A high-concealment signal detection method, system, device and medium based on a digital afterglow map

Abstract

A high-concealment signal detection method, system, device and medium based on a digital residual image, the sampled signal IQ data is preprocessed, that is, the key parameters including the observation bandwidth (that is, the sampling rate), the frame length, the overlap rate and the time and frequency resolution are set; at the same time, each frame of signal is subjected to windowing filtering processing, and the system automatically determines the segmentation number according to the signal length; then, each signal segment is subjected to fast Fourier transform, and the calculation result is stored in a bitmap database; then, the frequency spectrum data is continuously superimposed in the bitmap database, based on the total superposition number, the proportion of each cell is calculated to form a probability density distribution matrix, the probability density distribution matrix is subjected to normalization processing, and all probability values are ensured to be located between 0 and 1; by using a self-defined mapping table, the normalized probability density matrix is converted into a digital residual image through linear mapping; finally, by using a method combining Gamma correction and median filtering, the digital residual image obtained through mapping is further enhanced, and a final clear and smooth digital residual image is obtained; in the application, the Gamma correction can effectively enhance the darker part to improve the image contrast, the median filtering can effectively remove the salt and pepper noise generated by image mapping, and the important edges and details of the signal in the image are retained; due to the low calculation complexity, the application is suitable for real-time processing and meets the low complexity requirement.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of electronic reconnaissance and signal processing, and particularly relates to a high-concealment signal detection method, system, device and medium based on a digital phosphor image. BACKGROUND

[0002] With the rapid development of digital and intelligent communication technology, today's electromagnetic combat environment has become increasingly complex, and the tasks and challenges faced by electronic reconnaissance are also increasingly severe. Effective detection of high-concealment parasitic signals and burst signals is crucial, especially considering the widespread use of high-speed frequency hopping, spread spectrum, time division multiplexing and complex modulation communication technologies. Concealed parasitic communication has become the main means of secure information transmission by the enemy, and space-based reconnaissance systems often cause time-frequency aliasing of multiple signals due to their large geographical coverage, which may result in missed detection of important signals. Therefore, accurate capture of high-concealment signals is extremely critical for space-based electronic reconnaissance systems. Digital phosphor technology (DPX) generates a digital phosphor image to visually display signal changes by accumulating a large amount of spectral data for a short time, which is particularly suitable for revealing concealed and rapidly changing signals.

[0003] DPX technology is an innovative technology in the field of electronic reconnaissance and signal processing, and is particularly suitable for complex electromagnetic environments. Influenced by the traditional cathode ray tube (CRT) phosphor layer technology, this technology can simultaneously display multiple signals within the same frequency band, and improve the level and color representation of signals through digital processing, thereby highlighting and identifying different signals on multiple time levels. The principle and process of generating a digital phosphor image using this technology is as follows: the radio frequency data collected by the front-end device is fed into a hardware digital processing device (DSP or FPGA), and according to the parameters of the incoming data, the hardware digital processing device sequentially performs Fourier transform and statistics on the incoming data in a pipeline manner, forming a frequency spectrum bitmap matrix based on data blocks, and finally converting the statistical results into a digital phosphor image and displaying it.

[0004] DPX technology can capture the transient dynamics and concealment of signals through continuous accumulation of spectral data, with high resolution and high dynamic range, and is particularly suitable for detection of short-time burst signals and co-frequency interference signals. This technology reflects the frequency and intensity of signals through color intensity, significantly improving the visualization of transient signal changes and enhancing the processing capability of signals in complex electromagnetic environments. This technology is applied in the fields of aviation radar interference investigation, signal monitoring and radar reconnaissance, effectively processing co-frequency interference and signals in complex electromagnetic environments.

[0005] Defects and deficiencies of the prior art:

[0006] Although the DPX technology has many advantages, the existing technical solutions mainly use DSP or FPGA to perform spectrum graph operation and cumulative statistics. The scheme is based on the hardware digital processing device (DSP or FPGA) architecture, which has high cost, poor portability, complex peripheral circuit, high price of components, and the calculation capacity of the algorithm is based on the hardware digital processing device itself. Once the hardware device needs to process larger data volume or more complex signal types, if the digital processing capacity of the hardware device cannot meet the calculation requirements after upgrading, the peripheral circuit must be redesigned, and the upgrade is difficult. Moreover, the upgrade of the hardware digital processing device itself will also bring the same problem, which increases the cost of the overall device. SUMMARY

[0007] In order to overcome the shortcomings of the prior art, the purpose of the present application is to provide a high-concealment signal detection method, system, device and medium based on digital residual lobe diagram, which uses fast Fourier transform overlapping frame technology combined with probability density superposition interpolation mapping technology to effectively convert the collected IQ data into digital residual lobe diagram form. By introducing Gamma correction and median filtering technology, the generated digital residual lobe diagram is enhanced, the contrast and definition of the image are optimized, and the time domain, frequency domain and power information of the short-time burst signal and the concealed signal are obtained to realize visualization and high-reliability detection.

