Cross-shaped self-calibration constant false alarm rate target detection method based on millimeter waves
Through the cross-calibration constant false alarm rate target detection method, the false alarm probability and offset factor calculation are used, combined with the secondary corrugated comparator, the problems of false alarm and missing alarm in the existing technology are solved, and the problem of false alarm and missing alarm is realized in complex environments Adaptive false alarm rate setting and target point recognition improve the accuracy of millimeter wave point cloud generation.
Patent Information
- Application Number
- CN202411000884.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-08
AI Technical Summary
The existing millimeter wave constant false alarm rate algorithm is prone to false alarms and missed alarms in complex environments, and cannot adapt to different environments and scenarios, resulting in difficulty in setting false alarm probability.
Using the cross-calibration constant false alarm rate target detection method based on millimeter wave, through data acquisition, signal preprocessing, distance-Doppler matrix generation, false alarm probability calculation, offset factor calculation, reference unit selection and threshold calculation, and two The use of the stage ripple comparator, adaptively set false alarm probability and threshold values to reduce false alarm rate.
Effectively identify target points, reduce false alarm rates and missed alarm rates, improve the accuracy and adaptability of target detection, and is suitable for millimeter wave point cloud generation in complex environments.
Smart Images

Figure CN118914993A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of millimeter wave constant false alarm rate detection, in particular to a millimeter wave-based cross self-calibration constant false alarm rate target detection method. Background Art
[0002] With humanity's exploration of wireless intelligent sensing technology, the application of high-frequency, highly sensitive millimeter waves is becoming increasingly widespread. Millimeter-wave-based technologies such as vehicle reversing detection, obstacle detection, gesture recognition, body position recognition, and respiratory rate recognition have developed and become a key focus of scientific research. The first step in millimeter-wave applications is point cloud generation. Target point detection is a key step in generating target point clouds using millimeter-wave radar. The accuracy of target detection directly impacts the quality of target point cloud generation, and thus the accuracy of subsequent algorithm analysis and calculations. As the requirements for target detection accuracy in millimeter-wave radar continue to increase, a constant false alarm rate (CFAR) algorithm that can adapt to complex environments, discern clutter edges, and accurately detect target point clouds is under development.
[0003] There are three main types of constant false alarm rate algorithms currently available: one-dimensional constant false alarm rate, two-dimensional constant false alarm rate, and two-dimensional constant false alarm rate. The one-dimensional constant false alarm rate is to expand the data matrix into a one-dimensional array along the distance dimension and detect the target points through a sliding window. The two-dimensional constant false alarm rate is to expand the Doppler dimension of the target points detected by the one-dimensional constant false alarm rate into a one-dimensional array and perform one-dimensional constant false alarm rate detection again. The two-dimensional constant false alarm rate is to detect the target points one by one in the range-Doppler matrix through a two-dimensional sliding window. These existing algorithms may miss false alarms due to interference from clutter in the calculation of the threshold, or they may not be able to adapt to a variety of different environments because the false alarm probability needs to be manually set for different environments and scenarios.
[0004] Therefore, a constant false alarm rate algorithm is needed that can adapt to the environment and effectively suppress the false alarm rate and missed alarm rate. Summary of the Invention
[0005] In order to solve the above technical problems, the present invention provides a millimeter-wave-based cross self-calibration constant false alarm rate target detection method, which can solve the problem of false alarm and missed alarm in target detection in an environment with complex clutter interference, and realize the setting of adaptive false alarm probability.
