Passive signal enhancement method, device and equipment under low signal-to-noise ratio and medium

Abstract

The application relates to a passive detection signal enhancement method, device, equipment and medium under low signal-to-noise ratio, which comprises the following steps: receiving passive detection signals by an interferometer antenna to perform preprocessing, obtaining time-frequency data and interference data respectively, adopting a noise unit CFAR segmentation method, utilizing the random characteristics of noise phase to segment interference phase data, obtaining noise regions and non-noise regions represented by binary images, calculating noise power estimation of the noise regions according to the time-frequency amplitude data, calculating the amplitude change dynamic range of the signal according to the noise regions and the non-noise regions based on the time-frequency amplitude data and the noise power estimation, generating a guide image through amplitude nonlinear compression, guiding and filtering the interference data by using the guide image, and obtaining an enhanced signal. Then, the enhanced signal can be used for target detection, so that the accuracy of target detection is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of passive detection sensing, in particular to a passive detection signal enhancement method and device under low signal-to-noise ratio, equipment and medium. BACKGROUND

[0002] In the current field of unmanned aerial vehicle (UAV) detection, passive detection, recognition and positioning of UAVs are carried out by passively receiving and processing electromagnetic signals emitted by the UAVs, which has become one of the most important technical means.

[0003] With the rapid development of UAV technology, its application scenarios are constantly expanding, which brings about a series of safety hazards, making effective detection of UAVs crucial. Compared with active detection technology, passive detection technology has unique advantages. It does not actively emit signals, but relies on receiving electromagnetic signals radiated by the UAV itself to carry out detection work, which gives it good concealment and does not interfere with other radio equipment around, effectively avoiding signal conflicts and exposure risks that may be caused by active signal emission.

[0004] However, UAV passive detection technology faces many challenges in practical application. In particular, small and light UAVs have low transmission power, generally only 20-30 dBmW. Moreover, the UAV radiation signals received by the passive detection device will decay with the square of the distance. Taking a 2.4 GHz, 20 dBm radiation power UAV as an example, at a distance of 5 km, the power received by the passive detection device will be lower than -100 dBm. Considering that most UAV communication signals are in the form of wideband pulses, their power spectral density is even smaller. This results in extremely weak UAV signals being received at a long distance, which are easily overwhelmed by noise.

[0005] In order to improve the effective range of UAV passive detection, signal detection under low signal-to-noise ratio becomes a key point to overcome. In such a low signal-to-noise ratio environment, the energy difference between signal and noise is extremely small, and traditional signal detection methods often fail to accurately distinguish between signal and noise, resulting in missed detection or false detection. Therefore, how to extract effective UAV signal features from weak signals in a complex electromagnetic environment to achieve accurate detection and recognition has become an important problem to be solved in current UAV passive detection technology. SUMMARY

[0006] Therefore, it is necessary to provide a passive detection signal enhancement method, device, equipment and medium under low signal-to-noise ratio, which can improve the target detection accuracy of passive detection signals.

[0007] A passive detection signal enhancement method under low signal-to-noise ratio, the method comprising:

[0008] Obtaining passive detection signals respectively received by interferometer antennas;

[0009] Pretreating the passive detection signals to respectively obtain time-frequency data and interference data;

[0010] By adopting a noise cell CFAR segmentation method, the random characteristics of noise phase are used to segment the interference phase data to obtain noise regions and non-noise regions represented by binary images, and the noise power estimation of the noise regions is calculated according to the time-frequency amplitude data;

[0011] Based on the time-frequency amplitude data and the noise power estimation, the amplitude variation dynamic range of the signal is calculated according to the noise regions and the non-noise regions respectively, and a guide image is generated by amplitude nonlinear compression;

[0012] The interference data is guided filtered by using the guide image to obtain an enhanced signal.

