Early warning method and device for dangerous driving behavior and computer readable storage medium

Abstract

The invention discloses an early warning method for dangerous driving behaviors, and the method comprises the steps: obtaining point cloud data collected by a laser radar, dividing the point cloud data into ground point cloud data and non-ground point cloud data, recognizing a ground target object according to the ground point cloud data, and recognizing a non-ground target object according to the non-ground point cloud data. According to the movement change of the vehicle relative to the ground target object and / or the non-ground target object, whether a driver of the vehicle has a dangerous driving behavior or not is judged, and when the dangerous driving behavior occurs, real-time early warning is carried out. The invention further provides a device for implementing the method and a computer readable storage medium. Whether a driver has a dangerous driving behavior or not can be judged through target object detection of the laser radar so as to carry out safety early warning.

Description

Technical Field

[0001] This invention relates to the field of driving assistance technology, and in particular to a method, device, and computer-readable storage medium for warning of dangerous driving behavior. Background Technology

[0002] As living standards improve, more and more people are buying cars, and autonomous vehicles are attracting increasing attention, mainly because these cars are equipped with a variety of sensors to assist driving and improve driving safety to a certain extent.

[0003] However, traditional cars lack the sensors needed for driver assistance, relying entirely on the driver's subjective judgment. In situations like traffic congestion, accidents, or heavy traffic, drivers are easily influenced by these conditions and may engage in poor driving behavior. Poor driving behavior can easily lead to subsequent traffic accidents, but drivers often fail to realize the potentially serious consequences. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method, device and computer-readable storage medium for warning of dangerous driving behavior. By detecting and tracking the relative motion of surrounding targets, the method obtains the possible changes in driving behavior and surrounding targets, and provides warnings for possible dangerous driving behavior, thereby improving driving safety and cultivating good driving habits of drivers.

[0005] An embodiment of the present invention provides a method for warning of dangerous driving behavior. The method includes: acquiring point cloud data collected by lidar; distinguishing the point cloud data into ground point cloud data and non-ground point cloud data; identifying ground targets and non-ground targets based on the ground point cloud data and non-ground point cloud data respectively; analyzing the driving behavior of the driver of the vehicle based on the vehicle's movement relative to the ground targets and the changes in the movement of the ground targets; determining whether the driving behavior is dangerous driving behavior; and issuing a safety warning when the driving behavior is determined to be dangerous driving behavior.

[0006] An embodiment of the present invention also provides a warning device for dangerous driving behavior, including a processor; and a memory for storing a computer program, which, when executed by the processor, causes the processor to implement the warning method for dangerous driving behavior.

[0007] An embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned method for warning of dangerous driving behavior.

[0008] Compared with the prior art, the present invention provides a method, device and computer-readable storage medium for warning of dangerous driving behavior, which can detect ground targets and non-ground targets through point cloud data collected by lidar, and determine whether the driver of the vehicle is engaging in dangerous driving behavior based on the motion changes of the vehicle relative to the ground targets and non-ground targets, so as to provide real-time warning. Attached Figure Description

[0009] Figure 1 This is a flowchart of a method for warning of dangerous driving behavior according to an embodiment of the present invention.

[0010] Figure 2 This is a flowchart of a method for predicting road rage in drivers according to an embodiment of the present invention.

[0011] Figure 3 This is a block diagram of a warning device for dangerous driving behavior according to an embodiment of the present invention.

[0012] Explanation of main component symbols

[0013]

[0014] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0015] To facilitate understanding and implementation of this invention by those skilled in the art, the invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that this invention provides many applicable inventive concepts, which can be implemented in various specific forms. Those skilled in the art can utilize the details described in these or other embodiments, as well as other available structural, logical, and electrical variations, to implement the invention without departing from its spirit and scope.

[0016] This specification provides different embodiments to illustrate the technical features of different implementations of the invention. The configuration of elements in the embodiments is for illustrative purposes only and is not intended to limit the invention. Furthermore, the repetition of some reference numerals in the embodiments is for simplification and does not imply any correlation between different embodiments. The same element numbers used in the illustrations and specification represent the same or similar components. The illustrations in this specification are simplified and not drawn to scale.