[0008] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0009] A high-concealment signal detection method based on digital residual lobe diagram, the sampled signal IQ data is preprocessed, and key parameters including observation bandwidth (i.e. sampling rate), frame length, overlap rate and time and frequency resolution are set. At the same time, each frame of signal is subjected to window filtering processing, and the system automatically determines the number of segments according to the signal length. Then, FFT calculation is performed on each signal segment, and the calculation results are stored in a bitmap database. Then, the spectrum data is continuously superimposed in the bitmap database, based on the total number of superpositions, the proportion of each cell is calculated to form a probability density distribution matrix, and the probability density distribution matrix is normalized to ensure that all probability values are between 0 and 1. Then, using a self-defined mapping table, the normalized probability density matrix is converted into a digital residual lobe diagram through linear mapping. Finally, using the method of combining Gamma correction and median filtering, the digital residual lobe diagram obtained by mapping is further enhanced to obtain the final clear and smooth digital residual lobe diagram for detecting high-concealment signals.

[0010] A high-concealment signal detection method based on digital residual lobe diagram, specifically comprising the following steps:

[0011] Step one, probability density superposition calculation: by setting the initialization parameters, according to the input IQ data, the maximum segment number is calculated, each segment of data is windowed and the maximum segment number is calculated and the fast Fourier transform is carried out by using the FFT frame overlap technology and is written into the bitmap database, and then the probability density matrix is obtained by calculating and normalizing the probability density;

[0012] Step two, digital afterglow map mapping implementation: self-defined linear interpolation mapping is used to map the probability density matrix in the bitmap database into a color image and an infinite afterglow mode, so as to obtain a digital afterglow map;

[0013] Step three, digital afterglow map enhancement and visualization degree adjustment: the digital afterglow map obtained by mapping is further enhanced by using the method combining Gamma correction and median filtering, so as to obtain a final smooth and clear digital afterglow map. In the enhancement process, the value of Gamma can be adjusted according to the requirement to visually adjust the mapped image, so as to obtain digital afterglow maps with different clarity degrees and effects.

[0014] The specific method of step one is:

[0015] 1.1. Maximum segment number calculation

[0016] The sampled signal IQ data is preprocessed, and key parameters including the observation bandwidth (i.e. the sampling rate), the frame length, the overlap rate and the time and frequency resolution are set. At the same time, each frame of signal is subjected to windowing filter processing. After defining the signal IQ data segment length and the step length, the digital afterglow map generation system automatically determines the segment number according to the length of the signal IQ data, and the calculation method is as follows:

[0017]

[0018] Wherein, Overlap nfft =nfft*P overlap , N is the signal length, nfft is the frame length, Overlap nfft is the step length, and P overlap is the overlap rate; it can be known from the above formula that the number of times of probability density superposition of IQ data of a certain length is used to calculate the probability density of each cell in the bitmap database.

[0019] 1.2. Amplitude resolution and frequency resolution calculation

[0020] The amplitude value of each frequency point is traversed by using the amplitude resolution and the frequency resolution to determine which cell in the bitmap database is hit;

[0021] The amplitude resolution calculation method is as follows:

[0022]

[0023] where Max fft and Min fft are the maximum and minimum of each frame FFT result in the loop, respectively, and Pixel x is the width of the bitmap data.

[0024] The frequency resolution is calculated as follows:

[0025]

[0026] where Span is the scanning bandwidth, and Pixel y is the length of the bitmap data.

[0027] In the process of traversing each cell, the amplitude and frequency of a piece of IQ data obtained by fast Fourier transform are obtained. Once the corresponding cell is hit, the value of the statistical hit number in the cell is increased by 1. In this way, all frames are superimposed to obtain the final frequency spectrum hit number bitmap database for probability density calculation.

[0028] 1.3. Probability density calculation

[0029] After obtaining the frequency spectrum hit number bitmap database based on step 1.2, the probability density of each cell is calculated, and the calculation method is as follows:

[0030]

[0031] where N(m,n) is the statistical hit number of the cell in the mth row and nth column of the bitmap database, N FFT is the maximum number of segments, i.e., the total number of frames. The percentage of all cells in the entire bitmap database is obtained by calculation of formula (4), thereby obtaining the entire probability density matrix. It is worth noting here that the sum of the probability densities of all cells in each column is 1. Then, all the probability densities are normalized to the range of 0-1 using the normalization formula, thereby facilitating subsequent color mapping processing:

[0032]

[0033] where P(m,n) is the probability calculated for the cell, Min Pos and Max Pos are the minimum and maximum probabilities in the bitmap database, respectively.

