A vehicle driving intention recognition method based on laser radar and visible light communication

Abstract

The present invention relates to a vehicle driving intention recognition method based on laser radar and visible light communication, comprising: scanning vehicles within a detection range by laser radar to generate point cloud data; the point cloud data includes ground point cloud data; the detection range is the current driving lane of the vehicle and the adjacent lane; according to the point cloud data, detecting vehicles within the detection range to obtain a target vehicle; tracking the target vehicle to generate target vehicle tracking data and target vehicle detection status; judging whether basic information of the target vehicle is obtained to generate a first judgment result; judging the driving intention according to the first judgment result to generate a second judgment result; generating an early warning according to the second judgment result; the present invention can judge the vehicle's lane changing intention in advance when the vehicle is about to change lanes and turns on the turn signal, and accurate information perception is conducive to scientific and accurate decision-making and safety control of intelligent vehicles, thereby improving on-the-way operation safety.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle assisted driving and automatic driving, and in particular to a vehicle driving intention recognition method based on laser radar and visible light communication. Background Art

[0002] With the development of science and technology, vehicle assisted driving and autonomous driving technologies are becoming more and more mature, among which information perception technology is an important part of assisted driving and autonomous driving. As a vehicle-mounted sensor, LiDAR has the characteristics of high precision and high resolution, and is widely used for vehicle perception; LiDAR target detection technology is to scan the surrounding environment through the LiDAR, calculate the distance and position information of the object based on the reflected laser beam, and use algorithms and models to detect and identify the target. It has a strong ability to perceive the environment around the vehicle and can accurately identify the position of vehicles and obstacles on the road; the more laser beams the LiDAR emits, the more point cloud data is obtained, the higher the perception accuracy, but the more expensive it is; because the expensive high-beam LiDAR is difficult to commercialize, the low-beam LiDAR has poor detection effect, and the refraction or scattering of the laser beam caused by rain and smoke will have a certain impact on the detection results, resulting in certain detection errors, and there are certain limitations in using a single LiDAR;

[0003] To this end, the reliability of environmental perception can be further improved by fusing data with LiDAR and other sensors, such as joint perception with cameras, millimeter-wave radars, and ultrasonic radars. Joint perception technology based on LiDAR combines LiDAR with multiple other sensors to jointly perceive the target vehicle in front to achieve better perception effects. Currently, commonly used joint perception technologies based on LiDAR include joint perception of LiDAR with sensors such as cameras, millimeter-wave radars, and ultrasonic radars. The data sensed by the sensors are fused to achieve multi-dimensional, more accurate, and more reliable perception of a target vehicle. Although joint perception can improve perception performance to a certain extent, the vehicle speed, position, distance, and other information sensed by these sensors need to be calculated, which requires more computing power, and there will be certain errors in the calculation process.

[0004] With the development of wireless transmission technology, visible light communication technology has gradually attracted people's attention. It uses visible light as a data transmission medium, adopts certain modulation technology to convert the electrical signal containing data into a corresponding optical signal, transmits the optical signal through the optical transmitter, and then receives the optical signal at the receiving end, decodes and processes the signal to obtain the original data; however, the use of this technology requires the lights to be in a flashing state, such as the brake lights flashing when the vehicle brakes, and the turn signals flashing when the vehicle changes lanes, and the light signals can be transmitted to the surrounding vehicles through the lights; if the vehicle does not turn on the turn signal when changing lanes, the vehicle information cannot be transmitted to the surrounding vehicles through this technology. Summary of the invention

[0005] To this end, the technical problem to be solved by the present invention is to overcome the problem in the prior art that the detection results are affected by the environment such as rain and fog, which consumes more computing power and produces certain errors in the calculation process; the visible light communication technology can only be used to perceive the vehicle when the LED lights of the vehicle are flashing, otherwise the problem of no perception can be achieved; to solve the above technical problems, the present invention provides a vehicle driving intention recognition method based on lidar and visible light communication, which uses lidar to perform active perception, so that the vehicle actively scans the road ahead and obtains the information of the vehicle ahead; and passively receives the information of the vehicle ahead through the visible light communication receiver mounted on the vehicle, and merges it with the lidar perception data to achieve a more accurate perception effect, so that the vehicle can make more accurate decisions.

[0006] In order to achieve the above-mentioned purpose, the main technical scheme adopted by the present invention includes: generating point cloud data by scanning the vehicles within the detection range through the laser radar; the point cloud data includes ground point cloud data; the detection range is the current lane of the vehicle and the adjacent lane; according to the point cloud data, the vehicles within the detection range are detected to obtain the target vehicle; the target vehicle is tracked to generate target vehicle tracking data and target vehicle detection status; it is judged whether the basic information of the target vehicle is obtained, and a first judgment result is generated; according to the first judgment result, the driving intention is judged to generate a second judgment result; according to the second judgment result, an early warning is generated.

