Big data-based side-scan sonar detection method and device

Abstract

The invention discloses a side-scan sonar detection method and device based on big data, and the method comprises the steps: collecting detection performance parameters of a side-scan sonar technology, and carrying out the classification of the detection performance parameters, and obtaining a classification performance parameter corresponding to each evaluation detection performance index; obtaining a target analysis model corresponding to the target evaluation and detection performance index, and analyzing the classification performance parameters according to the target analysis model to obtain a performance index corresponding to each evaluation and detection performance index; performing qualification degree evaluation on each performance index to obtain a performance qualification degree corresponding to each evaluation detection performance index; obtaining a weight value corresponding to each performance qualification degree, and obtaining a detection performance comprehensive index according to each performance qualification degree and the corresponding weight value; and judging whether the detection performance comprehensive index exceeds a preset allowable range, and if yes, sending out an alarm signal. The mass sonar data is deeply analyzed and mined through the big data technology, accurate feature extraction and classification of the sonar data are achieved, and the data processing capacity and the detection efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of side-scan sonar, and in particular to a side-scan sonar detection method and device based on big data. Background Art

[0002] Side-scan sonar technology is an underwater detection technology that uses the principle of echo sounding to image underwater targets and landforms. By measuring the propagation time of sound waves and the strength of echo signals, the distance, size and shape of the seabed topography or underwater targets can be calculated. It is widely used in many fields such as marine science, seabed resource exploration, underwater archaeology, port and waterway maintenance, etc.

[0003] Traditional side-scan sonar technology usually has a large working distance, resulting in low spatial resolution of images, which in turn leads to reduced image quality. In addition, side-scan sonar images in complex environments are seriously polluted by noise, which interferes with the accurate identification and positioning of target objects. At the same time, the amount of data collected by the side-scan sonar system is extremely large, and the data often contains a large number of noise points, requiring automatic filtering and manual intervention of point cloud data. It lacks the ability to efficiently process and analyze data, making it difficult to meet the increasingly complex needs of ocean exploration. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a side-scan sonar detection method and device based on big data to improve the accuracy of side-scan sonar detection.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A side-scan sonar detection method based on big data, comprising:

[0007] Collect the detection performance parameters of the side scan sonar technology and classify them to obtain the classified performance parameters corresponding to each evaluation detection performance indicator;

[0008] Acquire a target analysis model corresponding to the target evaluation and detection performance index, analyze the classification performance parameters according to the target analysis model, and obtain a performance index corresponding to each of the evaluation and detection performance indicators;

[0009] Performing qualification evaluation on each of the performance indexes to obtain the performance qualification corresponding to each of the evaluation and detection performance indicators;

[0010] Obtaining a weight value corresponding to each of the performance qualifications, and obtaining a comprehensive indicator of the detection performance according to each of the performance qualifications and its corresponding weight value;

[0011] Determine whether the comprehensive index of the detection performance exceeds the preset allowable range, and if so, issue an alarm signal.

[0012] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0013] A side-scan sonar detection device based on big data comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step in the above-mentioned side-scan sonar detection method based on big data is implemented.

[0014] The beneficial effects of the present invention are as follows: by collecting the detection performance parameters of the side-scan sonar technology and classifying them to obtain the classified performance parameters corresponding to each evaluation detection performance indicator, that is, by using big data technology, it is possible to deeply analyze and mine massive sonar data, realize accurate feature extraction and classification of sonar data, and improve data processing capabilities and detection efficiency; and, by successively analyzing the classified performance parameters corresponding to each evaluation detection performance indicator and evaluating the degree of qualification to obtain the performance qualification, and then combining all the performance qualifications to obtain the comprehensive detection performance index to make a comprehensive judgment on the side-scan sonar performance, when the comprehensive detection performance index shows an abnormality, the management personnel are immediately prompted to handle it, thereby improving the accuracy of the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of the steps of a side scan sonar detection method based on big data in an embodiment of the present invention;

