WO2021217795A1 - Traffic radar and target detection method therefor and apparatus thereof, and electronic device and storage medium - Google Patents

Abstract

Disclosed are a traffic radar and a target detection method therefor and apparatus thereof, and an electronic device and a storage medium. Improvements are made on the basis of traditional CA-CFAR, influence factors such as strong interference in a background and mutual interference between multiple targets are eliminated by means of a deletion factor, and clutter in a scene of a traffic intersection is accumulated, such that the influence of factors such as the clutter and interference in the traffic intersection on target detection is reduced, thereby improving the detection performance.

Description

Traffic radar and its target detection method, device, electronic equipment and storage medium Technical field

The invention relates to the field of traffic detection, in particular to a method for detecting a target by a traffic radar.

Background technique

The basic concept of radar was formed in the early 20th century until World War II. Due to war needs, radar technology developed extremely rapidly. As a weapon, radar was mainly used to detect air planes, missiles and other aircraft. After the war, the use of radar was developed and used in the fields of weather forecasting, resource detection, celestial research, automobiles and transportation.

In the application of radar in various fields, the radar must use Constant False-Alarm Rate (CFAR) to detect targets. The constant false alarm detection technology is a technology for the radar system to distinguish between the signal and noise output by the receiver under the condition of keeping the false alarm probability constant to determine whether the target signal exists.

The constant false alarm detection technology uses adaptive threshold estimation technology to automatically detect the target. The detection threshold is related to the average power of the local environmental noise or clutter. Therefore, a better CFAR judge is designed to detect background noise or clutter. The statistics of waves are very important. Under normal circumstances, they obey various specific distributions, such as Rayleigh distribution, lognormal distribution, Weibull distribution, k distribution and so on.

Among the various applications of radar, the application platform of radar is divided into two types: fixed and mobile. The radar faces the clutter background, which is relatively fixed and constantly changing. Radar is applied to fixed platforms. For example, ground-based radar detects aircraft in the air. There are mainly clouds, rain and other backgrounds in the sky. The reflection from the background of clouds, rain, etc. is relatively obvious, so the constant false alarm detection algorithm used by ground-based radars to detect targets can meet the requirements as long as it can shield the effects of clouds, rain and other backgrounds on the target. Radar is applied to mobile platforms. For example, when radar is used for car collision warning, adaptive cruise and other functions, because the car platform is moving, the constant false alarm detection algorithm used by the car radar to detect targets must shield the background from changing during car movement. Impact on target detection.

The constant false alarm (CFAR) algorithm in the above two application scenarios is not suitable for the application scenarios of traffic intersections. In the application scenario of traffic intersections, the radar is placed on an electric police pole to emit electromagnetic waves to illuminate the road surface and detect various vehicles on the road surface. The application scenario of traffic intersection has its own characteristics. For example, there are metal fences on both sides of the intersection, roadside trees, and green belts between lanes. At the same time, there are many non-motorized lanes, large numbers of vehicles, small spacing, and large vehicles. Cover all kinds of complicated situations such as small cars, pedestrians, slow or stationary vehicles. Therefore, this scene of traffic intersection has many targets and uneven clutter, causing the background noise power level to deviate from the actual value, which will cause the false alarm rate and the detection rate to deviate, which brings challenges to target detection and affects the reliability of the detection effect. .

Summary of the invention

The purpose of the present invention is to provide a traffic radar target detection method, which is used to solve the technical problem that the false alarm rate and the detection rate of radar target detection at traffic intersections deviate, and the reliability of the detection effect is affected.

In order to achieve the above objective, the present invention proposes the following technical solutions:

The target detection method of traffic radar,

Receiving all echo signals in one pulse period in the TDM-MIMO working mode; in the TDM-MIMO working mode, N transmitting antennas transmit signals sequentially, and M receiving antennas receive signals synchronously;

Preprocess the echo signal to obtain a discrete digital signal;

Perform FFT transformation on the discrete digital signal along the sampling sequence number direction to obtain the transformed signal;

Combine all transformed signals into virtual channel data:

[S _0,0 ,S _0,1 ,…,S _0,M-1 ,S _1,0 ,S _1,1 ,…,S _1,M-1 ,…S _nm …S _N-1,0 , S _N-1,1 ,…,S _N-1,M-1 ],

The S _nm represents all data transmitted by the n-th transmitting antenna and received by the m-th receiving antenna in one period;