[0007] In one embodiment of the present invention, point cloud data is generated by scanning vehicles within a detection range through a laser radar; the point cloud data includes ground point cloud data; the detection range is the current lane of the vehicle and adjacent lanes; based on the point cloud data, vehicles within the detection range are detected to obtain a target vehicle; the target vehicle is tracked to generate target vehicle tracking data and target vehicle detection status; it is determined whether basic information of the target vehicle is obtained, and a first judgment result is generated; based on the first judgment result, the driving intention is determined, and a second judgment result is generated; based on the second judgment result, an early warning is generated.

[0008] In one embodiment of the present invention, the specific steps of tracking the target vehicle and generating target vehicle tracking data and target vehicle detection status include: generating a target vehicle number according to the longitudinal distance and lateral distance between the target vehicle and the vehicle; generating the target vehicle tracking data and the target vehicle detection status through a Kalman filter algorithm according to the target vehicle number and the target vehicle detection frame; the target vehicle detection status is:

[0009]

[0010] Wherein, Cdi(t) represents the state of the target vehicle i detected at time t, represents the coordinate of the center of the detection frame of the target vehicle i in the X-axis direction in the geodetic coordinate system; represents the coordinate of the center of the detection frame of the target vehicle i in the Y-axis direction in the geodetic coordinate system; represents the speed of the target vehicle i; represents the acceleration of the target vehicle i; represents the yaw angle of the target vehicle i; represents the yaw angular velocity of the target vehicle i.

[0011] In one embodiment of the present invention, the specific steps of determining whether the basic information of the target vehicle is obtained and generating a first judgment result include: if so, generating initial target vehicle real-time information and initial target vehicle real-time status according to the target vehicle basic information; generating fusion data according to the initial target vehicle real-time information, the target vehicle number and the target vehicle tracking data; generating the target vehicle real-time status according to the initial target vehicle real-time status; wherein the first judgment result is the fusion data and the target vehicle real-time status; if not, the first judgment result is the target vehicle tracking data and the target vehicle detection status.

[0012] In one embodiment of the present invention, the fusion data is generated according to the initial target vehicle real-time information, the target vehicle number and the target vehicle tracking data; the specific steps of generating the target vehicle real-time status according to the initial target vehicle real-time status include: matching the data at the same time according to the initial target vehicle real-time information and the target vehicle tracking data to generate the first fusion data; matching the data of the same target vehicle through the error range of the target vehicle according to the initial target vehicle real-time information, the target vehicle number and the target vehicle tracking data to generate the second fusion data; generating the fusion data according to the first fusion data and the second fusion data; generating the target vehicle real-time status according to the initial target vehicle real-time status.

[0013] In one embodiment of the present invention, the real-time state of the initial target vehicle is:

[0014]

[0015] Wherein, Cgj(t) represents the state of the target vehicle j at time t, represents the coordinate of the center of the target vehicle j in the X-axis direction in the geodetic coordinate system; represents the coordinate of the center of the detection frame of the target vehicle j in the Y-axis direction in the geodetic coordinate system; represents the speed of the target vehicle j; represents the acceleration of the target vehicle j; represents the heading angle of the target vehicle j; Indicates the flashing light type of the target vehicle j;

[0016] The real-time status of the target vehicle is:

[0017]

[0018] Wherein, Cgi(t) represents the state of the target vehicle i at time t, represents the coordinate of the center of the target vehicle i in the X-axis direction in the geodetic coordinate system; represents the coordinate of the center of the detection frame of the target vehicle i in the Y-axis direction in the geodetic coordinate system; represents the speed of the target vehicle i; represents the acceleration of the target vehicle i; represents the heading angle of the target vehicle i; Indicates the flashing light type of the target vehicle i.

[0019] In one embodiment of the present invention, the error range is:

[0020]

[0021] in, represents the coordinate of the center of the detection frame of the target vehicle i at time t in the X-axis direction in the geodetic coordinate system, represents the coordinate of the center of the detection frame of the target vehicle i at time t in the Y-axis direction in the geodetic coordinate system, represents the coordinate of the center of the target vehicle j in the X-axis direction in the geodetic coordinate system at time t, represents the coordinate of the center of the detection frame of the target vehicle j at time t in the Y-axis direction in the geodetic coordinate system, and D1 represents the preset coordinate deviation threshold.

[0022] In one embodiment of the present invention, if the first judgment result is the target vehicle tracking data and the target vehicle detection status, a second judgment result is generated based on the target vehicle detection status; if the first judgment result is the fusion data and the target vehicle real-time status, a second judgment result is generated based on the target vehicle real-time status.