[0016] Figure 2 Schematic diagram of the structure of a side-scan sonar detection device based on big data in an embodiment of the present invention;

[0017] Figure 3 Schematic diagram of a module of a side-scan sonar detection device based on big data in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Please refer to Figure 1 , a side scan sonar detection method based on big data, comprising:

[0020] Collect the detection performance parameters of the side scan sonar technology and classify them to obtain the performance parameters corresponding to each evaluation detection performance indicator;

[0021] Acquire a target analysis model corresponding to the target evaluation and detection performance index, analyze the classification performance parameters according to the target analysis model, and obtain a performance index corresponding to each of the evaluation and detection performance indicators;

[0022] Performing a qualification evaluation on each of the classification performance indexes to obtain a performance qualification corresponding to each of the evaluation and detection performance indicators;

[0023] Obtaining a weight value corresponding to each of the performance qualifications, and obtaining a comprehensive indicator of the detection performance according to each of the performance qualifications and its corresponding weight value;

[0024] Determine whether the comprehensive index of the detection performance exceeds the preset allowable range, and if so, issue an alarm signal.

[0025] From the above description, it can be seen that the beneficial effects of the present invention are: by collecting the detection performance parameters of the side-scan sonar technology and classifying them to obtain the classified performance parameters corresponding to each evaluation detection performance indicator, that is, through big data technology, it is possible to deeply analyze and mine massive sonar data, realize accurate feature extraction and classification of sonar data, and improve data processing capabilities and detection efficiency; and, by analyzing the classified performance parameters corresponding to each evaluation detection performance indicator in turn, evaluating the qualification to obtain the performance qualification, and then combining all the performance qualifications to obtain the comprehensive detection performance index to make a comprehensive judgment on the side-scan sonar performance. When the comprehensive detection performance index shows an abnormality, the management personnel are immediately prompted to handle it, thereby improving the accuracy of the detection.

[0026] Furthermore, the evaluation detection performance indicators include imaging clarity, target detection efficiency, target detection dynamic stability and intelligent classification recognition accuracy;

[0027] The classification performance parameters corresponding to each evaluation detection performance indicator obtained by classification include:

[0028] The detection performance parameters are classified into imaging clarity parameters, target detection efficiency parameters, target detection dynamic stability parameters and intelligent classification recognition accuracy parameters.

[0029] From the above description, it can be seen that by setting imaging clarity, target detection efficiency, target detection dynamic stability and intelligent classification and recognition accuracy as evaluation indicators for detection performance; by analyzing the imaging quality evaluation parameters, the ability of side-scan sonar technology to retain image clarity and details during underwater imaging can be measured; by analyzing the underwater target detection capability parameters, the ability and efficiency of side-scan sonar technology to detect targets underwater can be measured; by analyzing the detection stability parameters under the influence of dynamic environment, the ability of side-scan sonar technology to maintain image stability and continuity in a dynamic environment can be measured; by analyzing the intelligent classification and recognition parameters, the accuracy and efficiency of side-scan sonar technology combined with big data and intelligent algorithms for target classification can be measured, thereby realizing the analysis of the various capabilities of side-scan sonar technology.

[0030] Furthermore, the target analysis model includes an imaging clarity model:

[0031]

[0032] Among them, ICI represents the imaging clarity index, Indicates the influencing factor of imaging quality assessment; PD represents the image pixel density, W_ip represents the number of pixels in the image width, H_ip represents the number of pixels in the image height, W_sa represents the scanning area width; SNG represents the image signal-to-noise ratio gain; SNR aft Represents the signal-to-noise ratio after processing, SNR ori represents the original signal-to-noise ratio, Ps represents the image signal power, and Pn represents the image noise power; ES represents the image edge sharpness; G_x represents the gradient component of the image in the x direction, and G_y represents the gradient component of the image in the y direction; ICL represents the image compression loss, QI ori Represents the original image quality index, QI aft The quality index of compressed images, PSNR represents the peak signal-to-noise ratio of the image, and SSIM represents the structural similarity of the image.

