Active sonar imaging data online reverberation elimination and target detection method

Abstract

The invention discloses an active sonar imaging data online reverberation elimination and target detection method. The method comprises the steps that a sonar obtains a data frame at the current moment; judging whether the current state is an initialization stage or not; updating a reverberation base by using the current observation frame and the target component; the reverberation subspace weight is updated by the updated reverberation signal base; re-estimating a target component, and taking the target component as estimation of a detection target to realize online real-time reverberation elimination and target detection; according to the method, the problems of high calculation complexity, large resource consumption and the like caused by the fact that a conventional low-rank-sparse decomposition method does not consider time sequence data characteristics are solved.

Description

Technical Field

[0001] The invention relates to the field of underwater acoustic information processing and electronic information technology, in particular to an active sonar imaging data online reverberation elimination and target detection method. Background Art

[0002] Ensuring underwater port safety is of great significance to my country's coastal areas. In active detection in shallow waters, typical targets include small submersibles and frogmen. These targets pose a serious threat to surface ships, ports, docks and military facilities due to their good concealment and strong surprise maneuverability. Active sonar is the main means of detecting intrusion targets such as underwater frogmen and small unmanned vehicles. Usually, active sonar works in a noisy or reverberant environment, where reverberation is formed by the scattering superposition of active detection signals by a large number of ocean scatterers. In shallow water environments, seabed reverberation is the main component of reverberation. Since the strong scattering signal of the seabed is very similar to the target echo signal, the strong reverberation reduces the performance of the active sonar detection system, so it is difficult to effectively detect the target through a single frame of sonar data. The low-rank-sparse decomposition method based on multiple imaging frames is an effective means of reverberation suppression and target signal detection. This method decomposes the steady-state part of the reverberation into a low-rank matrix, and decomposes the target signal and the reverberation fluctuation part into a sparse matrix. However, existing technologies for eliminating reverberation in sonar imaging based on low-rank-sparse decomposition mostly use offline data models, which store a large amount of sonar data as multiple frames of images, and at each moment, it is necessary to re-perform low-rank-sparse decomposition using newly acquired data frames and stored data frames. This method does not consider the temporal nature of sonar image data acquisition, has high complexity, and is not conducive to the low-complexity deployment of sonar target detection algorithms. Therefore, it is of great significance to establish an online and real-time reverberation suppression algorithm based on low-rank-sparse decomposition for the collected data. Summary of the invention

[0003] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide an online reverberation elimination and target detection method for active sonar imaging data. The present invention solves the problems of high computational complexity and large resource consumption caused by the conventional low-rank-sparse decomposition method not considering the characteristics of time series data.

[0004] To achieve the above object, the technical solution adopted by the present invention is: an online reverberation elimination and target detection method for active sonar imaging data, comprising the following steps:

[0005] Step 1, sonar obtains the current time data frame;

[0006] Step 2, determine whether it is currently in the initialization stage: when it is in the initialization stage, determine whether a sufficient number of data frames for initialization have been accumulated, if not, continue to obtain data frames; if enough data frames have been obtained, calculate the mean of the data frames, and remove the mean of the data frames; regard the data in the initialization stage as reverberation background data, and calculate its feature space dimension and basis; when it is not in the initialization stage, if the initialization has just ended and the current frame is the first frame after initialization, subtract the estimated reverberation background data from the data frame to preliminarily estimate the target component; if it is not the first frame after initialization, use the result of the previous moment to preliminarily estimate the target component;

[0007] Step 3: Update the reverberation base using the current observation frame and the target component; and update the reverberation subspace weight using the updated reverberation signal base;

[0008] Step 4: Re-estimate the target component and use it as the estimation of the detected target to achieve online real-time reverberation elimination and target detection; obtain the next data frame and repeat steps 1 to 4.

[0009] As a further improvement of the present invention, the step 1 is specifically as follows:

[0010] In the sonar system using a circular receiving array, after the transmitting element emits a linear frequency modulation signal, the reflected signals at different angles are collected through beamforming and related matching is performed; for each set receiving angle, a vector of length P is formed according to the echo delay; the above process is performed for a total of Q angles to form a P×Q azimuth-range sonar imaging matrix; if the above operation is continued, a series of azimuth-range sonar imaging matrices at different times are formed, that is, a frame of data; the azimuth-range matrix at a specific time t is recorded as M t ; Concatenate the columns of the matrix to form a PQ×1 sonar data vector m t .