[0023] In one embodiment of the present invention, the second judgment result includes:

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030] Among them, t is the time, Δt is the time interval; represents the yaw angle of the vehicle at time t; It represents the coordinate of the center of the vehicle in the X-axis direction in the geodetic coordinate system at time t; represents the coordinate of the center of the vehicle in the Y-axis direction in the geodetic coordinate system at time t; D2 is the preset yaw rate threshold; D3 is the preset lateral spacing threshold; Indicates that the second judgment result is cruising; Indicates that the second judgment result is acceleration; Indicates that the second judgment result is deceleration; Indicates that the second judgment result is left lane change; Indicates that the second judgment result is right lane change;

[0031] It indicates that the second judgment result is that the target vehicle changes lanes and invades the driving path of the own vehicle.

[0032] In one embodiment of the present invention, the second judgment result further includes:

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039] in, Indicates that the second judgment result is cruising; Indicates that the second judgment result is acceleration; Indicates that the second judgment result is deceleration; Indicates that the second judgment result is left lane change; Indicates that the second judgment result is right lane change;

[0040] It indicates that the second judgment result is that the target vehicle changes lanes and invades the driving path of the own vehicle.

[0041] The above technical solution of the present invention has the following advantages compared with the prior art:

[0042] The present invention discloses a method for identifying vehicle driving intention based on laser radar and visible light communication. Visible light communication technology is used on the basis of laser radar perception. After the front vehicle sends vehicle information through the headlights, the visible light communication receiving end on the vehicle receives the information and merges it with the laser radar perception data. At this time, the information obtained is the data of the front vehicle sensor, which is more accurate than the information obtained by calculation. More accurate information on the running state of the front vehicle can be obtained, and a more accurate perception effect can be achieved. The present invention can obtain detailed information of the vehicle running in front, which is conducive to the vehicle making more accurate and timely decisions. Different from the need to make an accurate judgment on the vehicle driving intention only after the vehicle state changes, the present invention can judge the vehicle's lane change intention in advance when the vehicle is about to change lanes and turn on the turn signal. Accurate information perception is conducive to scientific and accurate decision-making and safety control of intelligent vehicles, and improves on-the-road operation safety; thereby improving vehicle driving safety. At present, LED headlights have been widely used. The visible light communication technology used in the present invention only needs to be modified on the basis of the original LED headlights to achieve. With the widespread use of vehicle-mounted laser radars and the reduction of production costs, the present invention will be easier to be widely promoted in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein

[0044] Figure 1 It is a schematic diagram of the overall process of a vehicle driving intention recognition method based on laser radar and visible light communication according to a preferred embodiment of the present invention;

[0045] Figure 2 yes Figure 1 A schematic diagram of the process of obtaining a target vehicle in the vehicle driving intention recognition method based on laser radar and visible light communication;

[0046] Figure 3 yes Figure 1 A schematic diagram of a process for generating target vehicle tracking data and target vehicle detection status of the vehicle driving intention recognition method based on laser radar and visible light communication;

[0047] Figure 4 yes Figure 1 A schematic diagram of a process of generating a first judgment result of the vehicle driving intention recognition method based on laser radar and visible light communication;

[0048] Figure 5 yes Figure 1A schematic diagram of a process of generating a second judgment result of the vehicle driving intention recognition method based on laser radar and visible light communication;

[0049] Figure 6 yes Figure 1 An interactive logic diagram of the vehicle driving intention recognition method based on laser radar and visible light communication;

[0050] Figure 7 yes Figure 1 Schematic diagram of the effect of the vehicle driving intention recognition method based on laser radar and visible light communication. DETAILED DESCRIPTION

[0051] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below in conjunction with the accompanying drawings through specific embodiments. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a clearer and more thorough understanding of the present invention and to be able to fully convey the scope of the present invention to those skilled in the art.

[0052] Reference Figures 1 to 7 As shown, a preferred embodiment of the present invention provides a vehicle driving intention recognition method based on laser radar and visible light communication, comprising:

[0053] S1, scan the vehicles within the detection range through the laser radar to generate point cloud data; the point cloud data includes ground point cloud data; the detection range is the current lane of the vehicle and the adjacent lanes;

[0054] Specifically, the on-board laser radar actively senses the road environment in front of the vehicle and generates point cloud data. The laser radar detection range is the current lane of the vehicle and the adjacent lanes;

[0055] Laser radar scans the surrounding environment by emitting laser beams, and calculates the distance and position information of objects based on the reflected laser beams. It has a strong ability to perceive the environment around the vehicle and can accurately identify the positions of vehicles and obstacles on the road. It can also be combined with sensors such as cameras and ultrasonic radars to achieve better perception effects. Therefore, laser radar has been widely used in fields such as assisted driving and automatic driving. However, it is affected by the environment, such as the scattering and reflection of rain, which will cause certain detection errors in laser radar. To achieve assisted driving or automatic driving in the scenario of high-speed vehicle driving, the on-board sensors need to have high sensitivity and accuracy. The combined perception method of laser radar and other sensors can improve certain perception accuracy. The present invention uses laser radar and visible light communication technology to jointly perceive the driving intention of the vehicle ahead, thereby providing technical support for assisted driving.