[0034] Subsequently, based on the superimposed number of each cell in the entire bitmap database obtained in steps 1.1 and 1.2, the digital residual image generation system will perform windowing and fast Fourier transform of each frame in a loop: the IQ data is continuously superimposed into the bitmap database based on the frequency spectrum data obtained by fast Fourier transform; after the superposition of all frames is completed, the system performs probability density calculation and normalization processing on each cell in the bitmap database to obtain the final probability density matrix.

[0035] The method of step two is specifically:

[0036] First, a set of RGB color nodes are defined, which cover eight intermediate tones from black to red and their intermediate tones, and the number of gradient steps between these key color nodes is calculated to form a smooth transition color mapping table; then the linear interpolation method is used to calculate the color corresponding to each normalized probability density in the matrix, and the calculation method is:

[0037]

[0038] This formula constructs a straight line between two points, and uses this straight line as a model to predict the value of y at x; this method assumes that the change between the two known points is uniform, i.e. the slope of the straight line (x1-x0) / (y1-y0) is constant, which is a linear interpolation mapping method;

[0039] After the obtained RGB color value is brought into the above formula, linear interpolation is used to smoothly transition between two color nodes; the method is as follows: each color node is represented by an RGB color value, i.e. (R0, G0, B0) and (R1, G1, B1), then the color (R, G, B) at any position p between the two color nodes is calculated by performing linear interpolation on each color channel independently:

[0040]

[0041] Where p is a value from 0 to 1, representing the probability density value of the cell calculated above; this method can generate smooth and continuous color gradient effects, and is widely used in data visualization and graphics rendering;

[0042] By linear interpolation mapping method, the normalized probability density matrix obtained in step one is mapped to digital residual image, this process maps each value in the probability density matrix to a color in the mapping table, accordingly, the numerical data is converted into a color image.

[0043] The conversion process of the step two of converting numerical data into a color image adopts a data visualization tool to automatically select a corresponding color according to a probability density value, generate an intuitive color digital residual flare image, and intuitively display the probability distribution of data, wherein different color depths and types reflect the height and change of the probability density.

[0044] In the step three, the digital residual flare image is enhanced by combining Gamma correction and median filtering in the field of image enhancement, and the combination is specifically as follows:

[0045] 3.1, first, Gamma correction is adopted, and the basic formula of Gamma correction is as follows:

[0046] O = I γ (8)

[0047] Wherein, O is an output pixel value, I is an input pixel value (between 0 and 1), and γ is a Gamma value; the present application uses a value less than 1, because when γ < 1, it is a nonlinear relationship output, enhances dark details, and improves overall brightness of the image;

[0048] 3.2, then, median filtering is adopted to further filter out salt and pepper noise in the digital residual flare image and retain its edge features, so as to highlight the spectral characteristics and color characteristics of the signal; and the specific formula is as follows:

[0049] I'(x, y) = median{I(x+i, y+j)} (9)

[0050] Wherein, I(x+i, y+j) is a pixel value in the neighborhood of pixel (x, y), I'(x, y) is a pixel value processed by median filtering, and median is a median operation, that is, a median of all pixel values in the neighborhood is selected;

[0051] The image mapped according to the adjustment of the value of Gamma is visually adjusted to obtain digital residual flare images with different mapping degrees, and the specific steps are as follows:

[0052] When adjusting, the value of γ is set to a value between 0.1 and 1, every 0.1 is a step, the residual flare image is corrected, and then the digital residual flare image obtained is subjected to median filtering, so that digital residual flare images with different mapping degrees are obtained.

[0053] A high-concealment signal detection system based on a digital residual flare image, comprising:

[0054] The data processing module is used in step one, and IQ data can be converted into frequency spectrum data by FFT fast Fourier transform, so as to realize data preprocessing.

[0055] The spectrum density calculation module is used in step one to calculate the number of hits after superimposing each segment of spectrum data into the bitmap database. It converts the number of hits in each cell into a percentage probability density by normalizing the number of hits in each cell, and realizes the probability density superposition statistics.

[0056] The digital afterglow image mapping module is used in step two to map the probability density bitmap database into a digital afterglow image using a linear interpolation mapping method based on the probability density obtained in step one, thereby realizing the mapping of the digital afterglow image.

[0057] The digital afterglow image enhancement module is used in step three. By performing image enhancement processing on the digital afterglow image obtained in step two, a digital afterglow image with clearer and more obvious signal characteristics can be obtained, thus realizing the enhancement processing of the digital afterglow image.

[0058] A highly concealed signal detection device based on digital afterglow maps includes:

[0059] Memory, used to store computer programs;

[0060] A processor is used to implement the high-coverage signal detection method based on digital afterglow map described in steps one to three when executing the computer program.

[0061] A computer-readable storage medium storing a computer program, which, when executed by a processor, is capable of performing high-coverage signal detection based on digital afterglow maps, according to the high-coverage signal method based on digital afterglow maps described in steps one to three.

