to-noise ratio result at the ith frequency, W snr represents the weight coefficient of sonar data, STB j represents the image stability result in the jth region, and W stb represents the weight coefficient of visual image data, n represents the frequency number of sonar signal, and m represents the region number of visual image.

[0067] The relative influence of each parameter is analyzed, and a reference coefficient is introduced. The weight parameter is set with reference to the reference value of marine acoustic signal-to-noise ratio and the reference value of marine optical image stability. The average value of sonar data signal-to-noise ratio at different frequencies is taken as the weight parameter W snr . The calculation formula is: The weight parameter W stb of visual image data is taken as the average value of color stability of each region of the image. The formula is: ​In the formula, n is the frequency number of the sonar signal, m is the region number of the visual image, and then the modal data weight coefficient WS is calculated according to the above weight parameters. If the frequency number of the sonar signal is 4, the region number of the visual image is 3, the sonar signal-to-noise ratio results are 20, 18, 15 and 22 in turn, and the image stability results are 0.12, 0.15 and 0.10 in turn, the calculation process is as follows: ;

[0068] ;

[0069] The numerical result of the modal data weight coefficient is 17.3, which indicates that the data fusion weight of the current sonar signal and visual image is in the upper middle interval, which can be further adjusted according to the specific application scene.

[0070] Please refer to Figure 3 , the environmental monitoring module comprises:

[0071] The light monitoring sub-module monitors the light intensity of the underwater area based on the modal data weight coefficient, analyzes the light intensity fluctuation in the time sequence according to the data collected by the light sensor, identifies the average light intensity and its standard deviation, and obtains the light intensity analysis result;

[0072] First, a plurality of light sensors are arranged in the underwater area, the sensors are distributed at different depths and horizontal positions to ensure covering the overall range of the monitoring area, for example, 5 light sensors are arranged at depths of 3 meters, 5 meters and 7 meters respectively to form a grid arrangement, the sensors record light intensity values at fixed time intervals, assuming that the sampling interval is 10 seconds, each sensor can record 6 data points in 1 minute, all data points are arranged in time sequence and labeled with time tags, according to the recorded data, the light intensity curve of each sensor is extracted, a sliding window average method is used for the curve, for example, a window width of 30 seconds is used to calculate the sliding average of the light data of each sensor, thereby smoothing the light intensity curve, and the extreme points (i.e. the highest and lowest points of light intensity) in the data are screened out, on the basis of excluding extreme abnormal points, the average light intensity of each sensor is calculated, and the average light intensity of the overall area is calculated according to the average light intensity of all sensors, and the standard deviation of the light intensity in the area is further calculated. The standard deviation can be calculated by the formula: , wherein is the standard deviation of the light intensity, N is the total number of measurement points, x i is the light intensity value of the i-th measurement point, is the mean value of the light intensity of the measurement points. Assuming that in a certain measurement, the mean values of the light intensities of the 5 sensors are 120, 110, 115, 118 and 125 lx respectively, the standard deviation of the light intensity is calculated as 5.74 lx, and the light intensity analysis result is obtained.

[0073] The particle concentration analysis submodule analyzes the turbidity particle data recorded by the turbidity sensor according to the light intensity analysis result, and uses the formula: ;

[0074] The adjusted turbidity particle concentration CH t, The turbidity data with large light influence is optimized, wherein, is the average light intensity, is the turbidity particle concentration of the bth measurement point, is the total number of measurement points, is the maximum value of the turbidity particle concentration during the measurement;

[0075] The turbidity sensor is arranged near the light sensor, records the turbidity particle concentration value at the same time interval, and labels the time tag. According to the turbidity particle data of each sensor, the time period with significant decrease in light intensity is screened out, and the turbidity particle concentration data in this time period is taken as the key analysis target. For the screened data segment, the average turbidity particle concentration of each measurement point is extracted. In a certain period, the average value of the light intensity of the four sensors is 112, and the corresponding turbidity particle concentration data is 6.5, 7.0, 8.0 and 7.5 mg / L, and the maximum turbidity particle concentration value is 8.0 mg / L. Then, the formula is calculated as:

[0076] ;

[0077] ;

[0078] The adjusted turbidity particle concentration CH mg / L, and the corrected turbidity particle concentration value is obtained. This result indicates that after adjusting the turbidity particle concentration, the actual turbidity degree can be better reflected in the period with large light intensity fluctuation.