[0033] From the above description, it can be seen that by comprehensively considering multiple factors that directly affect the resolution and detail expression of the image, such as pixel density, signal-to-noise ratio gain, edge sharpness, and image compression loss, the imaging quality evaluation parameters can be effectively analyzed and the image clarity and detail retention ability can be accurately judged.

[0034] Furthermore, the target analysis model includes a target detection efficiency model:

[0035]

[0036] Among them, TDE represents the target detection efficiency index; represents the target detection impact factor; MD maxrepresents the maximum detection depth of the side scan sonar; MD0 represents the theoretical detection depth; PLC represents the average percentage of penetration loss of the side scan sonar signal; n represents the number of types of different target media; Sat i represents the received signal strength of the i-th medium, Sor represents the transmitted signal strength; FPR represents the false positive rate, Ne represents the number of misidentified non-targets, and Iq represents the number of identified targets; DTE represents the detection time efficiency, S represents the target detection area, and t sao represents the side scan sonar scanning time, t chu Indicates data processing time, and DTE0 indicates theoretical detection time efficiency.

[0037] From the above description, it can be seen that comprehensive consideration of multiple parameters related to target detection capabilities, such as maximum detection depth, penetration loss ratio, false positive rate, and detection time efficiency, helps to improve target detection capabilities in complex environments.

[0038] Furthermore, the target analysis model includes a target detection dynamic stability model:

[0039]

[0040]

[0041] Where DSC represents the target detection dynamic stability index, Represents the influencing factor of the dynamic environment; ACA represents the attitude correction offset, θ_cal represents the corrected ship attitude data, and θ_ref represents the reference ship attitude data; VSI represents speed fluctuation suppression, σ_wave represents the standard deviation of the image quality of the scanning result under speed fluctuation, and σ_stable represents the standard deviation of the image quality of the scanning result under stable speed; ANI represents dynamic environment noise suppression, SNR_pro represents the processed side scan sonar signal-to-noise ratio, and SNR_ori represents the side scan sonar original signal-to-noise ratio; IC represents image continuity, S_ove represents the area of ​​the image overlapping area, and S_tot represents the total image area.

[0042] From the above description, it can be seen that comprehensive consideration of multiple factors affecting image quality, such as attitude correction accuracy, speed fluctuation suppression, environmental noise suppression, and image continuity, helps to improve the signal-to-noise ratio and clarity of the image, as well as maintain the continuity and stability of the image.

[0043] Furthermore, the target analysis model includes an intelligent classification recognition accuracy model:

[0044]

[0045] Among them, SCA represents the intelligent classification recognition accuracy index, represents the influencing factor of intelligent classification and recognition; BCA represents the basic recognition accuracy, Iq represents the number of recognitions, Cq represents the number of targets correctly recognized by target detection; AOG represents the algorithm optimization gain, CA represents aft represents the classification accuracy of target detection based on big data, CA ori represents the accuracy of original target detection classification; CDS represents the classification decision speed, N_i represents the number of image classification processes, t chu represents the image processing time, and CDS0 represents the original image classification decision speed.

[0046] From the above description, it can be seen that comprehensive consideration of multiple parameters related to data processing, such as basic recognition accuracy, algorithm optimization gain, and classification decision speed, helps to process and analyze data faster and improve processing efficiency and real-time performance.

[0047] Further, the qualification evaluation of each performance index includes:

[0048] Each of the performance indexes is processed by the performance qualification evaluation model to obtain the performance qualification corresponding to each of the evaluation and detection performance indicators, specifically:

[0049]

[0050] in, represents the performance qualification corresponding to the performance index, X0 represents the set value corresponding to the performance index, and X represents the performance index.

