Traffic target static slow-moving detection method and device and storage medium

Abstract

The invention discloses a traffic target static slow-moving detection method, which comprises the following steps: establishing a radar sensing area coordinate system, scanning a monitoring area through a radar and a camera, and measuring position information and speed information of a target vehicle; determining a lane where the target vehicle is located according to the x coordinate value of the target vehicle in the radar sensing area coordinate system; according to the y coordinate value of the target vehicle in the radar sensing area coordinate system, whether the instant speed of the target vehicle closest to the stop line is larger than or equal to a preset threshold value or not is judged; if yes, the target vehicle closest to the stop line is defined as a static target, and subsequent vehicles are counted in the y-axis direction. According to the invention, the traffic data collected by the millimeter wave radar and the camera are fused, so that the static and slow-down state of the vehicle at the actual road intersection can be better detected. The Leiyu fusion perception technology can be used for extracting and detecting multi-feature data and tracking a motion trail in the aspect of traffic incident detection, and can be used for realizing the acquisition of a queuing state at a road intersection and the high-precision detection of a road state.

Description

Technical Field

[0001] The present invention relates to a method, device, and storage medium for detecting stationary and slow-moving traffic targets. Background Art

[0002] The detection of the queue state at road intersections is an important part of the urban intelligent transportation system. It can visually observe the characteristics of road traffic conditions. Making good use of the queue data at road intersections can improve urban traffic efficiency and effectively reduce road congestion. Currently, the queue detection at road intersections mostly uses models or cross-section detection data for estimation and calculation. There are also perception methods based on video, infrared, floating cars, and multi-source data. Limited by theoretical model assumptions, the random and uneven distribution of sample sizes, the insufficient accuracy and difficult analysis of multi-source data, and the interference of external weather conditions such as visibility, there are deficiencies in the scientific perception and detection methods for the queue state at road intersections, making it difficult to accurately determine multiple traffic targets on the road comprehensively.

[0003] Currently, the mainstream method for detecting road vehicles is single-sensor detection, where the single sensor is mainly a millimeter-wave radar and a camera. The millimeter-wave radar can detect road traffic conditions all-weather and has advantages in depth information, harsh environments, and long-distance perception. However, it has poor visualization effects, poor target classification performance, is susceptible to ground clutter, and is prone to losing the detection of low-speed stationary targets. Moreover, the target point cloud is relatively sparse, and the accurate target shape and size cannot be obtained, resulting in the problem of difficult detection of stationary vehicles during the traffic target tracking process. The camera has advantages in stationary target detection, angular resolution, target information recognition and classification, but has problems such as inaccurate ranging and speed measurement, performance degradation under harsh conditions such as rain, snow, fog, strong light, and low light at night, and in the case of large targets or targets close to the camera, multiple target tracking frames may appear when the video detects the target, that is, the target splitting problem, and the road environment is complex and changeable, resulting in false detection problems during video processing. Summary of the Invention

[0004] The main objective of the present invention is to provide a method, device, and storage medium for detecting stationary and slow-moving traffic targets, aiming to solve the above technical problems.

[0005] To achieve the above objective, the present invention provides a method for detecting stationary and slow-moving traffic targets.

[0006] The method for detecting stationary and slow-moving traffic targets includes the following steps:

[0007] Establish a radar sensing area coordinate system and scan the monitoring area through the radar and the camera to measure the position information and speed information of the target vehicle;

[0008] Determine the lane where the target vehicle is located according to the x coordinate value of the target vehicle in the radar sensing area coordinate system;

[0009] According to the y - coordinate value of the target vehicle in the radar sensing area coordinate system, determine whether the instantaneous speed of the target vehicle closest to the stop line is greater than or equal to a preset threshold;

[0010] If so, define the target vehicle closest to the stop line as a stationary target and count the subsequent vehicles along the y - axis direction.

[0011] In one embodiment, the step of establishing the radar sensing area coordinate system and scanning the monitoring area through the radar and camera to measure the position information and speed information of the target vehicle includes:

[0012] Obtain the detection signals of the radar and the camera respectively;

[0013] Convert the radar detection signal and the camera detection signal;

[0014] Associate the converted radar detection signal and camera detection signal based on the target vehicle trajectory.