[0079] The data fusion adjustment submodule integrates the water depth and water temperature based on the adjusted turbidity particle concentration, analyzes the influence of environmental parameters, optimizes the fusion process of sonar and visual data, matches the current underwater environment, and obtains the data fusion strategy ratio.

[0080] The water depth and water temperature data are integrated. The water depth data is obtained by an ultrasonic water depth sensor. The sensor is arranged at three measurement points along the water depth direction, which are located at 2 meters, 4 meters and 6 meters below the water surface, respectively, to record the water depth fluctuation at different depths. The water temperature data is obtained by a temperature sensor. The temperature sensor is arranged at the same three measurement points to record the water temperature fluctuation at each measurement point. Based on the obtained water depth and water temperature data, the water temperature and water depth variation coefficients at each sensor position are calculated. The variation coefficient formula is: , wherein CV is the variation coefficient, a standard deviation of water temperature or water depth, a mean value of water temperature or water depth, if the mean values of water temperature data at three measuring points are 12.5℃, 14.0℃ and 15.5℃ respectively, and the standard deviations are 0.8℃, 0.5℃ and 0.6℃ respectively, then the coefficients of variation are: , , , the data of the measuring point with the smallest coefficient of variation is selected as the reference, the fusion process of the sonar and the visual data is optimized, the water temperature and water depth data of the reference measuring point are combined with the corrected turbidity particle concentration value to match the current underwater environment, the weight coefficients of the light, turbidity particle concentration and visibility distance data are adjusted, and a data fusion strategy ratio is obtained.

[0081] Please refer to Figure 4 , the feature fusion module comprises:

[0082] The sonar feature extraction submodule monitors the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the short-time changes of the frequency and amplitude in each data segment, calculates the characteristic fluctuation index of the sonar signal, and obtains the sonar characteristic fluctuation value;

[0083] The underwater sonar array device is used for signal monitoring of the target water area. The sonar array is laid in a water area with stable acoustic characteristics, and the spacing of the sonar sensors is moderate to cover the acoustic signals in the monitoring range. For example, 8 sonar sensors are laid in a water area with a diameter of 100 meters, and the working frequency of each sensor is set to a medium-low frequency acoustic wave range of 10-200 Hz. The sonar signal data received by each sensor is continuously monitored at a time interval of 1 second. After the data monitoring is completed, the time series data of the sonar signal is divided into multiple short time data segments according to the time axis, and the time division interval can be set to 1-5 seconds. The appropriate interval is selected according to different signal strengths to ensure the effectiveness of the data. The divided data segments contain all the sonar signal data points in the time range. The frequency and amplitude data in each data segment need to be processed respectively. First, the frequency data is processed by point-by-point difference to calculate the change value. For example, the frequency data in a data segment is 12 Hz, 13 Hz, 14 Hz, 15 Hz and 16 Hz. The frequency change of each adjacent data point is calculated, and the change is 1 Hz, 1 Hz, 1 Hz and 1 Hz. Further, the average value of the frequency change in the data segment is taken as the frequency fluctuation index of the data segment. The calculation result of the above example data is 1 Hz. The processing method of the amplitude data is similar. The amplitude change is calculated by point-by-point difference. For example, the amplitude data of a data segment is 3 dB, 3.5 dB, 4 dB, 4.5 dB and 5 dB. The amplitude change of adjacent data points is 0.5 dB, 0.5 dB, 0.5 dB and 0.5 dB. Further, the average value of the amplitude change in the data segment is calculated to obtain the amplitude fluctuation index of the data segment. Then, the frequency fluctuation index and the amplitude fluctuation index of each short time data segment are summarized to form the characteristic fluctuation index of the sonar signal as a whole. Further, the characteristic fluctuation index is summarized according to the interval to obtain the characteristic fluctuation trend of the sonar signal as a whole, and the sonar characteristic fluctuation value is obtained.

[0084] The visual feature analysis submodule collects the edge gradient change in the visual data, analyzes the gradient value of each image unit, calculates the average of the gradient value, and associates it with the geometric center position of each image unit to determine the geometric center distribution of the image unit and obtain the visual geometric distribution difference.