[0062] Compared with the prior art, the present invention has the following advantages:

[0063] 1) Low complexity and high efficiency: This invention effectively reduces the dependence on hardware resources and improves data processing efficiency by employing Fast Fourier Transform overlapping frame technology and probability density superposition interpolation mapping technology. This makes the generation process of digital afterglow maps not only fast but also more economical in terms of hardware requirements.

[0064] 2) Enhanced Image Quality: By combining Gamma correction with median filtering, this invention significantly optimizes the contrast and sharpness of the generated digital afterglow image and reduces its dependence on longer IQ data. This image enhancement process not only improves the visual effect but also enhances the identifiability of hidden signals in the image, making the detection of short-term burst signals and hidden parasitic signals more accurate.

[0065] 3) High sensitivity signal detection: The technical solution improves the detection ability of low-intensity and same-frequency interference signals through fine spectrum data processing. This is particularly important for signal monitoring and radar reconnaissance in complex electromagnetic environments, and can effectively distinguish and capture weak signals that are difficult to detect by conventional methods.

[0066] 4) Flexible visualization adjustment: By adjusting the Gamma value and using the pseudo-color image mapping method, the invention allows users to adjust the visualization degree of the digital afterglow map according to different monitoring needs, so as to better adapt to different analysis and interpretation needs.

[0067] 5) Reduce cost and improve portability: Compared with traditional solutions that rely on high-cost DSP or FPGA, the method of the invention avoids peripheral circuit design, saves component cost, and presents in the form of software algorithm, easy to deploy on different platforms, improving the portability and applicability of the system.

[0068] In summary, the invention not only improves the generation and processing capability of digital afterglow map, but also has significant advantages in economy and practicality, especially suitable for use in high-demand situations such as electronic reconnaissance and complex electromagnetic environment signal analysis. BRIEF DESCRIPTION OF DRAWINGS

[0069] Figure 1 is the low-complexity digital afterglow map generation flowchart based on the DPX technology of the invention.

[0070] Figure 2 is the probability density superposition flowchart based on the DPX technology of the invention.

[0071] Figure 3 is the digital afterglow map mapping flowchart based on the self-defined mapping table and linear interpolation of the invention.

[0072] Figure 4 is the digital afterglow map enhancement flowchart based on Gamma correction and median filtering of the invention.

[0073] Figure 5 is the Gamma correction curve used by the invention.

[0074] Figure 6 is the digital afterglow map not generated by the self-defined mapping table and linear interpolation method of the invention.

[0075] Figure 7 is the digital afterglow map generated based on the low-complexity algorithm of the invention.

[0076] Figure 8 is the digital afterglow map generated based on the image enhancement method of the invention.

[0077] Figure 9This is a schematic diagram of a simulation scenario for short-term burst signals and concealed parasitic signals according to the present invention.

[0078] Figure 10 This is a digital afterglow image of a short-term burst signal generated by a low-complexity algorithm according to the present invention.

[0079] Figure 11 This is a digital afterglow image of a hidden parasitic signal generated by a low-complexity algorithm according to the present invention. Detailed Implementation

[0080] The present invention will now be described in further detail with reference to the accompanying drawings.

[0081] A highly concealed signal detection method based on digital afterglow maps, reference Figure 1 The low-complexity digital afterglow image generation process based on DPX technology includes the following steps:

[0082] Step 1, Probability density superposition calculation: By setting initialization parameters, the maximum number of segments is calculated based on the input IQ data. Each segment of data is windowed and the maximum number of segments is calculated and fast Fourier transform is performed using FFT overlapping frame technology and written into the bitmap database. Then, the probability density matrix is ​​obtained by calculating and normalizing the probability density.

[0083] 1.1. Calculation of the maximum number of segments

[0084] The sampled IQ signal data is preprocessed, with key parameters including observation bandwidth (i.e., sampling rate), frame length, overlap rate, and time and frequency resolution set. Simultaneously, window filtering is applied to each frame. The segment length of the IQ signal data is defined as 1024 points, and the step size as 512 points. Then, the digital persistence map generation system automatically determines the number of segments based on the length of the IQ signal data, typically ranging from tens of thousands to hundreds of thousands. Next, a Fast Fourier Transform is performed on each signal segment, and the calculation results are stored in a 1024×1024 resolution bitmap database. The calculation method is as follows:

[0085]

[0086] Among them, Overlap nfft =nfft*P overlap N is the signal length, nfft is the frame length, and Overlap nfft P is the step length. overlap The overlap rate is calculated using the above formula, which determines the number of times a certain length of baseband IQ data can be superimposed on probability density. This is used to calculate the probability density of each cell in the bitmap database.

[0087] 1.2. Calculation of Amplitude Resolution and Frequency Resolution

[0088] The amplitude value of each frequency point is traversed by amplitude resolution and frequency resolution to determine which cell in the hit bitmap database;

[0089] The amplitude resolution is calculated as follows:

[0090]

[0091] Where Max fft , Min fft represent the maximum and minimum values of the fast Fourier transform result of each frame in the loop, respectively, and Pixel x is the width in the bitmap data.

