Underwater target detection method based on dynamic channel impact response analysis

Abstract

The invention discloses an underwater target detection method based on dynamic channel impact response analysis, and relates to the technical field of underwater target detection and identification based on dynamic channel impact response analysis. The method comprises the following steps: determining an abnormal region according to the difference between an instantaneous channel impact response and a reference channel impact response; generating a refined image of the abnormal region through a mode of adaptively filling pixel points of the abnormal region and adjusting the sampling step length of the abnormal region; according to the channel impact response of the refined image, performing difference extraction through a reference channel impact response corresponding to the refined image to obtain a difference feature; and determining a target contour according to the difference characteristics. According to the method, the channel impact response of the refined image is matched with the reference channel impact response, the target difference characteristics are determined, and the reconstructed contour which is very similar to the real contour of the target can be obtained according to the difference characteristics, so that the geometric morphology of the underwater target is highly restored, and the detection precision is remarkably improved.

Description

Technical Field

[0001] This application relates to the field of underwater target detection and identification technology based on dynamic channel impulse response analysis, and specifically to an underwater target detection method based on dynamic channel impulse response analysis. Background Technology

[0002] Currently, with the rapid development of underwater unmanned equipment technology, the underwater security situation is becoming increasingly severe. To address the underwater security needs of ports, anchorages, and high-value marine platforms, detection systems are required to accurately identify various "low," "small," "slow," and stealthy targets that suddenly intrude.

[0003] Existing underwater acoustic detection technologies still face many serious challenges when dealing with such complex scenarios, specifically: Traditional active sonar detection relies mainly on the strong emission of sound waves by the target. However, modern detection targets (such as miniature submarines coated with sound-absorbing materials or miniature unmanned underwater vehicles using sound-transmitting materials) have extremely strong acoustic stealth characteristics. The targets cause weak echoes, which are easily submerged in the environmental reverberation, making it difficult to accurately reconstruct the outline features of the target.

[0004] It is known that the existing underwater target detection accuracy based on active sonar is relatively low. Summary of the Invention

[0005] In view of the deficiencies in the existing technology, the technical problem to be solved by this application is: how to improve the detection accuracy of underwater targets.

[0006] To achieve the above objectives, in a first aspect, embodiments of this application provide an underwater target detection method based on dynamic channel impulse response analysis, the method comprising the following steps: A time-varying channel background model is constructed based on the historical channel frequency response of the water area to be detected. The water area to be detected is scanned, and the instantaneous channel impulse response of the current scanned area is obtained based on the scanning information of the current scanned area at the current moment; The reference channel impulse response of the current scanned region at the current time is obtained based on the time-varying channel background model. The abnormal region is determined based on the difference between the instantaneous channel impulse response and the reference channel impulse response; A refined image of the abnormal region is generated by adaptively filling pixels in the abnormal region and adjusting the sampling step size of the abnormal region. Based on the channel impulse response of the refined image, differential features are obtained by performing differential extraction on the reference channel impulse response corresponding to the refined image; the target contour is determined based on the differential features.

[0007] In conjunction with the first aspect, in one implementation method, the process of constructing a time-varying channel background model based on the historical channel frequency response of the water area to be detected includes: According to the set sampling step size, the water area to be detected is scanned in grids to obtain the channel state information of the water area to be detected; Determine the channel frequency response of the water area to be detected based on channel state information; The time-varying channel background model is obtained based on the channel frequency response.

[0008] In conjunction with the first aspect, in one implementation method, the formula for calculating the channel frequency response is: ; in represent t The frequency of time is f The channel frequency response; This represents the received signal; Represents a known transmitted signal; Representing the i The amplitude attenuation of the path; Representing the i The propagation delay of each path; The generation process of the time-varying channel background model includes: inputting the channel frequency response into a deep learning network for training, and then establishing a time-varying channel background model.

[0009] In conjunction with the first aspect, in one implementation method, the process for determining the abnormal region includes: All grids in the water area to be explored are scanned, and the instantaneous channel impulse response of the current scanned area is determined based on the scan information at the current moment. Predict the reference channel impulse response of the current scanned area at the current moment using a time-varying channel background model; Determine the scalar correlation coefficient between the instantaneous channel impulse response and the reference channel impulse response; Determine if the scalar correlation coefficient is greater than a preset threshold. If it is, determine that the current scanned area is a normal area; otherwise, determine that the current scanned area is an abnormal area.

