Airborne radar moving clutter suppression method and device based on mask filtering

Abstract

The invention relates to the technical field of airborne meteorological radars, and discloses an airborne radar moving clutter suppression method and device based on mask filtering, which combine frequency domain and image domain signal processing methods, utilize an image segmentation technology to process a distance Doppler graph, combine inverted S-shaped curve characteristics of wind shear, and suppress the moving clutter of the airborne radar. A connected domain of wind shear is extracted from the distance Doppler map and used for constructing a mask filter binary mask, finally, the mask filter binary mask is used for filtering distance-Doppler domain data, non-meteorological targets such as motion clutters are filtered out, the distance Doppler map for suppressing ground clutters is obtained, and the method is simple, effective, low in calculation amount and high in reliability. The method is easy to implement, can effectively suppress motion clutters of expressways and the like near an airport runway, and reduces the probability of wind shear alarm missing of an airborne wind shear radar caused by the motion clutters of the expressways and the like.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of airborne weather radar, and discloses an airborne radar dynamic clutter suppression method and device based on mask filtering. BACKGROUND

[0002] Clutter can be divided into static clutter and dynamic clutter. Static clutter is generated by the backscattering of stationary ground targets, and the energy is mainly concentrated near zero frequency in the range-doppler diagram, which is the main target of clutter suppression and can be filtered out by high-pass filtering, space-time adaptive algorithm and other methods. The related research is quite mature at present. Dynamic clutter is generated by the backscattering of moving targets, such as cars or other moving objects on highways and high-flow ground. When the highway and wind shear targets overlap in space, such as Figure 1 Because the backscattering of ground clutter is stronger, according to the Doppler effect, the speed of the spatially overlapping region will eventually be analyzed into the speed of the moving clutter, and the gradient distribution of the “inverted S” speed curve of the wind shear in the direction will be destroyed, so the calculation of the danger factor for measuring the severity of the wind shear based on the speed gradient will be affected. When the highway lane is wide enough, it will affect the wind speed gradient of the wind shear warning area, resulting in wind shear false alarm.

[0003] In summary, airborne wind shear weather radar needs to have the function of suppressing dynamic clutter, especially the dynamic clutter caused by the highway parallel to the runway. However, there is little research on the suppression of moving clutter for airborne wind shear weather radar in China at present, and the current suppression method for ground clutter and discrete clutter cannot completely suppress the dynamic clutter caused by the highway. The current foreign research mainly suppresses the moving clutter through adaptive filtering, which has the following shortcomings: (1) Adaptive filtering is based on modern spectrum estimation, such as AR, MA, ARMA, etc., and uses LMS algorithm to complete the suppression of moving clutter. However, the performance of this method is related to the order of modern spectrum estimation and the number of algorithm iterations, and the calculation amount is very large. (2) LMS and other adaptive filtering algorithms need to be based on the reference signal to complete the suppression of moving clutter. Foreign researchers have conducted multiple collection experiments on highways at different time points to obtain the speed distribution characteristic curve of dynamic clutter near the corresponding airport over time, and based on the reference signal of the characteristic curve to suppress the moving clutter at different time points. However, this method requires research on the characteristics of moving clutter near each airport, and the characteristics of dynamic clutter at each airport are not exactly the same, which means that the characteristics of moving clutter need to be researched again for each new scene, which makes this method difficult to operate in engineering and not good at adapting to the environment. SUMMARY

[0004] The present application aims to provide a mask filtering based airborne radar moving clutter suppression method and device, which can effectively suppress the moving clutter such as highway near the airport runway, reduce the probability of wind shear false alarm of the airborne wind shear radar caused by the moving clutter such as highway, and has less calculation and is easy to implement.

[0005] In order to achieve the above technical effects, the technical scheme adopted by the present application is: A mask filtering based airborne radar moving clutter suppression method, comprising: performing FFT processing on the original echo data of the airborne wind shear radar, and performing static clutter suppression and interpolation compensation to obtain a range Doppler map M0; determining the signal noise level of the range Doppler map M0, and performing denoising processing on the range Doppler map M0 according to the signal noise level to obtain a denoised range Doppler map M0'; performing connected domain feature calculation on the denoised range Doppler map M0' to obtain a connected domain map M1 containing moving clutter and low-altitude wind shear weather; determining the connected domain feature corresponding to the low-altitude wind shear weather, and selecting a binary mask M2 containing only the low-altitude wind shear weather from the connected domain map M1 according to the connected domain feature; multiplying the binary mask M2 and the range Doppler map M0 before denoising to obtain a range Doppler map M3 after moving clutter suppression.

