A time-frequency domain interference suppression method and device combined with morphological processing

Abstract

The present application relates to a kind of time-frequency domain interference suppression method and device of combined morphological processing, comprising: the echo signal of the target received is carried out short-time Fourier transform, and time-frequency diagram is obtained;Initial interference mask is obtained according to time-frequency diagram;Initial interference mask is handled by inflation operation, and target interference mask is obtained;Time-frequency diagram is handled by target interference mask, and the time-frequency diagram after interference suppression is obtained.The present application obtains the time-frequency diagram of echo signal by short-time Fourier transform, and interference mask is obtained in combination with time-frequency diagram, so that the interference edge lower than target echo energy in time-frequency diagram is effectively suppressed, high interference suppression ratio is realized, so that radar can identify and detect real target, the suppression ability of interference edge is improved by inflation operation, interference suppression ability is improved, and high interference suppression ratio and low signal loss are combined.

Description

Technical Field

[0001] This invention relates to the field of radar interference suppression technology, and more specifically, to a time-frequency domain interference suppression method and apparatus that combines morphological processing. Background Technology

[0002] In modern radar electronic warfare systems, with the development of digital radio frequency storage technology, jammers can flexibly set active jamming parameters, enabling the generation of diverse false targets on the RD (Range-Doppler) map. Because active jamming can achieve matched filtering and coherent processing gains, its jamming effect is significant, making target detection in active jamming environments extremely difficult. Therefore, to improve radar anti-jamming capabilities, echo signals are typically mapped to higher dimensions to extract the characteristic differences between active jamming and the actual signal. Filters are then constructed to remove the jamming, preserving the true target signal, thus allowing the radar to detect the real target.

[0003] Existing methods mainly fall into two categories: one type extracts interference parameters in the time-frequency domain and suppresses interference through interference reconstruction / time-frequency filtering. However, this type of method is computationally intensive and struggles to handle complex and varied interference types. The other type utilizes the difference between interference and signal energy to construct a time-frequency domain filter. However, because the difference between interference edge energy and signal energy is small, existing methods struggle to achieve both high interference suppression ratio and low signal loss. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, this invention provides a time-frequency domain interference suppression method and apparatus that combines morphological processing.

[0005] According to a first aspect of the present invention, a time-frequency domain interference suppression method combining morphological processing is provided, the method comprising:

[0006] Perform a short-time Fourier transform on the received echo signal from the target location to obtain the time-frequency diagram;

[0007] The initial interference mask is obtained based on the time-frequency diagram.

[0008] The initial interference mask is processed by dilation to obtain the target interference mask;

[0009] The time-frequency map is processed by the target interference mask to obtain the time-frequency map after interference suppression.

[0010] Optionally, obtaining the initial interference mask based on the time-frequency diagram includes:

[0011] Obtain the probability density distribution function of the time-frequency graph;

[0012] The initial interference mask is obtained based on the probability density distribution function of the time-frequency diagram.

[0013] Optionally, obtaining the probability density distribution function of the time-frequency graph includes:

[0014] The energy of the time-frequency diagram is obtained from the time-frequency diagram.

[0015] The energy of the time-frequency graph is normalized to obtain the energy of the normalized time-frequency graph;

[0016] The probability density distribution function is obtained from the energy of the normalized time-frequency graph.

[0017] Optionally, obtaining the initial interference mask based on the probability density distribution function of the time-frequency diagram includes:

[0018] The target segmentation threshold between interference and other energy in the echo signal is obtained based on the probability density distribution function of the time-frequency diagram.

[0019] An initial interference mask is obtained using the target segmentation threshold.

[0020] Optionally, the probability density distribution function is represented as follows:

[0021] Q(r) = imhist(E(t,f));

[0022] Where Q(r) represents the probability density distribution function, imhist(·) represents the quantization of E(t,f), E(t,f) represents the energy of the normalized time-frequency graph, t represents time, and f represents frequency.

[0023] Optionally, the target segmentation threshold is represented as follows:

[0024] Ψ′=argminQ(r);

[0025] Where Ψ′ represents the target segmentation threshold, Q(r) represents the probability density distribution function, and r s <r≤r j r j r represents the energy peak of the disturbance distribution on the probability density function. s The peak value of the target distribution within noise and interference.

