Millimeter wave radar speckle removal method, device, medium and program product based on spectrum matching

Abstract

The application provides a millimeter wave radar clutter point removal method and device based on spectrum matching, a medium and a program product. The method comprises the following steps: acquiring a multi-channel echo signal, constructing a measured angle spectrum reflecting the energy distribution of each direction through beam forming and angle traversal processing; determining each effective target angle of the measured angle spectrum; based on the geometric structure parameters of the millimeter wave radar array and the effective target angle, weighting and superimposing each target response based on a theoretical echo signal model to obtain a theoretical angle spectrum; performing spectrum matching analysis on the theoretical angle spectrum and the measured angle spectrum, performing point-by-point translation processing on the theoretical angle spectrum along the angle axis, and calculating the cross-correlation coefficient distribution of the two at different angle offsets; when the cross-correlation reaches a maximum value at zero angle offset and is greater than a threshold value, it is determined that the corresponding measured angle spectrum is effective, otherwise it is regarded as a clutter point and removed. The application realizes accurate identification of real targets by using the change characteristics of the cross-correlation coefficient at different angle offsets.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar signal processing, in particular to a millimeter wave radar clutter point removal method and device based on spectrum matching, a medium and a program product. BACKGROUND

[0002] With the rapid development of automatic driving and intelligent perception technology, vehicle-mounted radars have become one of the important sensors for realizing environmental perception, path planning and target detection. The radar can obtain the distance, speed and angle information of the surrounding objects by transmitting and receiving electromagnetic waves, and output the environmental perception results in the form of point cloud. However, in actual application, the vehicle-mounted radar point cloud data often contains non-real reflection points caused by factors such as multipath reflection, sidelobe interference or system noise. These clutter points cause the spatial distribution characteristics of the point cloud data to deviate, which seriously affects the accuracy and reliability of the point cloud, and further reduces the accuracy of target recognition, path planning and obstacle avoidance control.

[0003] In the prior art, the radar point cloud clutter point removal method mainly includes algorithms based on spatial clustering and trajectory tracking. For example, the commonly used DBSCAN clustering method realizes clutter point rejection by counting the adjacent density of the point cloud, and the method based on track tracking relies on the clustered point cluster to establish the target trajectory. However, these methods all rely on multiple spatially adjacent sampling points in the point cloud, and when the target only produces a single reflection point (such as a weak target or a long-distance target), clustering cannot be formed, resulting in the real target being mistakenly deleted.

[0004] Therefore, under complex environment, it is urgent to accurately identify and effectively filter the clutter points in the vehicle-mounted radar point cloud, so as to improve the stability and reliability of the point cloud data. SUMMARY

[0005] In view of the deficiencies in the prior art, the present application provides a millimeter wave radar clutter point removal method and device based on spectrum matching, a medium and a program product, at least to solve the problem that it is difficult to effectively distinguish real targets and noise points from the signal level in the prior art, resulting in insufficient radar point cloud clutter point determination accuracy and limited weak target recognition capability.

[0006] In order to achieve the above-mentioned purpose and other advantages, some embodiments of the present application provide the following aspects:

[0007] In a first aspect, some embodiments of the present application provide a millimeter wave radar clutter point removal method based on spectrum matching, comprising:

[0008] Obtaining the multi-channel echo signal of the millimeter wave radar array, and generating a measured angle spectrum reflecting the power response of each angle direction through beam forming and angle traversal processing;

[0009] perform peak searching on the measured angle spectrum to determine effective target angles corresponding to each power peak;

[0010] According to geometric structure parameters of the millimeter wave radar array and the effective target angles, each target response is weighted and superimposed based on a theoretical echo signal model to obtain a theoretical angle spectrum for representing ideal target angles and peak response characteristics;

[0011] The theoretical angle spectrum and the measured angle spectrum are subjected to spectrum matching analysis, the theoretical angle spectrum is subjected to point-by-point translation processing along an angle axis within a preset angle offset range, and a cross-correlation coefficient distribution of the two at different angle offsets is calculated.

[0012] Based on the cross-correlation coefficient distribution, a cross-correlation coefficient corresponding to each measured angle spectrum is determined, when the cross-correlation coefficient reaches a maximum at zero angle offset and is greater than a preset threshold, it is determined that the corresponding measured angle spectrum is effective, otherwise it is determined to be a spur and is subjected to removal processing.

[0013] In a second aspect, some embodiments of the present application further provide an electronic device, which comprises:

[0014] one or more processors; and a memory storing computer program instructions which, when executed, cause the processor to perform the millimeter wave radar spur removal method based on spectrum matching as described above.

[0015] In a third aspect, some embodiments of the present application further provide a computer readable storage medium having stored thereon computer programs and / or instructions which, when executed by a processor, implement the millimeter wave radar spur removal method based on spectrum matching as described above.

[0016] In a fourth aspect, some embodiments of the present application further provide a computer program product comprising computer programs and / or instructions which, when executed by a processor, implement the millimeter wave radar spur removal method based on spectrum matching as described above.

[0017] Compared with the prior art, in the scheme provided by the embodiment of the application, the frequency spectrum matching relationship between the theoretical angle spectrum and the measured angle spectrum is constructed, and the phase consistency of the target echo in the array dimension is judged by using the cross-correlation coefficient distribution, thereby significantly improving the distinguishing ability of the real target and the noise speckle in the millimeter wave radar point cloud. Since the theoretical angle spectrum is constructed based on the array geometry and the target angle, it is not affected by the environmental noise, multipath reflection and sidelobe interference, and the measured angle spectrum is obtained by beamforming of the multi-channel echo and can reflect the phase law of the real target, therefore whether a significant correlation peak is formed at the zero offset between the two can directly reflect the matching degree of the target echo and the theoretical model. The application utilizes the change characteristics of the cross-correlation coefficient at different angle offsets to realize accurate identification of the real target, so that the effective target can still be accurately retained in the weak reflection target, long distance target or sparse point cloud scene containing only a single point echo, while the speckle caused by random noise, sidelobe false alarm or multipath scattering is effectively filtered out, and the overall quality and reliability of the point cloud data are significantly improved. Thanks to the fact that the frequency spectrum matching does not depend on the point cloud density, clustering result or tracking trajectory, the method still maintains high stability and robustness in complex environments, multi-target scenes and dynamic working conditions, and can directly complete the speckle removal in the signal processing stage, thereby reducing the misjudgment rate of the subsequent algorithm and improving the overall processing efficiency of the system. In summary, the speckle removal method provided by the application can realize high-precision radar point cloud noise suppression in multiple types of scenes, and provide more reliable data input for target recognition, path planning and collision detection tasks in autonomous driving. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other embodiments can be obtained by those skilled in the art without creating any inventive labor.

