Ground dense static target rapid detection method and device
By filtering and iteratively calculating radar echo data, the detection difficulties in environments with dense ground targets and non-uniform clutter were solved, enabling rapid and accurate target detection and improving the helicopter's battlefield situational awareness capabilities.
Patent Information
- Application Number
- CN202511422578.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-30
- Publication Date
- 2025-12-05
AI Technical Summary
In helicopter-borne radar ground detection, dense ground targets interfere with each other and cannot be effectively detected. Target detection is difficult in non-uniform clutter environments. Existing improved algorithms have high complexity, leading to difficulties in target detection and missed detections.
A finite-length unit impulse response filter is used to smooth radar echo data, eliminating clutter and dense target interference. The true background power is obtained through multiple iterative calculations, and target detection is performed using signal-to-noise ratio and signal-to-clutter ratio.
It improves detection accuracy and speed under conditions of dense targets and uneven background power, reduces algorithm complexity, and enhances the helicopter's battlefield situational awareness capabilities.
Smart Images
Figure CN121069345A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of target detection, and particularly relates to a ground dense stationary target rapid detection method and device. BACKGROUND
[0002] The helicopter airborne radar ground detection technology needs to detect the ground stationary target from the echo and complete the tracking and display of the ground stationary target. Especially in the case that the number of ground targets is large and the ground environment is complex, the signal-to-clutter ratio is low and the targets interfere with each other, which leads to great difficulty in ground stationary target detection and resolution. Precise ground stationary target detection and analysis are beneficial to improve the battlefield situation awareness capability of the helicopter.
[0003] The most commonly used target detector in the helicopter airborne radar ground detection system is the constant false alarm rate (CFAR) detector. It adaptively changes the detection threshold according to the change of the noise and clutter intensity near the detection unit to obtain the maximum detection probability and constant false alarm probability. Although the current detector has been improved a lot in the early mean level (ML) CFAR detector, the performance of the detector will be greatly affected in the environment of dense target distribution and non-uniform clutter, and effective target detection cannot be achieved. The main problems are:
[0004] (1) When the ground targets are dense, the mutual interference cannot be effectively detected. The ground targets often appear in clusters. When the traditional detection method calculates the average power of the signal in the reference window, the average power of the signal in the reference window will be raised due to the dense distribution of the targets, which will lead to a decrease in the calculated signal-to-clutter ratio, and thus multiple targets cannot be detected.
[0005] (2) The target cannot be detected in the non-uniform clutter environment. The traditional detection method compares the power of the center point with the platform power in the reference window. In the non-uniform clutter environment, part of the strong clutter will significantly raise the average power in the reference window, which is easy to lead to that the target signal-to-clutter ratio does not meet the detection threshold, and thus the target is missed.
[0006] (3) The existing improved algorithm has high complexity. The current improved target detection method has the ordered statistical CFAR algorithm and the target detection algorithm based on deep learning. Although these detection methods can improve the performance of the target detection algorithm, the time complexity of these detection methods is high, and the overall algorithm operation burden is large. SUMMARY
[0007] In order to solve the above problems, the application provides a ground dense stationary target rapid detection method and device. The filter is used to smooth the data of the original echo, remove the clutter interference and dense target interference, and the processed reference data is used for target detection.
[0008] The first aspect of the present application provides a method for detecting ground dense static targets quickly, which mainly comprises:
[0009] Step S1, performing sliding window filtering processing on each sample point in the original echo data D ori to obtain background power data D ref . bak wherein the position of each sample point in the original echo data has a one-to-one correspondence with the spatial position of the scanning point in the radar scanning range;
[0010] Step S2, performing floating on each background power value in the background power data D bak based on a given background power floating threshold to obtain a background power floating value;
[0011] Step S3, when the power value of each sample point in the reference data D ref exceeds the background power floating value of the corresponding sample point, replacing the power value of the sample point in the reference data D ref with the background power value of the sample point to form new reference data, and after multiple iterations, obtaining the final background power data;
[0012] Step S4, calculating the signal-to-noise ratio and the signal-to-clutter ratio according to the original echo data D ori and the final background power data;
[0013] Step S5, determining the position of the ground static target based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold.