[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:
[0007] A millimeter-wave-based cross self-calibration constant false alarm rate target detection method comprises the following steps:
[0008] S1. Data acquisition: Use the IWR6843ISK millimeter-wave radar device to collect binary data from four antennas;
[0009] S2, signal preprocessing: convert binary data into decimal, distribute the real and imaginary parts of the data, and take the data from the first two receiving antennas to form a data sequence;
[0010] S3, data matrix generation: the data sequence is processed twice by fast Fourier transform to generate the range-Doppler matrix;
[0011] S4. Calculation of false alarm probability: By analyzing the characteristics of the range-Doppler matrix, the false alarm probability is calculated, which is used to calculate the threshold factor;
[0012] S5. Offset factor calculation: By analyzing the characteristics of the range-Doppler matrix, the offset factor is calculated. The offset factor is used for subsequent offset calculation; the offset is used to correct the threshold value and reduce the false alarm rate;
[0013] S6, reference unit selection and threshold calculation: select the cross calibration frame as the reference unit, calculate the threshold value and threshold offset, and input them into the first-level ripple comparator;
[0014] S7, the ripple difference calculated by the protection unit is compared with the set threshold in the secondary ripple comparator, and a binary value is output to reduce the false alarm rate;
[0015] S71, extracting the values of the two outermost circles of the protection unit to form two arrays, and taking the average value of the two arrays respectively;
[0016] S72. Since the target unit and the protection unit have ripple characteristics, that is, there are steps similar to ripples, the ripple calculation formula is as follows:
[0017] RDiff=|Rp1-Rp2|-|Rp1-DetectValue|;
[0018] Where: RDiff is the ripple difference; Rp1 is the average value of the protection unit of the inner circle, that is, ripple value 1; Rp2 is the average value of the protection unit of the outer circle, that is, ripple value 2; DetectValue is the value of the detection unit;
[0019] S73. In S6, the detection unit value in the primary ripple comparator is greater than the threshold value and is output to the secondary ripple comparator. It is defined that when the ripple difference of the detection unit is greater than the threshold value 0.1, the detection unit is output through the secondary ripple comparator.
[0020] Furthermore, the step S4 specifically includes:
[0021] S41. Normalize the range-Doppler matrix to prevent false alarms and missed alarms due to large data gaps. The calculation formula is as follows:
[0022]
[0023] Where: R norm Represents the normalized matrix, the matrix shape is N×M, indicating that the matrix has N rows and M columns; R rdm Represents the range-Doppler matrix generated by S3 processing. The matrix shape is N×M, which means the matrix has N rows and M columns. max(R rdm ) represents the matrix R rdm The matrix element with the largest value is used as the reference value, that is, the denominator;
[0024] S42. Obtain the range-Doppler matrix characteristics by taking a representative value for each row and column. That is, take the positive and negative diagonal elements of the matrix to form an array. The calculation formula for the diagonal array is as follows:
[0025] Diag1=R norm [i][i](0≤i<N);
[0026] Diag2=R norm [i][Mi](0≤i<N);
[0027] Where: Diag1, Diag2 represent the diagonal and anti-diagonal arrays; i represents the index value; R norm [i][i],R norm [i][Mi] represents the element under the corresponding matrix index;
[0028] S43. Calculate the value of the false alarm probability according to the number of protection units selected. The calculation formula is as follows:
[0029]
[0030] Where: P represents the value of false alarm probability, which represents the false alarm probability when the constant false alarm probability method is used; mean(Diag1[::3]) represents slicing the array Diag1, taking a value every 2 numbers to form an array, and taking the average of the array; mean(Diag2[::3]) represents slicing the array Diag2, taking a value every 2 numbers to form an array, and taking the average of the array; Unit p Is the protection unit size, that is, the size of the protection unit matrix, that is, the protection unit size is Unit p ×Unit p .
[0031] Furthermore, the step S5 specifically includes:
[0032] S51. Take the median of all positive numbers in the Diag1 and Diag2 arrays and calculate the offset factor. The calculation formula is as follows:
[0033] Corr factor =(Diag1m +Diag2 m )×2;
[0034] Where: Corr factor Indicates the offset factor, which is involved in the subsequent threshold offset calculation; Diag1 m and Diag2 m Represents the median of the positive sequence of the Diag1 and Diag2 arrays.
[0035] Furthermore, the step S6 specifically includes:
[0036] S61. Select a corresponding reference unit. The selected reference unit is a cross calibration frame. The elements surrounding the edge of the protection unit are used as the reference unit, i.e., the capture frame. The elements in the row and column corresponding to the detection unit are used as the reference unit for calculation.
