Dynamic target prompting method and device, electronic equipment and storage medium

Abstract

The invention relates to the field of vehicle intelligent control, in particular to a dynamic target prompting method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a dynamic target and motion data thereof in a vehicle driving environment; then, the driving direction of the vehicle is detected, the motion data is analyzed, a traversing dynamic target with a crossing trend is identified, and potential dangers are locked in advance. And then, detecting whether a collision risk target object exists between the transverse moving target and the transverse moving target based on the reverse driving direction, and if the collision risk target object exists, executing a safety prompt operation on the transverse moving target. Therefore, a set of complete active safety warning mechanism is constructed, other vehicles and pedestrians can timely perceive the danger, compared with the prior art, the system is not in the state of passively waiting for the danger, the safety of the vehicles and the pedestrians in the road environment is greatly improved, and the occurrence probability of traffic accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the field of vehicle intelligent control, and in particular to a dynamic target prompting method, device, electronic equipment and storage medium. Background Art

[0002] With the development of autonomous driving technology, vehicles, through the deployment of sensing devices such as cameras, ultrasonic radar, and millimeter-wave radar, have been able to achieve real-time detection of their surroundings and dynamic target tracking, significantly improving safety features such as collision warning and path planning. Currently, mainstream intelligent driving features (such as automatic emergency braking and lane departure warning) are centered around the driver, providing risk warnings through dashboard lighting, voice notifications, and steering wheel vibration. Their design logic is centered around the "vehicle," meeting only the one-way safety protection needs of the vehicle itself.

[0003] However, real-world traffic scenarios are interactive systems comprised of multiple entities, including the vehicle itself, pedestrians, two-wheeled vehicles, and other passenger vehicles. The movements and decision-making of each traffic participant influence each other. Existing technologies lack active safety warning mechanisms for non-vehicle traffic participants outside the vehicle, resulting in other vehicles and pedestrians being unable to perceive danger in a timely manner. This lack of information exchange can easily lead to collisions, especially in complex scenarios such as mixed traffic intersections and obstructed vision. Summary of the Invention

[0004] In view of this, an embodiment of the present invention provides a dynamic target prompting method, device, electronic device and storage medium to solve the problem that the existing technology lacks an active safety warning mechanism for non-vehicle traffic participants outside the vehicle, resulting in other vehicles / pedestrians being unable to perceive danger in a timely manner.

[0005] In a first aspect, an embodiment of the present invention provides a method for prompting a dynamic target, the method comprising:

[0006] Acquire a target sequence in a driving environment of the vehicle, wherein the target sequence includes a dynamic target and first motion data of the dynamic target;

[0007] detecting the current driving direction of the vehicle and analyzing the first motion data to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction;

[0008] If there is a crossing dynamic target that has a tendency to intersect the driving direction, detecting whether there is a target object that has a collision risk with the crossing dynamic target based on the opposite direction of the driving direction;

[0009] If there is a target object that has a collision risk with the crossing dynamic target, a safety prompt operation is performed on the crossing dynamic target.

[0010] Furthermore, analyzing the first motion data to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction includes:

[0011] Predicting a motion trajectory of the dynamic target using the first motion data;

[0012] Extracting the trajectory direction corresponding to the motion trajectory;

[0013] The trajectory direction and the driving direction of the vehicle are compared to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction.

[0014] Furthermore, before detecting the current driving direction of the vehicle, the method further includes:

[0015] Acquire point cloud data of the driving environment of the vehicle and a road section image of the road section on which the vehicle is currently traveling;

[0016] Analyzing environmental attributes of the driving environment using the road segment image and the point cloud data;

[0017] Querying the road section label of the road section where the vehicle is located;

[0018] Analyzing whether the current driving section has a crossing attribute based on the environmental attributes and the section label;

[0019] If the current driving section has a crossing attribute, the step of detecting the vehicle's driving direction is executed.

[0020] Furthermore, the detecting, based on the direction opposite to the driving direction, whether there is a target object with a collision risk with the crossing dynamic target includes:

[0021] Detecting a video stream in a direction opposite to the driving direction;

[0022] Based on the video stream, detecting whether there is a candidate object in the opposite direction of the vehicle's travel direction and having a travel speed greater than a preset speed;

[0023] If there is at least one candidate object whose travel speed is greater than a preset speed, predicting a target collision risk value between the candidate object and the crossing dynamic target;

[0024] Based on the comparison result of the target collision risk value and the first risk threshold, it is determined whether there is a target object with a collision risk with the crossing dynamic target; wherein, the candidate object whose target collision risk value is greater than or equal to the first risk threshold is determined as the target object with a collision risk with the crossing dynamic target.