[0011] As a further improvement of the present invention, in step 2, after the sonar is turned on, the initialization phase starts from time t=1 and continuously receives data, which can accumulate T c The batch initialization algorithm is used to estimate the reverberation background and parameters based on the frame data.

[0012] As a further improvement of the present invention, a batch initialization algorithm is used to estimate the reverberation background and parameters as follows:

[0013] The sonar data vector m t (t=1,…,T c ) are concatenated into a dimension of PQ×T c The matrix In the initialization phase, it is assumed that there is no target in the scene, so the data is meaned by column, and then the signal subspace is determined by singular decomposition;

[0014] Build T c ×T c Matrix of size Perform eigendecomposition on the smaller dimensional matrix to obtain:

[0015]

[0016] Total T c Group eigenvalues ​​and eigenvectors; Assume that the eigenvalues ​​are sorted from large to small, and obtain the first B eigenvalues ​​and corresponding eigenvectors that occupy a total ratio r (i=1,…,B); by Eigenvalues ​​and eigenvectors, obtain Z c The i-th group of singular values ​​and singular vectors:

[0017]

[0018]

[0019]

[0020] Construct the initialization of the reverberation background component of the sonar:

[0021]

[0022] As a further improvement of the present invention, in step 4, the online real-time reverberation elimination and target detection are specifically as follows:

[0023] At time t, the sonar observation data vector m t It can be expressed as:

[0024] m t =U t w t +g t +n t

[0025] Among them U t is the subspace where the reverberation signal is located, and its dimension is PQ×B, where B is determined by initialization; w t is the coefficient of the B×1 subspace; the vector g of PQ×1 t Represents the components of sparse targets; PQ×1 vector n t Represents the model noise component; online reverberation elimination uses the result U of the previous moment t-1 、w t-1 , g t-1 And the observation data m at this momentt Update variable U t 、w t and t ;

[0026] The following mathematical optimization objective function is used for online sonar reverberation elimination and target detection:

[0027]

[0028] The objective function is divided into three parts. The first part describes the accuracy of reconstructing the observed sonar data frame using the low-rank-sparse signal model at time t. The second part promotes the sparse characteristics of the target component. The third part promotes the closeness of the reverberation component subspace at time t to that at time t-1. t ∈Θ indicates that U t It should be located in a standard orthogonal subspace, q is a weighted value related to the pixel position, and non-negative η and τ are used to control the intensity of the second and third terms. After collecting a new data frame at time t, the time smoothing method is used to estimate the mean and perform de-meaning. The reverberation component elimination and target signal detection are performed online using the current sonar data observation frame and the reverberation subspace result estimated at the previous moment.

[0029] As a further improvement of the present invention, the online reverberation component elimination and target signal detection using the current sonar data observation frame and the reverberation subspace result estimated at the previous moment specifically include:

[0030] According to the reverberation background of the previous moment, that is, the reverberation subspace basis U t-1 and coefficient w t-1 Preliminary estimate of the target signal, namely:

[0031]

[0032] Available If the current frame is the first frame after initialization, L c Alternative U t-1 w t-1 ;

[0033] Update the reverberation subspace basis U by gradient method t-1 Get the updated value U′ t :

[0034]

[0035] Then U′ t Orthogonalize and normalize to get U t =Θ(U' t );

[0036] Solve the reverberation subspace coefficient w by solving the quadratic optimizationt :

[0037]

[0038] available

[0039] According to the updated U t 、w t Again, estimate the target signal g t :

[0040]

[0041] available

[0042] As a further improvement of the present invention, it also includes:

[0043] Each collection T e After the frame sonar data is collected and online calculations are performed, the initialization process is called again and the reverberation space dimensions and base are recalculated using an offline algorithm.

[0044] As a further improvement of the present invention, it also includes:

[0045] In a continuous period of time T s The reverberation subspace basis is not updated, and the reverberation estimation is only fine-tuned by adjusting the subspace weights.

