A high-efficiency two-dimensional constant false alarm detection method

Abstract

The application discloses a high-efficiency two-dimensional constant false alarm rate detection method, which comprises the following steps: after the velocity dimension data of the same distance unit is subjected to FFT processing, a threshold is updated; extreme value judgment is carried out on the data greater than the threshold to obtain an extreme value; a reference unit is selected in a distance-Doppler two-dimensional plane with the extreme value as the center to establish a rectangular reference window; the maximum data value in the velocity dimension data of the same distance unit in the rectangular reference window is selected to establish a new reference window; and the extreme value is compared with all the reference units in the new reference window by using an ordered statistical constant false alarm rate (OS_CFAR) algorithm to judge whether the extreme value is a target. The application can not only meet the detection requirements of multiple targets in a complex environment, but also reduce the time consumption in the sorting process during detection, realizes fast search of targets in a large airspace range, and improves the possibility of engineering implementation.

Description

Technical Field

[0001] This invention relates to millimeter-wave radar signal processing technology, and more particularly to a highly efficient two-dimensional constant false alarm rate (CFAR) detection method for traffic radar. Background Technology

[0002] Constant False Alarm Rate (CFAR) detection is a crucial component of radar signal processing. It distinguishes between the receiver's output signal and noise while maintaining a constant false alarm probability to determine the presence of a target, playing an indispensable role in radar echo signal detection. However, in practice, CFAR often faces the problem of excessively long detection times when searching for targets over large airspaces, degrading the effectiveness of traditional CFAR detection algorithms. Summary of the Invention

[0003] The purpose of this invention is to provide an efficient two-dimensional constant false alarm rate (CFAR) detection method that can meet the detection requirements of multiple targets in complex environments, reduce the time consumption during the sorting process, achieve rapid search of targets in a large airspace, and improve the feasibility of engineering implementation.

[0004] To achieve the above objectives, the present invention provides a highly efficient two-dimensional constant false alarm rate (CFAR) detection method, comprising:

[0005] After performing FFT processing on the velocity dimension data of the same distance cell, the threshold is updated;

[0006] For data exceeding a threshold, perform extreme value identification to obtain the extreme value;

[0007] In the distance-Doppler two-dimensional plane, a reference cell is selected centered on the extreme value to establish a rectangular reference window;

[0008] Within a rectangular reference window, select the velocity dimension data with the largest value in the same distance cell and create a new reference window;

[0009] The ordered statistics constant false alarm rate (OS_CFAR) algorithm is used to compare the extreme value with all reference cells in the new reference window to determine whether the extreme value is the target.

[0010] The method for updating the threshold includes: in the range-Doppler two-dimensional plane, the maximum number of points in the range dimension is R, and the maximum number of points in the velocity dimension is D. Let the amplitude value of the data unit corresponding to the r-th point in the range dimension and the d-th point in the velocity dimension on the two-dimensional amplitude spectrum be a. r,d The threshold updated after the i-th FFT at the distance r-th point is α. i,r After performing an FFT on all velocity dimension data at point r for the i-th time, we obtain a. r,1 ,a r,2 ,…,a r,D For data ar,1 ,a r,2 ,…,a r,D Summation is performed, and when data transmission ends, the threshold is updated using the mean and the threshold coefficient β.

[0011]

[0012] The method for determining extreme values ​​includes: using a "9-square grid" approach, taking the data that is greater than the threshold as the center, and comparing it with the other 8 data in the "9-square grid". Only when the center data is greater than all of the other 8 data is the center data determined to be an extreme value.

[0013] The method for establishing a rectangular reference window includes: setting the target determined to be an extreme value as the center, selecting N sequences on the left and right of the extreme value in the same velocity dimension as the extreme value as distance reference units, selecting units in the two velocity dimensions adjacent to the extreme value and sequences in the two velocity dimensions adjacent to the distance reference units with the same length as the distance reference units as Doppler reference units, and the distance reference units and Doppler reference units together form a rectangular reference window.

[0014] The method for determining whether an extreme value is a target includes: simultaneously extracting all reference cells X(a) in a new reference window, multiplying all these reference cells X(a) by the false alarm threshold coefficient T, and then determining the extreme value a. r,d Compare with the reference unit value X(a)·T after multiplication, when the extreme value a r,d When the value is greater than the reference cell value X(a)·T, it is counted as 1; otherwise, it is counted as 0. The pipeline operation is performed, and the decision results are accumulated to obtain the accumulated value m. When the extreme value a r,d After comparing with all reference unit values ​​X(a)·T, if the accumulated result m is greater than k and 0 < k ≤ 2N+1, then the extreme value a is determined. r,d If the cumulative result is less than or equal to k, then determine the extreme value a. r,d This is noise.

[0015] The present invention has the following beneficial effects:

[0016] 1. After performing FFT processing on the velocity dimension data, a pipeline approach is used to compare the velocity dimension data with the threshold and update the threshold. This eliminates the need to traverse the data again, speeding up the processing and preventing data smaller than the threshold from participating in subsequent processing, which would increase the additional time consumption.