[0014] Preferably, in step S1, the sliding window filtering processing comprises:
[0015] Step S11, performing processing on the reference data D ref through a preset sliding window W[k] to form preprocessed data;
[0016] Step S12, performing filtering processing on the preprocessed data using a k-order filter to obtain the background power data D bak .
[0017] Preferably, step S11 comprises:
[0018] Step S111, extracting the power values of k consecutive sample points in the reference data D ref with a step size k of the sliding window W[k];
[0019] Step S112, performing weighted calculation on the k sample points in the sliding window W[k] and the extracted k power values respectively to form the preprocessed data, wherein k is an odd number, and the k sliding window parameters in the sliding window W[k] are set to the reference data D refThe sliding window parameter is set to 0 for the location of the sample point to be processed and 0 or more locations around it, and the other sliding window parameters are set to 1.
[0020] Preferably, the step size k of the sliding window W[k] is 3, 5 or 7, and the proportion of sliding window parameters with a value of 1 is not less than 50%.
[0021] Preferably, in step S4, the signal-to-noise ratio D is calculated using the following formula. SNR :
[0022] D SNR =D ori p noise ;
[0023] The signal-to-noise ratio D is calculated using the following formula. SCR :
[0024] D SCR =D ori D bak ;
[0025] Where, p noise For noise power, D is the original echo data. ori The average power of each sampling point.
[0026] The second aspect of this application provides a rapid detection device for densely packed stationary targets on the ground, mainly comprising:
[0027] The background power data calculation module is used to calculate the raw echo data D. ori As reference data D ref The power value of each sampling point is subjected to sliding window filtering to obtain the background power data D. bak In this context, the position of each sampling point in the original echo data has a one-to-one correspondence with the spatial position of the scanning point within the radar scanning range;
[0028] The background power up-float module is used to adjust the background power data D based on a given background power up-float threshold. bak The background power values in the data are adjusted upwards to obtain the background power adjustment value;
[0029] Background power data update module, used when reference data D ref When the power value of each sampling point exceeds the background power surcharge value of the corresponding sampling point, the background power value of that sampling point is used to replace the reference data D. ref The power value at the sampling point is used to form new reference data. After multiple iterations of calculation, the final background power data is obtained.
[0030] The signal-to-noise ratio (SNR) and signal-to-noise ratio (SNR) calculation module is used to calculate the original echo data D. oriand the final background power data calculates signal-to-noise ratio and signal-to-clutter ratio;
[0031] The target position determination module is configured to determine the position of the ground stationary target based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold.
[0032] Preferably, the background power data calculation module comprises:
[0033] The sliding window preprocessing unit is configured to process the reference data D ref by a preset sliding window W[k] to form preprocessed data.
[0034] The filter processing unit is configured to perform filter processing on the preprocessed data using a k-order filter to obtain the background power data D bak .
[0035] Preferably, the sliding window preprocessing unit comprises:
[0036] The power extraction subunit is configured to extract power values of k consecutive sampling points in the reference data D ref at a step size k of the sliding window W[k];
[0037] The power parameter processing subunit is configured to perform weighted calculation on the k sliding window parameters in the sliding window W[k] and the k extracted power values respectively to form the preprocessed data, wherein k is an odd number, and the sliding window parameters corresponding to the position of the sample point to be processed in the reference data D ref and 0 or more positions around the position are set to 0, and other sliding window parameters are set to 1.
[0038] Preferably, the step size k of the sliding window W[k] is 3, 5 or 7, and the proportion of the sliding window parameters with a value of 1 is not less than 50%.
[0039] Preferably, in the signal-to-noise ratio and signal-to-clutter ratio calculation module, the signal-to-noise ratio D SNR is calculated by the following formula:
[0040] D SNR = D ori p noise ;
[0041] The signal-to-clutter ratio D SCR is calculated by the following formula:
[0042] D SCR = D ori D bak ;
[0043] Wherein, p noise is the noise power, and is the average value of the power of each sampling point in the original echo data D ori .
[0044] The application uses a finite-length unit impulse response filter (FIR filter) to perform multiple iterations of smoothing processing on the reference data, eliminates clutter interference and dense target interference, and obtains reference data reflecting the true background power, so that the algorithm can detect targets under the conditions of target density and uneven background power. While ensuring the operation speed, the adaptability and accuracy of the target detection algorithm in complex scenes are improved, and the helicopter battlefield situation awareness capability is improved. BRIEF DESCRIPTION OF DRAWINGS
[0045] Figure 1 is a flowchart of a preferred embodiment of the ground dense stationary target rapid detection method of the application.