[0010] In conjunction with the first aspect, in one implementation, the currently scanned region at the current time t Reference channel impulse response The calculation formula is: ; in This represents the inverse fast Fourier transform, used to convert frequency domain features into time domain impulse responses; Indicates the length of the historical observation sequence; The scalar correlation coefficient The calculation formula is: ; in The value range is [0,1]; Indicates complex conjugation.

[0011] In conjunction with the first aspect, in one implementation, the process of generating a refined image of the abnormal region by adaptively filling in the abnormal region pixels and adjusting the sampling step size of the abnormal region includes: Based on the spatial coordinates of the abnormal regions, a refined region encompassing all abnormal regions is determined; An adaptive interpolation algorithm is used to scan the refined region. During the scanning process, the sampling step size is adjusted according to the spatial channel feature gradient, which is inversely proportional to the sampling step size.

[0012] In conjunction with the first aspect, in one implementation, the spatial channel feature gradient The formula for calculating G is: ; in,( x,y,z () represents three-dimensional spatial coordinates; The adjusted sampling step size The calculation formula is: ; in, Indicates the initial sampling step size; This represents the sensitivity coefficient.

[0013] In conjunction with the first aspect, in one implementation, the process of determining the target contour includes: The channel impulse response of the refined image is differentially matched with the reference channel impulse response of the refined image, and the image regions that cannot be matched are used as difference features. By mapping the differential features back to three-dimensional coordinates, we obtain the probability distribution point set of the differential features in three-dimensional space. probability distribution point set The calculation formula is: ; in, This represents the sound field inverse mapping operator; Indicates a horizontal angle. Indicates a vertical angle; Cluster analysis is performed on the probability distribution point set to obtain the target contour.

[0014] In conjunction with the first aspect, in one implementation, the method further includes the following steps: After obtaining several target contours, the centroid velocity of the target contours is determined based on their coordinates and sampling period. Determine the Doppler frequency shift component based on the received signal from the target; The velocity vector of the target profile is obtained based on the centroid velocity and Doppler frequency shift component of the target profile.

[0015] In conjunction with the first aspect, in one embodiment, the centroid velocity of the target profile The calculation formula is: ; in The centroid coordinates representing the current target contour. Represents the sampling period. The centroid coordinates represent the target contour from the previous sampling. The Doppler frequency shift component The calculation formula is: ; Where c represents the speed of sound; This represents the carrier phase difference between two consecutive frames of signal; Represents the center frequency of the signal; Represents the pulse interval; This represents the angle between the target's trajectory and the sonar radial direction.

[0016] Compared with the prior art, the advantages of this application are: This application uses historical data of the water area to be explored (obtained by continuously learning and updating environmental information such as sound velocity profile, sea surface fluctuations, and seabed topography of the exploration area daily) to predict the reference channel impulse response during exploration.

[0017] Based on this, unlike existing technologies that often generate a large number of false signals (false alarms) due to changes in sea state, this application can automatically filter out environmental noise interference by implementing a comparative detection of the difference between the instantaneous channel impulse response and the reference channel impulse response, focusing on the abnormal response of non-cooperative targets. This separates the multipath micro-perturbations caused by the target (i.e., "channel fingerprint"), significantly reducing the probability of false alarms (by about 60%), and providing users with more reliable security early warning information.

[0018] Based on this, this application first uses a "coarse scan" method to quickly locate abnormal areas, and then uses a "fine scan" method to refine the abnormal areas to obtain a refined image.

[0019] As can be seen, compared with traditional uniform high-density scanning across the entire airspace, this application can not only significantly reduce the amount of computation, but also improve the speed of detecting sensitive targets with the same hardware computing power; this significantly shortens the time window from "detecting the target" to "confirming the outline", enabling it to meet the requirements for detecting high-speed moving targets (UUVs or frogmen).

[0020] Therefore, this application can significantly reduce the system's computational load while ensuring the accuracy of target contour sampling.

[0021] In this case, this application determines the target difference features by matching the channel impulse response of the refined image with the reference channel impulse response. Based on the difference features, a reconstructed contour that is very similar to the true contour of the target can be obtained, thereby highly restoring the geometry of the underwater target and significantly improving the detection accuracy. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the underwater target detection method based on dynamic channel impulse response analysis in the embodiments of this application; Figure 2 This is an example of the ability to perceive weak targets in a low signal-to-noise ratio environment in the embodiments of this application; Figure 3 This is a schematic diagram of the hierarchical scanning strategy in an embodiment of this application; Figure 4 This is a schematic diagram comparing the target outline obtained in the embodiments of this application with the real target; Figure 5 This is a comparison between the estimated target velocity vector and the actual velocity vector in the embodiments of this application; Figure 6 This is a schematic diagram of the hardware structure of an underwater target detection device based on dynamic channel impulse response analysis, which is involved in the embodiments of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0026] First, a brief description of the research and development principles in this application will be given.