[0017] Furthermore, in describing some embodiments of the present invention, the specification describes the method and / or procedure of the present invention in a specific order of steps. However, since the method and procedure are not necessarily performed according to the specific order of steps described, they are not limited to the specific order of steps. Those skilled in the art will understand that other orders are also possible implementations. Therefore, the specific order of steps described in the specification is not intended to limit the scope of the patent application. Moreover, the scope of the present invention for the method and / or procedure is not limited to the order of execution steps written therein, and those skilled in the art will understand that adjusting the order of execution steps does not depart from the spirit and scope of the present invention.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Some embodiments of the invention are described in detail below with reference to the accompanying drawings.

[0019] Please see Figure 1 The diagram shows a flowchart of a method for warning of dangerous driving behavior according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0020] Step S101: Obtain point cloud data collected by lidar.

[0021] LiDAR can be positioned in front of, behind, or other locations on a vehicle to scan its surroundings using lasers and generate point cloud data. This point cloud data can then be used to identify other vehicles, obstacles, pedestrians, trees, and traffic signs in the vehicle's environment.

[0022] Step S102: Divide the point cloud data into ground point cloud data and non-ground point cloud data.

[0023] In this embodiment, in order to improve the computational efficiency and save computation time in the subsequent target detection process, the point cloud data is first classified.

[0024] In one example, point cloud data can be transformed from a point cloud coordinate system to a vehicle coordinate system. Based on this transformation, the point cloud data can be further categorized into ground point cloud data and non-ground point cloud data. Specifically, point cloud height can be used as the distinguishing criterion. Point cloud data with a height less than a preset height threshold is classified as ground point cloud data, while point cloud data with a height greater than or equal to the preset height threshold is classified as non-ground point cloud data. For example, the preset height threshold could be 10 centimeters.

[0025] Step S103: Identify ground targets and non-ground targets based on ground point cloud data and non-ground point cloud data, respectively.

[0026] In one example, a ground object detection model is used to process ground point cloud data to obtain the identification results of ground objects. These ground objects include lane lines, lane speed markings, and lane directions. The ground object detection model is trained using ground point cloud sample data.

[0027] In one example, a non-ground object detection model is used to process non-ground point cloud data to obtain the identification results of non-ground objects. These non-ground objects include cars, motor vehicles, pedestrians, and lampposts. The non-ground object detection model is trained using non-ground point cloud sample data.

[0028] Regarding the detection of non-ground targets, since the identification of non-ground targets generally takes a long time, in one embodiment, a bird's-eye view can be used to monitor the distance of non-ground targets in order to improve the timeliness requirements for the identification and tracking of non-ground targets.

[0029] Taking a LiDAR system installed at the front of a vehicle as an example, the raw point cloud data acquired by this LiDAR is not conducive to detecting non-ground targets. Therefore, firstly, the non-ground point cloud data is converted into a bird's-eye view. Next, the bird's-eye view is divided into equally sized grid cells, each corresponding to a region above the ground. The number of emission points within each region varies depending on the target. Each point cloud data includes the distance, angle, height, and speed of the non-ground target relative to the vehicle. By clustering the gridded non-ground point cloud data using a clustering algorithm, the shapes of the main non-ground targets can be obtained, thus yielding the detection results for the non-ground targets.

[0030] Based on the aforementioned steps, it is also possible to further track non-ground targets.

[0031] Specifically, for each point cloud corresponding to a non-ground target, the point cloud closest to the vehicle can be extracted based on the distance between the target and the vehicle, serving as the two-dimensional feature of that non-ground target. Then, tracking of the non-ground target is performed based on distance monitoring using these two-dimensional features. Since the distance between the same point and the vehicle changes over time during vehicle movement, the non-ground target is continuously tracked when its point cloud distance is less than or equal to the vehicle's safe braking distance; otherwise, tracking stops when the distance exceeds the vehicle's safe braking distance. The distance between the point cloud and the vehicle can be calculated using the following formula: Distance c is the speed of light, T is the lidar pulse period, and Δφ is the phase difference between the transmitted and reflected pulses. The distance d changes with time t as follows: When d(t) > l, tracking stops. Here, l represents the vehicle's safe braking distance. The safe braking distance varies depending on the vehicle's current speed. Research shows that the average driver's braking reaction time (including triggering, sensing, judging, releasing the accelerator, braking, and initiating effective braking) is approximately 1.6 seconds. Based on this, at 40 km / h, the braking distance to a stop is 17.6 meters; at 50 km / h, it is 22.4 meters; and at 60 km / h, it is 27.2 meters. The safe braking distance can be based on the stopping distance plus a preset distance. In different embodiments, the vehicle's current speed can also be multiplied by 0.5 to obtain the safe braking distance. For example, the safe braking distance at 50 km / h is 25 meters, and at 60 km / h, it is 30 meters.