[0056] In this embodiment, the vehicle is marked as C0 and is traveling in the middle lane of a three-lane road. There are three target vehicles in front of the vehicle, of which C1 is target vehicle 1, which is in the left lane change state with the left turn signal flashing; C2 is target vehicle 2, which is in the deceleration state with the brake lights flashing; C3 is target vehicle 3, which is in the cruise state;

[0057] In this embodiment, the detection range is C0 current driving lane and the left lane and the right lane.

[0058] S2, detect the vehicles within the detection range according to the point cloud data and obtain the target vehicle;

[0059] Specifically, after acquiring the point cloud data, the point cloud data needs to be processed to detect the target vehicle, including:

[0060] S21, segmenting the ground point cloud data from the point cloud data to generate filtered point cloud data;

[0061] Specifically, in order to accurately detect the vehicle ahead and reduce the amount of calculation, the ground point cloud data needs to be segmented; the ground point cloud segmentation methods that can be used include the height threshold method, the road surface plane fitting method, and the method based on field information.

[0062] In this embodiment, a height threshold method is adopted, that is, a threshold of the point cloud height is set, and point cloud data below the threshold is filtered out.

[0063] S22, clustering the filtered point cloud data of the same vehicle according to the filtered point cloud data to generate clustered point cloud data;

[0064] Specifically, after filtering out the ground point cloud data, the point cloud data of the same vehicle needs to be clustered for subsequent detection; the methods that can be used to cluster the target vehicle point cloud data include K-means algorithm, DBSCAN algorithm, STING algorithm, Euclidean algorithm, etc.

[0065] In this embodiment, the DBSCAN algorithm is used to cluster the point cloud data. The algorithm is a density-based clustering algorithm with the characteristics of strong robustness and automatic determination of the number of clusters.

[0066] S23, constructing a target vehicle detection frame according to the clustered point cloud data to obtain the target vehicle;

[0067] Specifically, after clustering the point cloud data of the target vehicle, a detection frame of the vehicle is constructed according to the boundary points of the clustered point cloud;

[0068] In this embodiment, the detection frames of the target vehicles C1, C2, and C3 are obtained.

[0069] S3, tracking the target vehicle, generating target vehicle tracking data and target vehicle detection status;

[0070] Specifically, the target vehicle is tracked, and target vehicle tracking data and target vehicle detection status are generated, including:

[0071] S31, generating a target vehicle number according to the longitudinal distance and the lateral distance between the target vehicle and the own vehicle;

[0072] Specifically, when the target vehicle is in the lane of the own vehicle or the adjacent lane, the vehicle is sorted and numbered; first, the target vehicle is sorted from near to far according to the longitudinal distance between the target vehicle and the own vehicle; if the longitudinal distances between the two target vehicles in front and the own vehicle are the same, the target vehicle is sorted from near to far according to the lateral distance between the target vehicle and the own vehicle;

[0073] In this embodiment, the numbers C1, C2, and C3 of the three vehicles are determined based on the longitudinal distance between the target vehicle and the own vehicle.

[0074] S32, generating target vehicle tracking data and target vehicle detection status through a Kalman filter algorithm according to the target vehicle number and the target vehicle detection frame;

[0075] Specifically, the Kalman filter algorithm is used to calculate the speed and acceleration of the target vehicle, and the target vehicle trajectory is tracked according to the center point position of the target vehicle detection frame to obtain the target vehicle tracking data, and the angle between the direction of the forward and backward movement of the center point of the detection frame and the X-axis of the earth coordinate system is used as the yaw angle; when calculating the speed and acceleration of the target vehicle, the distance measurement values ​​from the center point of the tail of the continuous target frame to the laser radar are used as input, and the Kalman filter algorithm is used to iteratively update to obtain accurate target vehicle speed estimation values ​​and acceleration estimation values;

[0076] At this time, the target vehicle detection state is as follows:

[0077]

[0078] Among them, Cdi(t) represents the state of the detected target vehicle i at time t, Represents the coordinate of the center of the detection frame of the target vehicle i in the X-axis direction in the geodetic coordinate system; Represents the coordinate of the center of the detection frame of the target vehicle i in the Y-axis direction in the geodetic coordinate system; represents the speed of target vehicle i; represents the acceleration of target vehicle i; represents the yaw angle of target vehicle i; represents the yaw rate of the target vehicle i;

[0079] In this embodiment, the tracking data and detection status of the target vehicles C1, C2, C3, namely Cd1(t), Cd2(t) and Cd3(t), are obtained.