[0092] The frequency resolution is calculated as follows:

[0093]

[0094] Where Span is the scanning bandwidth, and Pixel y is the length of the bitmap data.

[0095] In the process of traversing each cell, the amplitude and frequency obtained by fast Fourier transform of a piece of IQ data are hit to the corresponding cell, and the value of the statistical hit number in the cell is increased by 1. In this way, all frames are superimposed to obtain the final frequency spectrum hit number bitmap database for probability density calculation.

[0096] 1.3. Probability density calculation, refer to Figure 2 :

[0097] After obtaining the frequency spectrum hit number bitmap database based on step 1.2, the probability density of each cell is calculated, and the calculation method is as follows:

[0098]

[0099] Where N(m,n) is the statistical hit number of the cell in the mth row and nth column in the bitmap database, N FFT is the maximum number of segments, i.e. the total number of frames. The percentage of all cells in the entire bitmap database is obtained by formula (4), thereby obtaining the entire probability density matrix. It is worth noting here that the sum of the probability densities of all cells in each column is 1. Then, all the probability densities are normalized to the range of 0-1 using the normalization formula, thereby facilitating subsequent color mapping processing:

[0100]

[0101] Where P(m,n) is the probability calculated for the cell, Min Pos , MaxPos respectively the minimum and maximum values of probability in the bitmap database;

[0102] Subsequently, based on the superposition number of each cell in the entire bitmap database obtained in steps 1.1 and 1.2, the digital residual image generation system will perform windowing and fast Fourier transform for each frame in a loop: the IQ data is continuously superimposed into the bitmap database based on the frequency spectrum data obtained by FFT transformation; after the superposition of all frames is completed, the system performs probability density calculation and normalization processing on each cell in the bitmap database to obtain the final probability density matrix;

[0103] Reference Figure 3 , step two, digital residual image mapping implementation: custom linear interpolation mapping, mapping the probability density matrix in the bitmap database to a color image and infinite residual pattern to obtain a digital residual image;

[0104] comprising the following steps:

[0105] First, define a set of RGB color nodes that cover the colors from black to red and eight intermediate tones, calculate the number of gradient steps between these key color nodes to form a smooth transition color mapping table; then use linear interpolation method to calculate the color corresponding to each normalized probability density in the matrix, the calculation method is:

[0106]

[0107] The core idea of this formula is to construct a straight line between two points and use this straight line as a model to predict the value of y at x. This method assumes that the change between the two known points is uniform, that is, the slope of the straight line (x1-x0) / (y1-y0) is constant, which is a linear mapping method.

[0108] Linear interpolation can be used to smoothly transition between two color nodes. Assuming each color node is represented by an RGB color value, (R0, G0, B0) and (R1, G1, B1), the color (R, G, B) at any position p between these two color nodes can be calculated by performing linear interpolation on each color channel independently:

[0109]

[0110] Where p is a value from 0 to 1, representing the probability density value of the cell calculated above. This method can generate smooth and continuous color gradient effects and is widely used in data visualization and graphics rendering.

[0111] The normalized probability density matrix is mapped into a digital residual image by a linear interpolation mapping method, which corresponds each value in the probability density matrix to a color in the mapping table, and converts the numerical data into a color image in this way. This conversion process needs to use a data visualization tool (such as MATLAB) to automatically select the corresponding color according to the probability density value, generate an intuitive color digital residual image, which intuitively shows the probability distribution of the data, and different color depth and type reflect the probability density.

[0112] Step three, digital residual image enhancement and visualization degree adjustment: using the method of combining Gamma correction and median filtering, the digital residual image obtained by mapping is further enhanced to obtain the final smooth and clear digital residual image. In the enhancement process, the value of Gamma can be adjusted according to the demand to visualize the mapped image, and digital residual images with different clarity and effects are obtained.

[0113] Reference Figure 4 And Figure 5 , specifically including the following steps:

[0114] The digital residual image is enhanced by combining Gamma correction and median filtering in the field of image enhancement. The basic formula of Gamma correction is:

[0115] O = I γ (8)

[0116] Where O is the output pixel value, I is the input pixel value (between 0 and 1), and γ is the Gamma value. The invention uses the value of γ less than 1, because when γ < 1, it is a nonlinear relationship output, which enhances the dark details and improves the overall brightness of the image.

[0117] Then median filtering is used to further filter out the salt and pepper noise in the digital residual image and retain its edge features, thereby highlighting the spectral and color features of the signal. The specific formula is:

[0118] I'(x, y) = median{I(x+i, y+j)} (9)

[0119] Where I(x+i, y+j) is the pixel value in the neighborhood of pixel (x, y), I'(x, y) is the pixel value after median filtering, and median is the median operation, that is, the median of all pixel values in the neighborhood. In the processing of digital residual image, median filtering can effectively remove the salt and pepper noise generated by image mapping, while retaining the important edges and details of the signal in the image. Due to its low computational complexity, it is suitable for real-time processing applications and meets the low complexity requirement.