[0026] Optionally, the time-frequency diagram after interference suppression is represented as follows:

[0027] TF anti (t,f)=~mask t,f (t,f)*TF(t,f);

[0028] Among them, TFanti (t,f) represents the time-frequency diagram after interference suppression, mask t,f (t,f) represents the target interference mask obtained after performing dilation operation on the initial interference mask, TF(t,f) represents the time-frequency diagram, ~ represents inversion, and * represents Hadamard product.

[0029] According to a second aspect of the present invention, a time-frequency domain interference suppression device combining morphological processing is provided, the device comprising:

[0030] The transformation module is used to perform a short-time Fourier transform on the received echo signal from the target to obtain a time-frequency diagram;

[0031] The mask acquisition module is used to obtain an initial interference mask based on the time-frequency diagram;

[0032] The dilation module is used to process the initial interference mask through dilation operations to obtain the target interference mask;

[0033] An interference suppression module is used to process the time-frequency map using the target interference mask to obtain an interference-suppressed time-frequency map.

[0034] The technical solution provided by this invention may include the following beneficial effects:

[0035] This invention obtains the time-frequency map of the echo signal by performing a short-time Fourier transform, and combines the time-frequency map to obtain an interference mask, so that the interference edges in the time-frequency map that are lower than the target echo energy are effectively suppressed, achieving a high interference suppression ratio. This enables the radar to identify and detect the real target. The invention also improves the interference edge suppression capability through dilation calculation, thus improving the interference suppression capability and achieving both a high interference suppression ratio and low signal loss.

[0036] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0037] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings:

[0038] Figure 1 This is a flowchart illustrating a time-frequency domain interference suppression method using combined morphological processing, according to an exemplary embodiment.

[0039] Figure 2 This is a schematic diagram illustrating a probability density distribution function according to an exemplary embodiment.

[0040] Figure 3aThis is a schematic diagram of the time-frequency domain of the signal before interference suppression, according to an exemplary embodiment.

[0041] Figure 3b This is a schematic diagram illustrating the location of interference according to an exemplary embodiment.

[0042] Figure 3c This is a schematic diagram of the time-frequency domain of a signal after interference suppression by conventional methods, according to an exemplary embodiment.

[0043] Figure 4a This is a schematic diagram illustrating the location of interference after being determined by an initial segmentation threshold, according to an exemplary embodiment.

[0044] Figure 4b This is a schematic diagram illustrating the location of interference after being determined by a target segmentation threshold, according to an exemplary embodiment.

[0045] Figure 4c This is a schematic diagram of the time-frequency domain of a signal after interference suppression by the present invention, according to an exemplary embodiment.

[0046] Figure 5a This is a schematic diagram illustrating the coherent accumulation result of the echo signal according to an exemplary embodiment.

[0047] Figure 5b This is a schematic diagram illustrating the result of coherent accumulation after interference suppression using conventional methods, according to an exemplary embodiment.

[0048] Figure 5c This is a schematic diagram illustrating the result of coherent accumulation after interference suppression by the present invention, according to an exemplary embodiment.

[0049] Figure 6 This is a block diagram illustrating a time-frequency domain interference suppression device with combined morphological processing according to an exemplary embodiment. Detailed Implementation

[0050] Figure 1 This is a flowchart illustrating a time-frequency domain interference suppression method using joint morphological processing according to an exemplary embodiment, such as... Figure 1 As shown, the method includes the following steps:

[0051] S101. Perform a short-time Fourier transform on the received echo signal at the target location to obtain a time-frequency diagram.

[0052] It is understandable. However, the idea of ​​the short-time Fourier transform is to select a time-frequency localized window function, assume that the analysis window function W(t) is stationary within a short time interval, and move the window function so that S(t)W(t) is a stationary signal within different finite widths, thereby calculating the power spectrum at different times. S(t) represents the echo signal at the target, which includes the target signal, interference signal and Gaussian white noise.

[0053] A time-frequency graph can be represented as follows:

[0054]

[0055] Where η is obtained by the short-time Fourier transform of f, e is the natural exponent, j is the imaginary unit, t represents time, and f represents frequency.