[0080] Before the ego vehicle obtains the basic information of the target vehicle, the present invention divides the driving state of the target vehicle into a low-risk driving state and a high-risk driving state according to the degree of influence of the preceding vehicle on the ego vehicle. The low-risk driving state includes cruising and acceleration, which have little influence on the ego vehicle; the high-risk state includes deceleration, left lane change and right lane change. When the preceding target vehicle decelerates or the target vehicle in the adjacent lane changes lanes and invades the current lane of the ego vehicle, it will have a greater impact on the driving of the ego vehicle; when the driver of the target vehicle decelerates by pressing the brake pedal, the vehicle generates a brake light signal; when the driver of the target vehicle changes lanes and turns on the turn signal switch, the vehicle generates a turn signal Turn signal: if the driver changes lanes but does not turn on the turn signal switch, the vehicle has no turn signal; when the vehicle generates a brake light signal or a turn signal, the ECU (Electronic Control Unit) transmits the acquired vehicle information to the LED driver through the CAN (Controller Area Network) bus. The vehicle information includes the vehicle's latitude and longitude coordinates, speed, acceleration and heading angle. This information is converted into corresponding electrical signals through encoding, modulation, and digital-to-analog conversion, and drives the corresponding LED lights to flash at a frequency. At this time, the flashing LED lights are the transmitters of visible light communication;

[0081] Visible light communication technology uses visible light as a data transmission medium, adopts certain modulation technology to convert the electrical signal containing data into a corresponding optical signal, transmits the optical signal through an optical transmitter, and then receives the optical signal at the receiving end, decodes and processes the signal to obtain the original data. This technology has the characteristics of high transmission rate, strong security, energy saving and environmental protection, and has broad application prospects; currently LED car lights have been widely used in mass-produced vehicles, and LED lights are good transmitters in visible light communication technology; visible light communication technology uses LED to realize data transmission, the transmitting end converts the electrical signal into an optical signal, and encodes the data into the optical signal through modulation technology; the receiving end uses a photodiode or a light sensor to convert the optical signal into an electrical signal, and then decodes and restores the original data, which has the characteristics of high transmission rate, strong security, energy saving and environmental protection; the present invention uses visible light communication technology to realize the transmission of detailed information of the target vehicle, and then realizes data fusion, so the use of visible light communication technology on the vehicle for vehicle perception is low-cost and easier to promote, and detailed vehicle information can be obtained without calculation;

[0082] In this embodiment, the target vehicle C1 changes lanes to the left, and the left turn signal flashes; the target vehicle C2 slows down, and the brake light flashes; the target vehicle C3 cruises; when the vehicle generates a brake light signal or a turn signal signal, the ECU transmits the acquired vehicle information to the LED driver through the CAN bus. The vehicle information includes the vehicle's latitude and longitude coordinates, speed, acceleration and heading angle. This information is encoded, modulated, and converted into corresponding electrical signals and drives the corresponding LED lights to flash at a frequency. At this time, the flashing LED lights are the visible light communication transmitter; the target vehicle C1 transmits a visible light signal to the self-vehicle C0 through the left turn signal, the target vehicle C2 transmits a visible light signal to the self-vehicle C0 through the brake light, and the target vehicle C3 does not send a visible light signal; if the vehicle does not generate a turn signal signal when changing lanes, the vehicle information cannot be sent through the visible light communication transmitter.

[0083] S4, determining whether basic information of the target vehicle is obtained, and generating a first determination result; specifically comprising:

[0084] S41: If yes, please include:

[0085] S411, generating initial target vehicle real-time information and initial target vehicle real-time status according to the target vehicle basic information;

[0086] Specifically, after the brake lights or turn signals of the target vehicle ahead flash, the visible light communication transmitter of the target vehicle sends the basic information of the target vehicle. The PD (Photo Diode) receiver installed on the vehicle captures the optical signal and obtains the initial target vehicle basic information sent by the target vehicle ahead through analog-to-digital conversion, equalization, and demodulation. At this time, the real-time status of the initial target vehicle is shown in the following formula:

[0087]

[0088] Among them, Cgj(t) represents the state of target vehicle j at time t, Represents the coordinate of the center of the target vehicle j in the X-axis direction in the geodetic coordinate system; Represents the coordinate of the center of the detection frame of the target vehicle j in the Y-axis direction in the geodetic coordinate system; represents the speed of target vehicle j; represents the acceleration of the target vehicle j; represents the heading angle of the target vehicle j; Indicates the flashing light type of target vehicle j;

[0089] In this embodiment, the states of the target vehicles C1 and C2 are obtained, that is, the real-time states of the target vehicles Cg1(t) and Cg2(t).