Arranging the transformed signal into a data cube containing the above-mentioned virtual channel data and the three dimensions of distance and Doppler velocity;

Obtain the distance-azimuth spectrum matrix from the data cube;

Traverse each data unit in the distance-azimuth spectrum matrix, and perform the following processing on each data unit:

Dividing sub-regions and obtaining an average value of each sub-region, where the sub-region is each sub-region in the detection region where the current data unit is located;

Obtaining a sub-region average ratio, where the sub-region average ratio is obtained by dividing the sub-region average value and the background value accumulated by the current data unit;

Obtaining the mean value of the background sub-region, the background sub-region is all the sub-regions corresponding to the part whose mean value ratio of the sub-region is less than the deletion factor, and is obtained by taking the mean value of the sub-region mean of each sub-region that meets the above conditions;

Acquiring a threshold, the threshold being obtained by multiplying the mean value of the background sub-region and the threshold factor;

Compare the current data unit with the threshold, and judge the current data unit larger than the threshold as the target.

Further, in the present invention, the data cube obtains the range-azimuth spectrum matrix through the angle spectrum estimation algorithm of the capon algorithm.

Further, in the present invention, the detection area is rectangular.

Further, in the present invention, the detection area where the current data unit is located is divided into A×B sub-areas, the sub-areas is a matrix containing E×D data units, and the average value of each sub-area is

s=1,2,...,AE; l=1,2,...,BD

Among them, A represents the number of vertical sub-areas of the detection area, B represents the number of horizontal sub-areas of the detection area, a represents the serial number of the vertical sub-area of the detection area, b represents the serial number of the horizontal sub-areas of the detection area, and E represents the vertical data unit in the sub-area Quantity, D represents the number of horizontal data units in the sub-region;

x _sl represents the element value in the current background area.

Further, in the present invention, the deletion factor is a cumulative probability distribution function of the ratio of sub-region means under a uniform clutter background

The value of c corresponding to, where C is the average ratio of the first sub-region, M _{I_acc,ii} (f-1) is the background value accumulated by the data of f-1 frames before the location of the current data unit, and f represents the current data frame number.

Further, in the present invention, the background value accumulated by the current data unit is

M _{I_acc,ii} (f)=β·M _{I_acc,ii} (f-1)+(1-β)·Z

Among them, β is the background accumulation coefficient; Z is the mean value of the background sub-region.

Beneficial effects:

It can be seen from the above technical solutions that the technical solution of the present invention provides a traffic radar and its target detection method, which is improved on the basis of traditional CA-CFAR, and can accumulate clutter in the scene of traffic intersections and reduce The impact of traffic intersection clutter on target detection and improve detection performance.

Specifically, the following technical means are adopted to obtain good technical effects:

1. Using the capon algorithm to estimate the angle spectrum, get the range-azimuth spectrum, and do CFAR detection based on the range-azimuth spectrum, which can simultaneously obtain the target's distance and azimuth;

2. The reference unit data matrix is divided into blocks. The average value of the block matrix and the accumulated background value are used to obtain the average value ratio. According to the average value ratio, strong interference is eliminated to prevent strong interference from raising the background value and reducing the target detection rate;

3. The application scenarios of traffic radar are relatively fixed, and background accumulation eliminates the influence of fixed backgrounds such as guardrails and green belts at traffic intersections on target detection, and improves CFAR detection performance in multi-target environments;

4. In the range-azimuth spectrum estimation, the spectrum within the radar field of view of the intersection is obtained. The radar field of view observation range is basically the same as the human eye field of view observation range, which is convenient for users to visually observe the reflection of targets at various distances and azimuths. Strong and weak.

It should be understood that all combinations of the foregoing concepts and the additional concepts described in more detail below can be regarded as part of the inventive subject matter of the present disclosure as long as such concepts are not mutually contradictory.

The foregoing and other aspects, embodiments and features of the teachings of the present invention can be more fully understood from the following description with reference to the accompanying drawings. Other additional aspects of the present invention, such as the features and/or beneficial effects of the exemplary embodiments, will be apparent in the following description, or learned from the practice of the specific embodiments taught in accordance with the present invention.