[0025] Furthermore, the predicting of a target collision risk value between the candidate object and the traversing dynamic target includes:

[0026] Acquiring second motion data of the candidate object and third motion data of the vehicle;

[0027] predicting a first collision risk value between the candidate object and the traversing dynamic target based on the second motion data and the first motion data;

[0028] predicting a second collision risk value between the candidate object and the vehicle based on the second motion data and the third motion data;

[0029] If the second collision risk value is greater than a second risk threshold, predicting fourth motion data of the vehicle after being collided with the candidate object;

[0030] predicting, based on the fourth motion data and the first motion data, a third collision risk value of the vehicle with the traversing dynamic target after being collided with the candidate object;

[0031] A target collision risk value is calculated based on the first collision risk value and the third collision risk value.

[0032] Furthermore, the performing of a safety prompt operation on the traversing dynamic target includes:

[0033] monitoring the movement of the target object in real time and obtaining a first prompt strategy corresponding to the movement;

[0034] detecting a relative positional relationship between the crossing dynamic target and the vehicle;

[0035] Determining, based on the relative positional relationship, a prompting device in the vehicle for providing a safety prompt to the crossing dynamic object;

[0036] The prompt device is controlled to perform a safety prompt operation on the crossing dynamic target according to the first prompt strategy.

[0037] Furthermore, after performing a safety prompt operation on the traversing dynamic target, the method further includes:

[0038] Monitoring the motion change data of the crossing dynamic target after the safety prompt operation;

[0039] If the motion change data is lower than a preset data, obtaining the object type of the target object;

[0040] Obtaining a second prompt strategy corresponding to the object type;

[0041] Perform an avoidance prompt operation on the target object according to the second prompt strategy.

[0042] In a second aspect, an embodiment of the present invention provides a dynamic target prompting device, the device comprising:

[0043] An acquisition module, configured to acquire a target sequence in a driving environment in which the vehicle is located, wherein the target sequence includes a dynamic target and first motion data of the dynamic target;

[0044] an analysis module, configured to detect the current driving direction of the vehicle and analyze the first motion data to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction;

[0045] a detection module configured to, if there is a crossing dynamic target having a tendency to intersect the driving direction, detect, based on a direction opposite to the driving direction, whether there is a target object with a collision risk with the crossing dynamic target;

[0046] The sending module is configured to perform a safety prompt operation on the crossing dynamic target if there is a target object that has a collision risk with the crossing dynamic target.

[0047] In a third aspect, an embodiment of the present invention provides an electronic device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, computer instructions being stored in the memory, and the processor executing the method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0048] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the method of the first aspect or any corresponding embodiment thereof.

[0049] This application obtains dynamic targets and their motion data in the vehicle driving environment. Then, it detects the vehicle's driving direction and analyzes the motion data to identify crossing dynamic targets with a crossing trend and locks in potential dangers in advance. Then, based on the opposite direction of driving, it detects whether there is a target object with a collision risk with the crossing dynamic target. If it is determined that there is a target object with a collision risk, a safety prompt operation is performed on the crossing dynamic target. In this way, a complete set of active safety warning mechanisms is constructed, which can enable other vehicles and pedestrians to perceive danger in a timely manner. Compared with the existing technology, they are no longer in a state of passively waiting for danger to occur, which greatly improves their safety in the road environment and reduces the probability of traffic accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 is a flowchart of a method for prompting a dynamic target according to some embodiments of the present invention;

[0052] Figure 2 is a schematic diagram of a scenario simulation according to some embodiments of the present invention;

[0053] Figure 3 is a schematic diagram of a scenario simulation according to some embodiments of the present invention;

[0054] Figure 4 is a flowchart of another dynamic target prompting method according to some embodiments of the present invention;

[0055] Figure 5 is a flowchart of another dynamic target prompting method according to some embodiments of the present invention;

[0056] Figure 6 is a structural block diagram of a device for prompting a dynamic target according to an embodiment of the present invention;

[0057] Figure 7 FIG. 4 is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0059] According to an embodiment of the present invention, a method, device, electronic device and storage medium for prompting a dynamic target are provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0060] In this embodiment, a method for prompting a dynamic target is provided. Figure 1FIG. 1 is a flow chart of a method for prompting a dynamic target according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0061] Step S101 : acquiring a target sequence in a driving environment of a vehicle, wherein the target sequence includes dynamic targets and first motion data of the dynamic targets.