[0006] Further, the technical scheme adopted by the present application is: performing FFT processing on the original echo data, wherein: is the FFT processed radar echo frequency domain signal; is the radar original echo data, i.e. original IQ data Z_ori; is the frequency of the radar echo; is the imaginary unit.

[0007] Further, by performing static clutter suppression on the FFT processed radar echo data, wherein: is the filtered radar echo frequency domain signal; is a high-pass filter; is the FFT processed radar echo frequency domain signal.

[0008] Further, by performing interpolation compensation on the static clutter suppressed radar echo data, wherein: is a linear interpolation operation, is the high-pass filtered radar echo frequency domain signal.

[0009] Furthermore, the connected domain features include connected domain scale features, centroid location features, and angular features between the major axis and the horizontal axis of the connected domain, as well as the connected domain graph.

[0010] An airborne radar dynamic clutter suppression device based on mask filtering, used in the aforementioned airborne wind shear radar dynamic clutter suppression method, includes: The preprocessing module is used to perform FFT processing on the raw echo data of the airborne wind shear radar, and to perform static ground clutter suppression and interpolation compensation to obtain the range Doppler image M0. The range Doppler image denoising module is used to determine the signal noise level of the range Doppler image M0, and to denoise the range Doppler image M0 according to the signal noise level to obtain the denoised range Doppler image M0'. The feature calculation module is used to perform connected component feature calculation on the denoised range Doppler graph M0' to obtain a connected component graph M1 containing dynamic clutter and low-level wind shear meteorology. The mask acquisition module is used to determine the connected component features corresponding to low-level wind shear meteorology, and select a binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 based on the connected component features. The filtering module is used to multiply the binary mask M2 with the undenoised range Doppler image M0 to obtain the range Doppler image M3 after noise suppression.

[0011] Furthermore, adopt The raw echo data is processed using FFT, where: This is the radar echo frequency domain signal after FFT processing; This refers to the raw radar echo data, i.e., the raw IQ data Z_ori; The frequency of the radar echo; It is the imaginary unit.

[0012] Furthermore, through Static ground clutter suppression is performed on the radar echo data after FFT processing, where: This is the filtered radar echo frequency domain signal; It is a high-pass filter; This is the radar echo frequency domain signal after FFT processing.

[0013] Compared with the prior art, the beneficial effects of this invention are: This invention combines frequency domain and image domain signal processing methods, utilizes image segmentation technology to process range-Doppler images, and incorporates the "inverted S" curve characteristics of wind shear to extract the connected components of wind shear from the range-Doppler images. These components are then used to construct a binary mask for a mask filter. Finally, the range-Doppler domain data is filtered using this binary mask to remove non-meteorological targets such as moving clutter, resulting in a range-Doppler image that suppresses ground clutter. This method is simple, effective, computationally inefficient, and easy to implement. It can effectively suppress moving clutter such as highways near airport runways, reducing the probability of missed wind shear warnings by airborne wind shear radar due to moving clutter such as highways. Attached Figure Description

[0014] Figure 1 This is a flowchart of the airborne radar dynamic clutter suppression method based on mask filtering in the embodiment; Figure 2 Range Doppler image before ground clutter suppression; Figure 3 Range Doppler image after ground clutter suppression; Figure 4 The distance Doppler image M0 obtained through interpolation compensation; Figure 5 The distance Doppler image M0' is obtained after denoising the distance Doppler image M0; Figure 6 M1 is a connected domain diagram containing dynamic clutter and low-level wind shear meteorology; Figure 7 M2 is a binary mask containing only low-level wind shear weather. Figure 8 M3 is the range Doppler image obtained after suppressing dynamic clutter. Figure 9 This is a schematic diagram of the airborne radar dynamic clutter suppression device based on mask filtering in the embodiment. Detailed Implementation