[0037] S62. Obtain several reference unit intervals and combine them into threshold value alternatives. The calculation formula is as follows:
[0038] ThresholdValue=[Cl1,Cl2,CB];
[0039] Where: ThresholdValue represents the threshold value alternative array; Cl1 represents the corresponding row reference cell array; Cl2 represents the corresponding column reference cell array; CB represents the capture box corresponding reference cell;
[0040] S621: Calculate the candidate threshold index. The calculation formula is as follows:
[0041] Index=|(1-P)×len(ThresholdValue)|;
[0042] Where: Index represents the index number; (1-P) represents the detection probability; the formula in len means the length of the alternative threshold array;
[0043] S622: Using the candidate threshold array index and the threshold offset proposed in S64, the calculation formula is as follows:
[0044] Value=ThresholdValue[Index]+Corr;
[0045] Where: Value represents the corrected threshold value; ThresholdValue[Index] represents the element in the array ThresholdValue located by the index value Index; Corr is the threshold offset;
[0046] S63. Calculate the reference unit: Calculate the threshold value using the reference unit. The calculation formula is as follows:
[0047]
[0048] Where: parameters N Indicates the threshold parameters paramerers2 and parameters2; Cl N Represents Cl1 and Cl2, CB mean Represents the mean of the capture frame reference cells; the formula in max means taking a value greater than or equal to 0; the calculation formula means taking the difference between each element of the reference cell array of the row and column and the capture frame mean, taking a value greater than or equal to 0 and averaging it;
[0049] S64. Calculate the threshold offset using the threshold value in S63 and the threshold offset factor in S51. The calculation formula is as follows:
[0050]
[0051] Where: Corr represents the threshold offset.
[0052] Compared with existing technologies, this invention calculates an adaptive false alarm probability and threshold offset factor by leveraging the overall characteristics of the range-Doppler matrix. By selecting reference cells in the shape of crosshairs and capture frames, the sum of the threshold and threshold offset calculated by the reference cells is unaffected by clutter, effectively identifying the target point. Furthermore, a secondary ripple comparator is introduced to calculate the ripple difference between the two layers of protection cells to obtain a ripple value. This ripple value is then compared with the threshold in the secondary ripple comparator to further determine whether the point is a target point, thereby reducing the false alarm rate. The superiority of this method will have a significant impact on the field of point cloud generation and promote the development of millimeter wave related fields.
[0053] In summary, this method can effectively extract target points from the range-Doppler matrix, reducing both false alarm and missed alarm rates. Point cloud generation has high application value in the field of millimeter-wave wireless sensing, and to some extent, it solves the problem of missed or false alarms at target points. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1 This is a flow chart of the millimeter-wave-based cross self-calibration constant false alarm rate target detection method in this application.
[0055] Figure 2 This is the structure of the cross self-calibration constant false alarm rate target detection method based on millimeter waves.
[0056] Figure 3 Flowchart for millimeter-wave signal preprocessing and range-Doppler matrix generation.
[0057] Figure 4 Select a schematic for the reference cell.
[0058] Figure 5 Schematic diagram of ripple difference calculation.
[0059] Figure 6 The results of the millimeter-wave-based cross self-calibration constant false alarm rate target detection method are presented in various environments. DETAILED DESCRIPTION
[0060] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. The specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0061] like Figure 1 As shown, this embodiment exemplarily shows a cross self-calibration constant false alarm rate target detection method based on millimeter waves. The structure of the cross self-calibration constant false alarm rate target detection system based on millimeter waves is shown in FIG. Figure 2 The specific detection process is as follows:
[0062] S1. Data acquisition: Use the IWR6843ISK millimeter-wave radar device to collect binary data from four antennas;
[0063] S2, signal preprocessing: convert binary data into decimal, distribute the real and imaginary parts of the data, and take the data from the first two receiving antennas to form a data sequence;
[0064] S3, data matrix generation: the data sequence is processed twice by fast Fourier transform to generate range-Doppler matrix; Figure 3 shown. Figure 3 Flowchart for millimeter-wave signal preprocessing and range-Doppler matrix generation.