[0051] It can be seen from the above description that by setting a corresponding setting value for each performance index, the performance qualification corresponding to each performance indicator can be accurately reflected.

[0052] Furthermore, the comprehensive index of detection performance obtained according to each performance qualification and its corresponding weight value includes:

[0053]

[0054] Among them, DCi represents the comprehensive index of detection performance; Indicates the performance qualification of the imaging clarity index; Indicates the performance qualification of the target detection effectiveness index; Indicates the performance qualification of the target detection dynamic stability index; It represents the performance qualification of the intelligent classification recognition accuracy index; ξ1, ξ2, ξ3 and ξ4 represent the weight values ​​corresponding to different performance qualification levels.

[0055] From the above description, it can be seen that by assigning a corresponding weight value to each performance qualification, all performance qualifications are comprehensively judged. Compared with evaluating the detection performance with a single indicator, comprehensive evaluation with multiple indicators can more comprehensively reflect the performance of the side scan sonar.

[0056] Another embodiment of the present invention provides a side scan sonar detection device based on big data, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, each step in the above-mentioned side scan sonar detection method based on big data is implemented.

[0057] The side-scan sonar detection method and device based on big data provided by the present invention can be applied to the side-scan sonar technology detection scene, which is described below through specific implementation methods:

[0058] Embodiment 1

[0059] Please refer to Figure 1 , a side scan sonar detection method based on big data, comprising:

[0060] S1. Collect the detection performance parameters of the side-scan sonar technology, and classify the detection performance parameters to obtain the classification performance parameters corresponding to each evaluation detection performance indicator. For example, in this embodiment, the evaluation detection performance indicators include imaging clarity, target detection efficiency, target detection dynamic stability, and intelligent classification recognition accuracy, and also include target area information and management personnel identity authentication information; the detection performance parameters are classified to obtain imaging clarity parameters, target detection efficiency parameters, target detection dynamic stability parameters, and intelligent classification recognition accuracy parameters.

[0061] Among them, the imaging quality (imaging clarity) evaluation parameters specifically include: image width pixel number, image height pixel number, scanning area width, image signal power, image noise power, image gradient component in the x direction, image gradient component in the y direction, image peak signal-to-noise ratio, and image structural similarity.

[0062] The underwater target detection capability (target detection efficiency) parameters specifically include: maximum detection depth of side-scan sonar, theoretical detection depth, number of types of different target media, received signal strength, transmitted signal strength, number of falsely identified non-targets, number of identifications, target detection area, side-scan sonar scanning time, and data processing time.

[0063] The detection stability parameters under the influence of dynamic environment (dynamic stability of target detection) specifically include: corrected ship attitude data, reference ship attitude data, standard deviation of image quality, signal-to-noise ratio, area of ​​image overlap, and total image area.

[0064] The intelligent classification and recognition rate parameters specifically include: the number of recognitions, the number of targets correctly recognized by target detection, the target detection classification accuracy, the number of image classification processes, and the image processing time.

[0065] S2. Obtain a target analysis model corresponding to the target evaluation and detection performance index, analyze the classification performance parameters according to the target analysis model, and obtain a performance index corresponding to each evaluation and detection performance index. The specific target analysis model and analysis method are as follows:

[0066] 1) Imaging clarity model:

[0067]

[0068] Among them, ICI represents the imaging clarity index; Indicates the influencing factor of imaging quality assessment. The value of the influencing factor depends on the actual detection situation. PD indicates the image pixel density, W_ip indicates the number of pixels in the image width, H_ip indicates the number of pixels in the image height, and W_sa indicates the scanning area width. SNG indicates the image signal-to-noise ratio gain. SNR aft Represents the signal-to-noise ratio after processing, SNR ori represents the original signal-to-noise ratio, Ps represents the image signal power, and Pn represents the image noise power. The signal power and noise power are obtained by performing Fourier transform on the original signal and the signal after noise suppression and calculating their respective power spectra; ES represents the image edge sharpness; G_x represents the gradient component of the image in the x direction, and G_y represents the gradient component of the image in the y direction. The edge in the image is detected by using edge detection algorithms such as the Sobel operator and the Canny edge detector, and the gradient of the edge area is obtained; ICL represents the image compression loss, QI ori Represents the original image quality index, QI aft The quality index of compressed images, PSNR represents the peak signal-to-noise ratio of the image, and SSIM represents the structural similarity of the image.