[0019] Figure 1 is one of the flowcharts of the millimeter wave radar speckle removal method provided by the embodiment of the application based on frequency spectrum matching;

[0020] Figure 2 is the second flowchart of the millimeter wave radar speckle removal method provided by the embodiment of the application based on frequency spectrum matching;

[0021] Figure 3 is the cross-correlation coefficient distribution between the noise point and the theoretical angle spectrum provided by the embodiment of the application;

[0022] Figure 4 is the cross-correlation coefficient distribution between the real target and the theoretical angle spectrum provided by the embodiment of the application;

[0023] Figure 5 is a performance curve diagram provided by the embodiment of the present application, which shows the change of the target judgment accuracy rate with the signal-to-noise ratio under different signal-to-noise ratio conditions.

[0024] Figure 6 is a structural schematic diagram of an electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0026] First Embodiment

[0027] The first embodiment of the present application relates to a millimeter wave radar clutter point removal method based on spectrum matching. As shown in Figure 1 、 Figure 2 , the method can include the following steps:

[0028] Step S1: Obtain the multi-channel echo signal of the millimeter wave radar array, and generate a measured angle spectrum reflecting the power response of each angle direction through beam forming and angle traversal processing.

[0029] In this embodiment, step S1 specifically includes:

[0030] Step S101: Obtain the echo signal received by the plurality of receiving antennas of the millimeter wave radar array;

[0031] Step S102: Weighted sum the echo signal to form a synthesized signal output corresponding to the angle;

[0032] Step S103: Traverse the scanning angle within a preset angle range, and dynamically adjust the weighting coefficient of each receiving antenna according to the scanning angle, and calculate the output power of the synthesized signal under each angle;

[0033] Step S104: Form the output power corresponding to each scanning angle into a distribution result of the power changing with the angle, as a measured angle spectrum, for characterizing the energy response characteristics of the echo signal in different directions.

[0034] Specifically, the millimeter wave radar adopts an antenna array structure composed of N elements, and the element spacing is d. When the target signal propagates to each element in space, different phases will be formed on each element due to the path difference. Let the azimuth angle of the target be The phase difference of the signal received at the nth array element relative to the reference array element can be represented as:

[0035]

[0036] wherein, is the phase difference, and λ is the wavelength of the signal.

[0037] Since the target echo is affected by different propagation path lengths when propagating in space to each array element of the array, the phases of the signals formed at different array elements are not the same. In other words, the signal received by each array element can be regarded as the original echo signal of the target superimposed with a phase delay caused by the difference in array element positions. Based on the definition of the phase delay, the echo signal received by the nth array element can be represented as:

[0038]

[0039] wherein, is the echo signal received by the nth array element; is the amplitude of the original echo signal of the target; is a phase factor varying with the array element number, used to reflect the relationship between the incident direction of the target and the array geometry structure.

[0040] Substituting the expression (1) of the aforementioned phase difference into the expression (2), the following expression can be obtained:

[0041]

[0042] This expression explicitly reveals the angle-dependent characteristics of the array received signal: when the incident angle of the target changes, the phases of the signals received by each array element will change synchronously according to the linear law of the array element positions. It is this phase characteristic varying with the angle that enables the subsequent beamforming processing to construct the directional response at different scanning angles, thereby realizing the target direction estimation and angle spectrum generation.

[0043] Based on the above array received signal model, in order to extract the direction information of the target from the signals received by multiple array elements, the signals of each array element need to be weighted and combined according to a preset scanning angle, so as to form an array output with direction selectivity. In the present embodiment, by designing a corresponding weighting coefficient for the scanning angle , the signals received by each array element can be coherently superimposed in this direction. Let the weighting coefficient at the nth array element be:

[0044]

[0045] wherein, is the weighting coefficient; ​Assuming the target is located at an angle The theoretical phase distribution that the time array should have.

[0046] By performing a weighted summation of the received signals, the array output corresponding to this scanning angle can be obtained:

[0047]

[0048] The weighted array output represents the scanning angle. The directional response of the lower array. This output reflects the target signal at the scanning angle. The coherent superposition result under the following conditions. When Compared with the true angle of incidence When they are in sync, the multi-element signals will achieve phase alignment, and the array output amplitude will reach its maximum. When they are out of sync, the phase superposition will partially cancel each other out, and the output amplitude will be weakened.

[0049] Substituting equations (2) and (4) into equation (5), we obtain the beam direction response model under single-target conditions:

[0050]

[0051] To make the formula more intuitive, you can... After proposing and combining the exponent terms, equation (6) can be rewritten equivalently as:

[0052]

[0053] As can be seen from equation (7), the directional response output after beamforming is... It can be equivalently represented as the target original echo amplitude. The product of the array factor and the array element, which is determined by the array structure. The magnitude of the array factor is influenced by the actual phase difference at each element. Corresponding scanning angle Preset estimated phase difference The difference between them determines the magnitude of the array factor. The smaller the difference, the larger the magnitude of the array factor, and the stronger the directivity of the array output.

[0054] At the array scanning angle true angle of incidence with the target When consistent, the estimated phase difference at each array element The true phase difference formed with the target incidence A perfect match, meaning that the following conditions are met:

[0055]

[0056] Under these conditions, the weighting coefficients The corresponding phase compensation can precisely offset the phase delay caused by the signal propagation between the array elements, so that the echo signals from all array elements are in phase before weighted summation. Since the array output signal remains coherent during spatial superposition, the amplitude of the array output reaches the theoretical extreme value, and the corresponding reaches the maximum value.

[0057] When there are K spatial targets in the detection range of the millimeter wave radar at the same time, each target will produce an independent echo signal to each receiving element of the array, and each echo carries a phase difference related to the incident direction of the target. Therefore, the echo signal received by the array at the nth receiving channel is no longer the contribution of a single target, but the superposition result of all target echoes at this element.

[0058] Specifically, let the echo intensity of the kth target be , and the spatial incident azimuth angle be , according to the array geometric relationship, the theoretical phase difference formed by it at the nth element is :

[0059]

[0060] Therefore, the received signal of the kth target at the nth element is :

[0061]

[0062] Under the condition of ignoring noise and other interference, the total received signal of the nth element of the array is the superposition of all target echoes:

[0063]

[0064] The actually collected data is composed of the phase superposition of target echoes at the element and noise interference, reflecting the receiving situation of the millimeter wave radar array in the real use scenario.