[0015] In one embodiment, the step of converting the radar detection signal and the camera detection signal includes:

[0016] Perform coordinate conversion on the radar detection signal and the camera detection signal, specifically using the following coordinate conversion formula:

[0017]

[0018] Where \((u r ,v r ,w r ,1) T is the coordinate of the target in the millimeter - wave radar coordinate, and the corresponding space coordinate system is \((u c ,v c ,w c ,1) T ;

[0019] (x,y,1) T is the coordinate of the target in the image pixel coordinate system;

[0020] R is a 3×3 unit orthogonal rotation matrix, T is a 3×1 translation matrix;

[0021] I is the identity matrix;

[0022] a x ,a y respectively represent the sizes of each pixel in the physical units in the horizontal and vertical directions;

[0023] f is the focal length of the camera;

[0024] k is the scaling factor.

[0025] In one embodiment, the step of converting the radar detection signal and the camera detection signal includes:

[0026] Mark the starting time points of the radar and the camera;

[0027] After aligning the starting time points, correspond the data frames according to the frame rates of the data collected by the radar and the camera.

[0028] In one embodiment, after the step of correlating the converted radar detection signal and camera detection signal based on the target vehicle trajectory, the method further includes:

[0029] Convert the discrete data points into continuous trajectories by curve fitting for the radar detection signal and the camera detection signal, define the trajectory curve obtained from the radar detection signal as S'(t), and define the trajectory curve obtained from the camera detection signal as P(t);

[0030] Calculate the similarity between the trajectory curve S'(t) and the trajectory curve P(t).

[0031] In one embodiment, the step of calculating the similarity between the trajectory curve S'(t) and the trajectory curve P(t) includes:

[0032] Calculate the set similarity between the target trajectory fitting curves detected by the sensors through the following formula:

[0033]

[0034] where cos(P(t), S'(t)) represents the covariance between the curve sampling coordinates;

[0035] represents the variance of the curve sampling coordinate points.

[0036] In one embodiment, the step of calculating the similarity between the trajectory curve S'(t) and the trajectory curve P(t) further includes:

[0037] Set the horizontal coordinate component u of the camera trajectory curve P(t), sample n times, and find the maximum value U max and the minimum value U min , and calculate the normalized value for each sampling point coordinate

[0038] Perform normalization processing on the radar projection trajectory curve S'(t) in the same way;

[0039] Let the Euclidean distance between the sampling points of the curves detected by the radar and the camera at this sampling be d;

[0040] Traverse the sampling points to obtain the maximum distance F of the sampling points of the two curves at the sampling number n n :

[0041] F n (p(t), s'(t)) = max{d(p(t)), s'(t))}, t ∈ [1, n]

[0042] Set the maximum number of sampling points N, and calculate the maximum distance F of the sampling points obtained at different sampling numbers n n , find the minimum value to obtain the distance proximity between the trajectory fitting curves:

[0043] The similarity between the curves is: R = pR u +(1 - p)R v , p ∈ [0, 1].

[0044] In addition, to achieve the above object, the present invention also provides a traffic target stationary and slow - moving detection method, the traffic target stationary and slow - moving detection method includes: a memory, a processor, and a traffic target stationary and slow - moving detection program stored on the memory and executable on the processor. When the traffic target stationary and slow - moving detection program is executed by the processor, the steps of the traffic target stationary and slow - moving detection method as described above are implemented.

[0045] In addition, to achieve the above object, the present invention also provides a computer - readable storage medium, on which a traffic target stationary and slow - moving detection program is stored. When the traffic target stationary and slow - moving detection program is executed by a processor, the steps of the traffic target stationary and slow - moving detection method as described above are implemented

[0046] The beneficial effects that the present invention can achieve: A traffic target stationary and slow - moving detection method proposed in an embodiment of the present invention can better detect the stationary and slow - moving states of vehicles at actual road intersections by fusing traffic data collected by millimeter - wave radars and cameras. The radar - vision fusion perception technology can extract and detect multi - feature data and track the movement trajectory in traffic event detection, and can achieve the acquisition of the queuing state at road intersections and the high - precision detection of road states. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a schematic structural diagram of the device of the hardware operating environment involved in the embodiment scheme of the present invention;