[0069] 2) Target Detection Performance Model:

[0070]

[0071]

[0072] Among them, TDE represents the target detection efficiency index; represents the target detection impact factor; MD max represents the maximum detection depth of the side scan sonar; MD0 represents the theoretical detection depth; PLC represents the average percentage of penetration loss of the side scan sonar signal; n represents the number of types of different target media; Sat irepresents the received signal strength of the i-th medium, Sor represents the transmitted signal strength; FPR represents the false positive rate, Ne represents the number of misidentified non-targets, Iq represents the number of identified targets, and 0.01 is added to avoid the denominator being 0; DTE represents the detection time efficiency, S represents the target detection area, and t sao represents the side scan sonar scanning time, t chu Indicates data processing time, and DTE0 indicates theoretical detection time efficiency.

[0073] 3) Target detection dynamic stability model:

[0074]

[0075] Where DSC represents the target detection dynamic stability index, Represents the influencing factor of the dynamic environment; ACA represents the attitude correction offset, θ_cal represents the corrected ship attitude data, θ_ref represents the reference ship attitude data, and the ship attitude data is collected by high-precision attitude sensors and inertial navigation systems, including pitch angle, roll angle, heading angle, etc.; VSI represents speed fluctuation suppression, σ_wave represents the standard deviation of the image quality of the scanning result under speed fluctuation, σ_stable represents the standard deviation of the image quality of the scanning result under stable speed, and 0.01 is added to avoid the denominator being 0; ANI represents dynamic environment noise suppression, SNR_pro represents the processed side scan sonar signal-to-noise ratio, and SNR_ori represents the side scan sonar original signal-to-noise ratio; IC represents image continuity, S_ove represents the area of ​​the image overlapping area, and S_tot represents the total image area.

[0076] 4) Intelligent classification and recognition accuracy model:

[0077]

[0078] Among them, SCA represents the intelligent classification recognition accuracy index, represents the influencing factor of intelligent classification and recognition; BCA represents the basic recognition accuracy, Iq represents the number of recognitions, Cq represents the number of targets correctly recognized by target detection; AOG represents the algorithm optimization gain, CA represents aft represents the classification accuracy of target detection based on big data, CA ori represents the accuracy of original target detection classification; CDS represents the classification decision speed, N_i represents the number of image classification processes, t chu represents the image processing time, and CDS0 represents the original image classification decision speed.

[0079] S3. Evaluate the qualification of each performance index to obtain the performance qualification corresponding to each evaluation and detection performance index; in an optional implementation, process each performance index through a performance qualification evaluation model to obtain the performance qualification corresponding to each evaluation and detection performance index. The expression of the performance qualification evaluation model is as follows:

[0080]

[0081] in, Indicates the performance qualification corresponding to the performance index, X0 indicates the set value corresponding to the performance index, and X indicates the performance index. For example, the imaging clarity index ICI, the target detection efficiency index TDE, the target detection dynamic stability index DSC, and the intelligent classification recognition accuracy index SCA are successively substituted into the above formula, and the following are obtained respectively:

[0082]

[0083]

[0084] in, Indicates the performance qualification of the imaging clarity index; Indicates the performance qualification of the target detection effectiveness index; Indicates the performance qualification of the target detection dynamic stability index; Indicates the performance qualification of the intelligent classification recognition accuracy index; ICI0, TDE0, DSC0, SCA0 represent the set values ​​corresponding to the index; it should be specifically stated that: if If the image quality is good, it means the image quality is good. If the image quality is bad, it means the image quality is abnormal. It indicates that the target detection performance is qualified; if It indicates that the dynamic stability of target detection is qualified; if This indicates that the intelligent classification recognition accuracy is qualified.