[0065] S4. Calculation of false alarm probability: By analyzing the characteristics of the range-Doppler matrix, the false alarm probability is calculated, which is used to calculate the threshold factor;
[0066] S41. Normalize the range-Doppler matrix to prevent false alarms and missed alarms due to large data gaps. The calculation formula is as follows:
[0067]
[0068] Where: R norm Represents the normalized matrix, the matrix shape is N×M, indicating that the matrix has N rows and M columns; R rdm Represents the range-Doppler matrix generated by S3 processing. The matrix shape is N×M, which means the matrix has N rows and M columns. max(R rdm ) represents the matrix R rdm The matrix element with the largest value is used as the reference value, that is, the denominator;
[0069] S42. Obtain the range-Doppler matrix characteristics by taking a representative value for each row and column. That is, take the positive and negative diagonal elements of the matrix to form an array. The calculation formula for the diagonal array is as follows:
[0070] Diag1=R norm [i][i](0≤i<N);
[0071] Diag2=R norm [i][Mi](0≤i<N);
[0072] Where: Diag1, Diag2 represent the diagonal and anti-diagonal arrays; i represents the index value; R norm [i][i],R norm [i][Mi] represents the element under the corresponding matrix index;
[0073] S43. Calculate the value of the false alarm probability according to the number of protection units selected. The calculation formula is as follows:
[0074]
[0075] Where: P represents the value of false alarm probability, which represents the false alarm probability when the constant false alarm probability method is used; mean(Diag1[::3]) represents slicing the array Diag1, taking a value every 2 numbers to form an array, and taking the average of the array; mean(Diag2[::3]) represents slicing the array Diag2, taking a value every 2 numbers to form an array, and taking the average of the array; Unit p Is the protection unit size, that is, the size of the protection unit matrix, that is, the protection unit size is Unit p ×Unit p .
[0076] S5. Offset factor calculation: By analyzing the characteristics of the range-Doppler matrix, the offset factor is calculated. The offset factor is used for subsequent offset calculation; the offset is used to correct the threshold value and reduce the false alarm rate;
[0077] S51. Take the median of all positive numbers in the Diag1 and Diag2 arrays and calculate the offset factor. The calculation formula is as follows:
[0078] Corr factor =(Diag1 m +Diag2 m )×2;
[0079] Where: Corr factor Indicates the offset factor, which is involved in the subsequent threshold offset calculation; Diag1 mand Diag2 m Represents the median of the positive sequence of the Diag1 and Diag2 arrays.
[0080] S6. Reference unit selection and threshold calculation: Select the cross calibration frame as the reference unit, such as Figure 4 As shown, Figure 4 Select a schematic diagram for the reference unit; and calculate the threshold value and threshold offset and input them to the first-level ripple comparator;
[0081] S61. Select a corresponding reference unit. The selected reference unit is a cross calibration frame. The elements surrounding the edge of the protection unit are used as the reference unit, i.e., the capture frame. The elements in the row and column corresponding to the detection unit are used as the reference unit for calculation.
[0082] S62. Obtain several reference unit intervals and combine them into threshold value alternatives. The calculation formula is as follows:
[0083] ThresholdValue=[Cl1, Cl2, CB];
[0084] Where: ThresholdValue represents the threshold value alternative array; Cl1 represents the corresponding row reference cell array; Cl2 represents the corresponding column reference cell array; CB represents the capture box corresponding reference cell;
[0085] S621: Calculate the candidate threshold index. The calculation formula is as follows:
[0086] Index=|(1-P)×len(ThresholdValue)|;
[0087] Where: Index represents the index number; (1-P) represents the detection probability; the formula in len means the length of the alternative threshold array;
[0088] S622: Using the candidate threshold array index and the threshold offset proposed in S64, the calculation formula is as follows:
[0089] Value=ThresholdValue[Index]+Corr;
[0090] Where: Value represents the corrected threshold value; ThresholdValue[Index] represents the element in the array ThresholdValue located by the index value Index; Corr is the threshold offset;
[0091] S63. Calculate the reference unit: Calculate the threshold value using the reference unit. The calculation formula is as follows:
[0092]
[0093] Where: parameters N Indicates the threshold parameters2 and parameters2; Cl N Represents Cl1 and Cl2, CB mean Represents the mean of the capture frame reference cells; the formula in max means taking a value greater than or equal to 0; the calculation formula means taking the difference between each element of the reference cell array of the row and column and the capture frame mean, taking a value greater than or equal to 0 and averaging it;
[0094] S64. Calculate the threshold offset using the threshold value in S63 and the threshold offset factor in S51. The calculation formula is as follows:
[0095]
[0096] Where: Corr represents the threshold offset.