[0046] The beneficial effects of the present invention are:

[0047] The present invention solves the problems of high computational complexity and large resource consumption caused by conventional low-rank-sparse decomposition methods that do not take into account the characteristics of time series data. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic diagram of signal construction in an embodiment of the present invention;

[0049] Figure 2 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0051] Example

[0052] An online reverberation elimination and target detection method for active sonar imaging data, comprising:

[0053] S1: Sonar obtains the current data frame;

[0054] S2: Determine whether it is currently in the initialization stage;

[0055] When in the initialization phase:

[0056] S2.1: Determine whether a sufficient number of data frames for initialization have been accumulated. If not, proceed to S1 to obtain data frames;

[0057] S2.2: If enough data frames have been obtained, the mean of the data frames is calculated, and the mean of the data frames is removed;

[0058] S2.3: Consider the data in the initialization phase as reverberation background data, and calculate its feature space dimension and basis;

[0059] When not in the initialization phase:

[0060] S3.1: If the initialization has just been completed and the current frame is the first frame after the initialization, the reverberation background estimated in S2.3 is subtracted from the data frame to preliminarily estimate the target component;

[0061] S3.2: If it is not the first frame after initialization, use the result of the previous moment to preliminarily estimate the target component;

[0062] S4: Update the reverberation base using the current observation frame and the target component in S3.1 or S3.2;

[0063] S5: Update the reverberation subspace weights using the current observation frame, the target component in S3.1 or S3.2, and the reverberation signal basis updated in S4;

[0064] S6: using the result in S5, re-estimate the target component and use this as the estimation of the detected target;

[0065] S7: Get the next data frame and repeat the process starting from S1.

[0066] The present embodiment is further described below:

[0067] Figure 1 101 is a schematic diagram of signal construction. As shown in 101, the sonar system using a circular receiving array, after the transmitting element emits a linear frequency modulation signal, collects the reflected signals at different angles through beam forming and performs correlation matching. For each set receiving angle, a vector of length P can be formed according to the echo delay. As shown in 102, the above process is performed on Q angles in total, and a P×Q azimuth-range sonar imaging matrix as shown in 103 can be formed. If the above operation is performed continuously, a series of azimuth-range sonar imaging matrices at different times can be formed, that is, a frame of data. The azimuth-range matrix at a specific time t is recorded as M t The columns of the matrix can be concatenated to form a PQ×1 vector m shown in 104. t .

[0068] Figure 2It is a full diagram of the implementation process of this embodiment.

[0069] Initialization phase of the algorithm: After the sonar is turned on, from time t = 1, if 201 continuously receives data, it can first accumulate T c The data of the frame is initialized by a batch algorithm (as shown in 202 and 203) to estimate the reverberation background and parameters. c The value of is related to the parameters P and Q of the sonar imaging settings and the dynamic characteristics of the reverberation. Preferably, it is recommended to be no less than 20.

[0070] First, the sonar data vector m t (t=1,…,T c ) are concatenated into a dimension of PQ×T c The matrix In the initialization phase, it can be assumed that there is no target in the scene, so the data can be meaned by column as in 204, and then the signal subspace is determined by conventional singular decomposition, but it is noted that PQ is usually much larger than T c , the following fast algorithm is used to determine step 205.

[0071] Build T c ×T c Matrix of size Perform eigendecomposition on the smaller dimensional matrix to obtain:

[0072]

[0073] Total T c The singular values ​​and singular vectors of the group are not general. It is assumed that the eigenvalues ​​are sorted from large to small, and the first B eigenvalues ​​and corresponding eigenvectors that occupy a total ratio r (preferably r = 95%) are obtained. By these The eigenvalues ​​and eigenvectors of Z can be obtained as follows c The i-th group of singular values ​​and singular vectors:

[0074]

[0075]

[0076]

[0077] The above quantities can be used to construct the initialization of the reverberation background component of the sonar:

[0078]

[0079] In an underwater environment, B is usually small, so the above method can quickly initialize the algorithm with low complexity. Online real-time reverberation elimination and target detection: At time t, the sonar observation data vector m t It can be expressed as:

[0080] m t =U t w t +g t +n t

[0081] Among them U t is the subspace where the reverberation signal is located, and its dimension is PQ×B, where B is determined by initialization; w t is the coefficient of the B×1 subspace; the vector g of PQ×1 t Represents the components of sparse targets; PQ×1 vector n t Represents the model noise component. Online reverberation elimination aims to use the result U of the previous moment t-1 、w t-1 , g t-1 And the observation data m at this moment t Update variable U t 、w t and t , thereby estimating the reverberation and the target using less data rather than all the accumulated data.