[0027] When a target enters the detection area, it causes a slight disturbance to the physical characteristics of the underwater acoustic channel (i.e., the time-varying nature of the channel impulse response). This dynamic change contains key information such as the target's position, outline, speed, and even attitude.

[0028] This application utilizes the aforementioned method to perform inverse inversion using this "channel fingerprint" change, making full use of the system's ability to perceive subtle environmental disturbances and its data depth mining capabilities, while preserving the channel information of the target's geometric features and motion state, thereby enabling a refined estimation of the target's outline and velocity.

[0029] By analyzing the dynamic channel impulse response, we can achieve high-sensitivity detection of weak targets while simultaneously estimating the target contour imaging and velocity vector.

[0030] Based on this, in order to make the objectives, technical solutions and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0031] In a first aspect, embodiments of this application provide an underwater target detection method based on dynamic channel impulse response analysis, see [link to relevant documentation]. Figure 1 As shown, the steps of this method include: S1: Environmental Background Modeling and Initialization: Based on the historical channel frequency response of the water area to be detected (i.e., the region of interest), a time-varying channel background model with environmental robustness is constructed. This model can predict the current reference channel impulse response over time and use it as a baseline for judging whether there are external targets. This enables adaptive updates to changes in tides, temperature layers, and sea states, thereby accurately filtering environmental background noise in subsequent processing.

[0032] S2: Global Fast Polling Coarse Scan: The water area to be detected is scanned, and the instantaneous channel impulse response of the current scanned area is obtained based on the scanning information of the current scanned area at the current moment; The reference channel impulse response of the current scanned region at the current time is obtained based on the time-varying channel background model. The abnormal region is determined based on the difference between the instantaneous channel impulse response and the reference channel impulse response.

[0033] S3: Abnormal Region Locking and Adaptive Mesh Refinement: By adaptively filling abnormal region pixels and adjusting the abnormal region sampling step size, the resolution of the abnormal region is improved, and a refined image of the abnormal region is generated.

[0034] S4: Based on the channel impulse response of the refined image, differential extraction is performed using the reference channel impulse response corresponding to the refined image to obtain the difference features; the target contour is determined based on the difference features.

[0035] In one embodiment, the process of S1 specifically includes: S101: Periodically scan the water area to be detected by grid according to the set sampling step size to obtain the channel state information (CSI) of the water area to be detected.

[0036] S102: Determine the channel frequency response (CFR) of the water area to be detected based on the channel state information.

[0037] Specifically, the formula for calculating the channel frequency response is as follows: ; in represent t The frequency of time is f The channel frequency response; This represents the received signal; Represents a known transmitted signal; Representing the i The amplitude attenuation of the path; Representing the i The propagation delay of each path; and It is extracted from the channel state information.

[0038] S103: Obtain a time-varying channel background model (CIR) based on the channel frequency response to generate the prediction benchmark.

[0039] The execution method of S103 is as follows: after inputting the channel frequency response into a deep learning network (in this embodiment, it is a long short-term memory network LSTM) for training, a time-varying channel background model is established.

[0040] The principle of S103 is as follows: For micro-submarines with sound-absorbing materials coated on their surfaces or underwater robots using sound-transmitting composite materials, traditional active / passive sonar often misses detection due to weak echoes. This application utilizes LSTM for in-depth mining of time-varying channel fingerprints, enabling it to keenly capture the multipath structural perturbations in the sound field caused by target intrusion. In environments with strong reverberation or low signal-to-noise ratio, this application increases the detection probability of low acoustic target intensity (TS) targets from 75% in existing technologies to over 92%, greatly enhancing the system's anti-stealth and anti-intrusion capabilities.

[0041] Meanwhile, the Fourier transform of the channel impulse response (which is used to improve efficiency) is the channel frequency response. This application uses the channel frequency response for calculation because it significantly reduces complexity, improves system efficiency, and lowers hardware costs compared to the channel impulse response. However, analyzing the channel impulse response makes it easier to observe micro-perturbation-related information and extract anomalous disturbances compared to the channel frequency response.