[0120] The digital afterglow image is enhanced by combining Gamma correction and median filtering in the field of image enhancement, and the specific method is as follows:

[0121] 3.1, first, Gamma correction is adopted, and the basic formula of Gamma correction is:

[0122] O = I γ (8)

[0123] Wherein, O is the output pixel value, I is the input pixel value (between 0 and 1), and γ is the Gamma value; The present application uses the value when γ is less than 1, because when γ < 1, it is a nonlinear relationship output, which enhances the dark details and improves the overall brightness of the image;

[0124] 3.2, then, median filtering is adopted to further filter out the salt and pepper noise in the digital afterglow image and retain its edge features, so as to highlight the spectral and color features of the signal; The specific formula is:

[0125] I'(x, y) = median {I(x+i, y+j)} (9)

[0126] Wherein, I(x+i, y+j) is the pixel value in the neighborhood of pixel (x, y), I'(x, y) is the pixel value after median filtering, and median is the median operation, that is, the median of all pixel values in the neighborhood is selected;

[0127] The value of Gamma is adjusted according to the requirement, the mapped image is visually adjusted, and digital afterglow images with different mapping degrees are obtained, and the specific steps are as follows:

[0128] When adjusting, the value of γ is set from 0.1 to 1, every 0.1 is a step, the afterglow image is corrected, and then the digital afterglow image is subjected to median filtering, so that digital afterglow images with different mapping degrees are obtained.

[0129] (2) Simulation verification analysis

[0130] 1) Digital afterglow image generation and enhancement result analysis

[0131] The present application can be used to map and enhance the digital afterglow image by setting the parameters in table 1 and the color node mapping table in table 2, and the specific results are shown in Figure 6 、 Figure 7 and Figure 8 .

[0132] Figure 6is the digital afterglow map based on the short IQ data collected without using the color mapping table, which is the digital afterglow map without color mapping, although multiple signals can be seen, but the profiles of the signals except signal 1 are not clear because the background is mainly blue, and the observation effect is relatively poor.

[0133] Figure 7 is the digital afterglow map based on the short IQ data collected by using the color mapping table, and different signals are marked with 1, 2, 3 and 4, which can clearly display signals 1 and 2, but because the data is short, the profiles of signals 3 and 4 with low frequency are blurred, and there is an unknown signal mixed with signal 2 on the left side of signal 2, both of which are difficult to clearly identify. Therefore, the digital afterglow needs to be enhanced.

[0134] Figure 8 is the digital afterglow map obtained by using Gamma correction with a gamma value of 0.5 and median filter set enhancement, through the combination of Gamma mapping and median filter processing, the contrast and brightness of the digital afterglow map are significantly improved, and the signal profile, color and layering effect are clearer. The processing method makes the characteristics of each signal more obvious, and can clearly observe part of the mixed phenomenon (such as 1 and 2, 1 and 6) and the hidden parasitic mixed phenomenon (such as 3, 4, 5 and 2).

[0135] Table 1 probability density superposition related parameters

[0136]

[0137]

[0138] Table 2 color node mapping table

[0139]

[0140]

[0141] In summary, the digital afterglow map generation enhancement algorithm based on the DPX technology proposed by the application effectively reduces the calculation complexity and the dependence on hardware resources by using the FFT overlapping frame technology and the probability density superposition interpolation mapping technology, and improves the image quality by combining the Gamma correction and the median filter technology, so that the short data can be effectively processed.

[0142] 2) Short-time burst signal and hidden parasitic signal observation effect analysis

[0143] Figure 9is the simulation scene established by the present application according to the characteristics of short-time burst signals and hidden parasitic signals. In the simulated communication interference scene, considering that the communication equipment of our side may encounter various interference signals from illegal radio stations or civil equipment, including short-time burst interference signals such as intercom signals and hidden eavesdropping signals, and weak signals that may be the same frequency as our signals, these signals may be hidden and superimposed in the frequently occurring signals, forming time-frequency superimposed signals, affecting the smoothness and safety of communication.

[0144] 2.1) Short-time burst signal monitoring effect analysis

[0145] For this scenario, the present application uses the parameters set in Table 3 to generate and enhance the digital residual image using the data of the frequently occurring signals and the short-time burst signals under non-time-frequency superposition in the same bandwidth simulated by the MATLAB platform, and obtains Figure 10 , the digital residual image of the short-time burst signal.

[0146] After digital residual image generation and enhancement processing, the short-time burst signal can be clearly seen, and the short-time burst signal often presents blue-green characteristics, which from the color point of view shows that its occurrence probability is low and has occasionality.