[0085] S4. Obtain the weight value corresponding to each performance qualification, and obtain the comprehensive index of detection performance according to each performance qualification and its corresponding weight value; in an optional implementation, the comprehensive index of detection performance is obtained through a comprehensive analysis model, and the comprehensive analysis model is as follows:

[0086]

[0087] Among them, DCi represents the comprehensive index of detection performance; Indicates the performance qualification of the imaging clarity index; Indicates the performance qualification of the target detection effectiveness index; Indicates the performance qualification of the target detection dynamic stability index; It indicates the performance qualification of the intelligent classification recognition accuracy index; ξ1, ξ2, ξ3 and ξ4 indicate the weight values ​​corresponding to different performance qualification levels, and the weight values ​​are generally determined according to the contribution of a single indicator to the evaluation of detection performance.

[0088] S5. Determine whether the comprehensive index of detection performance exceeds the preset allowable range. If so, an alarm signal is issued. That is, after analyzing the comprehensive index of detection performance, human-computer interaction is performed. When the comprehensive index of detection performance belongs to the set allowable range, it means that the detection performance of the side-scan sonar technology is good. If the comprehensive index of detection performance exceeds the allowable range, the management personnel are prompted to take timely measures. For example: optimize the signal acquisition and processing algorithm, improve the sampling rate and processing accuracy of the signal; optimize the sonar beamforming and scanning strategy, improve the detection range and depth, adjust the scanning parameters and strategies in real time according to the speed fluctuation, and strengthen the data preprocessing and feature extraction links to meet the increasingly complex and diverse underwater mission requirements.

[0089] Embodiment 2

[0090] Please refer to Figure 2 A side-scan sonar detection device based on big data includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step in a side-scan sonar detection method based on big data in Embodiment 1 is implemented. In a specific embodiment, the memory and the processor form a side-scan sonar technology detection data acquisition module, a side-scan sonar technology detection data analysis module, a side-scan sonar technology detection data evaluation module, a side-scan sonar technology detection comprehensive analysis module, and a side-scan sonar technology detection human-computer interaction module.

[0091] Please refer to Figure 3 The side scan sonar technology detection data acquisition module is connected to the side scan sonar technology detection data analysis module, the side scan sonar technology detection data evaluation module is respectively connected to the side scan sonar technology detection comprehensive analysis module and the side scan sonar technology detection data analysis module, and the side scan sonar technology detection comprehensive analysis module is respectively connected to the side scan sonar technology detection data evaluation module and the side scan sonar technology detection human-computer interaction module.

[0092] Among them, the side-scan sonar technology detection data acquisition module is used to collect the detection performance parameters of the side-scan sonar technology, obtain various parameters for evaluating the detection performance indicators, and transmit the various parameters to the side-scan sonar technology detection data analysis module.

[0093] The side scan sonar technology detection data analysis module is used to analyze various parameters to obtain various indexes of side scan sonar technology detection; for example, the corresponding detection performance parameters are analyzed respectively through the imaging quality evaluation index analysis unit, the underwater target detection capability index analysis unit, the detection stability index analysis unit under the influence of dynamic environment, and the intelligent classification and recognition index analysis unit to obtain the imaging clarity index ICI, the target detection efficiency index TDE, the target detection dynamic stability index DSC, and the intelligent classification and recognition accuracy index SCA as described in Example 1, and then the various indexes are transmitted to the side scan sonar technology detection data evaluation module.

[0094] The side-scan sonar technology detection data evaluation module is used to compare the various indexes of the side-scan sonar technology detection obtained by the above-mentioned data analysis module with the set values. For example, each performance index is processed by the performance qualification evaluation model in Example 1 to obtain the performance qualification corresponding to each evaluation detection performance indicator; and then the comparison and evaluation results are transmitted to the side-scan sonar technology detection comprehensive analysis module.