[0042] In one embodiment, see Figure 3 As shown, the process of S2 includes: S201: The transmitting array performs rapid polling scans of all grids in the water area to be detected, both vertically and horizontally, according to a preset timeliness priority principle; S202: Based on the scanning information (real-time calculation of the amplitude, phase, and angle of arrival of the received echo), determine the instantaneous channel impulse response of the current scanning area (current grid point) at the current moment.

[0043] S203: Predict the current scanned area at the current time using a time-varying channel background model. t Reference channel impulse response The calculation formula is: ; in This represents the inverse fast Fourier transform, used to convert frequency domain features into time domain impulse responses; Indicates the length of the historical observation sequence.

[0044] S204: Determine the scalar correlation coefficient (i.e., difference) between the instantaneous channel impulse response and the reference channel impulse response. The calculation formula is: ; in, The value range is [0,1]; Indicates complex conjugation.

[0045] S205: Determine whether the scalar correlation coefficient is greater than a preset threshold (in this embodiment, the preset threshold is 0.7~0.85): If so, the current scanning area is determined to be a normal area; at this time, the time-varying channel background model can be updated based on the instantaneous channel impulse response.

[0046] If not, the current scanning area is determined to be an abnormal area, indicating that an underwater target that needs to be detected has entered.

[0047] See Figure 2 As shown, Figure 2 This demonstrates the ability of this application to detect weak targets in low signal-to-noise ratio environments; see Figure (a) for background channel response, (b) for real-time response including the target, and (c) for extracted differential channel "fingerprint". As shown in (c), this application successfully separates multipath micro-perturbations (i.e., 'channel fingerprint') caused by the target by analyzing the differences in channel impulse responses. This proves that this application does not rely on strong target echoes, but achieves high-sensitivity detection by analyzing channel changes.

[0048] In one embodiment, the process of S3 includes: S301: Determine the refined region that includes all anomaly regions based on the spatial coordinates of all anomaly regions (all grid points).

[0049] S302: An adaptive interpolation algorithm (bicubic interpolation algorithm in this embodiment) is used to scan the refined region. During the scanning process, the sampling step size is adjusted according to the spatial channel feature gradient. The spatial channel feature gradient is inversely proportional to the sampling step size. This is to reduce the sampling step size to improve accuracy in locations where the spatial channel feature gradient changes drastically (i.e., target edges) and maintain a larger step size in areas where the spatial channel feature gradient is gentle to save computing power, thereby finely cutting the target contour in the abnormal region.

[0050] Specifically, spatial channel feature gradient The formula for calculating G is: ; in,( x,y,z () represents three-dimensional spatial coordinates.

[0051] Adjusted sampling step size The calculation formula is: ; in, Indicates the initial sampling step size; This represents the sensitivity coefficient.

[0052] See Figure 3 As shown, Figure 3 This demonstrates the hierarchical scanning strategy of this application (coarse scanning with fast polling of the entire area in S2 and precise scanning of abnormal areas in S3). Figure 3The sparse blue dots represent a coarse scan of the entire area, while the dense red dots represent an adaptive interpolation scan after the algorithm automatically identifies abnormal regions. The comparison shows that high-density computing resources are allocated only around the target (red area), while the environmental area is monitored only at a low frequency.

[0053] See Figure 3 As shown, compared with traditional full-space uniform high-density scanning, the computational load of this application is reduced by approximately 53.8%; under the same hardware computing power, the detection speed of sensitive targets is increased by more than 100%. This significantly shortens the time window from "target detection" to "confirmation of outline".

[0054] Therefore, compared with traditional full-area uniform high-density scanning, this application can significantly reduce the system's computational load while ensuring the target contour sampling accuracy.

[0055] In one embodiment, the specific process of S4 includes: S401: Image matching and feature extraction: The channel impulse response of the refined image is differentially matched with the reference channel impulse response (obtained from the time-varying channel background model in S1) to obtain the image regions that cannot be matched and use them as differential features.

[0056] S402: Inverse Inversion: Using the sound field propagation model, the differential features are inversely mapped back to three-dimensional coordinates to obtain the probability distribution point set of the differential features in three-dimensional space.

[0057] Specifically, probability distribution point set The calculation formula is: ; in, This represents the sound field inverse mapping operator; Indicates a horizontal angle. Indicates the vertical angle.

[0058] S403: Perform cluster analysis on the probability distribution point set to obtain the three-dimensional geometric contour of the target.