Description of the drawings

The drawings are not intended to be drawn to scale. In the drawings, each identical or nearly identical component shown in each figure may be represented by the same reference numeral. For the sake of clarity, not every component is labeled in every figure. Now, embodiments of various aspects of the present invention will be described by way of examples and with reference to the accompanying drawings, in which:

Figure 1 is a schematic diagram of the system configuration of a traffic radar in an embodiment of the present invention;

Figure 2 is a flow chart of signal processing of a traffic radar in an embodiment of the present invention;

Figure 3 is a signal processing module of a traffic radar in an embodiment of the present invention;

4 is a schematic diagram of the data structure before and after the Capon algorithm of the traffic radar in the embodiment of the present invention;

5 is a schematic diagram of the angular distance azimuth spectrum matrix H(r, θ) of the traffic radar in an embodiment of the present invention;

6 is a schematic diagram of a reference unit of a 2D-CA-CFAR rectangular detection window of a traffic radar in an embodiment of the present invention;

FIG. 7 is a schematic diagram of background accumulation of traffic radar in an embodiment of the present invention;

Fig. 8 is a schematic diagram of a 2D-CA-CFAR detector of a traffic radar in an embodiment of the present invention;

FIG. 9 is the CFAR detection result of the traffic radar in the embodiment of the present invention.

Detailed ways

In order to better understand the technical content of the present invention, specific embodiments are described in conjunction with the accompanying drawings as follows.

The original intention of the specific embodiments of the present invention is that the application scenarios of radar in the prior art are not suitable for traffic intersections, because the traffic intersections themselves have the characteristics of multiple targets and uneven clutter, which makes the background noise power level deviate from the actual These problems are not prominent in the mature application scenarios of radar, and under the requirement of detecting targets at traffic intersections, the above problems obviously increase the false alarm rate and reduce the detection rate.

Based on the analysis of the above problems, from the perspective of reducing the background noise power, the influence of the background noise power can be minimized to solve the above problems.

In the traffic intersection environment, the main reasons that affect the background noise power are the fixed scenes such as guardrails and green belts in the traffic intersection; the second is that the multiple targets are very close to each other, which raises the background value of the current detection unit . As long as these factors can be removed and a real clutter background can be restored, the detection rate can be effectively improved.

Based on the above-mentioned conception, the following specific embodiments of the present invention are derived.

Hardware condition

As shown in Figure 1, it is the composition of the traffic radar of the present invention. It consists of transmitting antenna, receiving antenna, radio frequency system, ARM and DSP processors, and common peripheral interfaces.

The function of the antenna is to transmit and receive millimeter waves; the main function of the radar radio frequency front end is to sample the intermediate frequency signal, and the function of the crystal oscillator is to provide a clock frequency of 40MHz as shown in the figure to support the radar radio frequency front end work;

The ARM processor is mainly used for the overall operation control of the radar;

DSP mainly performs digital signal processing functions on ADC sampled signals;

FLASH memory is used to store the results of program and data information processing;

Transmit to PC system or subsequent processing module through CAN port or serial port.

External data interfaces, such as CAN and CAN FD in the figure, are used to transmit the processed data to the outside;

The display module receives and displays the data transmitted from the external data interface. The display module is a graphical display interface GUI on the PC.

Embodiment one, method

The target detection method of traffic radar as shown in Fig. 2 includes the following steps:

S100. Receive all echo signals of one pulse period in the TDM-MIMO working mode; in the TDM-MIMO working mode, N transmitting antennas transmit signals sequentially, and M receiving antennas receive signals synchronously; as shown in Figure 1 As shown, the transmitting antenna is Tx, and the distance between two adjacent transmitting antennas is dt; the receiving antenna is Rx, and the distance between two adjacent receiving antennas is dr.

The signal receiving and sending process in the above-mentioned pulse period is as follows: N transmitting antennas start from number 0 and transmit sequentially until the end of the period after the transmission of number N-1 is completed, the working time of each transmitting antenna is T, and M receiving antennas receive numbers at the same time The echo signals generated by the transmitting antennas of 0, 1,..., N-1, that is, each receiving antenna can receive N echo signals, and M receiving antennas can receive a total of MN echo signals.

The entire elapsed time during the sending and receiving period of the above-mentioned pulse period and the signal processing period is referred to as a frame period. The embodiment of the present invention takes the processing procedure of each frame period as the object of discussion.

In each frame period, it specifically includes the following steps:

S101. Preprocess all echo signals to obtain discrete digital signals; this process is performed in the DSP, and is performed in real time as the echo signals are received.