[0032] Step S104: Analyze the driving behavior of the vehicle based on the changes in the vehicle's motion relative to ground targets and / or non-ground targets.

[0033] Step S105: Determine whether the driving behavior is a dangerous driving behavior.

[0034] If the driving behavior is determined to be dangerous, proceed to step S106; if the driving behavior is determined not to be dangerous, return to step S101 and continue collecting point cloud data.

[0035] In this embodiment, dangerous driving behaviors include driving in the wrong direction, driving outside designated roads, failing to maintain a safe distance, improper reversing, improper passing, improper overtaking, improper lane changing, continuous lane changing, and improper racing. The following details the analysis and judgment process of driving behaviors.

[0036] Analysis of dangerous driving behaviors (1) Driving against traffic or not driving on designated roads:

[0037] Analyze whether the lane direction in the ground target is opposite to the vehicle's travel direction. If the vehicle direction in the ground target is opposite to the vehicle's travel direction, the current driving behavior is judged as reversing traffic, a dangerous driving behavior.

[0038] When the ground target is a lane line that is not a double solid line, analyze whether the travel direction of multiple other vehicles in the adjacent lane is opposite to the travel direction of the vehicle. If the travel direction of multiple other vehicles in the adjacent lane is opposite to the travel direction of the vehicle, determine that the current driving behavior is dangerous driving behavior, specifically driving in the wrong direction.

[0039] Based on lane markings on ground targets, analyze whether the vehicle is traveling within the designated lane. If the vehicle is not traveling within the designated lane, the current driving behavior is judged as dangerous driving behavior, specifically driving outside the designated lane.

[0040] Analysis of dangerous driving behaviors (2) Failure to maintain a safe distance or improper reversing:

[0041] When there are other vehicles ahead of the vehicle and non-ground objects, analyze whether the distance between the vehicle and the other vehicles is greater than or equal to the safe braking distance. If the distance between the vehicle and the other vehicles is not greater than or equal to the safe braking distance, the current driving behavior is judged as dangerous driving behavior of failing to maintain a safe distance.

[0042] Randomly select a point on the lane line among ground targets outside the safe braking distance for distance monitoring. When the distance to that point is detected to gradually increase over time, the current driving behavior is judged as improper reversing, a dangerous driving behavior.

[0043] Analysis of dangerous driving behaviors (3) Improper passing or overtaking:

[0044] The analysis examines whether the road width gradually narrows over time. Road width is defined as the distance between lane lines on either side of a vehicle. If the road width gradually narrows over time, and the safe braking distance registers other vehicles traveling in different directions, and the vehicle does not change direction, the current driving behavior is judged as improper passing, a dangerous driving behavior.

[0045] Monitors the distance to other vehicles ahead. When the distance to other vehicles is detected to be less than the safe braking distance, and the vehicle changes lanes, the current driving behavior is judged as improper overtaking, a dangerous driving behavior. Specifically, if different lane markings appear on the ground within a default time frame and the vehicle's direction of travel does not change, it is judged that the vehicle has changed lanes.

[0046] Dangerous driving behaviors (4) Improper lane changes, continuous lane changes, or improper speeding:

[0047] When the analysis indicates that a vehicle has changed lanes and crossed a solid lane line, the current driving behavior is judged to be an improper lane change, which is considered dangerous driving behavior.

[0048] When the analysis shows that the vehicle is repeatedly changing lanes, the current driving behavior is judged to be dangerous driving behavior, namely continuous lane changing.