[0048] Figure 2 is a schematic flowchart of the first embodiment of the traffic target stationary and slow - moving detection method of the present invention;

[0049] Figure 3Schematic flowchart of the second embodiment of the traffic target stationary and slow-moving detection method of the present invention;

[0050] Figure 4 Schematic diagram of radar detection of the present invention;

[0051] Figure 5 Schematic diagram of time synchronization between radar and camera in the embodiment of the present invention;

[0052] Figure 6 Flowchart of target trajectory curve fitting in the embodiment of the present invention;

[0053] Figure 7 Flowchart of sensor detection target association in the embodiment of the present invention.

[0054] The implementation, functional features and advantages of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0055] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0056] As Figure 1 shown, Figure 1 is a schematic diagram of the terminal structure of the hardware operating environment involved in the embodiment of the present invention.

[0057] The terminal in the embodiment of the present invention may be a PC, or a movable terminal device with a display function such as a smart phone, a tablet computer, a portable computer, an e-book reader, etc.

[0058] As Figure 1 shown, the terminal may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard). Optionally, the user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0059] Optionally, the terminal may further include a camera, an RF (Radio Frequency) circuit, sensors, an audio circuit, a WiFi module, and so on. Among them, the sensors such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor. Among them, the ambient light sensor can adjust the brightness of the display screen according to the brightness of the ambient light, and the proximity sensor can turn off the display screen and / or the backlight when the mobile terminal is moved to the ear. As a kind of motion sensor, the gravity acceleration sensor can detect the magnitude of the acceleration in each direction (generally three axes). When stationary, it can detect the magnitude and direction of gravity, and can be used for applications that identify the posture of the mobile terminal (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc.; of course, the mobile terminal can also be configured with other sensors such as a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, etc., which will not be elaborated here.

[0060] Those skilled in the art can understand that Figure 1 the terminal structure shown in

[0061] does not limit the terminal, and may include more or fewer components than shown in the figure, or combine some components, or arrange different components. Figure 1 As shown in

[0062] In the Figure 1 shown terminal, the network interface 1004 is mainly used to connect to the background server and communicate with the background server for data; the user interface 1003 is mainly used to connect to the client (user side) and communicate with the client for data; and the processor 1001 can be used to call the traffic target stationary and slow-moving detection program stored in the memory 1005 and perform the following operations:

[0063] Establish a radar sensing area coordinate system and scan the monitoring area through the radar and the camera to determine the position information and speed information of the target vehicle;

[0064] Determine the lane where the target vehicle is located according to the x coordinate value of the target vehicle in the radar sensing area coordinate system;

[0065] According to the y coordinate value of the target vehicle in the radar sensing area coordinate system, determine whether the immediate speed of the target vehicle closest to the stop line is greater than or equal to the preset threshold;

[0066] If so, define the target vehicle closest to the stop line as a stationary target and count the subsequent vehicles along the y-axis direction.

[0067] Further, the processor 1001 may call the traffic target stationary and slow-moving detection program stored in the memory 1005 and further perform the following operations:

[0068] Obtain the detection signals of the radar and the camera respectively;

[0069] Convert the radar detection signal and the camera detection signal;

[0070] Associate the converted radar detection signal and camera detection signal based on the target vehicle trajectory.

[0071] Further, the processor 1001 may call the traffic target stationary and slow-moving detection program stored in the memory 1005 and further perform the following operations:

[0072] Perform coordinate transformation on the radar detection signal and the camera detection signal, specifically using the following coordinate transformation formula:

[0073]

[0074] where (u r , v r , w r , 1) T is the coordinate of the target in the millimeter-wave radar coordinate system, and the corresponding space coordinate system is (u c , v c , w c , 1) T ;

[0075] (x, y.1) T is the coordinate of the target in the image pixel coordinate system;

[0076] R is a 3×3 unit orthogonal rotation matrix, and T is a 3×1 translation matrix;

[0077] I is the identity matrix;

[0078] a x , a y respectively represent the sizes of each pixel in the physical units in the horizontal and vertical directions;

[0079] f is the focal length of the camera;

[0080] k is the scaling factor.