For example, for period i, the echo signal transmitted by the nth transmitting antenna and received by the mth receiving antenna is S _n,m (t,i), where 0≤i≤I-1, 0≤m ≤M-1, 0≤n≤N-1, t represents the fast time in period i, the echo signal S _n,m (t,i) is down-converted into an intermediate frequency signal, and then low-pass filtered and amplified After processing, the ADC is sampled and converted into a discrete digital signal S _n,m (k,i), where k is the ADC sampling sequence number at the sampling frequency fs;

S102: Perform FFT transformation on the discrete digital signal along the sampling sequence number direction to obtain a transformed signal;

For example, for the above-mentioned discrete digital signal S _n,m (k,i), the FFT transform of the distance dimension Nr is performed along the k direction to obtain the result Sr _n,m (k,i), Sr _n,m (k,i) is It is the transformed signal.

Preferably, Nr is the smallest value that is greater than or equal to the number of sampling points in the distance dimension and is an exponential power of 2. S103. Combine all transformed signals into virtual channel data:

[S _0,0 ,S _0,1 ,…,S _0,M-1 ,S _1,0 ,S _1,1 ,…,S _1,M-1 ,…S _nm …S _N-1,0 , S _N-1,1 ,…,S _N-1,M-1 ],

The S _nm represents all data transmitted by the n-th transmitting antenna and received by the m-th receiving antenna in one period;

S104. Since each S _nm has a corresponding distance and Doppler velocity parameter, a data cube is formed by three dimensions of virtual channel data, distance, and Doppler velocity; therefore, each frame period corresponds to a data cube.

As shown on the left side of Figure 5, the data cube Sr_cube(k,i,nm) established with virtual channel data, distance, and Doppler velocity as the two-by-two vertical coordinates (where i represents the pulse period number, 0≤i≤I -1, 0≤nm≤NM-1), the Z axis is the first virtual channel data, the X axis is the distance, and the Y axis is the Doppler velocity.

S105. Obtain the distance-azimuth spectrum matrix H(r, θ) from the data cube, where r is the distance from the pole of the polar coordinate, and θ is the angle with the positive direction. Specifically, in the specific embodiment of the present invention, the capon algorithm angle spectrum estimation algorithm is selected to obtain the two-dimensional matrix of Cartesian coordinates as shown on the right side of FIG. 4, where the Y axis is the azimuth angle, and the X axis is the distance; In 5, the distance-azimuth spectrum matrix is expressed in Cartesian coordinates and polar coordinates, respectively.

Through the above algorithm, the Doppler dimension information is accumulated while keeping the distance dimension unchanged. The original Doppler dimension has multiple sampled data, and after accumulation, it becomes one data, and the capon algorithm is applied to the virtual channel dimension. , Converted to azimuth dimension.

S106. Traverse each data unit in the distance-azimuth spectrum matrix, and perform the following processing on each data unit:

S1060. Divide sub-regions and obtain an average value of each sub-region, where the sub-region is each sub-region in the detection region where the current data unit is located;

There are many ways to divide sub-areas, which can be selected according to needs, as long as the current data unit is satisfied. When dividing sub-regions, it is necessary to consider factors such as the amount of calculation, how much computing time, and whether it can be processed in real time; for example, it can be 1 data unit as 1 sub-region, 2 data units as 1 sub-region, and 3 data units As a sub-area and so on.

As shown in FIG. 6, in terms of regular division and convenient calculation, a rectangle is used to divide the detection area to form the first sub-area. The detection area is a dark rectangle in FIG. 6, in which one unit is the current data unit, and the remaining units in the detection area are reference units.

There can be multiple sub-areas within the above rectangular area, which are divided according to the following methods:

The detection area where the current data unit is located is divided into A×B sub-areas, and the sub-areas is a matrix containing E×D data units.

For each sub-region, the average value of the sub-region can be obtained by averaging the data units in it:

s=1,2,...,AE; l=1,2,...,BD

Among them, A represents the number of vertical sub-areas of the detection area, B represents the number of horizontal sub-areas of the detection area, a represents the serial number of the vertical sub-area of the detection area, b represents the serial number of the horizontal sub-areas of the detection area, and E represents the vertical data unit in the sub-area Quantity, D represents the number of horizontal data units in the sub-region;

x _sl represents the element value of the background element x in the current background area, s represents the row number of the background element x in the current background area, and l represents the column number of the background element x in the current background area.