[0013] In one embodiment, the signals received by the interferometer antennas are down-converted to baseband IQ data, the IQ data is processed by using short-time Fourier transform to obtain time-frequency data, and the time-frequency data includes time-frequency amplitude data and time-frequency phase data;

[0014] The signals received by the pairs of interferometer antennas are conjugate correlation processed to obtain the interference data, and the interference data includes interference amplitude data and interference phase data.

[0015] In one embodiment, when the noise cell CFAR segmentation method is used to segment the interference phase data:

[0016] The interference phase data is differentially processed along the time dimension and unwound to obtain phase difference data, and the phase difference data is filtered by using a convolution window to obtain a multi-scale phase consistency feature matrix;

[0017] Pure noise phase data is simulated and generated, and multi-scale phase consistency features are extracted based on the pure noise phase data to obtain a pure noise feature matrix;

[0018] The mean and variance of the pure noise feature matrix are calculated to obtain a statistical parameter matrix, and the pure noise feature matrix is normalized and fused by using the statistical parameter matrix to obtain a two-dimensional fused pure noise feature matrix;

[0019] The mean and root mean square of the fused pure noise feature matrix are used to estimate the CFAR threshold in combination with the false alarm probability.

[0020] normalizing the multi-scale phase coherence feature matrix by using the statistical parameter matrix and feature fusion, to obtain a fused and normalized phase coherence feature matrix;

[0021] segmenting based on the fused and normalized phase coherence feature matrix by using the CFAR threshold, to obtain the noise region and the non-noise region.

[0022] In one embodiment, when normalizing the pure noise feature matrix or the multi-scale phase coherence feature matrix by using the statistical parameter matrix and feature fusion processing:

[0023] normalizing the pure noise feature matrix or the multi-scale phase coherence feature matrix according to the statistical parameter matrix, to obtain normalized data;

[0024] performing top-k average or maxout fusion processing on the normalized data along the feature channel direction, to obtain feature fused data.

[0025] In one embodiment, when generating a guide image:

[0026] obtaining a preliminary guide image by performing amplitude nonlinear compression mapping on the amplitude variation dynamic range;

[0027] synthesizing the guide image by directly using single-channel interference phase data and performing phase preservation in the preliminary guide image.

[0028] In one embodiment, using the guide image to guide filtering of the interference data to obtain an enhanced signal, including:

[0029] calculating a self-correlation covariance matrix of the guide image within a preset sliding window, and obtaining a self-correlation matrix determinant and a regularization parameter according to the self-correlation covariance matrix;

[0030] obtaining a cross-correlation matrix according to the self-correlation covariance matrix and complex form interference data;

[0031] performing guide filtering by using the self-correlation covariance matrix, the cross-correlation matrix, and the regularization parameter, to obtain the enhanced signal.

[0032] The application also provides a passive detection signal enhancement device under low signal-to-noise ratio, the device comprising:

[0033] a passive detection signal receiving module, configured to acquire passive detection signals, the passive detection signals being respectively received by an interferometer antenna;

[0034] a preprocessing module configured to preprocess the passive detection signal to obtain time-frequency data and interference data respectively;

[0035] a noise region segmentation module configured to segment the interference phase data by using a noise cell CFAR segmentation method based on the random characteristics of noise phase to obtain noise regions and non-noise regions represented by a binary graph, and calculate noise power estimation of the noise regions according to the time-frequency amplitude data;

[0036] a guide image generation module configured to calculate amplitude variation dynamic ranges of signals according to the noise regions and non-noise regions respectively based on the time-frequency amplitude data and the noise power estimation, and generate a guide image by amplitude nonlinear compression;

[0037] a signal enhancement module configured to perform guided filtering on the interference data by using the guide image to obtain an enhanced signal.

[0038] A computer device including a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the above passive detection signal enhancement method under low signal-to-noise ratio when the computer program is executed.

[0039] A computer readable storage medium having a computer program stored thereon, and the computer program implementing the steps of the above passive detection signal enhancement method under low signal-to-noise ratio when the computer program is executed by a processor.