[0015] The present invention will now be described in further detail with reference to the embodiments and accompanying drawings. However, this should not be construed as limiting the scope of the above-described subject matter of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0016] Example 1 See Figures 1-8 A method for suppressing dynamic clutter in airborne radar based on mask filtering, comprising: The raw echo data of the airborne wind shear radar is processed by FFT, and static ground clutter suppression and interpolation compensation are performed to obtain the range Doppler image M0. The signal noise level of the distance Doppler image M0 is determined, and the distance Doppler image M0 is denoised according to the signal noise level to obtain the denoised distance Doppler image M0'. Connectivity feature calculation is performed on the denoised distance Doppler graph M0' to obtain the connected component graph M1 containing dynamic clutter and low-level wind shear meteorology; Determine the connected component features corresponding to low-level wind shear meteorology, and select a binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 based on the connected component features; Multiply the binary mask M2 with the undenoised range Doppler image M0 to obtain the range Doppler image M3 after dynamic clutter suppression.

[0017] This embodiment presents an airborne radar clutter suppression method based on mask filtering. Combining frequency and image domain signal processing methods, it utilizes image segmentation to process the range-Doppler image. By incorporating the "inverted S" curve characteristics of wind shear, the connected components of wind shear are extracted from the range-Doppler image and used to construct a binary mask for a mask filter. Finally, the range-Doppler domain data is filtered using this binary mask to remove moving clutter and other non-meteorological targets, resulting in a range-Doppler image with suppressed ground clutter. This method is simple, effective, computationally inefficient, and easy to implement. It effectively suppresses moving clutter such as highways near airport runways, reducing the probability of missed wind shear warnings caused by highway clutter on airborne wind shear radar. Example 2 See Figure 1 A method for suppressing dynamic clutter in airborne radar based on mask filtering, comprising: Step 1: Perform FFT processing on the raw echo data of the airborne wind shear radar, and perform static ground clutter suppression and interpolation compensation to obtain the range Doppler image M0.

[0018] Specifically, firstly, adopt The original echo data is processed using FFT, which transforms the original radar echo data through Fourier transform to obtain the processed frequency domain signal. The processed frequency domain signal, This refers to the raw radar echo data, i.e., the raw IQ data Z_ori. The frequency of the radar echo. It is the imaginary unit.

[0019] Next, ground clutter suppression is performed, that is, the frequency domain signal is suppressed. High-pass filtering is performed to remove static ground clutter, obtaining the filtered frequency domain signal. The filtering expression is: ; in, This is a high-pass filter. For example... Figure 2 and Figure 3 As shown, Figure 2 The range Doppler image before ground clutter suppression. Figure 3 This is the range Doppler image after ground clutter suppression.

[0020] Then, linear interpolation compensation is performed on the low-frequency portion of the signal that was filtered out by the high-pass filter to obtain the range Doppler map M0, which is expressed as: ; in, For linear interpolation operations, This is the frequency domain signal after high-pass filtering. For example... Figure 4 As shown, Figure 4 The distance Doppler image M0 is obtained through interpolation compensation.

[0021] Step 2: Determine the signal noise level of the distance Doppler image M0, and perform denoising processing on the distance Doppler image M0 according to the signal noise level to obtain the denoised distance Doppler image M0'.

[0022] Specifically, firstly, select an area with no significant weather or clutter at a distance of M0 from the Doppler graph. The size of this area is... , , Given the length and width of the region, then... Calculate the average amplitude of all elements within the region, where, For the region The noise level within a given area is equal to the average amplitude of all elements within that area, and can represent the noise level over the entire distance Doppler curve M0. The function for finding the average value. This represents the amplitude of each element within the region.

[0023] Then, based on the noise level The meteorological signal and other non-static clutter are extracted to obtain the binarized data, i.e., the denoised range Doppler image M0'. The expression for extracting the meteorological signal and other non-static clutter is as follows: ; in, Let M0 be the coordinates of each element on the Doppler graph. These are the coordinates of each element on the denoised distance Doppler map M0'. For example... Figure 5 As shown, Figure 5 The distance Doppler image M0' is obtained after denoising the distance Doppler image M0.