[0046] Figure 2 is a schematic diagram of the target detection result before optimization
[0047] Figure 3 is a schematic diagram of the target detection result after optimization
[0048] Figure 4 is a schematic diagram of the target detection result before optimization
[0049] Figure 5 is a schematic diagram of the target detection result after optimization DETAILED DESCRIPTION
[0050] In order to make the purpose, technical scheme and advantages of the application clearer, the technical scheme in the embodiment of the application will be described in more detail below in combination with the drawings in the embodiment of the application. In the drawings, the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The described embodiments are part of the embodiments of the application, not all embodiments. The embodiments described below by reference to the drawings are exemplary and are intended to explain the application, and cannot be understood as limiting the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application. The embodiments of the application will be described in detail below in combination with the drawings.
[0051] The first aspect of the application provides a ground dense stationary target rapid detection method, as shown in Figure 1 mainly includes:
[0052] Step S1, the original echo data D ori is taken as reference data D ref , and the power value of each sampling point in it is processed by sliding window filtering to obtain background power data D bakwherein the position of each sampling point in the original echo data has a one-to-one correspondence with the spatial position of the scanning point in the radar scanning range;
[0053] Step S2, based on the given background power up threshold, the background power data D bak is up floated to obtain the background power up value.
[0054] Step S3, when the power value of each sampling point in the reference data D ref exceeds the background power up value of the corresponding sampling point, the power value of the sampling point in the reference data D ref is replaced by the background power value of the sampling point to form new reference data, and after multiple iterations, the final background power data is obtained.
[0055] Step S4, the signal-to-noise ratio and the signal-to-clutter ratio are calculated according to the original echo data D ori and the final background power data.
[0056] Step S5, based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold, the position of the ground stationary target is determined.
[0057] The present application takes the original data of radar echo as input to obtain initial reference data, first processes the reference data through a finite-length unit impulse response filter (FIR filter) to obtain background power data, removes the target or clutter interference in the reference data by comparing the reference data and the background data, iterates the above processing to obtain reference data that can reflect the true background power in the original radar echo signal, and finally calculates the signal-to-noise ratio and the signal-to-clutter ratio using the original data and the reference data to complete one-dimensional CFAR detection.
[0058] In step S1, first, the radar original echo data D ori is used to initialize the generation of reference data D ref , wherein the original echo data D ori ={d1, d2, …, d L}, the initialization D ref =D ori . Wherein L is the number of sampling points of the radar original echo. Set the maximum number of iterations K of the algorithm. d L is the echo power value at point L.
[0059] Then, the background power data D bak is obtained.
[0060] In some optional embodiments, in step S1, the sliding window filtering processing includes:
[0061] Step S11, the reference data D refThe processing is performed to form pre-processed data;
[0062] In step S12, the pre-processed data is filtered using a k-order filter to obtain background power data D bak .
[0063] In some optional embodiments, step S11 includes:
[0064] In step S111, the power values of k consecutive sampling points in the reference data D ref are extracted in a step of k of a sliding window W[k];
[0065] In step S112, the k sliding window parameters in the sliding window W[k] are respectively weighted with the k extracted power values to form the pre-processed data, wherein k is an odd number, and the sliding window parameter values of the positions corresponding to the sample points to be processed in the reference data D ref and the surrounding 0 or more positions in the sliding window W[k] are set to 0, and the other sliding window parameter values are set to 1.
[0066] In this embodiment, the calculation process of step S112 is as shown in Figure 2 . Figure 2 The design parameters of the normalized reference statistical sample vector for the FIR filter are designed, and the formula of the k-order FIR filter is:
[0067] y(n) = h(1)x(n) + h(2)x(n-1) + h(3)x(n-2)…+h(k)x(n-k+1);
[0068] Wherein, h(k) is the weight of each order of the filter, and X[n] = [x(n-k+1), x(n-k+2),…x(n)] is the result of sliding window taking values from the vector D using the sliding window W. Wherein, W refers to the vector composed of the sliding window parameters W[k], and D refers to the vector composed of d n-k+1 to d n The sliding window parameters W[k] are 0 or 1, with 0 in the middle and 1 on both sides. In some optional embodiments, the step k of the sliding window W[k] is 3, 5 or 7, and the proportion of the sliding window parameters with a value of 1 is not less than 50%.