[0065] In the digital beam forming process, the total signal obtained by each receiving unit of the array is applied with a preset weighting coefficient , and the weighted signal is coherently superimposed in space, so as to obtain the output response of the array in the direction of , that is, the output signal after beam forming is the weighted summation of all antenna receiving signals, which can be expressed as:

[0066]

[0067] Applying each array element received signal to a weighting coefficient under a scanning angle Substitute equation (11) into equation (12), where is the phase difference of the scanning angle , and sum the weighted signals to obtain the array output corresponding to the scanning angle:

[0068]

[0069] By rearranging the summation order in equation (12), the array output can be expressed as the superposition of the individual responses to each target:

[0070]

[0071] where the summation term in the brackets represents the array response or array factor, which describes the gain of the radar array to the signal from the direction when the beam is directed at .

[0072] For ease of description, equation (13) is further written as follows:

[0073]

[0074] where

[0075]

[0076] represents the single-target beamforming output for the true target direction when the array scanning direction is . Therefore, the overall array output in a multi-target scenario is equivalent to the linear superposition of multiple single-target directional responses.

[0077] According to the above directional response model, when the scanning angle is close to the true azimuth angle of a target, the array factor in equation (15) will reach a large amplitude, resulting in a significant power peak in the overall output . Based on this characteristic, the present embodiment performs point-by-point scanning of within a preset angle interval and calculates the corresponding directional response output amplitude or power to obtain the measured angle spectrum as a function of angle.

[0078] Specifically, the scanning range can be set to cover the azimuth angle interval in which the target may appear (e.g., [-90°, 90°]), and the directional response output is calculated for each candidate scanning angle ​The direction response power under each scanning angle is arranged in angle sequence, that is, an output power sequence varying with angle is formed. The power sequence represents the echo energy distribution of the millimeter wave radar array in different spatial directions, and can be used to construct a measured angle spectrum reflecting the directivity characteristics of the array.

[0079] In the measured angle spectrum, the target real azimuth angle usually corresponds to an obvious main lobe peak value, and the sidelobes inherent in the array structure can be expressed as smaller secondary peaks. By performing peak value searching on the angle spectrum, the angle positions corresponding to each power peak can be identified, and the valid peak values are further screened based on the power difference between the main lobe and the sidelobe, so as to determine the azimuth angle estimation results of multiple targets.

[0080] Through the processing of steps S101 to S104, the spatial phase difference characteristics and the digital beam forming mechanism of the millimeter wave radar array are used to construct the energy response distribution varying with the scanning angle on the basis of the multi-channel echo signal, and the incident direction of the target is accurately characterized. By dynamically adjusting the weighting coefficients of each array element in the preset angle range and coherently superimposing the multi-array element signals, a power response curve with obvious direction discrimination can be formed, which presents a significant peak value at the real direction of the target, so that the identification of the target direction is more intuitive and reliable.

[0081] Further, by traversing the full range of scanning angles, a stable measured angle spectrum can be constructed without relying on the point cloud density, the number of clusters or the track initialization condition. The angle spectrum can accurately reflect the energy distribution of the echo signal in each direction in space. Since the noise signal does not have stable phase consistency in the spatial domain, it generally presents a low-amplitude or irregular response in the angle spectrum, so it is easier to distinguish from the real target in subsequent processing, effectively enhancing the accuracy of the clutter filtering and the robustness of the overall processing link.

[0082] Step S2: performing peak value searching on the measured angle spectrum to determine the effective target angles corresponding to each power peak.

[0083] In this embodiment, step S2 specifically includes:

[0084] Step S201: traversing the power distribution curve of the measured angle spectrum to identify the angle positions corresponding to the local maximum values of the power;

[0085] Step S202: taking the peaks within a preset threshold range from the global maximum value of the power as candidate target angles;

[0086] Step S203: comparing the amplitudes of each candidate target angle, when the power difference between adjacent peaks is less than a preset sidelobe suppression threshold, retaining the angle corresponding to the main peak and suppressing the sidelobe peak, and taking the angle corresponding to the peak after the retention processing as the effective target angle.

[0087] Specifically, the amplitude or power of the array output corresponding to each calculated scanning angle is taken out, and a set of power sequences varying with angle is formed in the order of scanning angle. The power sequence reflects the echo energy distribution of the millimeter wave radar array in different directions, denoted as the measured angle spectrum . Wherein, The value range of ∈[-π / 2,π / 2] can be set to cover the direction interval where the target may appear, for example

[0088] The measured angle spectrum is traversed in the order of scanning angle from small to large, and the local variation gradient of each angle point is calculated. When the power value of a certain angle point is greater than the power values of its left and right adjacent angle points at the same time, the angle point is identified as a local peak. For each local peak position, the angle corresponding to the peak value can be recorded, for example, the main peak angle is denoted as , and then the second peak angle , the third peak angle , and the th peak angle can be obtained in turn, thereby forming a candidate peak angle set.

[0089] In order to distinguish the true target peak value from the noise and the pseudo-peak caused by the sidelobe, the relative threshold strategy is adopted in this embodiment. First, the global maximum power value in the measured angle spectrum is calculated , and a magnitude decay threshold (for example, 3dB) is set. For all local peaks, if the peak power satisfies: , it is considered that the peak value has sufficient energy contribution and may correspond to a real target, and the angle is taken as a candidate target angle; if the power is lower than the threshold, it is ignored as a noise peak or a sidelobe. Through the above threshold screening, the pseudo-peak caused by the array structure sidelobe can be effectively excluded, and the reliability of target recognition is improved.

[0090] In order to avoid multiple adjacent candidate peaks from producing repeated targets due to spatial correlation, the sidelobe suppression processing is performed on the above candidate peaks. Specifically, the peaks can be sorted in descending order of peak power, and the main peak angle with the highest power is first determined as the effective target angle. Then, the power difference between the remaining candidate peaks and the main peak is judged in turn whether it is greater than a preset sidelobe suppression threshold (for example, 6dB or an array design related index). If the power difference between a certain candidate peak and the nearest larger peak is not enough to distinguish, it is determined as a sidelobe peak and is excluded; otherwise, it is retained as a new effective target angle. Through the above processing, it can be ensured that the effective target angles obtained finally all correspond to real physical targets, and false angles caused by sidelobes are avoided.

[0091] Through the peak searching process of steps S201 to S203, the present application can accurately retain the main peaks in the measured angle spectrum, while reliably suppressing the sidelobe peaks and noise peaks, thereby significantly improving the accuracy and robustness of target angle extraction. The peak screening mechanism ensures that only real targets are subsequently subjected to theoretical angle spectrum construction and spectrum matching analysis, providing high-quality candidate target input for clutter removal and avoiding false peak propagation to subsequent stages to cause misjudgment, thereby improving the reliability and processing efficiency of the millimeter wave radar point cloud as a whole.