[0097] S7, the ripple difference calculated by the protection unit is compared with the set threshold in the secondary ripple comparator, and a binary value is output to reduce the false alarm rate; Figure 5 As shown, Figure 5 Schematic diagram of ripple difference calculation.
[0098] S71, extracting the values of the two outermost circles of the protection unit to form two arrays, and taking the average value of the two arrays respectively;
[0099] S72. Since the target unit and the protection unit have ripple characteristics, that is, there are steps similar to ripples, the ripple calculation formula is as follows:
[0100] RDiff=|Rp1-Rp2|-|Rp1-DetectValue|;
[0101] Where: RDiff is the ripple difference; Rp1 is the average value of the protection unit of the inner circle, that is, ripple value 1; Rp2 is the average value of the protection unit of the outer circle, that is, ripple value 2; DetectValue is the value of the detection unit;
[0102] S73. In S6, the detection unit value in the primary ripple comparator is greater than the threshold value and is output to the secondary ripple comparator. It is defined that when the ripple difference of the detection unit is greater than the threshold value 0.1, the detection unit is output through the secondary ripple comparator.
[0103] The cross self-calibration constant false alarm rate target detection method of the present invention is compared with other common constant false alarm rate methods to determine comparison indicators and evaluate them. The specific process is as follows:
[0104] 1) Define the evaluation criteria for the constant false alarm rate method, and define the false alarm rate. The false alarm rate is the ratio of the number of false alarm points to the number of true target points. The smaller the criterion, the better;
[0105] 2) Define the evaluation criteria for the constant false alarm rate method, and define the missed alarm rate. The missed alarm rate is the ratio of the number of missed alarm points to the number of true target points. The smaller the standard, the better;
[0106] 3) The cross self-calibration constant false alarm rate target detection method is compared with various secondary constant false alarm rate and general two-dimensional constant false alarm rate target detection methods in various clutter environments.
[0107] like Figure 6 As shown, Figure 6 The results show the performance of a cross-shaped self-calibration constant false alarm rate (CFAR) target detection method based on millimeter waves in various environments. The results show that the cross-shaped self-calibration constant false alarm rate (CFAR) target detection method can adapt to various environments, and the false alarm probability is determined by the matrix characteristics, eliminating the need for multiple experimental corrections.
[0108] The technical solution of the present invention is not limited to the above-mentioned specific embodiments. Any technical variations made according to the technical solution of the present invention fall within the protection scope of the present invention.
Claims
1. A millimeter-wave-based cross self-calibration constant false alarm rate target detection method, characterized in that: The following steps are involved: S1. Data collection: Use IWR6843ISK millimeter wave radar equipment to collect binary data including four antennas; S2, signal preprocessing: convert binary data into decimal, distribute the real and imaginary parts of the data, and take the data of the first two receiving antennas to form a data sequence; S3, data matrix generation: the data sequence is processed twice by fast Fourier transform to generate the range-Doppler matrix; S4, false alarm probability calculation: by analyzing the characteristics of the range-Doppler matrix, the false alarm probability is calculated, and the false alarm probability is used to calculate the threshold factor; S5, offset factor calculation: by analyzing the characteristics of the range-Doppler matrix, the offset factor is calculated, and the offset factor is used for subsequent offset calculation; the offset is used to correct the threshold value and reduce the false alarm rate; S6, reference unit selection and threshold calculation: select the cross calibration frame as the reference unit, calculate the threshold value and threshold offset, and input them into the primary ripple comparator; S7, the ripple difference calculated by the protection unit is compared with the set threshold in the secondary ripple comparator, and a binary value is output to reduce the false alarm rate; S71, extracting the values of the two outermost circles of the protection unit to form two arrays, and taking average values of the two arrays respectively; S72. Since the target unit and the protection unit have ripples, that is, there are steps similar to ripples, the ripple calculation formula is as follows: RDiff=|Rp1-Rp2|-|Rp1-DetectValue|; Where: Rdiff is the ripple difference; Rp1 is the average value of the protection unit of the inner circle, that is, the ripple value 1; Rp2 is the average value of the protection unit of the outer circle, that is, the ripple value 2; DetectValue is the value of the detection unit; S73. In S6, the detection unit value in the primary ripple comparator is greater than the threshold value and is output to the secondary ripple comparator. It is defined that when the ripple difference of the detection unit is greater than the threshold value 0.1, the detection unit is output through the secondary ripple comparator.