[0069] For example, when k = 3, W[1] = 1, W[2] = 0, W[3] = 1, and when k = 7, W[1] = 1, W[2] = 1, W[3] = 0, W[4] = 0, W[5] = 0, W[6] = 1, W[7] = 1. As mentioned before, d is the echo power value. For any n, for example n = 10, when k = 3, first form the vector D = [d8, d9, d 10 to d 10], by weighting vector D through sliding window W[3], we get X
[10] =[x(8),x(9),x(10)]=[d8,0,d 10 The filtered parameters y are calculated using the above k-order FIR filter formula. 10 ,like Figure 3 As shown.
[0070] It should also be noted that the data processed by the k-th order FIR filter will have a data length of [missing information - likely a variable value] compared to the original data. The time delay necessitates shifting the filter output forward. The background power data obtained after delay calibration of the data length is D bak .
[0071] Then, in step S2, the background power data is corrected, and a background power float threshold Thr is set. bak Determine D ref >D bak +Thr bak The data location. If there exists a data location that satisfies D. ref >D bak +Thr bak Data position n err , making D ref (n err ) = D bak (n err ).
[0072] After surfacing, the reference data initialized in step S1 is compared with the background power obtained in step S2. Abnormal data is filtered out from the reference data, and the target or clutter interference is replaced using the data at the corresponding position in the background power. Through multiple iterations up to the maximum number of iterations, the reference data after removing the target or clutter interference and the background power calculated from the reference data are finally obtained. Figure 5 The image shows a schematic diagram of the target detection results after removing target or clutter interference from the reference data. Correspondingly, Figure 4 The target detection results are shown before optimization.
[0073] Finally, in step S4, the signal-to-noise ratio (SNR) and signal-to-clutter ratio (SNR) thresholds for stationary target detection are set, the original echo data and the background power obtained from iterative processing are taken, the SNR and SNR of the signal are calculated, the target detection is completed, and the detection point is output.
[0074] In some alternative implementations, in step S4, the signal-to-noise ratio D is calculated using the following formula. SNR :
[0075] D SNR =D ori pnoise ;
[0076] The signal-to-clutter ratio D is calculated by the following formula SCR :
[0077] D SCR = D ori D bak ;
[0078] Where p noise is the noise power, and is the average of the power of each sample in the original echo data D ori .
[0079] In this embodiment, the signal-to-noise ratio detection threshold Thr SNR and the signal-to-clutter ratio detection threshold Thr SCR are set. The signal position n tar that satisfies is found. D ori (n tar ) is the original data of the detection point.
[0080] In a specific example, for a certain type of helicopter radar data, 6000 frames of echo data (each frame refers to the echo data scanned by the radar rotating an angle, from near to far, which constitutes the echo data D ori = {d1, d2, …, d L} of the present application), a total of 618 real targets are contained in the data. According to the detection method of the present application, 408 targets are correctly detected by the traditional method, and 544 targets are correctly detected by the present application. The detection accuracy of the present application is improved by 22.01% compared with the traditional algorithm; compared with the ordered statistical CFAR algorithm, the detection speed can be improved by more than 50%.
[0081] The method proposed in the present application has a greater performance improvement compared with the traditional CFAR detection method. First, by introducing the FIR filter and the method of removing outliers, the problem of the background power being raised by dense targets and abnormal clutter, resulting in a decrease in the signal-to-clutter ratio and the inability to detect, is avoided, so that multiple target detection can also be completed in the case of dense target distribution; second, the target detection speed of the present method is fast, and a general design idea is adopted, when the number of iterations is 1, the time complexity of the method is the same as that of the traditional method, when the number of iterations is greater than 1, the time complexity of the algorithm is linearly related to the number of iterations, and the algorithm has a faster operation speed compared with the improved target detection algorithm; third, the background power removed from the multiple target interference and the non-uniform clutter can be directly obtained in the algorithm, so the reference window in the traditional CFAR detection method is no longer needed, and the detection result of the algorithm is no longer limited by the size of the reference window.