[0092] Step S3: Based on the geometric structure parameters of the millimeter wave radar array and the effective target angles, the theoretical model of each target response is weighted and superimposed based on the theoretical echo signal model to obtain a theoretical angle spectrum for characterizing the ideal target angle and peak response characteristics.

[0093] In this embodiment, step S3 specifically includes:

[0094] Step S301: According to the antenna layout and element spacing of the millimeter wave radar array, a theoretical model for describing the array response characteristics of a single target at different incident angles is established;

[0095] Step S302: The effective target angles are combined with the theoretical model to construct a theoretical model matrix for describing the theoretical array response of multiple effective targets at their corresponding angle positions;

[0096] Step S303: Based on the theoretical model matrix, the actual echo intensity parameters of each target are calculated in combination with the measured angle spectrum;

[0097] Step S304: The echo intensity parameters are associated with the corresponding effective target angles as amplitude weights, and the theoretical array responses of each target are weighted and superimposed to form an accumulated theoretical model of the multi-target echo;

[0098] Step S305: The energy distribution of the accumulated theoretical model in the angle domain is taken as the theoretical angle spectrum.

[0099] Specifically, after determining the geometric structure of the millimeter wave radar array (including the number of elements N, the element spacing d, the carrier frequency, and the wavelength λ), the present embodiment first constructs a theoretical model of the array response characteristics of a single target at different incident angles based on the spatial layout of the array antenna (corresponding to formula (15)). This theoretical model describes the directional gain distribution of the array formed by the target echo from the incident angle when the array beam is directed at the incident angle, and is the basic model for constructing the theoretical angle spectrum.

[0100] According to the K effective target angles obtained in step S2, the theoretical model of step S301 is combined with the K effective target angles to construct a theoretical model matrix for describing the theoretical array response of the K effective targets at their corresponding angle positions. ​​​​, …, Each angle is substituted into equation (15) to construct a theoretical model matrix for describing the theoretical array response of K effective targets at their corresponding angle positions:

[0101]

[0102] where each column corresponds to an effective target direction, describing the array output characteristics of the target at different scanning angles, is the target direction The contribution vector of the entire angle domain theoretical echo distribution. Each row corresponds to a scanning angle, reflecting the theoretical gain distribution of the array to all candidate target directions at that scanning angle.

[0103] Therefore, using the measured effective target angle as the angle input for theoretical modeling makes the constructed theoretical model matrix accurately describe the array phase distribution characteristics of each target at its real azimuth angle position. If the angle used in the theoretical model does not match the actual target angle, the corresponding array response will not match the real physical scene, ultimately leading to the inability of the theoretical angle spectrum to form effective spectral matching with the measured angle spectrum, thereby failing to correctly distinguish real targets from noise points.

[0104] Theoretical model matrix has described the single-target directional response of the "unit strength target" at each effective target angle , , …, When the actual echo strength of the first target is , the contribution of this target to the array output at each angle can be considered as the first column scaled by the amplitude ; when K targets exist simultaneously, the total angular response is the weighted superposition of each column by . Therefore, the measured angle spectrum sampling value can be expanded in matrix form at K effective target angles and the following equation relationship is established between the theoretical model matrix :

[0105]

[0106] where , , …, indicates the angle position corresponding to the K effective targets identified in step S2; the of the matrix first column is the theoretical directional response when the real azimuth angle of the target is , at the scanning angle the values of the measured angle spectrum at the K effective target angles, corresponding to the theoretical single-target directional response obtained by formula (16); and the actual echo intensity parameters (i.e., target echo amplitudes) of the K targets located at angles , , …, .

[0107] Formula (16) shows that the values of the measured angle spectrum at the K effective target angles can be regarded as the result of the weighted superposition of the theoretical directional responses of each target according to the echo intensity parameters. Based on this linear relationship, the actual echo intensity parameters of each target can be solved . After obtaining the echo intensity parameters, to construct a theoretical angle spectrum that can cover the entire angle domain, the embodiment will scan the angle in a preset search range (for example, the values are taken point by point, and each scanning angle is substituted into the accumulated theoretical model of the multi-target echo constructed based on formula (14) to calculate the output amplitude or power at the angle. By sorting the output sequence obtained by traversing the above angles according to the scanning angle, a complete theoretical angle spectrum can be formed, which is used to describe the energy distribution of the multi-target echo in the direction domain under ideal conditions.

[0108] Through the processing of steps S301 to S305, a multi-target theoretical angle spectrum consistent with the physical characteristics of the array can be constructed under the condition that the array parameters are known and the preliminary angle of the target is determined. The theoretical angle spectrum can accurately describe the directional response characteristics that the multiple targets should present in the array dimension under ideal conditions, so that it obtains a stable reference mode that is not disturbed by environmental factors such as noise, multipath, sidelobe, etc. With the help of the theoretical reference angle spectrum, subsequent frequency domain matching can be performed between the measured angle spectrum and the theoretical reference angle spectrum, thereby significantly enhancing the distinguishability between different targets, improving the recognition ability of sidelobe false peaks, isolated noise points and weak echo interference, and making the angle and energy distribution of the real target clear.

[0109] Step S4: performing frequency spectrum matching analysis on the theoretical angle spectrum and the measured angle spectrum, performing point-by-point translation processing on the theoretical angle spectrum along the angle axis within a preset angle offset range, and calculating the cross-correlation coefficient distribution of the two at different angle offsets.

[0110] In the embodiment, step S4 specifically includes:

[0111] Step S401: establishing a cross-correlation function based on the theoretical angle spectrum and the measured angle spectrum, specifically including:

[0112] Step S4011: taking the theoretical angle spectrum as a reference input and the measured angle spectrum as a comparison input, and defining the comparison relationship between the two in the angle domain;

[0113] Step S4012: Introduce an angle offset parameter in the comparison relationship as an independent variable to represent the angle displacement of the theoretical angle spectrum relative to the measured angle spectrum;

[0114] Step S4013: According to the angle response relationship between the theoretical angle spectrum and the measured angle spectrum, an intercorrelation function is established with the angle offset parameter as the input and the correlation coefficient as the output.

[0115] Specifically, the theoretical angle spectrum obtained in step S3 is denoted as , and the measured angle spectrum obtained in step S1 is denoted as Since the theoretical angle spectrum reflects the directional response of multiple targets under ideal conditions, while the measured angle spectrum is affected by noise, sidelobes and multipath, it is necessary to establish a comparison relationship between the two to measure the similarity of the two in the angle domain. The theoretical angle spectrum is regarded as the reference signal, and the measured angle spectrum is regarded as the signal to be matched, and the point multiplication relationship between the two in the angle domain is taken as the basis for subsequent intercorrelation calculation.