[0095] The side scan sonar technology detection comprehensive analysis module is used to perform a comprehensive analysis on the results calculated by the above data evaluation module; for example, a comprehensive analysis is performed through the comprehensive analysis model in Example 1 to obtain a comprehensive index of detection performance, and the comprehensive index of detection performance is output to the side scan sonar technology detection human-computer interaction module.

[0096] The side-scan sonar technology detection human-computer interaction module is used to output the comprehensive detection performance indicators transmitted by the above modules to the management personnel information terminal.

[0097] In summary, the side-scan sonar detection method and device based on big data provided by the present invention analyzes the imaging quality evaluation parameters, underwater target detection capability parameters and detection stability parameters under the influence of dynamic environment respectively, and evaluates the image clarity and detail retention ability of the side-scan sonar technology in the underwater imaging process based on the imaging quality index. The pixel density directly affects the resolution and detail expression of the image, and a clear image helps to reduce misjudgment and missed judgment, and improve the accuracy and reliability of target recognition; the ability and efficiency of the side-scan sonar technology in underwater target detection is evaluated based on the underwater target detection capability index, which helps to accurately identify and locate underwater target objects, thereby helping users to make more accurate judgments, helping to quickly process a large amount of sonar data, improve work efficiency, and reduce detection costs; and the ability of the side-scan sonar technology to maintain image stability and continuity in a dynamic environment is evaluated based on the detection stability index under the influence of dynamic environment, which helps to adapt to these environmental changes, maintain stable detection performance, avoid detection result errors caused by the influence of dynamic environment, and thus improve the quality of detection data. Through big data technology, we can deeply analyze and mine massive amounts of sonar data, achieve accurate feature extraction and classification of sonar data, improve data processing capabilities and detection efficiency, and thus improve detection accuracy, providing more scientific decision-making support for marine resource development, environmental protection, underwater archaeology and other fields to meet the increasingly complex and diverse needs of underwater missions.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0099] The above contents are merely examples and explanations of the concept of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A side-scan sonar detection method based on big data, characterized in that: include: Collecting detection performance parameters of the side scan sonar technology, and classifying the detection performance parameters to obtain classified performance parameters corresponding to each evaluation detection performance indicator; Acquire a target analysis model corresponding to the target evaluation and detection performance index, analyze the classification performance parameters according to the target analysis model, and obtain a performance index corresponding to each of the evaluation and detection performance indicators; Performing a qualification evaluation on each of the performance indexes to obtain a performance qualification corresponding to each of the evaluation and detection performance indicators; Obtaining a weight value corresponding to each of the performance qualifications, and obtaining a comprehensive indicator of the detection performance according to each of the performance qualifications and its corresponding weight value; Determine whether the comprehensive index of the detection performance exceeds the preset allowable range, and if so, issue an alarm signal.

2. The side-scan sonar detection method based on big data according to claim 1, characterized in that: The evaluation and detection performance indicators include imaging clarity, target detection efficiency, target detection dynamic stability and intelligent classification and recognition accuracy; The classification performance parameters corresponding to each evaluation detection performance indicator obtained by classification include: The detection performance parameters are classified into imaging clarity parameters, target detection efficiency parameters, target detection dynamic stability parameters and intelligent classification recognition accuracy parameters.

3. The side-scan sonar detection method based on big data according to claim 2, characterized in that: The target analysis model includes an imaging clarity model: Among them, ICI represents the imaging clarity index, Indicates the influencing factor of imaging quality assessment; PD represents the image pixel density, W_ip represents the number of pixels in the image width, H_ip represents the number of pixels in the image height, W_sa represents the scanning area width; SNG represents the image signal-to-noise ratio gain; SNR aft Represents the signal-to-noise ratio after processing, SNR ori represents the original signal-to-noise ratio, Ps represents the image signal power, and Pn represents the image noise power; ES represents the image edge sharpness; G_x represents the gradient component of the image in the x direction, and G_y represents the gradient component of the image in the y direction; ICL represents the image compression loss, QI ori Represents the original image quality index, QI aft The quality index of compressed images, PSNR represents the peak signal-to-noise ratio of the image, and SSIM represents the structural similarity of the image.