[0017] 2. By selecting the unit with the largest data value in the velocity dimension at the same distance from the rectangular reference window, a new one-dimensional reference window is created, which reduces the number of reference units during sorting while retaining the ability to distinguish between targets at the same distance and from the opposite direction.

[0018] 3. By using the binary accumulation method, the process of sorting reference cells to find the k-th reference cell in the original ordered statistics constant false alarm rate OS_CFAR and then comparing it with the detection cell is replaced by directly comparing the detection cell with the reference cell. This method reduces the amount of computation and improves the processing speed of the algorithm. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the constant false alarm rate (CFAR) detection method used in this invention.

[0020] Figure 2 This is a schematic diagram of a two-dimensional plane with distance-Doppler.

[0021] Figure 3 This is a schematic diagram of the rectangular reference window used in this invention.

[0022] Figure 4 This is a schematic diagram of the OS-CFAR process implemented using binary accumulation in this invention. Detailed Implementation

[0023] The following is based on Figures 1-4 The preferred embodiments of the present invention will be described in detail below.

[0024] To meet the requirements of traffic radar for detecting multiple targets in complex environments, while reducing the time consumed during the detection process, such as... Figure 1 As shown, this invention provides a highly efficient two-dimensional constant false alarm rate (CFAR) detection method, comprising the following steps:

[0025] Step 1: After performing FFT processing on the velocity dimension data of cells with the same distance, a pipeline method is used to compare the velocity dimension data with the threshold and update the threshold.

[0026] like Figure 2 As shown, in the range-Doppler two-dimensional plane, the maximum number of points in the range dimension is R, and the maximum number of points in the velocity dimension is D. Let the amplitude value of the data unit corresponding to the r-th point in the range dimension and the d-th point in the velocity dimension on the two-dimensional amplitude spectrum be a. r,d The threshold updated after the i-th FFT at the distance r-th point is α. i,r After performing an FFT on all velocity dimension data at point r for the i-th time, we obtain a. r,1 ,a r,2 ,…,a r,D In actual FPGA circuits, data results are output sequentially in a pipeline manner. Here, we will use a... r,1 ,a r,2 ,…,a r,D The threshold α updated after the (i-1)th FFT at the point r in distance dimension is respectively compared with the threshold α. i-1,r The comparison is performed only when the value is greater than the threshold α. i-1,r The data is then used in the calculation of step 2, and the threshold comparison is performed on data a.r,1 ,a r,2 ,…,a r,D Summation is performed, and when data transmission ends, the threshold is updated using the mean and the threshold coefficient β.

[0027]

[0028] Step 2: Perform extreme value judgment on data exceeding the threshold, and only allow extreme values ​​to participate in the subsequent calculation process.

[0029] like Figure 2 As shown, a "9-square grid" method is used. The data that is greater than the threshold is taken as the center and compared with the other 8 data in the "9-square grid". Only when the center data is greater than all the other 8 data is the center data determined to be an extreme value and can participate in the subsequent calculation process.

[0030] Step 3: Select a reference cell in the distance-Doppler two-dimensional plane and establish a rectangular reference window.

[0031] The target identified as an extremum is set as the center. N sequences to the left and right of the extremum (the detected unit) in the same velocity dimension are selected as distance reference units. Units in the two velocity dimensions adjacent to the extremum and sequences in the two velocity dimensions adjacent to the distance reference units, having the same length as the distance reference units, are taken as Doppler reference units. These distance reference units and Doppler reference units together form a rectangular reference window, as shown below. Figure 3 As shown.

[0032] Step 4: Within the rectangular reference window, select the data with the largest velocity value among the cells at the same distance and create a new reference window.

[0033] right Figure 3 The rectangular reference window shown compares three Doppler reference cells of the same range dimension, and selects the largest Doppler reference cell to form a new reference window. This method reduces the number of reference cells that need to be sorted from (2N+1)×3 to 2N+1, thus reducing the computational load. It also retains the ability to distinguish between multiple targets at the same range, both approaching and departing, making it more suitable for traffic radar applications.

[0034] Step 5: Based on the new reference window, the binary accumulation method is used to compare the detection unit in the ordered statistic constant false alarm (OS_CFAR) with the k-th reference unit in the new reference window, thus completing the constant false alarm processing.

[0035] First, all reference cells X(a) in the new reference window are simultaneously extracted. Then, all reference cells X(a) are multiplied by the false alarm threshold coefficient T. Finally, the extreme value (the detected cell) a is calculated. r,d Compare with the reference unit value X(a)·T after multiplication, when the extreme value a r,dWhen the value is greater than the reference cell value X(a)·T, it is counted as 1; otherwise, it is counted as 0. The pipeline operation is performed, and the decision results are accumulated to obtain the accumulated value m. When the extreme value a r,d After comparing with all reference unit values ​​X(a)·T, if the accumulated result m is greater than k and 0 < k ≤ 2N+1, then the extreme value a is determined. r,d If the cumulative result is less than or equal to k, then determine the extreme value a. r,d For clutter, the process is as follows: Figure 4 As shown.