[0082] The application can be applied in the field of radar target detection, and the proposed ground stationary target detection method can be used to improve the traditional CFAR detection method.
[0083] The second aspect of the application provides a ground dense stationary target rapid detection device corresponding to the above method, mainly comprising:
[0084] The background power data calculation module is configured to calculate the background power data D ori The reference data D ref The power values of each sampling point are subjected to sliding window filtering processing to obtain the background power data D bak The positions of the sampling points in the original echo data have a one-to-one correspondence with the spatial positions of the scanning points in the radar scanning range.
[0085] The background power up module is configured to up the background power values in the background power data D bak based on a given background power up threshold to obtain background power up values.
[0086] The background power data update module is configured to replace the power value of the sampling point in the reference data D ref with the background power value of the sampling point when the power value of the sampling point in the reference data D ref exceeds the background power up value of the corresponding sampling point to form new reference data, and to obtain the final background power data after multiple iterations.
[0087] The signal-to-noise ratio and signal-to-clutter ratio calculation module is configured to calculate the signal-to-noise ratio and signal-to-clutter ratio according to the original echo data D ori and the final background power data.
[0088] The target position determination module is configured to determine the position of the ground stationary target based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold.
[0089] In some optional embodiments, the background power data calculation module comprises:
[0090] The sliding window preprocessing unit is configured to process the reference data D ref using a preset sliding window W[k] to form preprocessed data.
[0091] The filtering processing unit is configured to perform filtering processing on the preprocessed data using a k-order filter to obtain the background power data D bak .
[0092] In some optional embodiments, the sliding window preprocessing unit comprises:
[0093] The power extraction subunit is configured to extract the power values of the sampling points in the reference data Dref extracting power values of k continuous sampling points;
[0094] The power parameter processing subunit is configured to perform weighted calculation on the k sliding window parameters in the sliding window W[k] and the k extracted power values respectively to form the preprocessed data, wherein k is an odd number, and the sliding window parameters corresponding to the positions of the sample points to be processed and the 0 or more positions around the positions in the sliding window W[k] are set as 0, and the other sliding window parameters are set as 1. ref
[0095] In some optional embodiments, the step k of the sliding window W[k] is 3, 5 or 7, and the proportion of the sliding window parameters with the value of 1 is not less than 50%.
[0096] In some optional embodiments, in the signal-to-noise ratio and signal-to-clutter ratio calculation module, the signal-to-noise ratio D SNR :
[0097] D SNR = D ori p noise ;
[0098] The signal-to-clutter ratio D SCR is calculated by the following formula:
[0099] D SCR = D ori D bak ;
[0100] wherein p noise is the noise power, and D ori is the average value of the power of each sampling point in the original echo data.
[0101] The above merely describes specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A method for fast detection of ground dense stationary targets, characterized in that, The method comprises the following steps: Step S1, obtaining original echo data D ori as reference data D ref , and performing sliding window filtering processing on the power value of each of the sampling points to obtain background power data D bak wherein the position of each sampling point in the original echo data has a one-to-one correspondence with the spatial position of the scanning point in the radar scanning range; Step S2, based on the given background power up threshold, up the background power value in the background power data D bak to obtain the background power up value; Step S3, when the power value of each sampling point in the reference data D ref is higher than the background power up-floating value of the corresponding sampling point, replace the power value of the sampling point in the reference data D ref with the background power value of the sampling point, form new reference data, and obtain the final background power data after multiple iterations. Step S4, calculating the signal-to-noise ratio and signal-to-clutter ratio from the original echo data D ori and the final background power data. Step S5, determining the position of the ground stationary target based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold.
2. The method of claim 1, wherein the step of detecting the ground dense stationary target is characterized by, In step S1, the sliding window filtering processing comprises: Step S11, the reference data D is processed by a preset sliding window W[k] to form preprocessed data. ref ; Step S12, filtering processing is performed on the pre-processed data using a k-order filter to obtain background power data D bak .