[0116] To represent the relative displacement of the theoretical angle spectrum relative to the measured angle spectrum, the embodiment introduces an angle offset variable τ in the comparison relationship between the two, where τ represents the angle of the overall left or right shift of the theoretical angle spectrum. When τ>0, the theoretical angle spectrum is offset in the positive direction relative to the measured angle spectrum; when τ<0, it is offset in the negative direction. By continuously taking values of τ (for example, step by step in the range of -10° to +10°), the correlation variation between the theoretical angle spectrum and the measured angle spectrum under different offset conditions can be obtained.

[0117] Based on the comparison relationship constructed in steps S4011 and S4012, the embodiment defines the intercorrelation coefficient of the theoretical angle spectrum and the measured angle spectrum under the condition of angle offset τ as:

[0118]

[0119] Where E[·] is the expectation operator, which can be realized by angle domain integration or discrete summation, and is usually the average value; τ is the angle delay, which represents the offset of the theoretical model relative to the measured angle spectrum; R(τ) is the intercorrelation coefficient output, which is used to quantify the matching degree of the two under the τ offset.

[0120] According to the correlation theory, when the echo at a certain angle is completely contributed by the real target signal, the measured angle spectrum and the corresponding theoretical angle spectrum have consistent phase distribution and energy characteristics at that angle position. Therefore, in this case, the intercorrelation function between the theoretical angle spectrum and the measured angle spectrum reaches the maximum value at the angle offset τ=0. This phenomenon reflects the high correlation of the signal with itself, and is the theoretical basis for using the intercorrelation function to determine the consistency of the signal.

[0121] On the contrary, in the angle position dominated by noise or clutters, the phase and amplitude distribution of the measured angle spectrum usually do not satisfy the array direction response model, and thus there is no stable linear relationship between the theoretical angle spectrum and the measured angle spectrum. At this time, no matter what value the angle shift τ takes, the cross-correlation coefficient remains at a low level and cannot form a significant correlation peak.

[0122] Step S402: Within the preset angle shift range, the theoretical angle spectrum is subjected to angle shift processing, and correlation operation is performed at each angle shift position with the measured angle spectrum to obtain the corresponding cross-correlation coefficient.

[0123] To evaluate the matching degree of the theoretical angle spectrum and the measured angle spectrum under different angle shifts, first, the theoretical angle spectrum is subjected to point-by-point shift processing within the preset angle shift range. Specifically, the value range of the angle shift parameter τ can be set as , where is the maximum shift amount determined according to the angle resolution, which can be, for example, an angle range of 10°, 20°, or larger. Subsequently, τ is discretely traversed at a fixed angle step (such as 0.1°, 0.5°, or 1°), and for each discrete shift amount τ, the theoretical angle spectrum is shifted by τ along the angle axis and the product integral or dot product calculation is performed at the corresponding position with the measured angle spectrum, thereby obtaining the cross-correlation coefficient corresponding to the shift amount τ. Through the above processing, the corresponding cross-correlation coefficient sequence at all shift positions can be obtained.

[0124] Step S403: The cross-correlation coefficients are constituted into a cross-correlation coefficient distribution to reflect the matching relationship between the two in the entire angle domain.

[0125] All the cross-correlation coefficients obtained in step S402 are arranged in the order of the value of the angle shift τ, thereby constructing a distribution curve of the cross-correlation coefficient with respect to the shift angle. The cross-correlation coefficient distribution curve directly reflects the similarity between the theoretical angle spectrum and the measured angle spectrum under different angle shift amounts, where the larger the coefficient is, the higher the matching degree of the two at the shift position is. The cross-correlation coefficient distribution usually presents several peaks, and in an ideal case, the correlation peak corresponding to the real target should appear near the τ=0 position and have an amplitude significantly higher than that at other shift positions.

[0126] Step S404: The cross-correlation peak value and the corresponding angle shift position are determined according to the cross-correlation coefficient distribution.

[0127] The peak value search processing is performed on the sequence of the cross-correlation coefficient with respect to the angle shift, and by comparing the cross-correlation values at adjacent shift positions, the local maximum points in the curve are determined. For each local maximum point, the one with the largest amplitude is selected as the global cross-correlation peak value, and the corresponding angle shift parameter τ is recorded.

[0128] Through the processing of steps S401 to S403, by constructing the cross-correlation function and performing traversal operation on the angle offset range, the correlation distribution curve covering the entire angle domain can be obtained, thereby depicting the matching degree between the theoretical angle spectrum and the measured angle spectrum in the frequency domain dimension. Since the cross-correlation can reflect the similarity of two angle spectra at different angle offsets, the correlation distribution curve can clearly present the high correlation peak formed by the real target near the zero offset, and the low correlation characteristics of noise, sidelobe echoes or multipath interference in the full angle range. With the aid of this correlation distribution, not only the distinguishability between the real target and the clutter can be significantly enhanced, but also the random fluctuations of the power spectrum caused by measurement noise or environmental factors can be effectively suppressed, so that the effective target presents a stable and prominent peak structure in the correlation dimension, thereby providing a reliable basis for subsequent target effectiveness determination and clutter removal.

[0129] Step S5: Based on the cross-correlation coefficient distribution, the maximum cross-correlation coefficient corresponding to each measured angle spectrum is determined. When the cross-correlation coefficient reaches the maximum at the zero angle offset and is greater than a preset threshold value, it is determined that the corresponding measured angle spectrum is effective, otherwise it is determined as clutter and removed.

[0130] When the target actually exists in the measured angle spectrum, the theoretical angle spectrum and the measured angle spectrum have highly consistent amplitude distribution and phase structure at the actual direction of the target, so the cross-correlation function usually forms a significant main correlation peak at τ=0, that is, τ=0 or close to 0 (within the error range allowed by the angle resolution); R(τ)≥R th , where R th is a preset correlation determination threshold.

[0131] Based on the above characteristics, the present embodiment defines the following determination rule: if the cross-correlation coefficient of a certain measured angle spectrum reaches the global peak value at the zero offset position, and the peak amplitude is greater than the preset threshold R th , it is considered that the echo of the measured angle spectrum satisfies the array direction response model, and the directional characteristic has good consistency with the theoretical model, so the measured angle spectrum is determined as effective and retained. On the contrary, when any of the following situations occurs, it is considered that the measured angle spectrum belongs to clutter or noise:

[0132] The maximum cross-correlation peak is not located at τ=0 position, that is, there is no corresponding direction consistency between the theoretical angle spectrum and the measured angle spectrum;

[0133] The amplitude of the maximum cross-correlation peak is lower than the threshold R th , that is, the matching degree is insufficient and cannot reflect the structured directional characteristic of the real target.