[0059] See Figure 4 As shown, the green translucent sphere represents the target's actual physical boundary, and the purplish-red curved surface represents the target's three-dimensional geometric contour reconstructed in this application. It can be seen that the reconstructed contour has a high degree of spatial overlap with the actual contour, proving that this application can accurately reproduce the geometry of underwater stealth targets.

[0060] In one embodiment, the method further includes the following steps after S4: S5: Repeat S2 to S4 (usually once) to obtain several target contours.

[0061] S6: Estimate the velocity vector of the target contour based on the coordinates of different target contours and the sampling period of different target contours.

[0062] The S6 process includes: S601: Determine the centroid velocity of the target contour based on the coordinates of the target contour and the sampling period of different target contours.

[0063] Centroid velocity of the target profile The calculation formula is: ; in The centroid coordinates representing the current target contour. Represents the sampling period. The centroid coordinates of the target contour from the previous sampling.

[0064] S602: Determine the Doppler frequency shift component based on the received signal from the target.

[0065] Doppler frequency shift component The calculation formula is: ; Where c represents the speed of sound; This represents the carrier phase difference between two consecutive frames of signal; Represents the center frequency of the signal; Represents the pulse interval; This represents the angle between the target's trajectory and the sonar radial direction.

[0066] S603: Obtain the velocity vector of the target profile based on the centroid velocity and Doppler frequency shift component of the target profile.

[0067] See Figure 6 As shown, Figure 6 The accuracy of the velocity vector estimated in this application is demonstrated; the blue solid arrow in the figure represents the target's true velocity, while the red dashed arrow represents the estimated velocity after the algorithm's comprehensive displacement calculation and Doppler frequency shift. Both are essentially consistent in direction and magnitude, and the error is controlled within the allowable range, proving that this application can achieve real-time synchronous tracking of the target's motion state.

[0068] Secondly, embodiments of this application provide an underwater target detection device based on dynamic channel impulse response analysis. The underwater target detection device based on dynamic channel impulse response analysis can be a personal computer (PC), laptop computer, server, or other device with data processing capabilities.

[0069] Reference Figure 6 , Figure 6This is a schematic diagram of the hardware structure of an underwater target detection device based on dynamic channel impulse response analysis, as described in an embodiment of this application. In this embodiment, the underwater target detection device based on dynamic channel impulse response analysis may include a processor, memory, a communication interface, and a communication bus.

[0070] The communication bus can be of any type and is used to interconnect the processor, memory, and communication interface.

[0071] The communication interface includes input / output (I / O) interfaces, physical interfaces, and logical interfaces. These interfaces are used for interconnecting internal components of the underwater target detection equipment based on dynamic channel impulse response analysis, as well as for interconnecting the underwater target detection equipment with other devices (such as other computing devices or user equipment). Physical interfaces can be Ethernet interfaces, fiber optic interfaces, ATM interfaces, etc.; user equipment can be displays, keyboards, etc.

[0072] Memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0073] The processor can be a general-purpose processor, which can call the underwater target detection program based on dynamic channel impulse response analysis stored in memory and execute the underwater target detection method based on dynamic channel impulse response analysis provided in the embodiments of this application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed when the underwater target detection program based on dynamic channel impulse response analysis is called can be referred to the various embodiments of the underwater target detection method based on dynamic channel impulse response analysis in this application, and will not be repeated here.

[0074] Those skilled in the art will understand that Figure 6 The hardware structure shown does not constitute a limitation of this application and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0075] Thirdly, embodiments of this application also provide a computer-readable storage medium.

[0076] The computer-readable storage medium of this application stores an underwater target detection program based on dynamic channel impulse response analysis, wherein when the underwater target detection program based on dynamic channel impulse response analysis is executed by a processor, it implements the steps of the underwater target detection method based on dynamic channel impulse response analysis as described above.

[0077] The method implemented when the underwater target detection program based on dynamic channel impulse response analysis is executed can be referred to in various embodiments of the underwater target detection method based on dynamic channel impulse response analysis in this application, and will not be repeated here.

[0078] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0079] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0080] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0081] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0082] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0083] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0084] The above are merely specific embodiments of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the scope of the claims.