3. The rapid detection method for dense stationary targets on the ground as described in claim 2, characterized in that, Step S11 comprises: Step S111, extracting the power value of k continuous sampling points in the reference data D with the step k of the sliding window W[k]; ref Step S111, extracting the power value of k continuous sampling points in the reference data D with the step k of the sliding window W[k]; Step S112, the k window parameters in the sliding window W[k] are respectively weighted with the k extracted power values to form the preprocessed data, wherein k is an odd number, and the window parameters corresponding to the position of the sample point to be processed and the 0 or more positions around the position in the sliding window W[k] are set as 0, and the other window parameters are set as 1. ref The window parameters corresponding to the position of the sample point to be processed and the 0 or more positions around the position in the sliding window W[k] are set as 0, and the other window parameters are set as 1.
4. The method of claim 3, wherein the step of determining the presence of a ground dense stationary target comprises the steps of: determining the presence of a ground dense stationary target when the number of pixels in the first image is greater than the number of pixels in the second image. The step length k of the sliding window W[k] is 3, 5 or 7, and the proportion of the sliding window parameter with a value of 1 is not less than 50%.
5. The method of claim 1, wherein the step of detecting the ground dense stationary target is characterized by, In step S4, the signal-to-noise ratio D is calculated by the following equation SNR : D SNR = D ori p noise ; The signal-to-clutter ratio D is calculated by the following equation SCR : D SCR = D ori D bak ; where p noise is the noise power, and ori is the average of the power of each sample point in the original echo data D 6. A device for detecting ground dense stationary targets quickly, characterized in that, The method comprises the following steps: a background power data calculation module, configured to calculate background power data D ori as reference data D ref , and perform sliding window filtering on the power value of each of the sampling points to obtain the background power data D bak , wherein the positions of the sampling points in the original echo data have one-to-one correspondence with the spatial positions of the scanning points in the radar scanning range. A background power up module is configured to up the background power values in the background power data D bak based on a given background power up threshold to obtain background power up values. Background power data update module, used when reference data D ref When the power value of each sampling point exceeds the background power surcharge value of the corresponding sampling point, the background power value of that sampling point is used to replace the reference data D. ref The power value at the sampling point is used to form new reference data. After multiple iterations of calculation, the final background power data is obtained. A signal-to-noise ratio and signal-to-clutter ratio calculation module is configured to calculate the signal-to-noise ratio and signal-to-clutter ratio based on the original echo data D ori and the final background power data. The target position determination module is configured to determine the position of the ground stationary target based on the signal-to-noise ratio detection threshold and the signal-to-clutter ratio detection threshold.
7. The apparatus for fast detection of ground-hugging stationary targets of claim 6, wherein, The background power data calculation module comprises: A sliding window preprocessing unit is used to process reference data D through a preset sliding window W[k]. ref The data is processed to create preprocessed data. The filter processing unit is configured to perform filter processing on the pre-processed data using a k-order filter to obtain background power data D bak .
8. The apparatus for fast detection of ground-hugging stationary targets of claim 7, wherein, The sliding window preprocessing unit comprises: a power extraction sub-unit for extracting power values of k consecutive sampling points in the reference data D with a step k of the sliding window W[k] ref ; and The power parameter processing subunit is configured to perform weighted calculation on the k sliding window parameters in the sliding window W[k] and the k extracted power values respectively to form the preprocessed data, wherein k is an odd number, and the sliding window parameters corresponding to the position of the sample point to be processed and the 0 or more positions around the position are set as 0, and the other sliding window parameters are set as 1. ref The power parameter processing subunit is configured to perform weighted calculation on the k sliding window parameters in the sliding window W[k] and the k extracted power values respectively to form the preprocessed data, wherein k is an odd number, and the sliding window parameters corresponding to the position of the sample point to be processed and the 0 or more positions around the position are set as 0, and the other sliding window parameters are set as 1.
9. The apparatus for fast detection of ground-hugging stationary targets of claim 8, wherein, The step length k of the sliding window W[k] is 3, 5 or 7, and the proportion of the sliding window parameter with a value of 1 is not less than 50%.
10. The apparatus for fast detection of ground-hugging stationary targets of claim 6, wherein, In the signal-to-noise ratio and signal-to-interference ratio calculation module, the signal-to-noise ratio D is calculated by the following formula SNR : D SNR = D ori p noise ; The signal-to-clutter ratio D is calculated by the following equation SCR : D SCR = D ori D bak ; where p noise is the noise power, and D ori is the average of the power of each sample point in the original echo data.