As shown in Figure 7, an 8*8 detection area is divided into 2×2 sub-areas, and each sub-area is a matrix containing 4×4 data units, that is, A=2, B=2, C=4, D=4. In the detection area, the current data unit is x ₄₄ , and the remaining data units are reference units.

The average values of the above 2×2 sub-regions are

S1061. Obtain a sub-region average ratio, where the sub-region average ratio is obtained by dividing the sub-region average value with the background value accumulated by the current data unit, that is, the average ratio

M _{I_acc,ii} (f-1) represents the background value accumulated in the f-1 frame data before the current data unit location, which can represent the average level of the background in the current detection area, and _mab represents a certain sub-region after the current detection area is divided into sub-regions. The background mean of the area.

S1062. Obtain the mean value of the background sub-region. The background sub-region is all the sub-regions corresponding to the part where the mean ratio of the sub-region in S1061 is less than the deletion factor. The mean value of the sub-region of each sub-region that meets the above conditions can be obtained by taking the mean value. .

That is, if among the above m ₁₁ , m ₂₁ , m ₂₂ , and m ₁₂ , _{the mean ratio of m 12} is greater than the deletion factor, and the remaining mean ratios are less than the deletion factor, so _{take the average of m 11} , m ₂₁ , and m ₂₂ to get the background The average value of the sub-region.

The size of the sub-region mean ratio indicates the degree of deviation between the sub-region mean and the overall mean, and the overall mean represents the mean value of the clutter background when there is no interfering target in the entire detection area. When there is no interfering target in a sub-region, its sub-region mean is close to the overall mean. When there is an interfering target in a sub-region, the deviation of the sub-region average from the overall mean increases.

The sub-area mean ratio obeys a certain distribution. In a uniform clutter environment, it must have a relatively stable maximum value. When the sub-area mean ratio is greater than the above maximum value, there is interference in the sub-area. Therefore, the above maximum value is the deletion factor .

Then, delete the sub-regions whose average value of the sub-region meets the above conditions from all the sub-regions, and all the remaining sub-regions constitute the background sub-region, which can represent the sampling of the clutter power level, which is approximately the clutter in the uniform clutter environment. background.

Therefore, the part that can be classified into the background sub-areas is divided by the deletion factor, thereby removing the sub-areas with interfering targets, and the resulting background sub-areas is free of interfering targets, that is, it is approximately uniform clutter. The sub-area corresponding to the clutter background in the environment.

In the above step S1061, the background value accumulated by the current data unit is calculated as follows:

M _{I_acc,ii} (f)=β·M _{I_acc,ii} (f-1)+(1-β)·Z

Among them, β is the background accumulation coefficient, which generally ranges from 0.8 to 0.95; Z is the mean value of the background sub-region.

The above formula is an update formula for the background value accumulated in the current data unit. Since each frame period needs to go through the calculation process of S1062 above, that is, there is a corresponding Z value, therefore, the background value M _{I_acc accumulated by the current data unit, ii} (f) is not only related to the current Z value, but also related to the current data unit _{It is related to the background value M I_acc,ii} (f-1) accumulated in f-1 frame data before the location. Combining the above accumulation process as shown in Figure 8 to illustrate, that is, each frame period calculates the background value of the current data unit once and updates the corresponding distance-azimuth accumulation in the distance-azimuth background matrix accumulated on the left side of Figure 8 The background value is used as a factor that affects the accumulated background value of the current data unit of the next frame period in the next frame period.

In particular, for the first frame of data, that is, M _{I_acc, the} initial value of ii (f) is the mean value Z of the background sub-region corresponding to the first frame of data.

In the above step S1062, the selection of the deletion factor is related to the elimination of the first sub-region containing the interference background, and therefore directly affects the performance of the CFAR detection method.

The selection of the deletion factor is related to the deletion of sub-regions, and therefore directly affects the detection performance of the CFAR algorithm. Since the deletion factor is related to the clutter type and distribution parameters, Monte Carlo method can be used to approximate the solution. Under the background of uniform clutter, the cumulative probability distribution function of the mean ratio

In order to judge whether the clutter in the reference window is stable and get a good detection performance of a uniform background, the corresponding value c of the cumulative probability distribution function F(c)≈1 of the sub-region mean ratio is selected, and the value of c is equal to the deletion factor. Value, F(c)=1 is used for calculation. The clutter test method of probability density transformation is used in this patent to estimate the clutter distribution type and distribution parameters.