[0062] In the embodiment of the present application, during normal driving of the vehicle, the radar (millimeter-wave radar, lidar) and the camera (forward-view, surround-view camera) remain in continuous operation. The radar uses electromagnetic waves or laser beams to scan the surrounding space and calculates the distance, speed, and direction of the target through the reflected signal; the camera processes the collected images based on computer vision algorithms to identify traffic participants such as two-wheeled vehicles, pedestrians, and four-wheeled vehicles. For example, the millimeter-wave radar can monitor the speed changes of vehicles within a range of 200 meters in real time, and the camera uses the target detection algorithm to identify pedestrians in the driving environment.

[0063] Radar and camera data on detected targets is transmitted in real time to the central domain controller (DC) via the vehicle data bus. This data includes the target's point cloud coordinates (radar data), image pixel information, and preliminary recognition results (camera data), such as the target's 3D coordinates (X, Y, Z) in the radar coordinate system and the "pedestrian" category label annotated by the camera. The DC serves as the vehicle's "brain," receiving and storing multi-source data from various sensors for subsequent processing.

[0064] The central domain controller extracts and matches features from received radar and camera data. Using algorithms such as Kalman filtering and particle filtering, it eliminates errors and redundancy in sensor data and correlates information about the same target from different sensors. For example, it fuses the distance and speed of a vehicle detected by radar with the vehicle's appearance and position detected by the camera to form a complete dynamic description of the vehicle. Ultimately, this fused target information is generated into a target sequence in chronological order. Each sequence element contains target type (two-wheeled vehicle / pedestrian / four-wheeled vehicle), motion data such as position, speed, acceleration, and a timestamp.

[0065] Step S102 : Detect the current driving direction of the vehicle, analyze the first motion data, and determine whether there is a dynamic target crossing the driving direction in the target sequence.

[0066] In an embodiment of the present application, before detecting the current driving direction of the vehicle, the method also includes: obtaining point cloud data of the driving environment in which the vehicle is located, and a road section image of the road section in which the vehicle is currently traveling; using the road section image and point cloud data to analyze the environmental attributes of the driving environment; querying the road section label of the road section in which the vehicle is located; based on the environmental attributes and the road section label, analyzing whether the current driving section has a crossing attribute; if the current driving section has a crossing attribute, executing the step of detecting the vehicle's driving direction.

[0067] Specifically, first, the on-board radar collects point cloud data around the vehicle in real time to obtain three-dimensional coordinate information of road structure, obstacle locations, and dynamic targets. Simultaneously, the forward-looking camera captures images of the road ahead and extracts two-dimensional visual features such as lane lines and traffic signs. Secondly, the multi-source data is fused and processed: a point cloud clustering algorithm is used to detect road isolation facilities and terrain parameters (such as slope and curvature), and dynamic targets are marked. An image recognition algorithm is used to analyze crossing-related signs such as zebra crossings and intersection guide lines, generating environmental attribute labels such as "physical isolation status" and "presence of legal crossing signs."

[0068] Then, the system uses the vehicle's positioning information to query a high-precision map and retrieve predefined labels for the current road segment, such as road type (intersection / rural road) and regulatory restrictions (crossing permitted areas). If the environmental attributes include a zebra crossing without a barrier, or if the road segment is labeled "crosswalk," the crossing attribute is determined to be present. If the crossing attribute conditions are met, the real-time driving direction is further calculated using vehicle bus data (such as steering wheel angle and gyroscope).

[0069] After determining the vehicle's direction of travel, the first motion data in the target sequence is extracted. The position coordinates, velocity vector, and heading angle of each dynamic target are then extracted, and its motion trajectory for the next 3-5 seconds is predicted using algorithms such as Kalman filtering. The angle between each target's predicted trajectory and the vehicle's direction of travel is then calculated. If the angle is within the range of 60°-120°, the target's lateral velocity is greater than 1.2m / s (an empirical threshold), and the lateral distance difference between the target and the vehicle's travel path within the next 5 seconds is predicted to be less than a safety threshold (e.g., 3 meters), the target is determined to be a crossing dynamic target with a tendency to intersect the travel direction.