[0090] S412, matching the data at the same time according to the initial target vehicle real-time information and the target vehicle tracking data to generate first fused data;

[0091] Specifically, according to the time when the laser radar detects the target vehicle and the time when the visible light communication receiving end of the own vehicle receives the information of the preceding vehicle, the data timing is matched to ensure that the data at the same time is fused.

[0092] S413, according to the initial target vehicle real-time information, the target vehicle number and the target vehicle tracking data, matching the data of the same target vehicle through the error range of the target vehicle to generate second fusion data;

[0093] Specifically, the position information of a vehicle with a certain number in the target vehicle tracking data is compared with the vehicle coordinate information and longitude and latitude information of the visible light communication receiving end, that is, the two sets of position data with number i and number j are compared. If the comparison error of the two sets of position data is within a certain error range, it is judged that the two sets of data are the data of the same target vehicle. The error range is shown in the following formula:

[0094]

[0095] in, represents the coordinate of the center of the detection frame of the target vehicle i at time t in the X-axis direction in the geodetic coordinate system, represents the coordinate of the center of the detection frame of the target vehicle i at time t in the Y-axis direction in the geodetic coordinate system, represents the coordinate of the center of the target vehicle j at time t in the X-axis direction in the geodetic coordinate system, represents the coordinate of the center of the detection frame of the target vehicle j at time t in the Y-axis direction in the geodetic coordinate system, and D1 represents the preset coordinate deviation threshold;

[0096] In this embodiment, the laser radar scan obtains three sets of vehicle data and target vehicle detection status, namely Cd1(t), Cd2(t) and Cd3(t); two sets of initial target vehicle real-time information and initial target vehicle real-time status are obtained by receiving the target vehicle basic information, namely Cg1(t) and Cg2(t). According to the error range To determine whether the two sets of data match, D1 can be set to half the width of the vehicle; finally, it is assumed that Cd1(t) and Cg1(t) are paired, and Cd2(t) and Cg2(t) are paired.

[0097] S414, generating fused data according to the first fused data and the second fused data; generating a real-time state of the target vehicle according to the initial real-time state of the target vehicle.

[0098] Specifically, since the vehicle position, speed, acceleration, heading angle and other information perceived by the lidar are all inferred based on the point cloud data, and there will be perception errors due to the influence of the environment, the target vehicle tracking data obtained by the lidar scanning the vehicle in front has certain errors. After obtaining more accurate target vehicle information through visible light communication technology, the original target vehicle tracking data needs to be updated to the information obtained by the visible light communication receiving end, that is, the fused data; the updated target vehicle data, that is, the fused data, includes the type of headlights, vehicle position, vehicle speed, vehicle acceleration and vehicle heading angle; at this time, the real-time status of the target vehicle is shown in the following formula:

[0099]

[0100] Among them, Cgi(t) represents the state of target vehicle i at time t, Represents the coordinate of the center of target vehicle i in the X-axis direction in the geodetic coordinate system; Represents the coordinate of the center of the detection frame of the target vehicle i in the Y-axis direction in the geodetic coordinate system; represents the speed of target vehicle i; represents the acceleration of target vehicle i; represents the heading angle of target vehicle i; Indicates the flashing light type of target vehicle i;

[0101] At this time, the first judgment result is the fusion data and the real-time status of the target vehicle;

[0102] In this embodiment, the updated fusion data and the real-time status of the target vehicle are represented as Cg1(t), Cg2(t) and Cd3(t).

[0103] S42, if not, the first judgment result is the target vehicle tracking data and the target vehicle detection status.

[0104] S5, judging the driving intention according to the first judgment result, and generating a second judgment result;

[0105] Specifically, if the input data is target vehicle tracking data perceived by the laser radar, the driving intention is judged according to the detection status of the target vehicle; if the input data is updated fusion data of the target vehicle, the driving intention is judged according to the real-time status of the target vehicle; the judgment results include cruising, acceleration, deceleration and lane changing, where lane changing is divided into left lane changing and right lane changing; including:

[0106] S51, if the first judgment result is the target vehicle tracking data and the target vehicle detection state, then generate a second judgment result according to the target vehicle detection state;

[0107] Specifically, when the input data is Cdi(t), if The second judgment result is cruising, where Δt is the time interval; if The second judgment result is acceleration; if The second judgment result is deceleration; if The second judgment result is left lane change, where D2 is the yaw rate threshold; if The second judgment result is right lane change. When the target vehicle changes lanes and invades the driving path of the own vehicle, the second judgment result is to issue a warning, as shown in the following formula:

[0108]

[0109] In the formula represents the yaw angle of the vehicle at time t; It represents the coordinate of the center of the vehicle in the X-axis direction in the geodetic coordinate system at time t; It represents the coordinate of the center of the vehicle in the Y-axis direction in the geodetic coordinate system at time t; D3 is the lateral distance threshold between the target vehicle and the vehicle;