[0040] The above passive detection signal enhancement method, device, equipment and medium under low signal-to-noise ratio, by receiving passive detection signals from an interferometer antenna for preprocessing to obtain time-frequency data and interference data respectively, by using a noise cell CFAR segmentation method, using the random characteristics of noise phase to segment the interference phase data, to obtain noise regions and non-noise regions represented by a binary graph, and calculating noise power estimation of the noise regions according to the time-frequency amplitude data, based on the time-frequency amplitude data and the noise power estimation, calculating the amplitude variation dynamic ranges of signals according to the noise regions and non-noise regions respectively, and generating a guide image by amplitude nonlinear compression, using the guide image to perform guided filtering on the interference data to obtain an enhanced signal. The enhanced signal can then be used for target detection to improve the accuracy of target detection. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 A flowchart of a passive detection signal enhancement method under low signal-to-noise ratio in one embodiment;

[0042] Figure 2 A schematic diagram of the time-frequency data of an interferometer before and after enhancement using the method in one experiment, wherein, Figure 2(a) represents a schematic diagram before the interferometer time-frequency data is enhanced by the method, Figure 2 (b) represents a schematic diagram after the interferometer time-frequency data is enhanced by the method;

[0043] Figure 3 a structural block diagram of a passive detection signal enhancement device under low signal-to-noise ratio in an embodiment;

[0044] Figure 4 an internal structural diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0046] As shown in Figure 1 a low signal-to-noise ratio passive detection signal enhancement method is provided, which specifically includes the following steps:

[0047] Step S100, obtaining a passive detection signal, which is respectively received by an interferometer antenna.

[0048] Step S110, pre-processing the passive detection signal to obtain time-frequency data and interference data respectively.

[0049] Step S120, using the random characteristics of noise phase to segment the interference phase data by using a noise cell CFAR segmentation method, obtaining noise regions and non-noise regions represented by binary images, and calculating noise power estimates of the noise regions according to the time-frequency amplitude data.

[0050] Step S130, based on the time-frequency amplitude data and the noise power estimates, calculating the amplitude variation dynamic range of the signal according to the noise regions and the non-noise regions respectively, and generating a guide image through amplitude nonlinear compression.

[0051] Step S140, using the guide image to guide filter the interference data to obtain an enhanced signal.

[0052] In view of the problem of long-distance weak signal detection faced by unmanned aerial vehicle passive detection, in the present application, an interferometer array detection system is used, the distribution characteristics of communication signals in time-frequency interference correlation data are fully considered, and the methods of image processing and signal processing are comprehensively used, especially the image guided filtering algorithm is used, to enhance the weak signal while adaptively maintaining the edge characteristics of the strong signal.

[0053] In step S100, the passive detection signal is the signal radiated by other UAVs received by the interferometer antenna arranged on the UAV.

[0054] In step S110, the signal received by the single interferometer antenna is down-converted to baseband IQ data, the IQ data is processed by using short-time Fourier transform to obtain time-frequency data, the time-frequency data includes time-frequency amplitude data and time-frequency phase data, wherein the video data is two-dimensional data. The conjugate correlation processing is performed on the signal received by the pair of interferometer antennas to obtain interference data, the interference data includes interference amplitude data and interference phase data. The interference data is multi-channel data, also including amplitude and phase, respectively denoted as 、 , both are three-dimensional matrices, wherein 、 、 are respectively the number of time sampling points, the number of frequency sampling points and the number of channels.

[0055] In step S120, the noise region CFAR segmentation calculates the multi-scale phase consistency feature according to the random characteristics of the noise phase, compares the multi-scale phase consistency feature with the CFAR threshold, then locates the noise region and estimates the noise region power.