[0070] Step S103 : If there is a dynamic target crossing the driving direction, a target object with a collision risk with the dynamic target is detected based on the opposite direction of the driving direction.

[0071] In this embodiment, after determining the presence of a dynamic target crossing the direction of travel, a coordinate system is established based on the vehicle's own direction of travel, with the vehicle itself as the origin. This defines the area in the opposite direction of travel. Data from this area is collected using sensors such as the vehicle's rear-mounted camera and radar. The millimeter-wave radar focuses on acquiring information such as the target's distance, speed, and azimuth, while the rear-mounted camera uses image recognition technology to detect the target's type, outline, and motion state, providing a data foundation for subsequent analysis.

[0072] Then, the collected data is processed and analyzed. The millimeter-wave radar data and camera image information are fused, and the motion trajectory of the target object is predicted through algorithms such as Kalman filtering. A collision risk determination rule is set. If the predicted trajectory of the target object in the future (such as 3-5 seconds) intersects with the motion trajectory of the crossing dynamic target, and the relative speed and distance between the two meet certain threshold conditions (such as the relative speed is greater than 10m / s and the distance is less than the safe distance of 5 meters), then it is determined that the target object and the crossing dynamic target have a collision risk.

[0073] As an example, Figure 2 As shown in the figure, vehicle A is the detection subject, and its direction of travel is indicated by the arrow. When a two-wheeled vehicle is detected to be crossing the direction of vehicle A, the detection range is in the direction opposite to vehicle A's direction of travel (toward the rear of the vehicle). At this time, rear-mounted sensors (such as radars and cameras) are used to detect the black vehicle. By obtaining information such as the speed and distance of vehicle C and analyzing the motion trajectory of the two-wheeled vehicle, it is determined that within the next few seconds, vehicle C will encounter the two-wheeled vehicle at a certain point according to its current trajectory. In this case, there is a collision risk between vehicle C and the crossing dynamic target (two-wheeled vehicle).

[0074] Step S104 : If there is a target object that has a collision risk with the crossing dynamic target, a safety prompt operation is performed on the crossing dynamic target.

[0075] In an embodiment of the present application, when a target object with a collision risk with a crossing dynamic target is detected, the central domain controller sends an alarm signal. The external display screen immediately displays a striking warning pattern or text (such as "Danger, pay attention to avoid"), the external lights flash (such as the double flash lights flash at a high frequency), the external speakers emit a rapid alarm sound, and the voice player synchronously plays a voice warning (such as "Please pay attention, there is a risk of collision"), thereby performing a safety prompt operation for the crossing dynamic target. At the same time, the alarm conditions are continuously judged. If the collision risk is eliminated, that is, the alarm conditions are not met, each module will cancel the alarm; if the alarm conditions are still met, the next cycle of target recognition and alarm judgment will continue.

[0076] As an example, Figure 3As shown in the figure, a vehicle is traveling normally along the straight lane on a main urban road. The vehicle's radar and camera sensors are continuously operating. The radar sensor transmits and receives electromagnetic waves to obtain primary motion data such as the distance and speed of surrounding objects. The camera sensor captures images of the surrounding environment and identifies dynamic targets such as pedestrians and other vehicles. This information is combined into a target sequence and transmitted to the vehicle's central domain controller.

[0077] The central domain controller detects that the vehicle is currently moving straight ahead. When analyzing the first motion data of the target sequence, it finds a pedestrian walking from the right side of the road to the left side. The pedestrian's motion trajectory tends to intersect with the vehicle's direction of travel. The central domain controller determines that the pedestrian is a crossing dynamic target.

[0078] Therefore, the central domain controller performs detection based on the opposite direction of the vehicle's travel (i.e., the rear). Through sensor data, it finds that an electric vehicle behind is moving at a high speed and its driving trajectory may create a collision risk with the pedestrian. It determines that the electric vehicle is a target object with a collision risk with the crossing dynamic target.

[0079] Finally, the vehicle's central domain controller immediately sends instructions to the external display screen, external lights, and external speaker voice player. The external display screen displays a sign reminding pedestrians to pay attention to safety, the external lights flash as a warning, and the external speaker voice player plays a prompt voice "Please pay attention to vehicles coming from behind", performing safety reminder operations on the crossing dynamic target (pedestrian).