S1063. Obtain a threshold, which is obtained by multiplying the average value of the background sub-region and the threshold factor T; the threshold here represents a value for distinguishing whether it is a target, and usually the value is an empirical value.

S1064. Compare the current data unit with the threshold, and judge the current data unit larger than the threshold as a target, and judge the current data unit smaller than the current as no target.

Corresponding to the target points detected in the H(r,θ) matrix, assuming that there are a total of P target points after the detection, the distance-azimuth of the p-th target is (r _p ,θ _p );

Then, the Doppler estimation of the target is performed to obtain the target's velocity estimation v _p ; the parameters of the target's distance estimation, azimuth estimation, and velocity estimation are obtained, and the above-mentioned parameters of the target are sent to the subsequent processing module.

As shown in Figure 9, the black data unit in the matrix in the figure represents the location of the detected target, which is described by the orientation and distance.

Embodiment two, module

The specific embodiment of the present invention also provides a target detection device for traffic radar, as shown in FIG. 3, including

The receiving module is configured to receive all echo signals in one cycle in the TDM-MIMO working mode; in the TDM-MIMO working mode, N transmitting antennas transmit signals in sequence, and M receiving antennas receive signals synchronously;

The preprocessing module is used to preprocess the echo signal to obtain discrete digital signals;

The FFT transformation module is used to perform FFT transformation on the discrete digital signal in the direction of the sampling sequence number to obtain the transformed signal;

The virtual channel composing module is used to compose all the transformed signals into virtual channel data:

[S _0,0 ,S _0,1 ,…,S _0,M-1 ,S _1,0 ,S _1,1 ,…,S _1,M-1 ,…S _nm …S _N-1,0 , S _n-1,1 ,…,S _N-1,M-1 ],

The S _nm represents data transmitted by all the n-th transmitting antennas in one period and received by the m-th receiving antenna;

The data cube building module is used for arranging the transformed signal into a data cube containing the first virtual channel data, distance and Doppler velocity;

The distance-azimuth spectrum matrix building module is used to obtain the first distance-azimuth spectrum matrix from the data cube;

The first processing module is used to traverse each data unit in the distance-azimuth spectrum matrix and process each data unit;

The first processing module includes:

A sub-area division and sub-area average value acquisition module, configured to divide sub-areas and obtain the average value of each sub-area, where the sub-area is each sub-area in the detection area where the current data unit is located;

A sub-region mean value ratio obtaining module, configured to obtain a sub-region mean value ratio, the sub-region mean value ratio being obtained by dividing the sub-region mean value and the background value accumulated by the current data unit;

The background sub-region mean value obtaining module is used to obtain the mean value of the background sub-region. The background sub-region is all the sub-regions corresponding to the part whose mean value ratio of the sub-region is less than the deletion factor. Take the mean to get;

A threshold value acquiring module, configured to acquire a threshold value, the threshold value being obtained by multiplying the average value of the background sub-region and the threshold value factor;

The judgment module is used to compare the current data unit with the threshold, and judge the current data unit larger than the threshold as the target.

Further, in the foregoing embodiment, the distance-azimuth spectrum matrix construction module includes a capon algorithm angle spectrum estimation algorithm module, which is used to obtain the distance-azimuth spectrum matrix through the data cube through the capon algorithm angle spectrum estimation algorithm.

Further, in the above-mentioned embodiment, it further includes a deletion factor obtaining module, configured to obtain the deletion factor in the following manner: the deletion factor is the cumulative probability distribution function of the sub-region mean ratio under a uniform clutter background

The value of c corresponding to, where C is the ratio of sub-region averages, M _{I_acc,ii} (f-1) is the background value of data accumulated in f-1 frames before the location of the current data unit, and f represents the current data frame number.

Embodiment three, computer program product and computer readable storage medium

Another embodiment of the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the processor executes the computer program to realize the above-mentioned traffic Radar target detection method.

The processor is preferably, but not limited to, a central processing unit (CPU). For example, the processor may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (Field Programmable Gate Array, FPGA) or other Chips such as programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, or a combination of the above types of chips.