[0056] In the embodiment, when the noise unit CFAR segmentation method is used to segment the interference phase data: the interference phase data is differentiated along the time dimension and is unwrapped to obtain phase difference data, and the phase difference data is filtered by using a convolution window to obtain a multi-scale phase consistency feature matrix, pure noise phase data is simulated and generated, and multi-scale phase consistency features are extracted based on the pure noise phase data to obtain a pure noise feature matrix, the mean and variance of the pure noise feature matrix are calculated to obtain a statistical parameter matrix, and the pure noise feature matrix is normalized and feature fused by using the statistical parameter matrix to obtain a two-dimensional fused pure noise feature matrix, the mean and root mean square of the fused pure noise feature matrix are combined with the false alarm probability to estimate the CFAR threshold, the multi-scale phase consistency feature matrix is normalized and feature fused by using the statistical parameter matrix to obtain a fused normalized phase consistency feature matrix, and the noise region and the non-noise region are obtained by using the CFAR threshold to segment the fused normalized phase consistency feature matrix.

[0057] Specifically, the interference phase data is differentiated along the time dimension and is unwrapped to , and the phase difference data is mean filtered by using a convolution window, respectively , wherein 、 are respectively the window size along the time and frequency dimensions, and a feature map. The time-frequency correlation phase is a matrix with a size of After mean filtering, the multi-scale phase consistency feature is obtained, which is a three-dimensional matrix with a size of .

[0058] Specifically, pure noise is generated by simulation, and the multi-scale phase consistency feature is extracted for the phase of the pure noise, that is, the same extraction method as described above is used to obtain a pure noise feature matrix, and the mean and variance of each channel feature are estimated based on the pure noise feature matrix, denoted as , that is, a statistical parameter matrix.

[0059] Further, according to each feature dimension , the pure multi-scale phase consistency feature is normalized, and then top-k average or maxout fusion is performed along the feature channel direction to obtain a pure noise fusion phase consistency feature, which is a matrix. The mean and root mean square of the pure noise fusion phase consistency feature are calculated, denoted as , and the CFAR threshold is calculated accordingly.

[0060] In one embodiment, the mean and root mean square of the pure noise fusion phase consistency feature are calculated using the following formula:

[0061] .

[0062] Further, under a preset false alarm threshold , the CFAR threshold can be determined using the following formula:

[0063] ;

[0064] In the above formula, is the inverse function of the standard Gaussian distribution cumulative distribution function. Preferably, can be set to .

[0065] Further, the mean and variance of each channel feature are estimated based on the pure noise feature matrix, and the multi-scale phase consistency feature is normalized, and the process is represented as:

[0066] .

[0067] After normalization, the amplitude fusion feature The fused normalized phase-consistency feature matrix is matrix.

[0068] Further, according to the fused phase feature and the CFAR threshold , a noise prediction binary map is obtained In the binary map, "1" represents noise, thereby obtaining a noise region, and "0" represents a signal, i.e., a non-noise, thereby obtaining a non-noise region.

[0069] In the embodiment, according to the segmented noise region and the interference amplitude data , the data is converted into a single-channel matrix, and the mean value of the noise unit amplitude is taken as the noise power estimation, and the process is represented as:

[0070] .

[0071] In step S130, considering the multiplicative noise characteristics of the correlation data, the segmented noise region is used to perform nonlinear compression mapping on the correlation data to obtain a guide image, so that the guide image is relatively uniform inside the noise region and the signal region, and the edges between the signal regions and between the signal and the noise are maintained.

[0072] In the embodiment, when generating the guide image: by performing amplitude nonlinear compression mapping on the amplitude variation dynamic range, a preliminary guide image is obtained, and by directly using the single-channel interference phase data, phase preservation is performed in the preliminary guide image to synthesize the guide image.

[0073] Specifically, the lower limit of the dynamic range is the quantile of the noise region amplitude, and is represented as:

[0074] ;

[0075] In the above formula, represents the quantile of the data sample set . Preferably, takes a value of 0.95.

[0076] Specifically, the upper limit of the dynamic range is the quantile of the non-noise region amplitude, and is represented as:

[0077] ;

[0078] Further, by performing amplitude nonlinear compression mapping on the amplitude variation dynamic range, a preliminary guide image is obtained, and is represented as:

[0079] ;

[0080] Since the phase of the pilot image is the original phase in the original interferometric phase data, the pilot image is finally represented as:

[0081] .