[0049] Based on the vehicle's current speed and the current speeds of other surrounding vehicles, it is determined whether there is mutual acceleration. When both the vehicle's current speed and the speeds of other surrounding vehicles exceed a default speed threshold, the current driving behavior is judged as inappropriate speeding, a dangerous driving behavior. The speeds of other surrounding vehicles can be calculated by taking the vehicle's current speed and their speeds relative to the vehicle.

[0050] Step S106: Issue a safety warning.

[0051] Safety warnings are used to alert drivers to potential dangers. These warnings may include voice prompts, flashing lights, etc., to help drivers understand that their driving behavior is dangerous and to remind them to pay attention to driving safety. In different implementations, safety warnings may also be sent to the driver's emergency contacts or relevant traffic authorities.

[0052] In one embodiment, the number and frequency of dangerous driving behaviors can also be used to determine whether a driver has road rage.

[0053] Road rage is a symptom of uncontrolled emotions, commonly seen while driving in traffic jams or other instances of improper driving behavior. Drivers may exhibit anger, provocation, or abusive behavior. Drivers experiencing road rage are highly likely to intentionally drive in unsafe or dangerous ways, potentially endangering traffic safety. If drivers are unable to manage road rage on their own, they can seek professional support, such as psychological counseling or therapy, to reduce the likelihood of developing road rage.

[0054] Because drivers exhibiting road rage are highly likely to engage in the aforementioned dangerous driving behaviors, it is possible to predict whether a driver has road rage by statistically analyzing the types, frequency, and / or occurrence of these dangerous driving behaviors. This allows for timely warnings when the likelihood of road rage is predicted, thereby increasing driving safety.

[0055] Please see Figure 2 The diagram shows a flowchart of a method for predicting road rage in drivers according to an embodiment of the present invention. Figure 2 As shown, the method includes:

[0056] Step S201: Determine whether the driver engaged in dangerous driving behavior while the vehicle was in motion.

[0057] In one embodiment, step S201 can be implemented as follows: In steps S104 and S105, when the driving behavior is analyzed to be dangerous driving behavior, the type of dangerous driving behavior and the cumulative number of times the dangerous driving behavior occurs are recorded as the driver's driving behavior record. Based on the driver's driving behavior record, it is determined whether the driver engaged in dangerous driving behavior while the vehicle was in motion.

[0058] In one embodiment, road rage predictions for the driver can be made periodically each time the vehicle is started. Driving behavior records can be made each time the vehicle is started, or the driver's driving behavior can be continuously recorded and stored as historical data.

[0059] Step S202: When it is determined that at least one dangerous driving behavior exists, determine the road rage level corresponding to each dangerous driving behavior. The road rage level indicates the degree to which the driver is likely to have road rage.

[0060] It is understandable that during the period after the vehicle starts and is in motion, the driver may not have engaged in any dangerous driving behavior. In this case, it can be determined that the driver is not likely to have road rage. However, if dangerous driving behavior has occurred during the period of vehicle movement, the number of dangerous driving behaviors may be one or more.

[0061] In one embodiment, road rage levels can be pre-set based on the degree of harm to traffic safety posed by different types of dangerous driving behaviors. Thus, if it is determined that a driver has committed at least one dangerous driving behavior, for each dangerous driving behavior, the road rage level corresponding to that type can be queried to determine the road rage level for that dangerous driving behavior. For example, a road rage level of 1 represents a very low probability that the driver has road rage, while a road rage level of 10 represents a very high probability that the driver has road rage. Road rage levels are represented by numbers from 1 to 10. For example, the road rage level for illegal reversing can be set to 1, and the road rage level for failing to maintain a safe distance can be set to 10.

[0062] In different embodiments, the severity levels corresponding to different types of dangerous driving behaviors can be preset based on the degree of harm to traffic safety posed by each type of dangerous driving behavior. Simultaneously, the frequency of occurrence of each type of dangerous driving behavior is counted. The road rage level corresponding to each dangerous driving behavior is calculated based on its type, frequency, and severity. For example, a severity level of 1 represents not serious, and a severity level of 10 represents very serious; the severity level of illegal reversing is preset to 1, and the severity level of failing to maintain a safe distance is preset to 10. The cumulative frequency of illegal reversing is 0, and the cumulative frequency of failing to maintain a safe distance is 8. The road rage level corresponding to a type of dangerous driving behavior can be expressed by multiplying the severity level of that type of dangerous driving behavior by the frequency of its occurrence. That is, the road rage level for illegal reversing is 1 × 0 = 0, and the road rage level for failing to maintain a safe distance is 10 × 8 = 80.