[0056] S102. Obtain the initial interference mask based on the time-frequency diagram.

[0057] Optionally, S102 may include:

[0058] Obtain the probability density distribution function of the time-frequency plot;

[0059] The initial interference mask is obtained based on the probability density distribution function of the time-frequency diagram.

[0060] Optionally, the probability density distribution function of the time-frequency graph is obtained, including:

[0061] The energy of the time-frequency diagram is obtained from the time-frequency diagram.

[0062] The energy of the time-frequency graph is normalized to obtain the energy of the normalized time-frequency graph;

[0063] The probability density distribution function is obtained from the energy of the normalized time-frequency graph.

[0064] Understandably, the energy R(t,f) of the time-frequency graph can be expressed as follows:

[0065] R(t,f)=10log 10 (TF(t,f));

[0066] The energy of the time-frequency graph is normalized, and the normalized energy E(t,f) of the time-frequency graph is expressed as follows:

[0067]

[0068] Optionally, the probability density distribution function of each energy in the image can be calculated from the energy E(t,f) of the normalized time-frequency image, as shown below:

[0069] Q(r) = imhist(E(t,f));

[0070] Where Q(r) represents the probability density distribution function, and imhist(·) represents quantizing E(t,f) to a value between 1 and 256. Then, the frequency of each quantized data point is counted to obtain the probability density distribution function of the energy in the time-frequency graph. Figure 2 This is a schematic diagram of a probability density distribution function according to an exemplary embodiment, where E(t,f) represents the energy of the time-frequency graph after normalization.

[0071] Optionally, an initial interference mask is obtained based on the probability density distribution function of the time-frequency plot, including:

[0072] The target segmentation threshold between interference and other energy in the echo signal is obtained based on the probability density distribution function of the time-frequency diagram.

[0073] An initial interference mask is obtained using the target segmentation threshold.

[0074] It is understandable that, such as Figure 2 As shown, this probability density distribution function can be considered as the superposition of the probability density distribution functions of noise, target, and interference energies, assuming that each type of signal follows a gamma distribution. In the time-frequency graph, the noise region accounts for the highest proportion; therefore, the mode of the noise energy is the peak value of the probability density distribution function, as shown. Figure 2 As shown by the red dot, denoted as r n The interference energy is higher than that of the target and noise; therefore, the interference energy is distributed within the energy peak range of the probability density distribution function, such as... Figure 2 The area shown in green is denoted as r. j The targets are mostly distributed within the peak range of noise and interference, with the peak value denoted as r. s According to the minimum error Bayesian decision, the optimal separation threshold Ψ between the disturbance and the remaining energy is:

[0075] argmin[P(R∈jam|R(t,f)≤Ψ)+P(R∈else|R(t,f)>Ψ)];

[0076] The optimal threshold cannot be obtained directly. Instead, a suboptimal segmentation threshold can be found as the target segmentation threshold Ψ′. At the target segmentation threshold, the energy probability distribution of the time-frequency map reaches a local minimum, thereby making the error rate lower than 2Q(Ψ′).

[0077] Optionally, the target segmentation threshold can be represented as follows:

[0078] Ψ′=argminQ(r);

[0079] Where Ψ′ represents the target segmentation threshold, Q(r) represents the probability density distribution function, and r s <r≤r j rj r represents the energy peak of the disturbance distribution on the probability density function. s The peak value of the target distribution within noise and interference.

[0080] It is understandable that the initial interference mask is obtained by using the target segmentation threshold Ψ′, where Φ(t,f) is 1 to indicate interference, and 0 to indicate other types.

[0081]

[0082] S103. The initial interference mask is processed by dilation operation to obtain the target interference mask.

[0083] It is understandable that the initial interference mask obtained by using the target segmentation threshold can locate the area of ​​strong interference while ensuring the false positive rate. However, since the energy of the interference edge in the time-frequency map is weak and may be lower than the signal energy, there are limitations to using only the energy difference in the time-frequency map to distinguish between interference and other signals.

[0084] Dilation, as a mathematical morphological operation, allows a target region to expand outwards in a predetermined manner. Since the interference edge is located at the edge of a strong interference region, and dilation can expand the initial interference mask range, it can cover the weak-energy region at the interference edge, thereby improving the interference suppression ratio.