[0082] In step S140, referring to the mature pilot filtering algorithm, the real part and the imaginary part of the interferometric data are regarded as two channels of the image, the autocorrelation covariance of the pilot image, the cross-correlation data of the pilot image and the cross-correlation are calculated respectively, and the hyperparameter is obtained according to the autocorrelation statistics of the pilot image in the noise region.

[0083] In the embodiment, the pilot image is used to guide the filtering of the interferometric data to obtain an enhanced signal, including: in a preset sliding window, the autocorrelation covariance matrix of the pilot image is calculated, the determinant of the autocorrelation matrix and the regularization parameter are obtained according to the autocorrelation covariance matrix, the cross-correlation matrix is obtained according to the autocorrelation covariance matrix and the complex form of the interferometric data, and the pilot filtering is performed through the autocorrelation covariance matrix, the cross-correlation matrix and the regularization parameter to obtain the enhanced signal.

[0084] Specifically, the autocorrelation of the pilot image is the autocorrelation covariance of the data in the sliding window, denoted as , which is a matrix with a size of .

[0085] Further, the autocorrelation covariance is calculated. The determinant of each grid point is represented as:

[0086] ;

[0087] Then, the regularization parameter, i.e. the hyperparameter , is calculated according to the statistics in the noise region, and is represented as:

[0088] ;

[0089] Then, in the sliding window, the autocorrelation covariance is cross-correlated with the pilot image, denoted as , which is a matrix with a size of .

[0090] Finally, the pilot image is used to guide the filtering of the interferometric data, and the process is represented as:

[0091] ;

[0092] When solving the above formula, since the above formula is The matrix inversion can be directly deduced to combine and accelerate the calculation of the analytic expression.

[0093] After obtaining the enhanced signal, target detection is realized according to the signal, and the target detection accuracy can be effectively improved.

[0094] In this paper, the effectiveness of the application is also proved by experiments, as shown in Figure 2 As shown in Figure 2 (a) is shown as the interferometer time-frequency data, it can be seen that there is a relatively weak figure signal in the data, and Figure 2 (b) is shown as the interferometer time-frequency data enhanced by the method, the original interferometer time-frequency data enhanced by other methods and the enhanced schematic diagram, it can be seen that the signal that is almost invisible by the method becomes significant.

[0095] In the above low signal-to-noise ratio passive detection signal enhancement method, the passive detection signal received by the interferometer antenna is preprocessed to obtain time-frequency data and interference data, the noise unit CFAR segmentation method is used to segment the interference phase data by using the random characteristics of the noise phase, the noise region and the non-noise region represented by the binary graph are obtained, the noise power estimation of the noise region is calculated according to the time-frequency amplitude data, the amplitude change dynamic range of the signal is calculated according to the noise region and the non-noise region based on the time-frequency amplitude data and the noise power estimation, and the guide image is generated by amplitude nonlinear compression, the interference data is guided and filtered by using the guide image, and the enhanced signal is obtained. The enhanced signal can be used for target detection to improve the accuracy of target detection. At the same time, under the condition that the hardware equipment does not change, the software algorithm enhancement processing is used to effectively improve the action distance of the unmanned aerial vehicle passive detection system to the unmanned aerial vehicle.

[0096] It should be understood that, although Figure 1 The steps in the flowchart of the application are displayed in sequence according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this paper, the execution of these steps has no strict order limitation, and these steps can be executed in other order. Moreover, Figure 1 At least part of the steps in the application can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed in rotation or alternation with other steps or sub-steps or stages of other steps.