[0085] The image data collected needs to be preprocessed, including image graying and noise reduction processing. The graying can be achieved by converting the original color image into a gray image to eliminate color interference. The noise reduction processing can be achieved by mean filtering or median filtering to remove random noise points in the image to improve the image quality. After the image preprocessing is completed, the edge gradient change in the image data is analyzed. Each image unit (such as a 3*3 pixel area) is taken as an analysis unit, and the gradient value in each unit is calculated. The gradient value can be calculated by using the Sobel operator method. The gradient values in the horizontal direction and the vertical direction of the image unit are calculated respectively, and the square root of the sum of squares of the two is taken as the gradient value of each image unit. For example, the gradient value of a certain image unit in the horizontal direction is 8, and the gradient value in the vertical direction is 6. The total gradient value of the image unit is calculated as follows: Next, the gradient mean value of each image unit is calculated, and the gradient mean value of each image unit is associated with its geometric center position to obtain the geometric center distribution feature. The geometric center position can be determined by the image coordinate system. Assuming that the image size is 640*480 pixels, the coordinates of the center position of the image are (320, 240). The center coordinates of each image unit are calculated by the row and column numbers of its position. For example, a certain image unit is located at the 150th row and the 200th column, and the center coordinates are (200, 150). After obtaining the geometric center position and the gradient mean value of each image unit, the offset degree of the geometric center position is further calculated. Specifically, the difference between the center coordinates of each image unit and the center coordinates of the whole image is calculated to reflect the geometric center deviation of the unit. The geometric center deviation of each image unit is averaged by region to obtain the visual geometric distribution difference.

[0086] The fusion density calculation sub-module calculates the feature fusion density based on the sonar feature fluctuation value and the visual geometric distribution difference by using the formula: to obtain the feature fusion density index DR, wherein FR u represents the u-th sonar feature fluctuation value, GR u represents the u-th visual geometric distribution difference value, n dr represents the total sample size of the data.

[0087] If the sonar feature fluctuation values are 1.2, 1.5, 1.3 and 1.4, and the visual geometric distribution difference values are 1.1, 1.6, 1.2 and 1.3, the calculation is as follows:

[0088] ​​The result shows that the feature fusion density index is small, indicating that the feature fluctuation of the sonar signal and the visual geometric distribution are small, and it can be further judged that the feature correlation of the two is high, and there is strong correlation between the data, and the feature fusion density index is obtained.

[0089] Please refer to Figure 5 , the target feature extraction module comprises:

[0090] The feature density analysis submodule calls the feature fusion density index, extracts the reflection intensity and position information of the pixel points according to the target region data in the sonar image, analyzes the spatial distribution, marks the difference density partition, and generates the density distribution feature;

[0091] According to the target region data in the sonar image, first, the target region data in the sonar image is obtained, and the data acquisition of the sonar image can be based on the change of the pulse echo signal intensity. The position coordinates of each point in the sonar data are determined by the angle and position coordinates of the scanning device. Each coordinate point should contain its reflection signal intensity value as a key parameter. Assuming that the coordinate system of the sonar image is defined by X-Y axis, the coordinates of each pixel point can be expressed as , wherein represents the i-th pixel point, and the reflection signal intensity is , in the case of known sonar image resolution of 300*300 pixels, after obtaining the data of each point, the reflection intensity value of each point and the spatial distribution density between adjacent points are calculated. The density index calculation can be completed by using the local point reflection intensity value mean value method. For each pixel point, a 5*5 neighborhood window (i.e. 25 pixel points including the current point) is set, and the reflection signal intensity mean value of each point in the neighborhood window is calculated , and the ratio of the reflection signal intensity value of each point to the reflection signal intensity mean value in the neighborhood is taken as the density index of the point, that is, the density index can be calculated as: In this calculation, if the density index of a point is higher than the set threshold, the point is identified as a high-density point, otherwise it is a low-density point. After calculating each point in the sonar image in this way, the adjacent high-density points are clustered into a density interval using clustering method. Assuming that the clustering standard is that the Euclidean distance between points does not exceed 2 pixel points, the adjacent high-density point set will be classified as a group of density partitions. Further, the average density index value of each density partition region in the sonar image is marked as the feature density index of the partition. Assuming that the density partition region is the k-th partition, the feature density index of the partition can be expressed as: , wherein, n kThe density distribution index of each partition is obtained, and the density distribution index is taken as a density distribution feature.