[0110] S52, if the first judgment result is the fusion data and the real-time state of the target vehicle, generating a second judgment result according to the real-time state of the target vehicle;

[0111] Specifically, when the input data is Cgi(t), if The second judgment result is cruising; if The second judgment result is acceleration; if The second judgment result is deceleration; if and is a left turn signal, the second judgment result is left lane change, where D4 is the heading angle threshold; if and is a right turn signal, the second judgment result is a right lane change; when the target vehicle changes lanes and invades the driving path of the own vehicle, the second judgment result is to issue a warning, as shown in the following formula:

[0112]

[0113] In this embodiment, the input data of target vehicles C1 and C2 are updated target vehicle data, and the input data of target vehicle C3 is the original tracking data. and is the left turn signal, the target vehicle C1 is judged to be in the left lane change state, where D4 can be set to 30°. , the target vehicle C2 is judged to be in a decelerating state. It is determined that the target vehicle C3 is in a cruising state.

[0114] S6. Generate a warning based on the second judgment result.

[0115] Specifically, after judging the driving intention of the target vehicle ahead, different degrees of warning can be made according to the risk level of the target vehicle's driving status; when the target vehicle slows down or changes lanes and affects the normal driving of the vehicle, the driver will be reminded; when the target vehicle is cruising or accelerating, no warning will be issued;

[0116] In this embodiment, when the target vehicle C1 satisfies the following conditions, an early warning is issued.

[0117]

[0118] D3 can be set to 3.5 meters.

[0119] The target vehicle C2 is in a deceleration state and issues a warning; the target vehicle C3 is in a cruising state and does not issue a warning.

[0120] The present invention actively senses the vehicle ahead through the laser radar installed on the vehicle. After forming the point cloud data, the ground point cloud is segmented, the target vehicle point cloud data is clustered, and the target vehicle detection frame is constructed to achieve the perception of the vehicle ahead. Then the target vehicles are numbered, and the numbered vehicles are tracked using the Kalman filter algorithm to obtain the target vehicle tracking data; according to the degree of influence of the front vehicle on the vehicle, the target vehicle state is divided into a low-risk driving state and a high-risk driving state, wherein the low-risk driving state includes cruising and acceleration, and the high-risk state includes deceleration, left lane change, and right lane change. When the target vehicle decelerates, the vehicle generates a brake light signal; when the target vehicle changes lanes to the left, the driver turns the left turn signal switch, and the vehicle generates a left turn signal. If the driver does not turn the left turn signal switch, the vehicle does not generate a left turn signal; when the target vehicle changes lanes to the right, the driver turns the right turn signal switch, and the vehicle generates a right turn signal. If the driver does not turn the right turn signal switch, the vehicle does not generate a right turn signal; when the visible light communication receiving end of the self-vehicle receives the detailed information sent by the target vehicle, it is fused with the target vehicle tracking data, and the updated target vehicle data is obtained after data timing matching, data and vehicle number pairing and data update, including vehicle position, vehicle speed, vehicle acceleration and vehicle heading angle. Then the updated data is used to judge the driving intention, and the judgment results include cruising, acceleration, deceleration and lane change. Finally, the driving intention is input into the collision warning module to prevent the self-vehicle from colliding with the target vehicle. If the visible light communication receiving end of the self-vehicle does not receive the information of the target vehicle, only the target vehicle tracking data is used to judge the driving intention.

[0121] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include these modifications and variations.

[0123] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may alter, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A vehicle driving intention recognition method based on laser radar and visible light communication, characterized in that: include: Scan the vehicles within the detection range through LiDAR to generate point cloud data; The point cloud data includes ground point cloud data; the detection range is the current lane of the vehicle and the adjacent lanes; Detecting vehicles within the detection range according to the point cloud data to obtain a target vehicle; Tracking the target vehicle to generate target vehicle tracking data and target vehicle detection status; Determine whether basic information of the target vehicle is obtained, and generate a first determination result; According to the first judgment result, judging the driving intention, and generating a second judgment result; generating an early warning according to the second judgment result; The specific steps of determining whether the basic information of the target vehicle is obtained and generating the first determination result include: If yes, then generate initial target vehicle real-time information and initial target vehicle real-time status according to the target vehicle basic information; generate fusion data according to the initial target vehicle real-time information, target vehicle number and target vehicle tracking data; generate target vehicle real-time status according to the initial target vehicle real-time status; wherein the first judgment result is the fusion data and the target vehicle real-time status; If not, the first judgment result is the target vehicle tracking data and the target vehicle detection status.

2. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 1 is characterized in that: According to the point cloud data, the specific steps of detecting the vehicle within the detection range and obtaining the target vehicle include: Segmenting the ground point cloud data from the point cloud data to generate filtered point cloud data; Clustering the filtered point cloud data of the same vehicle according to the filtered point cloud data to generate clustered point cloud data; A target vehicle detection frame is constructed according to the clustered point cloud data to obtain the target vehicle.

3. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 2 is characterized in that: The specific steps of tracking the target vehicle and generating target vehicle tracking data and target vehicle detection status include: Generate a target vehicle number according to the longitudinal distance and the lateral distance between the target vehicle and the own vehicle; Generate the target vehicle tracking data and the target vehicle detection status through a Kalman filter algorithm according to the target vehicle number and the target vehicle detection frame; The target vehicle detection state is: in, Indicates that the target vehicle is detected exist The state of the moment, Indicates the target vehicle The coordinate of the center of the detection frame in the X-axis direction in the geodetic coordinate system; Indicates the target vehicle The coordinate of the center of the detection frame in the Y-axis direction in the geodetic coordinate system; Indicates the target vehicle Speed ​​of vehicle; Indicates the target vehicle acceleration; Indicates the target vehicle The yaw angle; Indicates the target vehicle The yaw angular velocity.

4. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 1 is characterized in that: generating fusion data according to the initial target vehicle real-time information, the target vehicle number and the target vehicle tracking data; According to the initial target vehicle real-time state, the specific steps of generating the target vehicle real-time state include: According to the real-time information of the initial target vehicle and the tracking data of the target vehicle, matching the data at the same time to generate first fusion data; According to the real-time information of the initial target vehicle, the target vehicle number and the tracking data of the target vehicle, the data of the same target vehicle is matched through the error range of the target vehicle to generate second fusion data; Generate fusion data according to the first fusion data and the second fusion data; and generate a real-time state of the target vehicle according to the initial real-time state of the target vehicle.

5. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 4 is characterized in that: The real-time status of the initial target vehicle is: in, Indicates the target vehicle exist The state of the moment, Indicates the target vehicle The coordinate of the center in the X-axis direction in the geodetic coordinate system; Indicates the target vehicle The coordinate of the center of the detection frame in the Y-axis direction in the geodetic coordinate system; Indicates the target vehicle Speed ​​of vehicle; Indicates the target vehicle acceleration; Indicates the target vehicle The heading angle; Indicates the target vehicle Type of flashing lights; The real-time status of the target vehicle is: in, Indicates the target vehicle exist The state of the moment, Indicates the target vehicle The coordinate of the center in the X-axis direction in the geodetic coordinate system; Indicates the target vehicle The coordinate of the center of the detection frame in the Y-axis direction in the geodetic coordinate system; Indicates the target vehicle Speed ​​of vehicle; Indicates the target vehicle acceleration; Indicates the target vehicle The heading angle; Indicates the target vehicle Type of flashing lights.

6. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 5 is characterized in that: The error range is: in, express The target vehicle at time The coordinate of the center of the detection frame in the X-axis direction in the geodetic coordinate system, express The target vehicle at time The coordinate of the center of the detection frame in the Y-axis direction in the geodetic coordinate system, express The target vehicle at time The coordinate of the center in the X-axis direction in the geodetic coordinate system, express The target vehicle at time The coordinate of the center of the detection frame in the Y-axis direction in the geodetic coordinate system, Indicates the preset coordinate deviation threshold.

7. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 6 is characterized in that: If the first judgment result is the target vehicle tracking data and the target vehicle detection state, generating a second judgment result according to the target vehicle detection state; If the first judgment result is the fused data and the real-time state of the target vehicle, a second judgment result is generated according to the real-time state of the target vehicle.

8. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 7 is characterized in that: The second judgment result includes: Among them, t is the time, is the time interval; Indicates that the car is The yaw angle at the moment; express The coordinate of the center of the vehicle in the X-axis direction in the geodetic coordinate system at the moment; express The coordinate of the center of the vehicle in the Y-axis direction in the geodetic coordinate system at the moment; is the preset yaw angular velocity threshold; is the preset horizontal spacing threshold; Indicates that the second judgment result is cruising; Indicates that the second judgment result is acceleration; Indicates that the second judgment result is deceleration; Indicates that the second judgment result is left lane change; Indicates that the second judgment result is right lane change; It indicates that the second judgment result is that the target vehicle changes lanes and invades the driving path of the own vehicle.

9. The vehicle driving intention recognition method based on laser radar and visible light communication according to claim 8 is characterized in that: The second judgment result also includes: in, Indicates that the second judgment result is cruising; Indicates that the second judgment result is acceleration; Indicates that the second judgment result is deceleration; Indicates that the second judgment result is left lane change; Indicates that the second judgment result is right lane change; It indicates that the second judgment result is that the target vehicle changes lanes and invades the driving path of the own vehicle.

Images