[0063] In different embodiments, the road rage level corresponding to each dangerous driving behavior can also be set or calculated based on other parameters. For example, the road rage level corresponding to the dangerous driving behavior can be calculated based on the number of occurrences, frequency of occurrence, severity, and ease of detection of the dangerous driving behavior.

[0064] Step S203: Predict the driver's road rage status based on the road rage level corresponding to all dangerous driving behaviors.

[0065] Specifically, the road rage level corresponding to dangerous driving behavior is used as an observation value and input into a trained Hidden Markov Model (HMM) to obtain the road rage state prediction output by the HMM. Based on the road rage state prediction, the future road rage state of the driver at a future moment while the vehicle is in motion is predicted. Here, the future moment is any moment within a time period after the current moment.

[0066] The road rage state prediction output by the Hidden Markov Model is the driver's current or future road rage state obtained through the Hidden Markov Model.

[0067] In one embodiment, the Viterbi algorithm (a dynamic programming algorithm) of a Hidden Markov Model can be used to predict the driver's road rage state. The Hidden Markov Model applies basic statistical tools to attempt to reconstruct the state sequence for a given set of observations.

[0068] In this embodiment, the changes in a driver's road rage state, consisting of several road rage states and transition probabilities, constitute a Markov chain with Markov properties, while the dangerous driving behavior adopted by the driver is a stochastic process. Each dangerous driving behavior has a certain impact on the transition of the driver's road rage state. Among them, the impact of dangerous driving behaviors with different road rage levels is different. Thus, the sequence of road rage levels corresponding to each dangerous driving behavior and the driver's road rage state constitute a Hidden Markov Model.

[0069] Understandably, a hidden Markov model consists of hidden states, observable states, and triples (π, A, B).

[0070] The implicit states can include multiple preset road rage states, each corresponding to a different degree of road rage. For example, preset road rage states can include mild road rage, moderate road rage, and severe road rage, with progressively increasing severity. That is, S = {L, M, H}, where S represents the set of implicit states, L represents mild road rage, M represents moderate road rage, and H represents severe road rage. Assuming a driver's road rage state at a certain time t is qt, then qt ∈ S, and the driver's road rage states at multiple times constitute a road rage state sequence.

[0071] The observable states can include multiple observations, each with a default road rage level. The default road rage level can be a pre-set road rage level for each type of dangerous driving behavior, i.e., V = {v1, v2, ..., vN}, where V represents the set of observable states (i.e., the set of observations), vN represents the Nth observation, and N is an integer greater than 1. Assuming the road rage level corresponding to the type of dangerous driving behavior a driver exhibits at a certain time t is ot, then ot ∈ V. The road rage levels corresponding to the types of dangerous driving behaviors a driver exhibits at multiple times constitute an observation sequence.

[0072] Wherein, π is the initial state probability matrix, including the probability distribution of each preset road rage state at the initial time; A is the state transition probability matrix, including the probability of transitioning between each preset road rage state; B is the observation probability matrix, including the probability of generating each road rage level under the condition of being in each preset road rage state at any time. For example, the element in the j-th column of the initial state probability matrix represents the probability of being in the j-th preset road rage state at the initial time. For example, the element in the i-th row and j-th column of the state transition probability matrix represents the probability of transitioning to the j-th preset road rage state at the next time, given that the current state is the i-th preset road rage state. For example, the element in the i-th row and j-th column of the observation probability matrix represents the probability of generating the j-th road rage level, given that the current state is the i-th preset road rage state.

[0073] π and A determine the road rage state sequence, and B determines the observation sequence. Therefore, the Hidden Markov Model can be represented using ternary notation. π, A, and B are called the three elements of the Hidden Markov Model.