[0134] For the above case, the embodiment will correspond to the measured angle spectrum judgment as a scatter point, and perform rejection processing, so as to avoid noise, sidelobe false alarm and multipath reflection to interfere with the quality of point cloud.

[0135] Finally, by judging the consistency of the measured angle spectrum based on the amplitude and position of the zero offset correlation peak, the embodiment can accurately distinguish the real target and the scatter point in the angle domain dimension, and form the final effective target set.

[0136] On the basis of the above embodiment, a millimeter wave radar scatter point rejection method based on two-dimensional cross-correlation joint determination mechanism is further provided, which is used to simultaneously utilize the cross-correlation distribution of the horizontal angle and the elevation angle two angle dimensions, and improve the reliability of the scatter point identification. That is, the method can also include the following steps:

[0137] A1: jointly determining in the cross-correlation coefficient distribution of the effective target angle in the horizontal angle direction and the elevation angle direction;

[0138] A2: in the cross-correlation coefficient distribution, for each effective target angle, searching the correlation coefficient value of the zero angle offset position corresponding to the effective target angle in the horizontal angle direction and the elevation angle direction;

[0139] A3: when the same effective target angle forms a correlation peak at the zero angle offset in the horizontal angle direction and the elevation angle direction, and the peak value of the correlation peak is greater than the preset correlation threshold, the target echo located in the direction of the effective target angle is marked as the final effective target;

[0140] A4: if any effective target angle only forms a one-sided correlation peak satisfying the correlation threshold in the horizontal angle direction or the elevation angle direction, the target echo located in the direction of the effective target angle is determined as a scatter point and is rejected.

[0141] Specifically, in order to determine the direction information of the millimeter wave radar detection target in the three-dimensional space, the array output after the beam forming processing is traversed and scanned in a preset angle search region. With the point-by-point change of the scanning angle, the amplitude or power of the array direction response is recorded. When the beam forming output at a certain angle reaches the global maximum value, it indicates that the angle direction is most consistent with the true incident direction of the target. Therefore, by identifying the position of the maximum response, the spatial incident direction of the target can be determined, and the horizontal angle and the elevation angle of the target can be obtained at the same time.

[0142] For each effective target angle determined in step S2, the embodiment constructs a corresponding theoretical angle spectrum in the horizontal angle direction and the elevation angle direction respectively, and calculates the cross-correlation coefficient distribution in the two-dimensional angle domain according to the method of step S4:

[0143] The horizontal direction cross-correlation distribution is denoted as: ,

[0144] The cross-correlation distribution in the pitch direction is denoted as: ,

[0145] wherein, and denote the angular offset in the horizontal direction and the pitch direction, respectively.

[0146] Based on the above cross-correlation distribution, in the horizontal angular direction, the cross-correlation coefficient of the zero offset τ = 0 position corresponding to the horizontal angle is searched , and it is judged whether the coefficient forms a main correlation peak; at the same time, in the pitch angle direction, the cross-correlation coefficient of the zero offset τ = 0 position corresponding to the pitch angle is searched , and it is judged whether the coefficient forms a main correlation peak.

[0147] If the same target in both directions shows that the peak value is located near τ = 0; and the peak value amplitudes are both greater than a preset correlation threshold, it is indicated that the echo at the direction and the theoretical model are consistent in two-dimensional directions, which is a real target; the system marks the target as a final effective target. If a target only forms a one-sided correlation peak in the horizontal angle direction or the pitch angle direction that meets the threshold condition, while the zero offset correlation peak in the other direction does not reach the threshold, it is indicated that the target does not meet the directional response law of the millimeter wave array in structural consistency, and such echo is usually caused by noise, random scattered reflection, multi-path false alarm or sidelobe response. The target is determined as a clutter point and is removed to avoid affecting the point cloud quality.

[0148] Since the spatial incident direction of the real target has a stable and predictable phase structure in both the horizontal angle and the pitch angle, the measured angle spectrum of the real target will form a significant correlation peak at the zero offset position in both directions, and the correlation coefficients will simultaneously exceed the preset threshold; the clutter points caused by noise, sidelobe or multi-path usually only present accidental local correlation peaks in a certain angular dimension, and it is difficult to meet the consistency requirement in two dimensions at the same time. Through the above joint determination mechanism based on the zero offset cross-correlation coefficients in the horizontal angle and the pitch angle, the false retention easily occurring in single-dimensional cross-correlation can be effectively avoided, thereby further reducing the false alarm rate and improving the precision of clutter point removal. In addition, the method can maintain stable performance in low signal-to-noise ratio scenes, sparse point cloud scenes or multi-target complex environments, so that the system forms a more strict and robust confirmation condition for real targets in the three-dimensional angle space, and provides reliable protection for target recognition and point cloud quality improvement of the millimeter wave radar.

[0149] To further verify the effectiveness of the above cross-correlation decision mechanism, the cross-correlation coefficient performance of real targets and noise clutter points under different signal-to-noise ratios is simulated and analyzed.

[0150] Specifically, a real target point with a signal-to-noise ratio of -8 dB and a false target point composed of random noise are constructed in the simulation scene, and the measured angle spectrum of each point is cross-correlated with the theoretical angle spectrum obtained in step S3. Figure 3 The cross-correlation coefficient distribution between the pure noise point and the theoretical angle spectrum is shown; from Figure 3 It can be seen that the cross-correlation coefficient of the noise point remains in a low-amplitude random fluctuation range at all angle shift ranges, and cannot form a stable peak value because the noise point does not satisfy the array direction response characteristics.

[0151] Figure 4 The cross-correlation coefficient distribution between the real target point located in the zero shift position (τ = 0) direction and the theoretical angle spectrum is shown. It can be observed that the cross-correlation coefficient significantly increases at τ = 0 and forms a prominent correlation main peak, and the peak value is obviously higher than that at other shift positions, which reflects the high consistency of the phase structure of the real target with the theoretical model. In contrast, the correlation coefficient at the non-zero shift position decays rapidly, further indicating that the cross-correlation peak has good direction uniqueness.

[0152] Based on Figure 3 Compared with Figure 4 It can be seen that when the threshold of cross-correlation is set to 0.2, the cross-correlation value of the noise point is always lower than the threshold, while the real target point can stably exceed the threshold near τ = 0. Therefore, the threshold can effectively distinguish the real target from the noise false alarm point.