[0081] Further, the processor 1001 may call the traffic target stationary and slow-moving detection program stored in the memory 1005 and further perform the following operations:

[0082] Mark the starting time points of the radar and the camera;

[0083] After aligning the starting time points, the data frames are corresponded according to the frame rates of the data collected by the radar and the camera.

[0084] Further, the processor 1001 can call the traffic target stationary and slow-moving detection program stored in the memory 1005 and further perform the following operations:

[0085] Convert the discrete data points into continuous trajectories by curve fitting for the radar detection signal and the camera detection signal. Define the trajectory curve obtained from the radar detection signal as S'(t), and define the trajectory curve obtained from the camera detection signal as P(t);

[0086] Calculate the similarity between the trajectory curve S'(t) and the trajectory curve P(t).

[0087] The specific embodiments of the present invention applying the data storage device are basically the same as those of the following embodiments of the traffic target stationary and slow-moving detection method, and will not be elaborated here.

[0088] Refer to Figures 2 - 4 , a first embodiment of the present invention provides a traffic target stationary and slow-moving detection method, and the traffic target stationary and slow-moving detection method includes:

[0089] Step S10, establish a radar sensing area coordinate system and scan the monitoring area through the radar and the camera to measure the position information and speed information of the target vehicle.

[0090] The radar is installed on the traffic light pole (the height of the traffic light pole is not limited), the radar irradiation direction is the y-axis direction, the black square is the vehicle target, the dotted line is the lane line, the stop line of the road intersection is set as the radar target detection stop line, and the target travels in the oncoming direction.

[0091] First, establish a radar sensing area coordinate system according to the actual installation position, and the traveling direction of the target vehicle is the negative y-axis direction of the radar.

[0092] Step S20, determine the lane where the target vehicle is located according to the x coordinate value of the target vehicle in the radar sensing area coordinate system.

[0093] The millimeter-wave radar scans the monitoring area at a frequency of 60 ms, and performs mixing calculation through the vehicle target reflected wave to obtain the speed and position information of the target vehicle.

[0094] Step S30, according to the y coordinate value of the target vehicle in the radar sensing area coordinate system;

[0095] Step S40, determine whether the instant speed of the target vehicle closest to the stop line is greater than or equal to the preset threshold.

[0096] Wherein, according to the y coordinate value of the target, it is determined whether the instantaneous speed of the vehicle target closest to the stop line is less than or equal to the pre-determined threshold. If the speed of the front row vehicle is less than or equal to the threshold, continue to step S20, if the speed is greater than the threshold, continue to step S50.

[0097] Step S50, defining the target vehicle closest to the stop line as a stationary target and counting subsequent vehicles along the y-axis direction.

[0098] The queue length is the y-coordinate value of the last vehicle in the queue minus the y-coordinate value of the stop line.

[0099] In this embodiment, by fusing the traffic data collected by the millimeter-wave radar and the camera, the stationary and slow-moving state of vehicles at actual road intersections can be better detected. In terms of traffic event detection, the radar-vision fusion perception technology can extract and detect multi-feature data and track motion trajectories, which can realize the collection of the queue status at road intersections and the high-precision detection of road conditions.

[0100] Further, see Figure 3 , based on the above Figure 2 In the embodiment shown, the steps of establishing a radar sensing area coordinate system and scanning the monitoring area by radar and camera to determine the position information and speed information of the target vehicle include:

[0101] Step S60, respectively acquiring detection signals of the radar and the camera;

[0102] Step S70, converting the radar detection signal and the camera detection signal;

[0103] Step S80, correlating the converted radar detection signal and camera detection signal based on the target vehicle trajectory.

[0104] In this embodiment, to realize the fusion of radar and camera, it is necessary to start from the data level and unify the data of radar and camera from the time dimension and space dimension.