Claims

1. An underwater target detection method based on dynamic channel impulse response analysis, characterized in that, The method includes the following steps: A time-varying channel background model is constructed based on the historical channel frequency response of the water area to be detected. The water area to be detected is scanned, and the instantaneous channel impulse response of the current scanned area is obtained based on the scanning information of the current scanned area at the current moment; The reference channel impulse response of the current scanned region at the current time is obtained based on the time-varying channel background model. The abnormal region is determined based on the difference between the instantaneous channel impulse response and the reference channel impulse response; A refined image of the abnormal region is generated by adaptively filling pixels in the abnormal region and adjusting the sampling step size of the abnormal region. Based on the channel impulse response of the refined image, differential features are obtained by performing differential extraction on the reference channel impulse response corresponding to the refined image; the target contour is determined based on the differential features.

2. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 1, characterized in that: The process of constructing a time-varying channel background model based on the historical channel frequency response of the water area to be detected includes: According to the set sampling step size, the water area to be detected is scanned in grids to obtain the channel state information of the water area to be detected; Determine the channel frequency response of the water area to be detected based on channel state information; The time-varying channel background model is obtained based on the channel frequency response.

3. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 2, characterized in that, The formula for calculating the channel frequency response is: ； in represent t The frequency of time is f The channel frequency response; This represents the received signal; Represents a known transmitted signal; Representing the i The amplitude attenuation of the path; Representing the i The propagation delay of each path; The generation process of the time-varying channel background model includes: inputting the channel frequency response into a deep learning network for training, and then establishing a time-varying channel background model.

4. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 1, characterized in that, The process for determining the abnormal region includes: All grids in the water area to be explored are scanned, and the instantaneous channel impulse response of the current scanned area is determined based on the scan information at the current moment. Predict the reference channel impulse response of the current scanned area at the current moment using a time-varying channel background model; Determine the scalar correlation coefficient between the instantaneous channel impulse response and the reference channel impulse response; Determine if the scalar correlation coefficient is greater than a preset threshold. If it is, determine that the current scanned area is a normal area; otherwise, determine that the current scanned area is an abnormal area.

5. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 4, characterized in that, The currently scanned area at the current moment t Reference channel impulse response The calculation formula is: ； in This represents the inverse fast Fourier transform, used to convert frequency domain features into time domain impulse responses; Indicates the length of the historical observation sequence; The scalar correlation coefficient The calculation formula is: ； in The value range is [0,1]; Indicates complex conjugation.

6. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 1, characterized in that, The process of generating a refined image of the abnormal region by adaptively filling in the abnormal region pixels and adjusting the sampling step size of the abnormal region includes: Based on the spatial coordinates of the abnormal regions, a refined region encompassing all abnormal regions is determined; An adaptive interpolation algorithm is used to scan the refined region. During the scanning process, the sampling step size is adjusted according to the spatial channel feature gradient, which is inversely proportional to the sampling step size.

7. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 6, characterized in that: The spatial channel feature gradient The formula for calculating G is: ； in,( x,y,z () represents three-dimensional spatial coordinates; The adjusted sampling step size The calculation formula is: ； in, Indicates the initial sampling step size; This represents the sensitivity coefficient.

8. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 1, characterized in that, The process for determining the target contour includes: The channel impulse response of the refined image is differentially matched with the reference channel impulse response of the refined image, and the image regions that cannot be matched are used as difference features. By mapping the differential features back to three-dimensional coordinates, we obtain the probability distribution point set of the differential features in three-dimensional space. probability distribution point set The calculation formula is: ； in, This represents the sound field inverse mapping operator; Indicates a horizontal angle. Indicates a vertical angle; Cluster analysis is performed on the probability distribution point set to obtain the target contour.

9. The underwater target detection method based on dynamic channel impulse response analysis as described in any one of claims 1 to 8, characterized in that, The method also includes the following steps: After obtaining several target contours, the centroid velocity of the target contours is determined based on their coordinates and sampling period. Determine the Doppler frequency shift component based on the received signal from the target; The velocity vector of the target profile is obtained based on the centroid velocity and Doppler frequency shift component of the target profile.

10. The underwater target detection method based on dynamic channel impulse response analysis as described in claim 9, characterized in that: The centroid velocity of the target contour The calculation formula is: ； in The centroid coordinates representing the current target contour. Represents the sampling period. The centroid coordinates represent the target contour from the previous sampling. The Doppler frequency shift component The calculation formula is: ； Where c represents the speed of sound; This represents the carrier phase difference between two consecutive frames of signal; Represents the center frequency of the signal; Represents the pulse interval; This represents the angle between the target's trajectory and the sonar radial direction.

Images