[0153] Figure 5 The target decision accuracy curve obtained by Monte-Carlo simulation of a large number of point targets under different signal-to-noise ratios under the above threshold condition is shown. The results show that when the signal-to-noise ratio is higher than -10 dB, the target decision accuracy based on the cross-correlation spectrum matching mechanism of the embodiment can stably reach more than 98%. It is verified that the clutter point removal method of the present application can still achieve efficient suppression of clutter points and stable identification of real targets in complex environments with strong noise, sparse echoes, and low signal-to-noise ratio.

[0154] The step division of the above methods is only for clear description, and can be combined into one step or split into multiple steps when implemented, as long as the same logical relationship is included, and all are within the protection scope of the present application; adding insignificant modifications or introducing insignificant designs in the algorithm or process, but not changing the core design of the algorithm and process are within the protection scope of the present application.

[0155] Furthermore, some embodiments of the application provide an electronic device. The electronic device can be any of various kinds of digital computers, such as a laptop computer, a desktop computer, a workstation, a personal digital assistant, a server, a blade server, a mainframe computer, and so on. The electronic device can also be any of various kinds of mobile devices, such as a personal digital assistant, a cellular phone, a smart phone, a wearable device, and other similar computing devices.

[0156] The electronic device includes one or more processors, and a memory storing computer program instructions that, when executed, cause the processors to perform a millimeter wave radar speckle removal method based on spectrum matching as provided by any one or more of the above embodiments. Figure 6 An exemplary structural diagram of the electronic device is disclosed. The electronic device includes one or more processors 1101, a memory 1102, and interfaces for connecting components, including high-speed interfaces and low-speed interfaces. The components are connected to each other by different buses, and can be mounted on a common motherboard or otherwise mounted as desired. The processor can process instructions executed within the electronic device, including instructions stored in the memory or on the memory to display a GUI on an external input / output device, such as a display device coupled to the interface. In some other embodiments, multiple processors and / or multiple buses can be used with multiple memories and multiple memories, if desired. Also, multiple electronic devices can be connected, each device providing part of the necessary operations. Among them, the components shown herein, their connections and relationships, and their functions are only examples, and are not intended to limit the implementation of the application described and / or claimed herein.

[0157] The electronic device can also include an input device 1103 and an output device 1104. The processor 1101, the memory 1102, the input device 1103 and the output device 1104 can be connected by bus or other means, Figure 6 The bus connection is taken as an example in the middle.

[0158] The input device 1103 can receive input digital or character information, and generate key signal input related to user settings and function control of the electronic device, such as touch screen, keypad, mouse, trackpad, touchpad, pointing stick, one or more mouse buttons, trackball, joystick, etc. The output device 1104 can include a display device, an auxiliary lighting device (e.g., LED), and a tactile feedback device (e.g., a vibration motor), etc. The display device can include, but is not limited to, a liquid crystal display, a light-emitting diode display, and a plasma display. In some embodiments, the display device can be a touch screen.

[0159] To provide for interaction with a user, the electronic device can be a computer. The computer has a display device (e.g., a cathode ray tube or LCD monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, etc.); and input from the user can be received in any form (e.g., acoustic input, speech input, tactile input, etc.).

[0160] In the embodiments of the present application, the computer program / instruction is stored on the computer readable medium, and the computer program / instruction is executed by the processor to implement the method for removing millimeter wave radar clutter points based on spectrum matching provided by any one or more of the above embodiments. The computer readable medium can be included in the electronic device described in the above embodiments, or can exist separately and not be assembled into the device. The computer readable medium carries one or more computer readable instructions.

[0161] The memory 1102 can be used to store non-transitory software programs, non-transitory computer executable programs and modules as a kind of non-transitory computer readable storage medium. The processor 1101 executes various functions and data processing of the server by running the non-transitory software programs, instructions and modules stored in the memory 1102, so as to implement the program instructions / modules corresponding to the method provided by any one or more of the above embodiments in the embodiments of the present application.

[0162] The memory 1102 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store data created according to the use of the electronic device, etc. In addition, the memory 1102 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory 1102 can optionally include a memory disposed remotely with respect to the processor 1101, and these remote memories can be connected to the electronic device through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0163] Note that the computer-readable medium described herein can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable medium can be, for example but not limited to, a system, a device, or a computer program product embodied in one or more computer readable media embodying computer readable instructions, data structures, program modules, or other data. Computer-readable storage media include, at least, volatile memory, non-volatile memory, removable or non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Examples of computer-readable storage media include, but are not limited to, phase-change memory, static random access memory, dynamic random access memory, other types of random access memory, read-only memory, electrically erasable programmable read-only memory, flash memory or other memory technology, compact disc read-only memory, digital versatile discs or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information for access by a computing device. In addition, computer-readable storage media can include any appropriate media, which can be used for storing data accessible by a computing device, including a hard disk drive, solid state drive, RAM, ROM, EEPROM, CD-ROM or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information for access by a computing device.

[0164] Computer-readable media includes permanent and non-permanent, removable and non-removable media, which can be implemented by any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change memory, static random access memory, dynamic random access memory, other types of random access memory, read-only memory, electrically erasable programmable read-only memory, flash memory or other memory technology, read-only optical disc, digital versatile disc or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device or any other non-transmission medium that can be used to store information for access by a computing device.

[0165] Computer program code for carrying out operations of the present application can be written in one or more programming languages or combinations of languages including object-oriented, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as the C programming language or similar programming languages. Program code can be executed entirely on a user computer, partially on a user computer, as a standalone software package, partially on a user computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user computer through any kind of network, including a local area network or a wide area network, or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0166] In the above-described embodiments, all or part of the embodiments can be implemented by software, hardware, firmware, or any combination thereof. For example, application specific integrated circuits, general purpose computers or any other similar hardware devices can be used. In some embodiments, the software programs of the present application can be executed by a processor to implement the above steps or functions. Similarly, the software programs of the present application (including related data structures) can be stored in a computer-readable recording medium, such as a RAM memory, a magnetic or optical drive or a soft disk and the like. In addition, some steps or functions of the present application can be implemented by hardware, such as a circuit cooperating with a processor to perform the respective steps or functions.

[0167] The computer program product provided by the embodiments of the present application includes one or more computer programs / instructions, which, when executed by a processor, generate all or part of the processes or functions described in the embodiments of the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be magnetic media (such as floppy disk, hard disk, magnetic tape), optical media (such as DVD), or semiconductor media (such as solid state disk) and the like.

[0168] The flowchart or block diagram in the drawings illustrates the architecture, functionality, and operation of possible implementations of devices, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagram can represent a module, a segment, or a portion of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks noted in succession can in fact be executed substantially concurrently, or the blocks can sometimes be executed in reverse order, depending on the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, or combinations of blocks in the block diagrams and / or flowchart illustrations, can be implemented by special-purpose hardware-based systems that perform the specified functions or operations, or combinations of special-purpose hardware and computer instructions.