[0036] The essence of the Ordered Statistics Constant False Alarm Rate (OS_CFAR) algorithm is to sort all reference cells in a new reference window in ascending order, find the k-th reference cell, multiply it by the false alarm threshold coefficient T, and compare this value with the extreme value (the detected cell) to determine whether the detected cell is a target. The value of k needs to be determined through multiple measurements, which is equivalent to an empirical value; the value of k is between 1 and 2N+1. By controlling the choice of k, the threshold size is controlled. The value of k is fixed, but since all reference cells are of different sizes, the threshold can be dynamically selected according to the magnitude of the target amplitude within different distance and velocity dimensions. Here, the detected cell is directly compared with the product of all reference cells and the false alarm threshold coefficient T, and the cumulative amount of the decision result is used as the detected cell a. r,d Whether it is the target or not saves the amount of sorting computation and greatly improves the processing speed of the algorithm.

[0037] The present invention has the following beneficial effects:

[0038] 1. After performing FFT processing on the velocity dimension data, a pipeline approach is used to compare the velocity dimension data with the threshold and update the threshold. This eliminates the need to traverse the data again, speeding up the processing and preventing data smaller than the threshold from participating in subsequent processing, which would increase the additional time consumption.

[0039] 2. By selecting the unit with the largest data value in the velocity dimension at the same distance from the rectangular reference window, a new one-dimensional reference window is created, which reduces the number of reference units during sorting while retaining the ability to distinguish between targets at the same distance and from the opposite direction.

[0040] 3. By using the binary accumulation method, the process of sorting reference cells to find the k-th reference cell in the original ordered statistics constant false alarm rate OS_CFAR and then comparing it with the detection cell is replaced by directly comparing the detection cell with the reference cell. This method reduces the amount of computation and improves the processing speed of the algorithm.

[0041] It should be noted that, in the embodiments of the present invention, the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the embodiments and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0042] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A highly efficient two-dimensional constant false alarm rate (CFAR) detection method, characterized in that, Include: After performing FFT processing on the velocity dimension data of the same distance cell, the threshold is updated; For data exceeding a threshold, perform extreme value determination to obtain the extreme value; In the distance-Doppler two-dimensional plane, a reference cell is selected centered on the extreme value to establish a rectangular reference window; Within a rectangular reference window, select the velocity dimension data with the largest value in the same distance cell and create a new reference window; The ordered statistics constant false alarm rate (OS_CFAR) algorithm is used to compare the extreme value with all reference cells in the new reference window to determine whether the extreme value is the target. The method for updating the threshold includes: in the range-Doppler two-dimensional plane, the maximum number of points in the range dimension is R, and the maximum number of points in the velocity dimension is D. Let the amplitude value of the data unit corresponding to the r-th point in the range dimension and the d-th point in the velocity dimension on the two-dimensional amplitude spectrum be a. r,d The threshold updated after the i-th FFT at the distance r-th point is α. i,r After performing an FFT on all velocity dimension data at point r for the i-th time, we obtain a. r,1 ,a r,2 ,…,a r,D For data a r,1 ,a r,2 ,…,a r,D Summation is performed, and when data transmission ends, the threshold is updated using the mean and the threshold coefficient β. The method for determining whether an extreme value is a target includes: simultaneously extracting all reference cells X(a) in a new reference window, multiplying all these reference cells X(a) by the false alarm threshold coefficient T, and then determining the extreme value a. r,d Compare with the reference unit value X(a)·T after multiplication, when the extreme value a r,d When the value is greater than the reference cell value X(a)·T, it is counted as 1; otherwise, it is counted as 0. The pipeline operation is performed, and the decision results are accumulated to obtain the accumulated value m. When the extreme value a r,d After comparing with all reference unit values ​​X(a)·T, if the accumulated result m is greater than k and 0 < k ≤ 2N+1, then the extreme value a is determined. r,d If the cumulative result is less than or equal to k, then determine the extreme value a. r,d For clutter, k is the k-th reference cell in the new reference window, and N is the range reference cell selected from the N sequences to the left and right of the extreme value, which are located in the same velocity dimension.

2. The efficient two-dimensional constant false alarm rate (CFAR) detection method as described in claim 1, characterized in that, The method for determining extreme values ​​includes: using a "9-grid" approach, taking the data greater than the threshold as the center, and comparing it with the other 8 data in the "9-grid". Only when the center data is greater than all the other 8 data is the center data determined to be an extreme value.

3. The efficient two-dimensional constant false alarm rate (CFAR) detection method as described in claim 2, characterized in that, The method for establishing a rectangular reference window includes: setting the target determined to be an extreme value as the center, selecting N sequences on the left and right of the extreme value in the same velocity dimension as the extreme value as distance reference units, selecting units in the two velocity dimensions adjacent to the extreme value and sequences in the two velocity dimensions adjacent to the distance reference units with the same length as the distance reference units as Doppler reference units, and the distance reference units and Doppler reference units together form a rectangular reference window.

Images