[0147] Table 3 Short-time burst signal simulation parameters

[0148] Parameter name Parameter value Sampling frequency Fs 20 MHz Sampling time T 1s Signal A signal-to-noise ratio SNR 20 dB Signal B interference-to-noise ratio INR 1 dB Signal A bandwidth BW A ]] 5 MHz Signal B bandwidth BW B ]]> 1 MHz Signal A carrier frequency Fc A ]] 0 MHz Signal B carrier frequency Fc B ]] 5 MHz Signal B burst time 20 ms Signal A modulation BPSK Signal B modulation BPSK

[0149] 2.2) Hidden parasitic signal monitoring effect analysis

[0150] For this scenario, the present application uses the parameters set in Table 4 to generate and enhance the digital residual image using the data of the frequently occurring signals and another weak signal under time-frequency superposition in the same bandwidth simulated by the MATLAB platform, and obtains Figure 11 , the digital residual image of the hidden parasitic signal.

[0151] After digital residual image generation and enhancement processing, the hidden parasitic superposition phenomenon can be clearly seen, including the host signal and the parasitic signal, and the parasitic signal often exhibits blue-green characteristics in the digital residual image due to its weakness and occasionality.

[0152] From the above two results, further analysis shows that the method of the present application not only improves the image processing speed and economy, but also improves the detection ability of low-intensity and same-frequency interference signals through fast spectrum data processing. The recognition effect of short-time burst signals and hidden parasitic signals is enhanced, and it is especially suitable for real-time electronic reconnaissance and signal analysis in complex electromagnetic environments.

[0153] Table 4 Hidden parasitic signal simulation parameters

[0154] Parameter name Parameter value Sampling frequency Fs 20 MHz Sampling time T 1s Signal A signal-to-noise ratio SNR 20 dB Signal B interference-to-noise ratio INR 1 dB Signal A bandwidth BW A ]]> 5 MHz Signal B bandwidth BW B ]]> 1 MHz Signal A carrier frequency Fc A ]]> 0 MHz Signal B carrier frequency Fc B ]] 0 MHz Signal A modulation BPSK Signal B modulation BPSK

Claims

1. A method for detecting highly concealed signals based on digital afterglow maps, characterized in that, Specifically, the following steps are included: Step 1, Probability density superposition calculation: By setting initialization parameters, the maximum number of segments is calculated based on the input IQ data. Each segment of data is windowed and the maximum number of segments is calculated and fast Fourier transform is performed using FFT overlapping frame technology and written into the bitmap database. Then, the probability density matrix is ​​obtained by calculating and normalizing the probability density. 1.

1. Calculation of the maximum number of segments The sampled IQ signal data is preprocessed, with key parameters including observation bandwidth (sampling rate), frame length, overlap rate, and time and frequency resolution. Simultaneously, each frame of signal is windowed and filtered. After defining the segment length and step length of the IQ signal data, the digital afterglow map generation system automatically determines the number of segments based on the length of the IQ signal data, calculated as follows: Among them, Overlap nfft =nfft*P overlap N is the signal length, nfft is the frame length, and Overlap nfft P is the step length. overlap The overlap rate; 1.

2. Calculation of Amplitude Resolution and Frequency Resolution The amplitude value of each frequency point is traversed using amplitude resolution and frequency resolution to determine which cell in the bitmap database to hit. The amplitude resolution is calculated as follows: Among them, Max fft Min fft Pixel represents the maximum and minimum values ​​of the FFT result for each frame within the loop. x This refers to the width in the bitmap data; Frequency resolution is calculated as follows: Where Span is the scan bandwidth, and Pixel is the pixel. y The length of the bitmap data; During the process of traversing each cell, the amplitude and frequency of a segment of IQ data are obtained by fast Fourier transform. Once the corresponding cell is hit, the statistical hit count value in that cell is incremented by 1. This process continues until all frames are overlaid with bitmaps to obtain the final spectrum hit count bitmap database for probability density calculation. 1.