[0024] Step 3: Perform connected component feature calculation on the denoised distance Doppler graph M0' to obtain the connected component graph M1 containing dynamic clutter and low-level wind shear meteorology.

[0025] Specifically, the connected component features of binarized data include connected component scale features, centroid location features, angular features between the major axis and the horizontal axis of the connected component, and the connected component graph M1; the expression for calculating the connected component features of the denoised distance Doppler graph M0' in this embodiment is as follows: ; in, The scale feature of the connected component. The centroid location feature of a connected domain. This represents the angular feature between the major axis and the horizontal axis of the connected domain. For a connected component graph, For connected component feature extraction operations, This is the denoised distance Doppler plot. It should be noted that there are multiple connected components in plot M1, including meteorological information such as dynamic clutter and low-level wind shear. Figure 6 As shown, Figure 6 M1 is a connected domain diagram containing dynamic clutter and low-level wind shear meteorology.

[0026] Step 4: Determine the connected component features corresponding to the low-level wind shear meteorology, and select a binary mask M2 containing only the low-level wind shear meteorology from the connected component graph M1 based on the connected component features.

[0027] Specifically, since the scale of wind shear meteorology is often larger than that of dynamic clutter, height clutter, and sidelobe clutter, the scale characteristics of the connected domain are used as reference features for screening wind shear meteorology. Simultaneously, because the centroid of wind shear meteorology is closer to zero frequency after low-frequency linear interpolation compensation, the centroid position characteristics of the connected domain are used as reference features for screening wind shear meteorology. Furthermore, due to the inverted S-shaped velocity curve of wind shear meteorology, its angle with the horizontal axis in the connected domain diagram M1 ranges from -20° to -80°; therefore, the angle between the major axis of the connected domain and the horizontal axis is also used as a reference feature for screening wind shear meteorology.

[0028] Therefore, based on the features obtained from the connected component feature calculation of the denoised distance Doppler graph M0' in step three, the scale features, centroid location features, and the angle features between the major axis and the horizontal axis of the connected component are used as reference features for screening wind shear meteorological data. Thus, a binary mask M2 containing only low-level wind shear meteorological data is selected from the connected component graph M1. For example... Figure 7 As shown, Figure 7 M2 is a binary mask containing only low-level wind shear weather.

[0029] Step 5: Multiply the binary mask M2 with the undenoised range-Doppler image M0 to obtain the range-Doppler image M3 after noise suppression, which is the filtered result. Its expression is: ; in, The coordinates of each element in the range Doppler plot M3 after noise suppression are given. Here are the coordinates of each element in the distance Doppler image M0 before denoising. These are the coordinates of each element in the binary mask M2. For example... Figure 8 As shown, Figure 8 M3 is the range Doppler image obtained after suppressing dynamic clutter.

[0030] Based on the same inventive concept, see [link to inventive concept] Figure 9 This embodiment also provides an airborne radar dynamic clutter suppression device based on mask filtering, used to implement the aforementioned airborne wind shear radar dynamic clutter suppression method, which includes: The range Doppler image generation module is used to perform FFT processing on the raw echo data of the airborne wind shear radar, and to perform static ground clutter suppression and interpolation compensation to obtain the range Doppler image M0. The range Doppler image denoising module is used to determine the signal noise level of the range Doppler image M0, and to denoise the range Doppler image M0 according to the signal noise level to obtain the denoised range Doppler image M0'. The connected domain graph module is used to calculate the connected domain features of the denoised distance Doppler graph M0' to obtain a connected domain graph M1 containing dynamic clutter and low-level wind shear meteorology; the connected domain features include the directional and scale features of each connected region within the connected domain.

[0031] A mask acquisition module is used to determine the connected component features corresponding to low-level wind shear meteorology, and select a binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 based on the connected component features; the method for selecting the binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 includes: In the connected domain graph M1, determine the directional characteristics, scale characteristics, and center point location of the connected regions corresponding to low-level wind shear meteorology; Using the directional features, the scale features, and the location of the center point of the connected region as filtering conditions, connected regions that meet the filtering conditions are selected in the connected region graph M1 and output as a binary mask M2.

[0032] The filtering module is used to multiply the binary mask M2 with the undenoised range Doppler image M0 to obtain the range Doppler image M3 after noise suppression.