4. The side scan sonar detection method based on big data according to claim 2 is characterized in that: The target analysis model includes a target detection performance model: Among them, TDE represents the target detection efficiency index; represents the target detection impact factor; MD max represents the maximum detection depth of the side scan sonar; MD0 represents the theoretical detection depth; PLC represents the average percentage of penetration loss of the side scan sonar signal; n represents the number of types of different target media; Sat i represents the received signal strength of the i-th medium, Sor represents the transmitted signal strength; FPR represents the false positive rate, Ne represents the number of misidentified non-targets, and Iq represents the number of identified targets; DTE represents the detection time efficiency, S represents the target detection area, and t sao represents the side scan sonar scanning time, t chu Indicates data processing time, and DTE0 indicates theoretical detection time efficiency.

5. The side scan sonar detection method based on big data according to claim 2 is characterized in that: The target analysis model includes a target detection dynamic stability model: Where DSC represents the target detection dynamic stability index, Represents the influencing factor of the dynamic environment; ACA represents the attitude correction offset, θ_cal represents the corrected ship attitude data, and θ_ref represents the reference ship attitude data; VSI represents speed fluctuation suppression, σ_wave represents the standard deviation of the image quality of the scanning result under speed fluctuation, and σ_stable represents the standard deviation of the image quality of the scanning result under stable speed; ANI represents dynamic environment noise suppression, SNR_pro represents the processed side scan sonar signal-to-noise ratio, and SNR_ori represents the side scan sonar original signal-to-noise ratio; IC represents image continuity, S_ove represents the area of ​​the image overlapping area, and S_tot represents the total image area.

6. The side scan sonar detection method based on big data according to claim 2 is characterized in that: The target analysis model includes an intelligent classification recognition accuracy model: Among them, SCA represents the intelligent classification recognition accuracy index, represents the influencing factor of intelligent classification and recognition; BCA represents the basic recognition accuracy, Iq represents the number of recognitions, Cq represents the number of targets correctly recognized by target detection; AOG represents the algorithm optimization gain, CA represents aft represents the classification accuracy of target detection based on big data, CA ori represents the accuracy of original target detection classification; CDS represents the classification decision speed, N_i represents the number of image classification processes, t chu represents the image processing time, and CDS0 represents the original image classification decision speed.

7. The side scan sonar detection method based on big data according to claim 1, characterized in that: The qualification evaluation of each performance index comprises: Each of the performance indexes is processed by a performance qualification evaluation model to obtain a performance qualification corresponding to each of the evaluation and detection performance indicators.

8. The side-scan sonar detection method based on big data according to claim 7, characterized in that: The performance qualification evaluation model includes: in, represents the performance qualification corresponding to the performance index, X0 represents the set value corresponding to the performance index, and X represents the performance index.

9. The side-scan sonar detection method based on big data according to claim 2, characterized in that: The comprehensive index of detection performance obtained according to each performance qualification and its corresponding weight value includes: Among them, DCi represents the comprehensive index of detection performance; Indicates the performance qualification of the imaging clarity index; Indicates the performance qualification of the target detection effectiveness index; Indicates the performance qualification of the target detection dynamic stability index; It represents the performance qualification of the intelligent classification recognition accuracy index; ξ1, ξ2, ξ3 and ξ4 represent the weight values ​​corresponding to different performance qualification levels.

10. A side-scan sonar detection device based on big data, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, each step of the side scan sonar detection method based on big data as described in any one of claims 1 to 9 is implemented.

Images