[0085] The expansion operation expands the data space using an n×n all-zero struct element c:

[0086]

[0087] (f+y)∈D Φ And (x,y)∈D c}

[0088] Among them, mask t,f (t,f) represents the target interference mask after performing dilation operation on Φ(t,f); Indicates the expansion operation; D Φ D represents the domain of the data space; c This indicates the domain of the structuring element c.

[0089] Specifically, the initial interference mask Φ(t,f) is first dilated along the time dimension. The signal energy outside the interference time dimension edge located by the time-frequency plot is progressively searched and determined outwards. When this energy equals the energy value corresponding to the peak of the noise distribution, it indicates that the interference edge has been fully identified, and the neighborhood length is extracted. Based on the extracted neighborhood length, a linear structuring element about the data symmetry center is created, and then dilation is performed to obtain the result of the initial interference mask Φ(t,f) after time dimension dilation, which is called mask. t(t,f).

[0090] Secondly, the interference masking results after time-dilation. t (t,f) undergoes frequency-dimensional dilation, analogous to the time-dimensional dilation method. Similarly, a linear structuring element about the data symmetry center is created based on the extracted neighborhood length, and dilation is performed to obtain the target interference mask result as mask. t,f (t,f).

[0091] S104. The time-frequency map is processed by the target interference mask to obtain the time-frequency map after interference suppression.

[0092] Optionally, the time-frequency diagram after interference suppression can be represented as follows:

[0093] TF anti (t,f)=~mask t,f (t,f)*TF(t,f);

[0094] Among them, TF anti (t,f) represents the time-frequency diagram after interference suppression, mask t,f (t,f) represents the target interference mask obtained after performing dilation operation on the initial interference mask, TF(t,f) represents the time-frequency plot, ~ represents inversion, and * represents Hadamard product.

[0095] Figure 3a This is a schematic diagram of the time-frequency domain of the signal before interference suppression, according to an exemplary embodiment. Figure 3b This is a schematic diagram illustrating the location of interference according to an exemplary embodiment. Figure 3c This is a schematic diagram in the time-frequency domain of a signal after interference suppression using conventional methods, according to an exemplary embodiment. Figure 4a This is a schematic diagram illustrating the location of interference after initial segmentation threshold determination, according to an exemplary embodiment. Figure 4b This is a schematic diagram illustrating the location of interference after being determined by a target segmentation threshold, according to an exemplary embodiment. Figure 4c This is a schematic diagram in the time-frequency domain of a signal after interference suppression according to an exemplary embodiment of the present invention. Figure 5a This is a schematic diagram illustrating the coherent accumulation result of the echo signal according to an exemplary embodiment. Figure 5b This is a schematic diagram illustrating the result of coherent accumulation after interference suppression using conventional methods, according to an exemplary embodiment. Figure 5c This is a schematic diagram illustrating the coherent accumulation result after interference suppression using the present invention, according to an exemplary embodiment. The algorithm calculates the target segmentation threshold to accurately locate the interference position, uses the dilation algorithm in mathematical morphology to increase the suppression rate of the interference edges, and then suppresses the interference by setting a filter. The result is shown in the figure below. Figure 4cAs shown, the time-domain representation of the interference-suppressed echo signal is obtained through short-time inverse Fourier transform, and the final result is as follows. Figure 5c As shown, comparison Figure 3b and Figure 4c It can be seen that the method proposed in this invention can accurately locate the interference position, and the dilation operation can suppress the interference more completely, with a significantly better suppression effect than the traditional method; comparing the final coherent accumulation results of the two, as shown... Figure 5b and Figure 5c As shown, traditional interference suppression algorithms result in significant loss to the target signal and do not completely suppress interference. In contrast, the method proposed in this invention results in very little loss to the target signal and almost complete suppression of interference.

[0096] Figure 6 This is a block diagram illustrating a time-frequency domain interference suppression device with combined morphological processing according to an exemplary embodiment, such as... Figure 6 As shown, the device 600 may include:

[0097] The transformation module 601 is used to perform a short-time Fourier transform on the received echo signal at the target to obtain a time-frequency diagram;

[0098] The mask acquisition module 602 is used to obtain an initial interference mask based on the time-frequency diagram.