[0092] The noise distribution calculation submodule calls the density distribution feature, analyzes the noise signal strength, compares the ratio of the reflection signal and the noise signal, identifies the key area of noise interference, calculates the correlation between the noise and the density, and obtains a noise interference ratio.

[0093] Based on the density distribution feature, the noise signal strength in each density partition area is obtained. The determination of the noise signal strength can be completed by a frequency spectrum energy analysis method. For each density partition, the total spectrum signal energy of each pixel point in the partition is taken as the noise signal strength. The specific calculation method is as follows: , wherein N k is the total noise signal strength in the kth partition, is the spectrum energy value of each pixel point. Further, in the calculation process, the ratio of the total reflection signal energy of each density partition to the total noise signal energy is taken as the noise interference ratio, and the calculation method is as follows: , wherein Q k is the noise interference ratio of the density partition. Further, the correlation between the noise signal and the characteristic density of the density partition is calculated. The total reflection signal energy is taken as the correlation reference, the distribution proportion of the noise signal strength in each density partition area is calculated, and it is judged whether there is a significant noise interference area in the area. If the noise interference ratio of a certain density partition is less than 0.5 (in experimental tests, the ratio of the sonar image signal reflection strength to the noise signal strength is usually about 0.5 as the noise interference reference value), it is considered that there is obvious noise interference in the area. In this way, the key area of noise interference is identified, and the noise interference ratio is obtained.

[0094] The edge feature optimization submodule selects the area with obvious edge features according to the noise interference ratio, analyzes the area edge, and adopts the formula: ; the target feature recognition degree is optimized to obtain an edge recognition intensity value SQ. The area with a value higher than a threshold value is selected as the target feature, wherein GQ p is the gradient intensity value of the pth pixel point, CQ p is the edge curvature, EQ p is the reflection intensity value, RQ P is the reflection feature of the area, NQ P is the noise signal strength, n sq is the total number of analyzed pixel points.

[0095] If the area contains 2 pixel points, the parameters take the following values:

[0096] 1st pixel point: , , , , ;

[0097] The second pixel point: , , , , ;

[0098] The above parameters are substituted into the formula for item-by-item calculation: ; ;

[0099] The results show that the edge recognition intensity value is about 0.829, and further screening the region with a value greater than the set threshold (assuming the threshold is 0.5) as the target feature, the edge recognition intensity value is obtained.

[0100] Referring to Figure 6 , the dynamic prediction recognition module comprises:

[0101] The edge recognition analysis submodule analyzes the node response value output by the DBN model based on the edge recognition intensity value, detects the edge recognition intensity of the node, identifies the intensity difference of the node response value, and obtains the edge recognition deviation data;

[0102] The node response values output by the DBN model at different time points are obtained, and the edge recognition intensity value of each node is extracted. When obtaining the edge recognition intensity value, the historical response data of each node is taken as a reference, the response amplitude change of each node at different time points is calculated, the maximum and minimum values of the response amplitude change are used to determine the response intensity interval value of the node, and the intensity threshold range of the node is further divided according to the upper limit and lower limit of the interval value. For example, if the historical response data of node 1 fluctuates within the interval [0.2, 0.8], the upper limit 0.8 of the interval is taken as the upper limit of the intensity threshold, and the lower limit 0.2 is taken as the lower limit of the intensity threshold. Then, the current response value of each node is compared with the corresponding intensity threshold, and the response value exceeding the upper limit or below the lower limit of the threshold is marked as a deviation point. The number of deviation points of each node at different time points and the deviation amplitude of each deviation point are recorded. The amplitude value of each deviation point is taken as the edge recognition deviation value, and the values are summarized to form the edge recognition deviation data of the node.