As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions/modules of the traffic radar target detection method in the embodiment of the present invention, The processor executes various functional applications and data processing of the processor by running non-transitory software programs, instructions, and modules stored in the memory, that is, the processor in the above method embodiment is preferably, but not limited to, a central processing unit ( Central Processing Unit, CPU). For example, the processor may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (Field Programmable Gate Array, FPGA) or other Chips such as programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, or a combination of the above types of chips.

As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as instructions/modules of a traffic radar target detection method in an embodiment of the present invention , The processor executes various functional applications and data processing of the processor by running non-transient software programs, instructions, and modules stored in the memory, that is, realizes the traffic radar target detection method in the above method embodiment.

Another embodiment of the present invention provides a computer-readable storage medium, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, realizes the above-mentioned traffic radar target detection method.

The memory may include a program storage area and a data storage area. The program storage area may store an operating system and an application program required by at least one function; the data storage area may store data created by the processor and the like. In addition, the memory is preferably but not limited to a high-speed random access memory. For example, it may also be a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may also optionally include a memory remotely arranged with respect to the processor, and these remote memories may be connected to the processor through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

Those skilled in the art can understand that the implementation of all or part of the procedures in the above-mentioned embodiment methods is a program that can be completed by a computer program instructing relevant hardware, which can be stored in a computer readable storage medium, and the program is executed during execution. At this time, it may include the procedures of the embodiments of the above-mentioned methods. Among them, storage media can be magnetic disks, optical disks, read-only memory (Read-Only Memory, ROM), random access memory (RAM), flash memory (Flash Memory), hard disk (Hard Disk Drive) , Abbreviation: HDD) or solid-state drive (Solid-State Drive, SSD), etc.; the storage medium may also include a combination of the foregoing types of memories.

Example four, products

The foregoing embodiments are integrated into a traffic radar, which is suitable for traffic intersections with complex environments and has a high target detection rate.

In this disclosure, various aspects of the present invention are described with reference to the accompanying drawings, in which numerous illustrated embodiments are shown. The embodiments of the present disclosure are not necessarily intended to include all aspects of the present invention. It should be understood that the various concepts and embodiments introduced above, as well as those described in more detail below, can be implemented in any of many ways, because the concepts and embodiments disclosed in the present invention are not Limited to any implementation. In addition, some aspects disclosed in the present invention can be used alone or in any appropriate combination with other aspects disclosed in the present invention.

Claims (12)

1. The target detection method of traffic radar is characterized by:

   Receiving all echo signals in one pulse period in the TDM-MIMO working mode; in the TDM-MIMO working mode, N transmitting antennas transmit signals sequentially, and M receiving antennas receive signals synchronously;

   Preprocess the echo signal to obtain a discrete digital signal;

   Perform FFT transformation on the discrete digital signal along the sampling sequence number direction to obtain the transformed signal;

   Combine all transformed signals into virtual channel data:

   [S _0,0 ,S _0,1 ,…,S _0,M-1 ,S _1,0 ,S _1,1 ,…,S _1,M-1 ,…S _nm …S _N-1,0 , S _N-1,1 ,…,S _N-1,M-1 ],

   The S _nm represents all data transmitted by the n-th transmitting antenna and received by the m-th receiving antenna in one period;

   Arranging the transformed signal into a data cube containing the above-mentioned virtual channel data and the three dimensions of distance and Doppler velocity;

   Obtain the distance-azimuth spectrum matrix from the data cube;

   Traverse each data unit in the distance-azimuth spectrum matrix, and perform the following processing on each data unit:

   Dividing sub-regions and obtaining an average value of each sub-region, where the sub-region is each sub-region in the detection region where the current data unit is located;

   Obtaining a sub-region average ratio, where the sub-region average ratio is obtained by dividing the sub-region average value and the background value accumulated by the current data unit;

   Obtaining the mean value of the background sub-region, the background sub-region is all the sub-regions corresponding to the part whose mean value ratio of the sub-region is less than the deletion factor, and is obtained by taking the mean value of the sub-region mean of each sub-region that meets the above conditions;

   Acquiring a threshold, the threshold being obtained by multiplying the mean value of the background sub-region and the threshold factor;

   Compare the current data unit with the threshold, and judge the current data unit larger than the threshold as the target.
2. The traffic radar target detection method according to claim 1, wherein the data cube obtains a range-azimuth spectrum matrix through a capon algorithm angle spectrum estimation algorithm.
3. The target detection method for traffic radar according to claim 1, wherein the detection area is a rectangle.
4. The traffic radar target detection method according to claim 1, wherein the detection area where the current data unit is located is divided into A×B sub-areas, and the sub-areas is a matrix containing E×D data units, each Mean of subregions