[0033] This invention combines frequency domain and image domain signal processing methods, utilizes image segmentation technology to process range-Doppler images, and incorporates the "inverted S" curve characteristics of wind shear to extract the connected components of wind shear from the range-Doppler images. These components are then used to construct a binary mask for a mask filter. Finally, the range-Doppler domain data is filtered using this binary mask to remove non-meteorological targets such as moving clutter, resulting in a range-Doppler image that suppresses ground clutter. This method is simple, effective, computationally inefficient, and easy to implement. It can effectively suppress moving clutter such as highways near airport runways, reducing the probability of missed wind shear warnings by airborne wind shear radar due to moving clutter such as highways.

[0034] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for suppressing moving clutter in airborne radar based on mask filtering, characterized in that, include: The raw echo data of the airborne wind shear radar is processed by FFT, and static ground clutter suppression and interpolation compensation are performed to obtain the range Doppler image M0. The signal noise level of the distance Doppler image M0 is determined, and the distance Doppler image M0 is denoised according to the signal noise level to obtain the denoised distance Doppler image M0'. Connectivity feature calculation is performed on the denoised distance Doppler graph M0' to obtain the connected component graph M1 containing dynamic clutter and low-level wind shear meteorology; Determine the connected component features corresponding to low-level wind shear meteorology, and select a binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 based on the connected component features; Multiply the binary mask M2 with the undenoised range Doppler image M0 to obtain the range Doppler image M3 after dynamic clutter suppression.

2. The airborne radar dynamic clutter suppression method based on mask filtering according to claim 1, characterized in that, use The raw echo data is processed using FFT, where: This is the radar echo frequency domain signal after FFT processing; This is the raw radar echo data; The frequency of the radar echo; It is the imaginary unit.

3. The airborne radar dynamic clutter suppression method based on mask filtering according to claim 2, characterized in that, pass Static ground clutter suppression is performed on the radar echo data after FFT processing, where: This is the filtered radar echo frequency domain signal; It is a high-pass filter; This is the radar echo frequency domain signal after FFT processing.

4. The airborne radar dynamic clutter suppression method based on mask filtering according to claim 3, characterized in that, pass Interpolation compensation is performed on the radar echo data after static ground clutter suppression, where: For linear interpolation operations, This is the radar echo frequency domain signal after high-pass filtering.

5. The airborne radar dynamic clutter suppression method based on mask filtering according to claim 1, characterized in that, The connected domain features include connected domain scale features, centroid location features, angular features between the major axis and the horizontal axis of the connected domain, and the connected domain graph.

6. An airborne radar dynamic clutter suppression device based on mask filtering, used to implement the airborne wind shear radar dynamic clutter suppression method according to any one of claims 1-5, characterized in that, include: The preprocessing module is used to perform FFT processing on the raw echo data of the airborne wind shear radar, and to perform static ground clutter suppression and interpolation compensation to obtain the range Doppler image M0. The range Doppler image denoising module is used to determine the signal noise level of the range Doppler image M0, and to denoise the range Doppler image M0 according to the signal noise level to obtain the denoised range Doppler image M0'. The feature calculation module is used to perform connected component feature calculation on the denoised range Doppler graph M0' to obtain a connected component graph M1 containing dynamic clutter and low-level wind shear meteorology. The mask acquisition module is used to determine the connected component features corresponding to low-level wind shear meteorology, and select a binary mask M2 containing only low-level wind shear meteorology from the connected component graph M1 based on the connected component features. The filtering module is used to multiply the binary mask M2 with the undenoised range Doppler image M0 to obtain the range Doppler image M3 after noise suppression.

7. The airborne radar dynamic clutter suppression device based on mask filtering according to claim 6, characterized in that, use The raw echo data is processed using FFT, where: This is the radar echo frequency domain signal after FFT processing; This is the raw radar echo data; The frequency of the radar echo; It is the imaginary unit.

8. The airborne radar dynamic clutter suppression method based on mask filtering according to claim 7, characterized in that, pass Static ground clutter suppression is performed on the radar echo data after FFT processing, where: This is the filtered radar echo frequency domain signal; It is a high-pass filter; This is the radar echo frequency domain signal after FFT processing.

Images