[0097] In one embodiment, as Figure 3As shown, a passive detection signal enhancement device under low signal-to-noise ratio is provided, comprising: a passive detection signal receiving module 200, a preprocessing module 210, a noise region segmentation module 220, a guide image generation module 230 and a signal enhancement module 240, wherein:

[0098] The passive detection signal receiving module 200 is configured to obtain a passive detection signal, wherein the passive detection signal is received by an interferometer antenna;

[0099] The preprocessing module 210 is configured to preprocess the passive detection signal to obtain time-frequency data and interference data, respectively;

[0100] The noise region segmentation module 220 is configured to segment the interference phase data by using a noise cell CFAR segmentation method based on the random characteristics of the noise phase to obtain noise regions and non-noise regions represented by a binary image, and calculate noise power estimates of the noise regions based on the time-frequency amplitude data;

[0101] The guide image generation module 230 is configured to calculate amplitude variation dynamic ranges of signals based on the time-frequency amplitude data and the noise power estimates according to the noise regions and the non-noise regions, respectively, and generate a guide image by amplitude nonlinear compression;

[0102] The signal enhancement module 240 is configured to perform guided filtering on the interference data by using the guide image to obtain an enhanced signal.

[0103] The specific limitations of the passive detection signal enhancement device under low signal-to-noise ratio can refer to the limitations of the passive detection signal enhancement method under low signal-to-noise ratio in the foregoing, which will not be repeated here. Each module in the above passive detection signal enhancement device under low signal-to-noise ratio can be realized by software, hardware and combinations thereof, in whole or in part. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to each module.

[0104] In one embodiment, a computer device is provided, which can be a terminal, and its internal structure diagram can be as shown in Figure 4As shown in the figure. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the computer device is used to communicate with the external terminal through the network connection. The computer program is executed by the processor to implement a low signal-to-noise ratio passive detection signal enhancement method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0105] Those skilled in the art can understand that, Figure 4 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0106] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the following steps:

[0107] Obtaining a passive detection signal, wherein the passive detection signal is received by an interferometer antenna;

[0108] Pretreating the passive detection signal to obtain time-frequency data and interference data respectively;

[0109] By using a noise cell CFAR segmentation method, the random characteristics of noise phase are used to segment the interference phase data to obtain noise regions and non-noise regions represented by binary images, and the noise power estimation of the noise regions is calculated according to the time-frequency amplitude data;

[0110] Based on the time-frequency amplitude data and the noise power estimation, the amplitude variation dynamic range of the signal is calculated according to the noise regions and the non-noise regions respectively, and a guide image is generated by amplitude nonlinear compression;

[0111] Using the guide image to guide filter the interference data to obtain an enhanced signal.

[0112] In one embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the following steps:

[0113] acquire passive detection signals, the passive detection signals being received by the interferometer antenna;

[0114] pre-process the passive detection signals to obtain time-frequency data and interference data respectively;

[0115] divide the interference phase data by using a noise cell CFAR division method, to obtain noise regions and non-noise regions represented by binary images, and calculate noise power estimates of the noise regions according to the time-frequency amplitude data;

[0116] calculate amplitude variation dynamic ranges of signals according to the noise regions and non-noise regions respectively based on the time-frequency amplitude data and noise power estimates, and generate a guide image by amplitude nonlinear compression;

[0117] filter the interference data by using the guide image to obtain an enhanced signal.