[0074] In the embodiments of this application, the parameters can be estimated using the Baum-Welch algorithm in related technologies to obtain the values ​​of the ternary symbols π, A, and B in the triplet, thereby enabling the learning of the Hidden Markov Model. The learned Hidden Markov Model can then be used to predict the driver's road rage state.

[0075] By inputting the road rage level corresponding to at least one type of dangerous driving behavior committed by the driver during vehicle operation as an observation into a Hidden Markov Model (HMM), the HMM can output at least one predicted road rage state, thus predicting the driver's current road rage state. The number of road rage states the driver is in can be the same as the number of dangerous driving behaviors committed during vehicle operation. If only one dangerous driving behavior occurs during vehicle operation, the road rage level corresponding to that behavior can be used as an observation, input into the HMM, and the HMM can output a predicted road rage state. If multiple dangerous driving behaviors occur during vehicle operation, the road rage levels corresponding to the multiple dangerous driving behaviors can be used as multiple observations. These observations can form an observation sequence, which is input into the HMM, allowing the HMM to output a road rage state sequence, including multiple predicted road rage states.

[0076] In one embodiment, the future road rage state of the final driver at a future moment in the vehicle can be predicted based on the predicted road rage state output by the Hidden Markov Model and the number of predicted road rage states, i.e. the number of dangerous driving behaviors committed by the driver while the vehicle is in motion.

[0077] For example, if a driver commits multiple dangerous driving behaviors, the Hidden Markov Model will also output multiple predicted road rage states. The predicted road rage state that occurs most frequently or has the highest probability can be identified as the future road rage state.

[0078] It is understandable that the driver may not have engaged in any dangerous driving behavior while the vehicle was in motion, and in this situation, it can be determined that the driver does not have road rage.

[0079] In one embodiment, different levels of warnings can be provided based on the driver's road rage state to ensure that the driver receives real-time alerts and corrects their driving behavior. These different levels of warnings include, for example, voice warnings, visual warnings, and a combination of voice and visual warnings. This embodiment does not limit the implementation method of the warnings.

[0080] Please see Figure 3 The diagram shown is a hardware block diagram of a dangerous driving behavior warning device 300 according to an embodiment of the present invention. The device 300 includes a processor 302, a memory 304, a sensing unit 306, a communication interface 308, and an alarm unit 310. Those skilled in the art should understand that... Figure 3 The composition of the device 300 shown does not constitute a limitation of the embodiments of the present invention. Figure 3The device 300 shown is simplified for ease of description, and in different embodiments it may include fewer or more components than shown.

[0081] In one embodiment, the processor 302 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits packaged with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 302 is the control unit of the device 300, connecting various components of the device 300 via various interfaces and lines. It executes computer programs or modules stored in the memory 302 and calls data stored in the memory 302 to perform various functions of the device 300 and process data, such as a method for warning of dangerous driving behavior.

[0082] In one embodiment, the memory 304 is used to store computer program code and various data, such as a method for warning of dangerous driving behavior, and to achieve high-speed, automatic access to programs or data during the operation of the device 300. The memory 304 includes read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable storage medium capable of carrying or storing data.

[0083] In one embodiment, the sensing unit 306 includes several sensors for sensing environmental information about the vehicle's surroundings. For example, the sensing unit 306 may include a positioning system, an inertial measurement unit, a lidar, and a camera. The sensing unit 306 may also include sensors that monitor internal vehicle systems, such as an in-vehicle air quality monitor, a fuel gauge, and an oil temperature gauge. In different embodiments, different sensors in the sensing unit 306 may be different, independent devices connected to the device 300 via wireless or wired communication. The lidar, using laser light as a medium, detects objects based on Time-of-Flight (TOF) or phase-shift methods, and detects the position, distance, and relative velocity of the detected objects.

[0084] In one embodiment, the communication interface 308 is composed of a communication circuit for communicating data or information with external devices.

[0085] In one embodiment, the alarm unit 310 includes features for providing visual or auditory warnings. The visual warning includes a light alert, and the auditory warning includes a voice alert or a specific audio alert. In different instances, the alarm unit 310 may also provide both visual and auditory warnings simultaneously.