[0105] First, coordinate conversion operations need to be performed on the radar and camera to achieve spatial synchronization. The spatial synchronization conversion model of the radar and camera can be expressed as:

[0106]

[0107] Where (u r ,v r ,w r ,1) T is the coordinate of the target under the millimeter wave radar coordinate, and the corresponding spatial coordinate system is (u c ,v c ,w c ,1) T ;

[0108] (x,y.1) T is the coordinate of the target in the image pixel coordinate system;

[0109] R is a 3×3 unit orthogonal rotation matrix and T is a 3×1 translation matrix;

[0110] I is the identity matrix;

[0111] a x ,a y respectively represent the sizes of each pixel in the physical units in the horizontal and vertical directions;

[0112] f is the focal length of the camera;

[0113] k is the scaling factor.

[0114] Secondly, in a further embodiment, the step of converting the radar detection signal and the camera detection signal further includes:

[0115] Mark the starting time points of the radar and the camera;

[0116] After aligning the starting time points, perform the correspondence of data frames according to the frame rates of the data collected by the radar and the camera. Mark the starting times of the two sensors.

[0117] After aligning the starting time points, perform the correspondence between data frames according to the frame rates of the data collected by the radar and the camera. The scanning frequency of the millimeter-wave radar is one scan every 60 milliseconds, and the camera can collect 60 frames of data per second. The schematic diagram of the time synchronization of the radar and the camera is as Figure 4 shown.

[0118] Please refer to Figures 5 - 7 , and use the method of trajectory target association to achieve the synchronization and fusion of the detection targets of the two sensors. By fusing the target tracking results of the radar and the camera, decision-level fusion at the target trajectory level between the radar and the video is achieved. The radar-vision integrated detector can obtain the regional traffic state and perform single-point intersection detection based on the data fusion of the radar and the camera.

[0119] Specifically, after the step of associating the converted radar detection signal and camera detection signal based on the target vehicle trajectory, the method further includes:

[0120] Convert the discrete data points into continuous trajectories by curve fitting for the radar detection signal and the camera detection signal. Define the trajectory curve obtained from the radar detection signal as S'(t), and define the trajectory curve obtained from the camera detection signal as P(t);

[0121] Calculate the similarity between the trajectory curve S'(t) and the trajectory curve P(t).

[0122] In this embodiment, the target trajectory data within a period window obtained by the radar sensor and the camera sensor respectively are discrete data points. According to the differences in sensor hardware, it is necessary to convert the discrete data points into continuous trajectories through curve fitting, which can make the data smoother, reduce the system error caused by the asynchrony of sensor data acquisition time, and also avoid the influence of the trajectory space granularity difference on the subsequent similarity calculation results.

[0123] Within the same time window, the trajectory curve S'(t) obtained by curve fitting of the projection points of the target detected by the radar sensor on the image, and the trajectory curve P(t) obtained by curve fitting of the target detected by the camera. The similarity R is calculated as a measure of the similarity degree of the two trajectory curves.

[0124] First, it is necessary to calculate the geometric similarity between the target trajectory fitting curves detected by the sensors through the following formula:

[0125]

[0126] Among them, cos(P(t), S'(t)) represents the covariance between the curve sampling coordinates;

[0127] represents the variance of the curve sampling coordinate points.

[0128] Secondly, it is necessary to calculate the distance proximity between the trajectory fitting curves. Specifically:

[0129] Set the horizontal coordinate component u of the camera trajectory curve P(t), sample n times, and find the maximum value U max and the minimum value U min , and calculate its normalized value for each sampling point coordinate

[0130] Perform normalization processing on the radar projection trajectory curve S'(t) in the same way;

[0131] Let the Euclidean distance between the sampling points of the curves detected by the radar and the camera at this sampling be d;

[0132] Traverse the sampling points to obtain the maximum distance F of the sampling points of the two curves at the sampling number n n :

[0133] F n (p(t), s'(t)) = max{d(p(t)), s'(t))}, t ∈ [1, n]

[0134] Set the maximum number of sampling points N, and calculate the maximum distance F of the sampling points obtained at different sampling numbers n n, find the minimum value to obtain the distance proximity between the trajectory fitting curves:

[0135] Calculate the similarity R as a measure of the similarity degree of two trajectory curves. The similarity between the curves is: R = pR u +(1 - p)R v , p ∈ [0, 1].

[0136] Among them, p is the weight of the geometric similarity. The closer the value of R is to 1, the greater the similarity between the two trajectory curves.