[0080] The embodiment of the present application constructs an active safety warning mechanism for non-vehicle traffic participants outside the vehicle from multiple dimensions, effectively solving the shortcomings of the existing technology. First, the target sequence and its motion data in the vehicle driving environment are fully acquired. Secondly, the vehicle driving direction is accurately detected, and the dynamic target motion data is deeply analyzed to capture the crossing dynamic targets with a crossing trend, so as to achieve early prediction of danger. On this basis, the collision risk target objects are further detected based on the opposite direction of driving, and safety prompt operations are performed on the crossing dynamic targets, and warning signals are issued in time with the help of various devices outside the vehicle. This set of coherent and systematic processes has changed the previous situation where other vehicles / pedestrians could only respond to danger passively, and actively provides them with safety warnings, so that they can perceive danger in time and respond in advance, which greatly improves the safety of road traffic, fills the gaps in the existing technology in active safety warnings, and effectively reduces the possibility of traffic accidents.

[0081] Figure 4 FIG. 1 is a flow chart of a method for prompting a dynamic target according to an embodiment of the present invention. Figure 4 As shown, the process includes the following steps:

[0082] Step S201 : Acquire a target sequence in a driving environment where a vehicle is located, wherein the target sequence includes dynamic targets and first motion data of the dynamic targets.

[0083] In an embodiment of the present application, the vehicle's surrounding environment is continuously scanned by on-board sensors (such as lidar, camera, millimeter-wave radar, etc.), the lidar obtains information such as the three-dimensional spatial position and reflectivity of the target, the camera collects the visual image of the target to identify the target type and appearance characteristics, and the millimeter-wave radar measures parameters such as the target's speed and distance.

[0084] The data collected by each sensor is fused and processed, and based on target detection and tracking algorithms (such as YOLO target detection and Kalman filter tracking), dynamic targets such as pedestrians and vehicles in the vehicle driving environment are identified, and their first motion data such as position, speed, and acceleration are extracted. Finally, a target sequence containing dynamic targets and their first motion data is constructed.

[0085] Step S202 : Detect the current driving direction of the vehicle, analyze the first motion data, and determine whether there is a dynamic target crossing the driving direction in the target sequence.

[0086] In an embodiment of the present application, the first motion data is analyzed to determine whether there is a dynamic target crossing the target sequence that has a cross-trend with the driving direction, including: using the first motion data to predict the motion trajectory of the dynamic target; extracting the trajectory direction corresponding to the motion trajectory; comparing the trajectory direction and the driving direction of the vehicle to determine whether there is a dynamic target crossing the target sequence that has a cross-trend with the driving direction.

[0087] Specifically, the target's initial motion data (position, velocity, acceleration, etc.) is fed into a motion prediction model, such as a deep learning-based LSTM or Transformer model. These models learn and analyze historical data, taking into account the target's inertia, dynamic characteristics, and surrounding environmental constraints. They predict the target's position over a period of time in the future and then sequentially connect multiple predicted positions to generate the target's trajectory.

[0088] After obtaining the motion trajectory of a dynamic target, analyze the trajectory. By calculating the tangent direction of the trajectory at a specific moment or fitting multiple consecutive points along the trajectory to obtain its trend direction, the corresponding trajectory direction can be determined. Vector calculations can be used to quantify the trajectory direction into an angle value, for example, with due east as 0° and increasing clockwise, thus clearly representing the trajectory's direction.

[0089] The vehicle's direction of travel is then determined, typically using an onboard gyroscope, electronic compass, or in combination with high-precision map data. The dynamic target's trajectory is compared with the vehicle's direction of travel, and a threshold for determining a crossing trend is set. If the angle between the trajectory and the vehicle's direction of travel is between 60° and 120°, and the trajectory is predicted to intersect with the vehicle's path within a certain period of time (e.g., 3-5 seconds), the dynamic target is considered a crossing target with a crossing trend; otherwise, the target is considered to have no crossing risk.

[0090] In step S203 , if there is a dynamic target crossing the driving direction, a target object with a collision risk with the dynamic target crossing the driving direction is detected based on the opposite direction of the driving direction.

[0091] In an embodiment of the present application, detecting whether there is a target object with a collision risk with a crossing dynamic target based on the direction opposite to the driving direction includes the following steps A1-A4:

[0092] Step A1: Detect the video stream in the opposite direction of the driving direction.

[0093] Specifically, at least two high-definition cameras are installed at the rear of the vehicle. One camera's field of view faces directly behind the vehicle's lane, capturing a first video stream of that lane; the other camera's field of view covers the area behind the adjacent lane, capturing a second video stream of that lane. The cameras continuously record at a frame rate of 25-30 frames per second and transmit the video data in real time via an onboard data transmission line to an onboard computing unit. The computing unit decodes and caches the video stream, providing raw data for subsequent target detection and analysis.