[0086] In summary, the dangerous driving behavior warning method, device, and computer-readable storage medium of the present invention can be used to detect dangerous driving behaviors of drivers and issue real-time warnings. Furthermore, based on the type of dangerous driving behavior, it can predict the driver's future tendency to experience road rage.

[0087] It is worth noting that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for early warning of dangerous driving behavior, characterized in that, The method includes: Acquire point cloud data collected by lidar; The point cloud data is divided into ground point cloud data and non-ground point cloud data; Identify ground targets and non-ground targets based on ground point cloud data and non-ground point cloud data respectively; Analyze the driver's behavior based on the vehicle's motion relative to ground and non-ground targets. Determine whether the driving behavior is dangerous. as well as When the driving behavior is determined to be dangerous, a safety warning is issued.

2. The method for early warning of dangerous driving behavior as described in claim 1, characterized in that, The types of dangerous driving behaviors include driving in the wrong direction and failing to maintain a safe distance.

3. The method for early warning of dangerous driving behavior as described in claim 1, characterized in that, The step of distinguishing the point cloud data into ground point cloud data and non-ground point cloud data further includes: The point cloud data is transformed from the point cloud coordinate system to the vehicle's coordinate system; Point cloud data with a height less than a preset height threshold are classified into the ground point cloud data; as well as Point cloud data with a height greater than or equal to a preset height threshold are classified as non-ground point cloud data.

4. The method for early warning of dangerous driving behavior as described in claim 1, characterized in that, The method further includes: Convert the non-ground point cloud data into a bird's-eye view; The bird's-eye view is divided into equal-sized grid cells, each corresponding to an area above the ground. The non-ground point cloud data after being gridded is clustered using a clustering algorithm to obtain the shape of the non-ground target object; Based on the distance between each non-ground object and the vehicle, the point cloud closest to the vehicle in each non-ground object is extracted as the two-dimensional feature of each non-ground object; as well as Based on the two-dimensional features of each non-ground target, each non-ground target whose distance relative to the vehicle is less than or equal to the vehicle's safe braking distance is tracked.

5. The method for early warning of dangerous driving behavior as described in claim 2, characterized in that, The determination of whether the driving behavior is dangerous driving behavior also includes: Analyze whether the lane direction in the ground target is opposite to the vehicle's travel direction; When the direction of the lane is opposite to the direction of travel of the vehicle, the driving behavior is determined to be reversing the direction of travel, which is considered a dangerous driving behavior.

6. The method for early warning of dangerous driving behavior as described in claim 2, characterized in that, The determination of whether the driving behavior is dangerous driving behavior also includes: When there are lane lines that are not double solid lines in the ground target, analyze whether the driving direction of multiple other vehicles in the non-ground targets in the adjacent lane is opposite to the driving direction of the vehicle. When the driving direction of the plurality of other vehicles is opposite to that of the vehicle, the driving behavior is determined to be reversing the direction of travel in the dangerous driving behavior.

7. The method for early warning of dangerous driving behavior as described in claim 2, characterized in that, The determination of whether the driving behavior is dangerous driving behavior also includes: When there are other vehicles ahead of the non-ground target, analyze whether the distance between the vehicle and the other vehicles is greater than or equal to the safe braking distance. When the distance between the vehicle and other vehicles is not greater than or equal to the safe braking distance, the driving behavior is determined to be a failure to maintain a safe distance in the dangerous driving behavior category.

8. The method for early warning of dangerous driving behavior as described in claim 1, characterized in that, The method further includes: To determine whether the driver engaged in the dangerous driving behavior while the vehicle was in motion; When it is determined that the driver has at least one of the dangerous driving behaviors, the road rage level corresponding to each dangerous driving behavior is determined, wherein the road rage level is used to indicate the degree to which the driver is likely to have road rage disorder; as well as Based on the road rage level corresponding to all dangerous driving behaviors, the driver's road rage status is predicted.

9. A warning device for dangerous driving behavior, characterized in that, include Processor; and A memory for storing a computer program that, when executed by the processor, causes the processor to implement a warning method for dangerous driving behavior as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a method for warning of dangerous driving behavior as described in any one of claims 1 to 8.

Images