[0137] Please refer to Figure 7 , assuming there are x target trajectories detected by the radar and y target trajectories detected by the camera. After calculating the similarity pairwise, a matrix R of x × y can be obtained x×y , where r i,j represents the similarity between the target trajectory curves detected by the i-th radar and the j-th camera. If its maximum value is r max , which is located in the m-th row and n-th column of the matrix R x×y , after removing it, a new matrix R is obtained (x-1)×(y-1) , if r max is greater than the set threshold P, then the radar and camera targets corresponding to this value are matched, and the matching relationship between the two targets is recorded. Repeat the operation until the matrix is empty.

[0138] In addition, an embodiment of the present invention also proposes a computer-readable storage medium. A traffic target stationary and slow-moving detection program is stored on the computer-readable storage medium. When the traffic target stationary and slow-moving detection program is executed by a processor, the following operations are implemented:

[0139] Establish a radar sensing area coordinate system and scan the monitoring area through the radar and the camera to measure the position information and speed information of the target vehicle;

[0140] Determine the lane where the target vehicle is located according to the x coordinate value of the target vehicle in the radar sensing area coordinate system;

[0141] According to the y coordinate value of the target vehicle in the radar sensing area coordinate system, determine whether the immediate speed of the target vehicle closest to the stop line is greater than or equal to the preset threshold;

[0142] If so, define the target vehicle closest to the stop line as a stationary target and count the subsequent vehicles along the y-axis direction.

[0143] Furthermore, when the traffic target stationary and slow-moving detection program is executed by the processor, the following operations are also implemented:

[0144] Obtain the detection signals of the radar and the camera respectively;

[0145] Convert the radar detection signal and the camera detection signal;

[0146] Associate the converted radar detection signal and camera detection signal based on the target vehicle trajectory.

[0147] Further, when the traffic target stationary and slow - moving detection program is executed by a processor, the following operations are also implemented:

[0148] Perform coordinate transformation on the radar detection signal and the camera detection signal, specifically using the following coordinate transformation formula:

[0149]

[0150] where \((u r ,v r ,w r ,1)\) T is the coordinate of the target in the millimeter - wave radar coordinate system, and the corresponding space coordinate system is \((u c ,v c ,w c ,1)\) T ;

[0151] (x,y,1) T is the coordinate of the target in the image pixel coordinate system;

[0152] R is a 3×3 unitary orthogonal rotation matrix and T is a 3×1 translation matrix;

[0153] I is the identity matrix;

[0154] a x ,a y respectively represent the sizes of each pixel in the physical units in the horizontal and vertical directions;

[0155] f is the focal length of the camera;

[0156] k is the scaling factor.

[0157] Further, when the traffic target stationary and slow - moving detection program is executed by a processor, the following operations are also implemented:

[0158] Deploy the real - time calculation of different clustering classifications as multiple instances.

[0159] Further, when the traffic target stationary and slow - moving detection program is executed by a processor, the following operations are also implemented:

[0160] Mark the starting time points of the radar and the camera;

[0161] After aligning the starting time points, perform the correspondence of data frames according to the frame rates of the data collected by the radar and the camera.

[0162] Further, when the traffic target stationary and slow-moving detection program is executed by a processor, the following operations are also implemented:

[0163] Convert the discrete data points into a continuous trajectory by curve fitting for the radar detection signal and the camera detection signal. Define the trajectory curve obtained from the radar detection signal as S'(t), and define the trajectory curve obtained from the camera detection signal as P(t);

[0164] Calculate the similarity between the trajectory curve S'(t) and the trajectory curve P(t)

[0165] The specific embodiments of the computer-readable storage medium of the present invention are basically the same as those of the above-mentioned traffic target stationary and slow-moving detection methods, and will not be elaborated here.

[0166] It should be noted that in this article, the terms "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or system. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or system including that element.

[0167] The serial numbers of the above-mentioned embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments.

[0168] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disc) as described above, and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0169] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for detecting a stationary or slow-moving traffic target, characterized in that: The method for detecting a stationary and slow-moving traffic target comprises the following steps: Establish the radar sensing area coordinate system and scan the monitoring area through radar and camera to determine the location and speed information of the target vehicle; Determine the lane where the target vehicle is located according to the x-coordinate value of the target vehicle in the radar sensing area coordinate system; According to the y coordinate value of the target vehicle in the radar sensing area coordinate system, determine whether the instantaneous speed of the target vehicle closest to the stop line is greater than or equal to a preset threshold; If so, the target vehicle closest to the stop line is defined as a stationary target and the following vehicles are counted along the y-axis direction.