[0099] The dilation module is used to process the initial interference mask through dilation operations to obtain the target interference mask 603;

[0100] The interference suppression module 604 is used to process the time-frequency map through the target interference mask to obtain the time-frequency map after interference suppression.

[0101] The preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the specific details of the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, and these simple modifications all fall within the protection scope of the present invention.

[0102] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

[0103] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed by the present invention.

Claims

1. A time-frequency domain interference suppression method combining morphological processing, characterized in that, The method includes: Perform a short-time Fourier transform on the received echo signal from the target location to obtain the time-frequency diagram; The initial interference mask is obtained based on the time-frequency diagram. The initial interference mask is processed by dilation to obtain the target interference mask; The time-frequency map is processed using the target interference mask to obtain the time-frequency map after interference suppression; The step of obtaining the initial interference mask based on the time-frequency diagram includes: Obtain the probability density distribution function of the time-frequency graph; The initial interference mask is obtained based on the probability density distribution function of the time-frequency diagram; The step of obtaining the probability density distribution function of the time-frequency graph includes: The energy of the time-frequency diagram is obtained from the time-frequency diagram. The energy of the time-frequency graph is normalized to obtain the energy of the normalized time-frequency graph; The probability density distribution function is obtained from the energy of the normalized time-frequency graph; The step of obtaining the initial interference mask based on the probability density distribution function of the time-frequency diagram includes: The target segmentation threshold between interference and other energy in the echo signal is obtained based on the probability density distribution function of the time-frequency diagram. The initial interference mask is obtained using the target segmentation threshold. The probability density distribution function is represented as follows: ； in, Denotes the probability density distribution function. Indicates to Quantify, This represents the energy of the time-frequency graph after normalization. Indicates time, Indicates frequency; The target segmentation threshold is represented as follows: ； in, This represents the target segmentation threshold. Denotes the probability density distribution function. , The energy peak of the disturbance distribution lies in the probability density function. The peak value of the target distribution within noise and interference.

2. The time-frequency domain interference suppression method with combined morphological processing according to claim 1, characterized in that, The time-frequency diagram after interference suppression is shown below: ； in, This represents the time-frequency diagram after interference suppression. This represents the target interference mask obtained after performing a dilation operation on the initial interference mask. This refers to the time-frequency diagram. Indicates negation, This indicates the execution of the Hadamarda accumulation.

3. A time-frequency domain interference suppression device combining morphological processing, characterized in that, The device includes: The transformation module is used to perform a short-time Fourier transform on the received echo signal from the target to obtain a time-frequency diagram; The mask acquisition module is used to obtain an initial interference mask based on the time-frequency diagram; The dilation module is used to process the initial interference mask through dilation operations to obtain the target interference mask; An interference suppression module is used to process the time-frequency map through the target interference mask to obtain an interference-suppressed time-frequency map. The step of obtaining the initial interference mask based on the time-frequency diagram includes: Obtain the probability density distribution function of the time-frequency graph; The initial interference mask is obtained based on the probability density distribution function of the time-frequency diagram; The step of obtaining the probability density distribution function of the time-frequency graph includes: The energy of the time-frequency diagram is obtained from the time-frequency diagram. The energy of the time-frequency graph is normalized to obtain the energy of the normalized time-frequency graph; The probability density distribution function is obtained from the energy of the normalized time-frequency graph; The step of obtaining the initial interference mask based on the probability density distribution function of the time-frequency diagram includes: The target segmentation threshold between interference and other energy in the echo signal is obtained based on the probability density distribution function of the time-frequency diagram. The initial interference mask is obtained using the target segmentation threshold. The probability density distribution function is represented as follows: ； in, Denotes the probability density distribution function. Indicates to Quantify, This represents the energy of the time-frequency graph after normalization. Indicates time, Indicates frequency; The target segmentation threshold is represented as follows: ； in, This represents the target segmentation threshold. Denotes the probability density distribution function. , The energy peak of the disturbance distribution lies in the probability density function. The peak value of the target distribution within noise and interference.

Images