3. Probability Density Calculation After obtaining the spectrum hit count bitmap database in step 1.2, the probability density of each cell is calculated as follows: Where N(m,n) represents the number of times the cells in the m-th row and n-th column of the bitmap database are hit, N FFT The maximum number of segments is the total number of frames. The percentage of all cells in the entire bitmap database is calculated using formula (4), thus obtaining the entire probability density matrix. The sum of the probability densities of all cells in each column is 1. Then, the normalization formula is used to normalize all probability densities to the range of 0-1, which facilitates subsequent color mapping processing. Where P(m,n) is the probability calculated for that cell, and Min Pos Max Pos These are the minimum and maximum probabilities in the bitmap database, respectively. Subsequently, based on the number of times each cell in the entire bitmap database is superimposed, obtained in steps 1.1 and 1.2, the digital afterglow image generation system will cyclically perform windowing and fast Fourier transform for each frame: continuously superimposing the spectral data obtained from the IQ data based on the fast Fourier transform into the bitmap database; after completing the superimposition of all frames, the system performs probability density calculation and normalization on each cell in the bitmap database to obtain the final probability density matrix. Step 2, Digital Afterglow Mapping Implementation: A custom linear interpolation mapping is used to map the probability density matrix in the bitmap database to a color image and an infinite afterglow mode to obtain a digital afterglow map; First, a set of RGB color nodes is defined, covering black to red and eight intermediate hues. The gradient steps between these key color nodes are calculated to form a smooth color mapping table. Then, linear interpolation is used to calculate the color corresponding to each normalized probability density in the matrix, using the following formula: This formula constructs a straight line between two points and uses this line as a model to predict the y value at x. This method assumes that the change between the two known points is uniform, that is, the slope (x1-x0) / (y1-y0) of the line is a constant, which is a linear interpolation mapping method. A smooth transition between two color nodes is achieved using linear interpolation. The method involves representing each color node with an RGB color value, i.e., (R0, G0, B0) and (R1, G1, B1). The color (R, G, B) at any position p between these two color nodes is then calculated by independently performing linear interpolation on each color channel. Where p is a value from 0 to 1, representing the probability density value of the calculated cell; this method can generate smooth and continuous color gradient effects and is widely used in data visualization and graphics rendering. The normalized probability density matrix obtained in step one is mapped to a digital afterglow image by using a linear interpolation mapping method. Each value in the probability density matrix is ​​mapped to a color in the mapping table, thereby converting the numerical data into a color image. Step 3, Digital Afterglow Image Enhancement and Visualization Adjustment: Using a combination of Gamma correction and median filtering, the mapped digital afterglow image is further enhanced to obtain a smooth and clear final digital afterglow image. During the enhancement process, the Gamma value is adjusted according to the needs to perform visualization adjustments on the mapped image, resulting in digital afterglow images with different levels of clarity and effects.

2. The method for detecting highly concealed signals based on digital afterglow maps according to claim 1, characterized in that, The second step, which converts numerical data into a color image, uses a data visualization tool to automatically select the corresponding color based on the probability density value, generating an intuitive color digital afterglow image that visually displays the probability distribution of the data. Different color depths and types reflect the level and changes in probability density.

3. The method for detecting highly concealed signals based on digital afterglow maps according to claim 1, characterized in that, In step three: The method combining Gamma correction and median filtering further enhances the mapped digital persistence image, specifically as follows: 3.1 First, Gamma correction is used. The basic formula for Gamma correction is: O=I γ (8) Where O is the output pixel value, I is the input pixel value, which is between 0 and 1, and γ is the Gamma value; γ is the value when it is less than 1, because when γ<1, it is a non-linear output that enhances dark details and improves the overall brightness of the image; 3.2 Next, median filtering is used to further remove salt-and-pepper noise from the digital persistence image and preserve its edge features, thereby highlighting the signal's spectral and color characteristics; the specific formula is as follows: I′(x,y)=median{I(x+i,y+j)} (9) Where I(x+i,y+j) is the pixel value in the neighborhood of pixel (x,y), I′(x,y) is the pixel value after median filtering, and median is the median operation, which selects the median of all pixel values ​​in the neighborhood; The process of adjusting the Gamma value according to requirements to perform visual adjustments on the mapped image, resulting in digital afterglow images with different levels of clarity and effects, involves the following steps: During adjustment, the γ value is set to a value between 0.1 and 1, with each step being 0.1, to correct the resulting afterglow image. Then, median filtering is applied to the obtained digital afterglow images to obtain digital afterglow images with different mapping degrees.

4. A high-coverage signal detection system based on digital afterglow maps, used to implement the high-coverage signal detection method based on digital afterglow maps as described in any one of claims 1 to 3, characterized in that, include: The data processing module is used in step one to convert IQ data into spectral data through fast Fourier transform, thereby achieving data preprocessing. The spectrum density calculation module is used in step one to overlay each segment of spectrum data onto the bitmap database and count the number of hits after overlay. It normalizes the number of hits in each cell and converts it into a percentage probability density, thus realizing the probability density overlay statistics. The digital afterglow image mapping module is used in step two to map the probability density matrix obtained in step one into a digital afterglow image using a linear interpolation mapping method. The digital afterglow image enhancement module is used in step three to perform image enhancement processing on the digital afterglow image obtained in step two to obtain a digital afterglow image with clearer and more obvious signal characteristics, thereby realizing the enhancement processing of the digital afterglow image.

5. A highly concealed signal detection device based on digital afterglow maps, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the high-coverage signal detection method based on digital afterglow maps as described in any one of claims 1 to 3 when executing the computer program.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it can perform high-coverage signal detection based on digital afterglow maps using the high-coverage signal detection method based on digital afterglow maps as described in any one of claims 1 to 3.

Images