2. The method for detecting a stationary and slow-moving traffic target according to claim 1, characterized in that: The steps of establishing a radar sensing area coordinate system and scanning the monitoring area with radar and camera to determine the position information and speed information of the target vehicle include: Obtain detection signals from radar and camera respectively; Convert radar detection signals and camera detection signals; The converted radar detection signal and camera detection signal are correlated based on the target vehicle trajectory.

3. The method for detecting a stationary and slow-moving traffic target according to claim 2, characterized in that: The step of converting the radar detection signal and the camera detection signal comprises: The radar detection signal and the camera detection signal are transformed into coordinates using the following coordinate transformation formula: Where (u r ,v r ,w r ,1) T is the coordinate of the target under the millimeter wave radar coordinate, and the corresponding space coordinate system is (u c ,v c ,w c ,1) T ; (x,y.1) T is the coordinate of the target in the image pixel coordinate system; R is a 3×3 unit orthogonal rotation matrix and T is a 3×1 translation matrix; I is the identity matrix; a x ,a y Respectively represent the size of each pixel in horizontal and vertical physical units; f is the focal length of the camera; k is the scaling factor.

4. The method for detecting a stationary and slow-moving traffic target according to claim 2, characterized in that: The step of converting the radar detection signal and the camera detection signal comprises: Mark the starting time points of the radar and camera; After aligning the starting time points, the data frames are matched according to the frame rates of the radar and camera data acquisition.

5. The method for detecting a stationary and slow-moving traffic target according to claim 2, characterized in that: After the step of associating the converted radar detection signal and the camera detection signal based on the target vehicle trajectory, the method further includes: The radar detection signal and the camera detection signal are converted into continuous trajectories by curve fitting. The trajectory curve obtained by the radar detection signal is defined as S'(t), and the trajectory curve obtained by the camera detection signal is defined as P(t). Calculate the similarity between the trajectory curve S'(t) and the trajectory curve P(t).

6. The method for detecting a stationary and slow-moving traffic target according to claim 5, characterized in that: The step of calculating the similarity between the trajectory curve S'(t) and the trajectory curve P(t) includes: The set similarity between the target trajectory fitting curves detected by the sensor is calculated by the following formula: Where cos(P(t),S'(t)) represents the covariance between the sampling coordinates of the curve; Represents the variance of the sampling coordinate points of the curve.

7. The method for detecting a stationary and slow-moving traffic target according to claim 5, characterized in that: The step of calculating the similarity between the trajectory curve S'(t) and the trajectory curve P(t) also includes: Set the horizontal coordinate component u of the camera trajectory curve P(t), sample n times, and find the maximum value U of the sampling point data max and the minimum value U min , for each sampling point coordinate, calculate its normalized value The radar projection trajectory curve S'(t) is normalized using the same method; Let the Euclidean distance between the sampling points of the curves detected by the radar and the camera be d; Traverse the sampling points to obtain the maximum distance F between the sampling points of the two curves under the sampling number n n : F n (p(t),s'(t))=max{d(p(t)),s'(t))},t∈[1,n] Set the maximum number of sampling points N and calculate the maximum distance F of the sampling points obtained under different sampling numbers n n , find the minimum value and get the distance closeness between the trajectory fitting curves: The similarity between the curves is: R = pR u +(1-p)R v ,p∈[0,1].

8. A traffic target stationary and slow-moving detection device, characterized in that: The traffic target stationary and slow-moving detection device includes: a memory, a processor, and a traffic target stationary and slow-moving detection program stored in the memory and executable on the processor. When the traffic target stationary and slow-moving detection program is executed by the processor, the steps of the traffic target stationary and slow-moving detection method as described in any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a stationary and slow-moving traffic target detection program, which, when executed by a processor, implements the steps of the stationary and slow-moving traffic target detection method according to any one of claims 1 to 6.

Images