[0094] Step A2: Based on the video stream, detect whether there is a candidate object in the opposite direction of the vehicle's travel direction and whose travel speed is greater than a preset speed.

[0095] Specifically, computer vision technology is used to process the collected first and second video streams, and target detection algorithms (such as YOLO, FasterR-CNN, etc.) are used to identify vehicles, fallen rocks and other objects in the video, and a unique identifier is assigned to each detected object. In continuous video frames, the position changes of the object are tracked, and the pixel displacement of the object between adjacent frames is calculated. The pixel displacement is converted into actual physical displacement by combining the calibration parameters of the camera and the motion information of the vehicle itself, and the actual driving speed of the object is calculated. The calculated object speed is compared with the preset speed (such as 40km / h). If the object speed is greater than the preset speed, it is determined to be a candidate object, otherwise it is excluded.

[0096] In step A3, if there is at least one candidate object whose travel speed is greater than a preset speed, a target collision risk value between the candidate object and the crossing dynamic target is predicted.

[0097] Specifically, when at least one candidate object with a travel speed greater than a preset speed is detected, a Kalman filter or deep learning prediction model is used to combine the current position, speed, acceleration and other motion data of the candidate object and the crossing dynamic target, as well as road environment information (such as lane width, curve curvature, etc.), to predict the motion trajectory of the two in the next 3-5 seconds. A risk assessment model is constructed by analyzing the intersection possibility, intersection time point, and relative speed and distance of the predicted trajectories. For example, a weighted scoring method is used to assign different weights to factors such as trajectory intersection probability, relative speed difference, and expected collision time (such as a trajectory intersection probability weight of 0.4, a relative speed difference weight of 0.3, and an expected collision time weight of 0.3) to calculate the target collision risk value between the candidate object and the crossing dynamic target.

[0098] In addition, an embodiment of the present application also provides a target collision risk value prediction process, which is as follows: obtaining the second motion data of the candidate object and the third motion data of the vehicle; based on the second motion data and the first motion data, predicting the first collision risk value between the candidate object and the crossing dynamic target; based on the second motion data and the third motion data, predicting the second collision risk value between the candidate object and the vehicle; if the second collision risk value is greater than the second risk threshold, predicting the fourth motion data of the vehicle after being collided with the candidate object; based on the fourth motion data and the first motion data, predicting the third collision risk value of the vehicle with the crossing dynamic target after being collided with the candidate object; based on the first collision risk value and the third collision risk value, calculating the target collision risk value.

[0099] First, real-time data is collected through onboard sensors such as radar and cameras. Radar continuously monitors the relative speed and acceleration of targets. Cameras utilize computer vision algorithms, feature point matching, and target tracking to identify the motion posture and contour changes of candidate objects and vehicles. This multi-source, heterogeneous data is spatially and temporally aligned and fused to extract secondary motion data such as the candidate object's position, velocity, and acceleration, as well as tertiary motion data such as the vehicle's speed, steering angle, and acceleration.

[0100] Secondly, the second motion data of the candidate object and the first motion data of the crossing dynamic target are input into the collision risk prediction model, which can be a dynamic model based on physical rules or a deep learning model (such as LSTM, Transformer). The model first analyzes the historical motion trajectories of the two and predicts the motion trajectories for the next 3-5 seconds through algorithms such as Kalman filtering. Then, the potential collision point is calculated based on the predicted trajectory. Combined with parameters such as relative speed and approach angle, the risk assessment function (such as risk = relative speed × collision probability) is used to calculate the possibility of collision between the candidate object and the crossing dynamic target, and finally the first collision risk value is obtained.

[0101] The second motion data of the candidate object and the third motion data of the vehicle are then input into the collision risk prediction model. Based on the vehicle's dynamic characteristics and the motion patterns of the candidate object, the model simulates the future motion states of both, taking into account constraints such as road boundaries and traffic regulations. By calculating the probability of intersection between the candidate object and the vehicle's path, as well as indicators such as relative speed and distance at the intersection, and combining them with a pre-defined risk assessment algorithm (such as a fuzzy logic comprehensive assessment), the model quantifies the risk of collision between the candidate object and the vehicle, thereby generating a second collision risk value.