[0103] The node response offset calculation submodule calculates the offset amplitude PF of the zth node based on the edge recognition deviation data using the formula:

[0104] The offset amplitude data of the zth node is obtained, wherein R z z,y ​the yth response value of the zth node, used to identify the activity level of the node at the target time point, the preset edge value of the yth response, which is the reference value, used to evaluate whether the node response reaches a certain specific edge identification standard, the total number of response events, the weight coefficient of the kth node, the total number of nodes;

[0105] If the response value of the 1st node is [0.25, 0.55, 0.45], the preset edge value is [0.30, 0.50, 0.40], the total number of response events is 3, the total number of nodes is 4, and the weight value of each node is [0.2, 0.3, 0.25, 0.25], then the offset amplitude is calculated as follows: ;

[0106] ;

[0107] The calculation result of the offset amplitude value of the 1st node is 0.0075, which indicates that the response value of the node at each time point deviates from the preset edge value by a small amplitude, and the node is stable in fluctuation. The response data is closer to the set reference range, and the result is further included in the node response offset amplitude data.

[0108] The target tracking judgment sub-module judges whether the node meets the target tracking condition based on the offset amplitude data, identifies the nodes that meet the condition, and obtains the target tracking analysis result;

[0109] The offset amplitude value of each node is extracted, the target tracking threshold range is set, and the offset amplitude value of each node is further compared with the target tracking threshold to determine whether each node meets the target tracking condition. When setting the target tracking threshold range, the median of the node offset amplitude value is taken as the reference, and the upper quartile and the lower quartile are combined to calculate the target tracking threshold interval. For example, in the sample data, the node offset amplitude values are [0.0075, 0.012, 0.015, 0.020], the median is 0.0135, the upper quartile is 0.018, and the lower quartile is 0.010. Nodes below the lower quartile or above the upper quartile are marked as offset abnormal nodes. Further, according to the distribution of abnormal nodes, the nodes that meet or do not meet the target tracking condition are marked, and the target tracking analysis result is obtained.

[0110] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. An underwater target tracking system based on DBN, characterized in that, The system includes: The multi-source data processing module acquires the amplitude and frequency values ​​from the sonar signal and the pixel brightness and color saturation values ​​from the visual image, performs a preliminary signal-to-noise ratio assessment, compares it with the current signal-to-noise ratio benchmark in the marine environment, judges the validity of the modal data, dynamically adjusts the weights of acoustic and visual data, and generates modal data weight coefficients. Based on the modal data weighting coefficients, the environmental monitoring module monitors the light intensity, turbidity particle concentration, and visibility distance in the underwater area, analyzes the impact of environmental parameters on the quality of sonar and visual data, adjusts the data fusion process, and obtains the data fusion strategy ratio. The feature fusion module processes the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the frequency and amplitude changes of short data segments, determines the key differences in the distribution of geometric centers, and obtains the feature fusion density index. The target feature extraction module calls the feature fusion density index to perform spatial and noise distribution analysis on the target region in the sonar image, calculates the interference ratio and reflection characteristics of the region, filters out regions with obvious edge features, and optimizes the target feature recognition to obtain the edge recognition intensity value. The dynamic prediction and recognition module performs difference analysis on the node response values ​​output by the DBN model based on the edge recognition strength value, calculates the offset amplitude of the node response, performs dynamic prediction based on the offset amplitude, determines whether the current response meets the target tracking conditions, and obtains the target tracking analysis result. The target tracking analysis results include node response offset, tracking prediction accuracy, and condition compliance.

2. The underwater target tracking system based on DBN according to claim 1, characterized in that, The modal data weighting coefficients include acoustic data weights and visual data weights; the data fusion strategy ratios include illumination adaptation ratios, turbidity adaptation ratios, and visibility adaptation ratios; the feature fusion density indices include frequency variation density, amplitude variation density, and visual edge density; and the edge recognition intensity values ​​include spatial interference ratios, noise reflection ratios, and edge recognition accuracy.

3. The underwater target tracking system based on DBN according to claim 1, characterized in that, The multi-source data processing module includes: The sonar data extraction submodule obtains the amplitude and frequency values ​​in the sonar signal, calculates the signal-to-noise ratio at each frequency based on environmental influencing factors and the range of amplitude and frequency data changes, and obtains the sonar signal-to-noise ratio result. The visual image extraction submodule obtains the pixel brightness and color saturation values ​​in the visual image, calculates the color stability of each image region based on different lighting conditions and color distribution, and obtains the image stability result. The data weight calculation submodule assigns weights to the sonar and visual image data based on the sonar signal-to-noise ratio results and image stability results, using the following formula: ; Obtain the modal data weighting coefficients WS, where SNR1 represents the sonar signal-to-noise ratio at the i-th frequency, and W... snr The STB represents the weighting factor of the sonar data. j W represents the image stability result within the j-th region. stb The weighting coefficients represent the visual image data, where n represents the frequency of the sonar signal and m represents the number of regions in the visual image.