[0118] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0119] Any combination of the technical features of the above embodiments can be made, and to make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0120] The above-described embodiments are merely illustrative of several embodiments of the present application, which are described in more detail and in a specific manner, but should not be construed as limiting the scope of the present application. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method for enhancing a signal in a low signal-to-noise ratio passive detection, characterized in that, The method comprises: acquiring passive detection signals respectively received by interferometer antennas; preprocessing the passive detection signals to respectively obtain time-frequency data and interference data; by using a noise cell CFAR segmentation method, using the random characteristics of noise phase to segment the interference phase data to obtain noise regions and non-noise regions represented by binary graphs, and calculating noise power estimates of the noise regions according to the time-frequency amplitude data, when the noise cell CFAR segmentation method is used to segment the interference phase data: performing difference and unwrapping processing on the interference phase data along the time dimension to obtain phase difference data, and filtering the phase difference data using a convolution window to obtain a multi-scale phase consistency feature matrix, simulating pure noise phase data, and performing multi-scale phase consistency feature extraction based on the pure noise phase data to obtain a pure noise feature matrix, calculating the mean and variance of the pure noise feature matrix to obtain a statistical parameter matrix, and using the statistical parameter matrix to normalize and fuse the pure noise feature matrix to obtain a two-dimensional fused pure noise feature matrix, estimating a CFAR threshold according to the mean and root mean square of the fused pure noise feature matrix and in combination with a false alarm probability, using the statistical parameter matrix to normalize and fuse the multi-scale phase consistency feature matrix to obtain a fused normalized phase consistency feature matrix, and using the CFAR threshold to segment the fused normalized phase consistency feature matrix to obtain the noise regions and non-noise regions; based on the time-frequency amplitude data and noise power estimates, calculating amplitude variation dynamic ranges of signals according to the noise regions and non-noise regions respectively, and generating a guide image through amplitude nonlinear compression; using the guide image to perform guided filtering on the interference data to obtain an enhanced signal.

2. The method of claim 1, wherein the signal enhancement is performed in a low signal-to-noise ratio (SNR) environment. IQ data of signals received by the interferometer antennas are down-converted to baseband, the IQ data is processed using a short-time Fourier transform to obtain time-frequency data, the time-frequency data includes time-frequency amplitude data and time-frequency phase data; the interference data is obtained by performing conjugate correlation processing on signals received by pairs of interferometer antennas, the interference data includes interference amplitude data and interference phase data.

3. The method of claim 2, wherein the signal enhancement is performed by a passive detection system. when the statistical parameter matrix is used to normalize and fuse the pure noise feature matrix or the multi-scale phase consistency feature matrix: the pure noise feature matrix or the multi-scale phase consistency feature matrix is normalized according to the statistical parameter matrix to obtain normalized data; the normalized data is processed along the feature channel direction to obtain feature fused data through top-k average or maxout fusion processing.

4. The method of claim 3, wherein the signal enhancement is performed by a passive detection system. when the guide image is generated: a preliminary guide image is obtained by performing amplitude nonlinear compression mapping on the amplitude variation dynamic range; the guide image is synthesized by directly using single-channel interference phase data to perform phase preservation in the preliminary guide image.

5. The method of claim 4, wherein the signal enhancement is performed by a passive detection system. The guided image is used to perform guided filtering on the interference data to obtain an enhanced signal, including: In a preset sliding window, a self-correlation covariance matrix of the guided image is calculated, and a self-correlation matrix determinant and a regularization parameter are obtained according to the self-correlation covariance matrix; A cross-correlation matrix is obtained according to the self-correlation covariance matrix and the complex form of the interference data; The self-correlation covariance matrix, the cross-correlation matrix and the regularization parameter are used to perform guided filtering to obtain the enhanced signal.

6. A low signal-to-noise ratio passive signal enhancement device, comprising: The device implements the passive detection signal enhancement method under low signal-to-noise ratio according to any one of claims 1-5, including: A passive detection signal receiving module is configured to obtain passive detection signals, which are respectively obtained by an interferometer antenna; A preprocessing module is configured to preprocess the passive detection signals to obtain time-frequency data and interference data respectively; A noise region segmentation module is configured to segment the interference phase data by using a noise cell CFAR segmentation method according to the random characteristics of noise phase to obtain noise regions and non-noise regions represented by a binary image, and to calculate noise power estimation of the noise regions according to time-frequency amplitude data; A guided image generation module is configured to calculate amplitude variation dynamic ranges of signals according to the noise regions and the non-noise regions respectively based on the time-frequency amplitude data and the noise power estimation, and to generate a guided image by amplitude nonlinear compression; A signal enhancement module is configured to use the guided image to perform guided filtering on the interference data to obtain an enhanced signal. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-6 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 5.

Images