   

   Among them, A represents the number of vertical sub-areas of the detection area, B represents the number of horizontal sub-areas of the detection area, a represents the serial number of the vertical sub-area of the detection area, b represents the serial number of the horizontal sub-areas of the detection area, and E represents the vertical data unit in the sub-area Quantity, D represents the number of horizontal data units in the sub-region;

   x _sl represents the element value in the current background area.
5. The traffic radar target detection method according to claim 1, wherein the deletion factor is a cumulative probability distribution function of the ratio of the mean value of the sub-region under the background of uniform clutter.

   

   

   The value of c corresponding to, where C is the average ratio of the first sub-region, M _{I_acc,ii} (f-1) is the background value accumulated by the data of f-1 frames before the location of the current data unit, and f represents the current data frame number.
6. The traffic radar target detection method according to claim 5, wherein the background value accumulated by the current data unit is

   M _{I_acc,ii} (f)=β·M _{I_acc,ii} (f-1)+(1-β)·Z

   Among them, β is the background accumulation coefficient; Z is the mean value of the background sub-region.
7. The target detection device of traffic radar is characterized in that it comprises:

   The receiving module is configured to receive all echo signals in one cycle in the TDM-MIMO working mode; in the TDM-MIMO working mode, N transmitting antennas transmit signals in sequence, and M receiving antennas receive signals synchronously;

   The preprocessing module is used to preprocess the echo signal to obtain discrete digital signals;

   The FFT transformation module is used to perform FFT transformation on the discrete digital signal in the direction of the sampling sequence number to obtain the transformed signal;

   The virtual channel composing module is used to compose all the transformed signals into virtual channel data:

   [S _0,0 ,S _0,1 ,…,S _0,M-1 ,S _1,0 ,S _1,1 ,…,S _1,M-1 ,…S _nm …S _N-1,0 , S _N-1,1 ,…,S _N-1,M-1 ],

   The S _nm represents data transmitted by all the n-th transmitting antennas in one period and received by the m-th receiving antenna;

   The data cube building module is used for arranging the transformed signal into a data cube containing the first virtual channel data, distance and Doppler velocity;

   The distance-azimuth spectrum matrix building module is used to obtain the first distance-azimuth spectrum matrix from the data cube;

   The first processing module is used to traverse each data unit in the distance-azimuth spectrum matrix and process each data unit;

   The first processing module includes:

   A sub-area division and sub-area average value acquisition module, configured to divide sub-areas and obtain the average value of each sub-area, where the sub-area is each sub-area in the detection area where the current data unit is located;

   A sub-region mean value ratio obtaining module, configured to obtain a sub-region mean value ratio, the sub-region mean value ratio being obtained by dividing the sub-region mean value and the background value accumulated by the current data unit;

   The background sub-region mean value obtaining module is used to obtain the mean value of the background sub-region. The background sub-region is all the sub-regions corresponding to the part whose mean value ratio of the sub-region is less than the deletion factor. Take the mean to get;

   A threshold value acquiring module, configured to acquire a threshold value, the threshold value being obtained by multiplying the average value of the background sub-region and the threshold value factor;

   The judgment module is used to compare the current data unit with the threshold, and judge the current data unit larger than the threshold as the target.
8. The traffic radar target detection device according to claim 7, characterized in that: the distance-azimuth spectrum matrix construction module includes a capon algorithm angle spectrum estimation algorithm module, which is used to obtain the data cube through the capon algorithm angle spectrum estimation algorithm Range-azimuth spectrum matrix.
9. The traffic radar target detection device according to claim 7, characterized in that it further comprises a deletion factor acquisition module, configured to acquire the deletion factor in the following manner: the deletion factor is the ratio of the sub-region average value under the uniform clutter background Cumulative probability distribution function

   

   

   The value of c corresponding to, where C is the ratio of sub-region averages, M _{I_acc,ii} (f-1) is the background value of data accumulated in f-1 frames before the location of the current data unit, and f represents the current data frame number.
10. The electronic equipment for target detection of traffic radar includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to achieve The method steps described in any one of 1-6 are required.
11. A computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the method steps according to any one of claims 1 to 6 are realized .
12. The traffic radar is characterized by having the electronic device according to claim 10.

Images