4. The underwater target tracking system based on DBN according to claim 1, characterized in that, The environmental monitoring module includes: The illumination monitoring submodule monitors the illumination intensity of the underwater area based on the modal data weighting coefficients. According to the data collected by the illumination sensor, it analyzes the illumination intensity fluctuations in the time series, identifies the average illumination intensity and its standard deviation, and obtains the illumination intensity analysis results. The particle concentration analysis submodule analyzes the turbidity particle data recorded by the turbidity sensor based on the light intensity analysis results, using the following formula: ; Obtain the adjusted concentration of turbid particles CH t, Optimize turbid data that is greatly affected by illumination, including LH avg It is the average light intensity, pH b n is the concentration of turbid particles at the b-th measurement point. ch It is the total number of measurement points. It is the maximum concentration of turbid particles during the measurement period; The data fusion adjustment submodule integrates water depth and water temperature based on the adjusted turbidity particle concentration, analyzes the influence of environmental parameters, optimizes the fusion process of sonar and visual data, matches the current underwater environment, and obtains the data fusion strategy ratio.

5. The underwater target tracking system based on DBN according to claim 1, characterized in that, The feature fusion module includes: The sonar feature extraction submodule monitors the time series data of the sonar signal according to the data fusion strategy ratio, analyzes the short-term changes in frequency and amplitude within each data segment, calculates the characteristic fluctuation index of the sonar signal, and obtains the sonar characteristic fluctuation value. The visual feature analysis submodule collects edge gradient changes in visual data, analyzes the gradient value of each image unit, calculates the average gradient value, and correlates it with the geometric center position of each image unit to determine the geometric center distribution of the image units and obtain the visual geometric distribution differences. The fusion density calculation submodule uses the formula based on the differences between the sonar feature fluctuation values ​​and the visual geometric distribution: ; Calculate the feature fusion density to obtain the feature fusion density index DR, where FR u CR represents the u-th sonar characteristic fluctuation value. u Represents the difference value of the u-th visual geometric distribution, n dr This indicates the total sample size of the data.

6. The underwater target tracking system based on DBN according to claim 1, characterized in that, The target feature extraction module includes: The feature density analysis submodule calls the feature fusion density index, extracts the reflection intensity and location information of pixels based on the target area data in the sonar image, analyzes the spatial distribution, marks the differential density partitions, and generates density distribution features; The noise distribution calculation submodule calls the density distribution characteristics to analyze the noise signal intensity. By comparing the ratio of the reflected signal to the noise signal, it identifies key areas of noise interference and calculates the correlation between noise and density to obtain the noise interference ratio. The edge feature optimization submodule filters regions with obvious edge features based on the noise interference ratio, analyzes the edges of these regions using the following formula: ; Optimize the target feature recognition level to obtain the edge recognition strength value SQ, where GQ p CQ is the gradient intensity value of the p-th pixel. p It's edge curvature, EQ p It is the reflection intensity value, RQ p It is the reflectance characteristic of the region, NQ p It is the noise signal strength, n sq It represents the total number of pixels analyzed.

7. The underwater target tracking system based on DBN according to claim 1, characterized in that, The dynamic prediction and recognition module includes: The edge recognition analysis submodule analyzes the node response values ​​output by the DBN model based on the edge recognition strength value, detects the edge recognition strength of the nodes, identifies the intensity differences of the node response values, and obtains edge recognition deviation data. The node response offset calculation submodule uses the edge recognition deviation data and the following formula: ; Calculate the offset magnitude PF of the z-th node. z, Obtain the offset magnitude data, where R z,y E represents the y-th response value of the z-th node. y n represents the preset edge value of the y-th response. pf WF represents the total number of response events. k The weight coefficient of the k-th node, m pf Represents the total number of nodes; The target tracking judgment submodule determines whether a node meets the target tracking conditions based on the offset amplitude data, identifies nodes that meet the conditions, and obtains the target tracking analysis results.

Images