2. The cross self-calibration constant false alarm rate target detection method based on millimeter wave according to claim 1 is characterized in that: The step S4 specifically includes: S41. Normalize the range-Doppler matrix to prevent false alarms and missed alarms due to large data gaps. The calculation formula is as follows: Where: R norm Represents the normalized matrix, the matrix shape is N×M, which means the matrix has N rows and M columns; R rdm represents the range-Doppler matrix generated by S3 processing. The matrix shape is N×M, which means the matrix has N rows and M columns. max(R rdm ) represents taking the matrix R rdm The matrix element with the largest value is used as the reference value, i.e. the denominator; S42, obtaining the range-Doppler matrix characteristics, taking a representative value for each row and column, that is, taking the positive and negative diagonal elements of the matrix to form an array, the calculation formula of the diagonal array is as follows: Diag1=R norm [i][i](0≤i<N); Diag2=R norm [i][M-i](0≤i<N); Where: Diag1, Diag2 represent the diagonal and anti-diagonal arrays; i represents the index value; R norm [i][i],R norm [i][Mi] represents the element under the corresponding matrix index; S43. Calculate the value of the false alarm probability according to the number of protection units selected. The calculation formula is as follows: Where: P represents the value of false alarm probability, which represents the false alarm probability when the constant false alarm probability method is performed; mean(Diag1[::3]) represents slicing the array Diag1, taking a value every 2 numbers to form an array, and taking the average value of the array; mean(Diag2[::3]) represents slicing the array Diag2, taking a value every 2 numbers to form an array, and taking the average value of the array; Unit p is the protection unit size, that is, the size of the protection unit matrix, that is, the protection unit size is Unit p ×Unit p .
3. The cross self-calibration constant false alarm rate target detection method based on millimeter wave according to claim 1, characterized in that: The step S5 specifically includes: S51. Take the median of all positive numbers in the Diag1 and Diag2 arrays and calculate the offset factor. The calculation formula is as follows: Corr factor =(Diag1 m +Diag2 m )×2; Where: Corr factor Represents the offset factor, which is involved in the subsequent threshold offset calculation; Diag1 m and Diag2 m Represents the median of the positive sequence of the Diag1 and Diag2 arrays.
4. The cross self-calibration constant false alarm rate target detection method based on millimeter wave according to claim 1, characterized in that: The step S6 specifically includes: S61, selecting a corresponding reference unit, the selected reference unit is a cross calibration frame, the shape of which is a circle of elements surrounding the edge of the protection unit as a reference unit, i.e., a capture frame, and calculating with the elements of the row and column corresponding to the detection unit as the reference unit; S62, obtain several reference unit intervals and combine them into threshold value alternatives. The calculation formula is as follows: ThresholdValue=[Cl1,Cl2,CB]; Where: ThresholdValue represents the threshold value alternative array; Cl1 represents the corresponding row reference cell array; Cl2 represents the corresponding column reference cell array; CB represents the capture box corresponding reference cell; S621: Calculate the candidate threshold index, the calculation formula is as follows: Index=|(1-P)×len(ThresholdValue)|; Where: Index represents the index number; (1-P) represents the detection probability; the formula in len means the length of the candidate threshold array; S622: By using the candidate threshold array index and combining the threshold offset proposed in S64, the calculation formula is as follows: Value=ThresholdValue[Index]+Corr; Where: Value represents the corrected threshold value; thresholdValue[Index] represents the element in the array ThresholdValue located by the index value Index; Corr is the threshold offset; S63, calculating the reference unit: calculating the threshold value through the reference unit, and the calculation formula is as follows: Where: parameters N Indicates the threshold parameters2 and parameters2; Cl N Indicates Cl1 and Cl2, CB mean Indicates the mean of the capture frame reference unit; the formula in max means taking a value greater than or equal to 0; the meaning of this calculation formula is to take the difference between each element of the reference unit array of the row and column and the capture frame mean and take the value greater than or equal to 0 and average it; S64, calculating the threshold offset by the value of the threshold in S63 and the threshold offset factor in S51, the calculation formula is as follows: Where: Corr represents the threshold offset.