[0082] This embodiment proposes to use the following mathematical optimization objective function for online sonar reverberation elimination and target detection:

[0083]

[0084] The objective function is divided into three parts. The first part describes the accuracy of reconstructing the observed sonar data frame using the low-rank-sparse signal model at time t, the second part promotes the sparse characteristics of the target component, and the third part promotes the closeness of the reverberation component subspace at time t to that at time t-1. t ∈Θ indicates that U t It should be located in a standard orthogonal subspace, q is a weighted value related to the pixel position, and non-negative η and τ are used to control the strength of the second and third terms. After a new data frame is collected at time t, a time smoothing method is used to estimate the mean, and the mean is removed (206). The following steps can be used to online remove the reverberation component and detect the target signal using the current sonar data observation frame and the reverberation subspace result estimated at the previous moment.

[0085] Step 1 (208, 209): Based on the reverberation background at the previous moment, i.e., the reverberation subspace basis U t-1 and coefficient w t-1 The target signal (209) is preliminarily estimated, namely:

[0086]

[0087] available If the current frame is the first frame after initialization, L c Alternative U t-1 w t-1 (208).

[0088] Step 2 (210): Update the reverberation subspace basis U by gradient method t-1 Get the updated value U′ t :

[0089]

[0090] Then U′ t Orthogonalize and normalize to get U t = Θ(U′ t ).

[0091] Step 3 (211): Solve the reverberation subspace coefficient w by solving the quadratic optimization t :

[0092]

[0093] available

[0094] Step 4 (212): Based on the updated U t 、w t Update the estimated target signal g again t :

[0095]

[0096] available

[0097] In order to further reduce the complexity of the algorithm and increase the stability of the algorithm, the following adjustments can be made in the above method, including:

[0098] Method 1: Every collection T e After the frame sonar data is collected and online calculation is performed, the initialization process is called again, and the offline algorithm is used to recalculate the reverberation space dimension and base to avoid the deficiency of the online algorithm in tracking the reverberation subspace for a long time and increase the stability of the system. e Initialization is required at this time".

[0099] Method 2: Considering the sonar reverberation background as a slow-changing process, it can be calculated over a continuous period of time T sThe reverberation subspace base is not updated through step 210, and only the subspace weight is adjusted through 211 to fine-tune the reverberation estimation, thereby reducing the overall complexity of the target detection system. That is, a judgment of "whether T s If yes, module 209 is still connected to module 210; if no, module 209 is directly connected to module 211.

[0100] The above-mentioned embodiments only express the specific implementation of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for online reverberation elimination and target detection of active sonar imaging data, characterized in that: The following steps are involved: Step 1, sonar obtains the current time data frame; Step 2, determine whether it is currently in the initialization stage: when in the initialization stage, determine whether a sufficient number of data frames for initialization have been accumulated, if not, continue to obtain data frames; if enough data frames have been obtained, calculate the mean of the data frames, and remove the mean of the data frames; The data in the initialization stage is regarded as the reverberation background data, and its feature space dimension and basis are calculated; when it is not in the initialization stage, such as when the initialization has just ended and the current frame is the first frame after the initialization, the estimated reverberation background data is subtracted from the data frame to preliminarily estimate the target component; If it is not the first frame after initialization, use the result of the previous moment to preliminarily estimate the target component; Step 3: Update the reverberation base using the current observation frame and the target component; and updating the reverberation subspace weights using the updated reverberation signal basis; Step 4: Re-estimate the target component and use it as the estimation of the detected target to achieve online real-time reverberation elimination and target detection; obtain the next data frame and repeat steps 1 to 4.

2. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 1, characterized in that: The step 1 is specifically as follows: In the sonar system using a circular receiving array, after the transmitting element emits a linear frequency modulation signal, the reflected signals at different angles are collected through beamforming and related matching is performed; for each set receiving angle, a vector of length P is formed according to the echo delay; the above process is performed for a total of Q angles to form a P×Q azimuth-range sonar imaging matrix; if the above operation is continued, a series of azimuth-range sonar imaging matrices at different times are formed, that is, a frame of data; the azimuth-range matrix at a specific time t is recorded as M t ; Concatenate the columns of the matrix to form a PQ×1 sonar data vector m t .

3. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 2, characterized in that: In step 2, the initialization phase starts from t=1 after the sonar is turned on and continuously receives data, which can accumulate T c The batch initialization algorithm is used to estimate the reverberation background and parameters based on the frame data.

4. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 3 is characterized in that: The batch initialization algorithm is used to estimate the reverberation background and parameters as follows: The sonar data vector m t (t=1,…,T c ) are concatenated into a dimension of PQ×T c The matrix In the initialization phase, it is assumed that there is no target in the scene, so the data is meaned by column, and then the signal subspace is determined by singular decomposition; Build T c ×T c Matrix of size Perform eigendecomposition on the smaller dimensional matrix to obtain: Total T c Group eigenvalues ​​and eigenvectors; Assume that the eigenvalues ​​are sorted from large to small, and obtain the first B eigenvalues ​​and corresponding eigenvectors that occupy a total ratio r Depend on Eigenvalues ​​and eigenvectors, obtain Z c The i-th group of singular values ​​and singular vectors: Construct the initialization of the reverberation background component of the sonar:

5. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 4, characterized in that: In step 4, online real-time reverberation elimination and target detection are as follows: At time t, the sonar observation data vector m t It can be expressed as: m t =U t w t +g t +n t Among them U t is the subspace where the reverberation signal is located, and its dimension is PQ×B, where B is determined by initialization; w t is the coefficient of the B×1 subspace; the vector g of PQ×1 t Represents the components of sparse targets; PQ×1 vector n t Represents the model noise component; online reverberation elimination uses the result U of the previous moment t-1 、w t-1 , g t-1 And the observation data m at this moment t Update variable U t 、w t and t ; The following mathematical optimization objective function is used for online sonar reverberation elimination and target detection: The objective function is divided into three parts. The first part characterizes the accuracy of reconstructing the observed sonar data frame using the low-rank-sparse signal model at time t, the second part promotes the sparse characteristics of the target component, and the third part promotes the closeness of the reverberation component subspace at time t to that at time t-1. Among them U t ∈Θ indicates that U t It should be located in a standard orthogonal subspace, q is a weighted value related to the pixel position, and non-negative η and τ are used to control the strength of the second and third terms; After collecting a new data frame at time t, the time smoothing method is used to estimate the mean and perform de-meaning. The reverberation component elimination and target signal detection are performed online using the current sonar data observation frame and the reverberation subspace result estimated at the previous moment.

6. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 5, characterized in that: Using the current sonar data observation frame and the reverberation subspace result estimated at the previous moment to perform reverberation component elimination and target signal detection online specifically includes: According to the reverberation background of the previous moment, that is, the reverberation subspace basis U t-1 and coefficient w t-1 Preliminary estimate of the target signal, namely: Available If the current frame is the first frame after initialization, L c Alternative U t-1 w t-1 ; Update the reverberation subspace basis U by gradient method t-1 Get the updated value U' t : Then U' t Orthogonalize and normalize to get U t =Θ(U' t ); Solve the reverberation subspace coefficient w by solving the quadratic optimization t : available According to the updated U t 、w t Again, estimate the target signal g t : available 7. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 6, characterized in that: Also includes: Each collection T e After the frame sonar data is collected and online calculations are performed, the initialization process is called again and the reverberation space dimensions and base are recalculated using an offline algorithm.

8. The method for online reverberation elimination and target detection of active sonar imaging data according to claim 6, characterized in that: Also includes: In a continuous period of time T s The reverberation subspace basis is not updated, and the reverberation estimation is only fine-tuned by adjusting the subspace weights.

Images