[0169] The scope of this application is defined by the appended claims rather than the foregoing description, and is therefore intended to encompass all variations falling within the meaning and scope of equivalents of the claims. No reference numerals in the claims should be construed as limiting the scope of the claims. Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices recited in a device claim may also be implemented by a single unit or device in software or hardware. Terms such as "first," "second," etc., are used only for distinguishing descriptions and do not indicate any particular order, nor should they be construed as indicating or implying relative importance.

[0170] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily made by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims, and the above embodiments should be regarded as exemplary and non-limiting.

Claims

1. A method for millimeter wave radar speckle removal based on spectrum matching, characterized in that, include: The multi-channel echo signals of the millimeter-wave radar array are acquired and processed by beamforming and angle ergonomics to generate a measured angle spectrum that reflects the power response in each angular direction. Peak search is performed on the measured angle spectrum to determine the effective target angle corresponding to each power peak; Based on the geometric parameters of the millimeter-wave radar array and the effective target angle, the responses of each target are weighted and superimposed according to the theoretical echo signal model to obtain a theoretical angle spectrum characterizing the ideal target angle and peak response features, including: Based on the antenna layout and element spacing of the millimeter-wave radar array, a theoretical model is established to describe the array response characteristics of a single target under different incident angles. By combining the effective target angles with the theoretical model, a theoretical model matrix is ​​constructed to describe the theoretical array response of multiple effective targets at their corresponding angular positions; Based on the theoretical model matrix, the actual echo intensity parameters of each target are calculated by combining the measured angle spectrum. The echo intensity parameter is used as an amplitude weight and associated with the corresponding effective target angle. The theoretical array responses of each target are weighted and superimposed to form a cumulative theoretical model of multi-target echoes. The energy distribution of the accumulated theoretical model in the angular domain is taken as the theoretical angular spectrum; The theoretical angle spectrum and the measured angle spectrum are subjected to spectral matching analysis. Within a preset angle offset range, the theoretical angle spectrum is translated point by point along the angle axis, and the cross-correlation coefficient distribution of the two under different angle offsets is calculated. Based on the cross-correlation coefficient distribution, the cross-correlation coefficient corresponding to each measured angle spectrum is determined. When the cross-correlation coefficient reaches its maximum at zero angle offset and is greater than a preset threshold, the corresponding measured angle spectrum is determined to be valid; otherwise, it is determined to be noise and is removed.

2. The method of claim 1, wherein, The step of acquiring the multi-channel echo signal of the millimeter-wave radar array, and generating a measured angle spectrum reflecting the power response in each angular direction through beamforming and angle ergonomic processing includes: Acquire echo signals received by multiple receiving antennas of a millimeter-wave radar array; The echo signals are weighted and summed to form a composite signal output corresponding to the angle; The scanning angle is traversed within a preset angle range, and the weighting coefficients of each receiving antenna are dynamically adjusted according to the scanning angle. The output power of the synthesized signal at each angle is calculated. The output power corresponding to each scanning angle is used to form the power distribution as a function of angle, which is used as the measured angle spectrum to characterize the energy response characteristics of the echo signal in different directions.

3. The method of claim 1, wherein, The step of performing peak search on the measured angle spectrum to determine the effective target angle corresponding to each power peak includes: The power distribution curve of the measured angle spectrum is traversed to identify the angular positions corresponding to the local power maxima. The peak value within the preset threshold range is used as the candidate target angle when the power reaches the global maximum value. The amplitude of each candidate target angle is compared. When the power difference between adjacent peaks is less than the preset sidelobe suppression threshold, the angle corresponding to the main peak is retained and the sidelobe peak is suppressed. The angle corresponding to the retained peak is taken as the effective target angle.

4. The method of claim 1, wherein, The step of performing spectrum matching analysis on the theoretical angle spectrum and the measured angle spectrum includes: Establishing a cross-correlation function based on the theoretical angle spectrum and the measured angle spectrum; Performing angle translation processing on the theoretical angle spectrum within a preset angle offset range, and performing correlation operation with the measured angle spectrum at each angle offset position to obtain a corresponding cross-correlation coefficient; The cross-correlation coefficient constitutes a cross-correlation coefficient distribution to reflect the matching relationship between the two in the entire angle domain; According to the cross-correlation coefficient distribution, determine the cross-correlation peak value and the corresponding angle offset position.

5. The method of claim 4, wherein, The step of establishing a cross-correlation function based on the theoretical angle spectrum and the measured angle spectrum includes: Define the comparison relationship between the two in the angle domain by taking the theoretical angle spectrum as the reference input and the measured angle spectrum as the comparison input; Introduce an angle offset parameter in the comparison relationship as an independent variable for representing the angle displacement of the theoretical angle spectrum relative to the measured angle spectrum; According to the angle response relationship between the theoretical angle spectrum and the measured angle spectrum, establish a cross-correlation function with the angle offset parameter as input and the cross-correlation coefficient as output.

6. The spectrum matching based millimeter wave radar clutter removal method of claim 1, wherein, Further comprising: Jointly determine the cross-correlation coefficient distribution of the effective target angle in the horizontal angle direction and the pitch angle direction respectively; In the cross-correlation coefficient distribution, search for the correlation coefficient value of the zero angle offset position corresponding to the effective target angle in the horizontal angle direction and the pitch angle direction for each effective target angle; When the same effective target angle forms a correlation peak at the zero angle offset position in the horizontal angle direction and the pitch angle direction, and the peak value of the correlation peak is greater than a preset correlation threshold, the target echo located in the direction of the effective target angle is marked as the final effective target; If any effective target angle only forms a one-sided correlation peak in the horizontal angle direction or the pitch angle direction that meets the correlation threshold, the target echo located in the direction of the effective target angle is determined as a clutter point and is removed.

7. An electronic device, comprising: The electronic device includes: One or more processors; and a memory storing computer program instructions that, when executed, cause the processors to perform the spectrum matching-based millimeter wave radar clutter removal method of any one of claims 1-6.

8. A computer readable storage medium having stored thereon a computer program and / or instructions, characterized in that, The computer program and / or instructions are executed by the processor to implement the spectrum matching-based millimeter wave radar clutter removal method of any one of claims 1-6.

9. A computer program product comprising computer programs and / or instructions, characterized in that, The computer program and / or instructions are executed by the processor to implement the spectrum matching-based millimeter wave radar clutter removal method of any one of claims 1-6.

Images