[0102] When the second collision risk value exceeds the pre-set second risk threshold (such as 0.7, indicating a high probability of collision), the post-collision vehicle motion prediction process is initiated. Using the vehicle collision dynamics model, the mass of the candidate object and vehicle, collision angle, collision speed and other parameters are input to simulate the collision process. Through Newton's laws of motion and the principle of conservation of energy, the vehicle's speed, acceleration, change in direction of motion and other data after the collision are calculated. Combined with factors such as the vehicle suspension system and tire friction, the vehicle's driving trajectory and sliding distance after the collision are predicted, and the fourth motion data after the vehicle is hit by the candidate object is finally determined.

[0103] Next, the fourth post-collision motion data of the vehicle and the first post-crossing dynamic object motion data are re-entered into the collision risk prediction model. Based on the vehicle's post-collision motion trend and the ongoing motion of the crossing dynamic object, the model re-predicts the future trajectory of both. By analyzing key indicators such as whether the new trajectory intersects, the time of intersection, and the relative speed at the time of intersection, a risk assessment method similar to that used to calculate the first collision risk value is used to calculate the probability of a secondary collision between the vehicle and the crossing dynamic object after being hit by the candidate object, resulting in a third collision risk value.

[0104] Finally, a weighted summation approach is used to comprehensively consider the first and third collision risk values. Different weights are assigned to the two risk values (e.g., a weight of 0.6 for the first collision risk value and 0.4 for the third collision risk value) based on the severity of single and secondary collisions in actual traffic scenarios. The final target collision risk value is calculated using the formula "target collision risk value = first collision risk value × 0.6 + third collision risk value × 0.4." This value comprehensively reflects the combined collision risk between candidate objects, vehicles, and crossing dynamic targets, providing a quantitative basis for subsequent safety decisions (such as emergency braking, avoidance path planning, and warning reminders).

[0105] Based on this, the embodiment of the present application has built a complete safety assessment system through multi-dimensional and multi-level collision risk prediction and analysis, which effectively improves the ability to predict risks in complex traffic scenarios. First, the motion data of candidate objects, crossing dynamic targets and vehicles are obtained, which comprehensively covers the dynamic information of road participants and provides accurate data support for risk assessment. The collision risk values of candidate objects and crossing dynamic targets, and candidate objects and vehicles are calculated separately, and the granularity of risk analysis is refined from different interactive relationships; when the collision risk between the vehicle and the candidate object exceeds the threshold, the motion data after the vehicle collision is further predicted, and the secondary collision risk with the crossing dynamic target after the collision is deduced, and the chain reaction of accidents is simulated, breaking through the traditional limitation of only considering a single collision. Finally, the risk values of each link are integrated to calculate the target collision risk value to form a comprehensive risk assessment result.

[0106] Step A4, based on the comparison result of the target collision risk value and the first risk threshold, determines whether there is a target object with a collision risk with the crossing dynamic target; wherein, the candidate object with a target collision risk value greater than or equal to the first risk threshold is determined as the target object with a collision risk with the crossing dynamic target.

[0107] Step S204: If there is a target object that has a collision risk with the crossing dynamic target, a safety prompt operation is performed on the crossing dynamic target.

[0108] In an embodiment of the present application, a safety prompt operation is performed on a crossing dynamic target, including: real-time monitoring of the movement of the target object, and obtaining a first prompt strategy corresponding to the movement; detecting the relative position relationship between the crossing dynamic target and the vehicle; determining a prompt device in the vehicle for providing a safety prompt for the crossing dynamic target based on the relative position relationship; and controlling the prompt device to perform a safety prompt operation on the crossing dynamic target according to the first prompt strategy.

[0109] Specifically, based on the target object's speed and directional changes, a pre-defined rule library is queried to obtain the corresponding first alert strategy. For example, if the target object is moving rapidly and gradually approaching the vehicle, the strategy is set to high-frequency alerts. Simultaneously, sensor data is used to calculate the relative position of the crossing dynamic object and the vehicle, determining whether the crossing dynamic object is in front of or to the side of the vehicle, and its distance from the vehicle. Based on this relative position, the appropriate alert device is selected. For example, if the crossing dynamic object is in front of the vehicle, the front display and front speakers are activated; if it is to the side, the side display and speakers on the corresponding side are activated.

[0110] Finally, according to the first prompt strategy obtained, a control instruction is sent to the selected prompt device to control the prompt device to perform corresponding safety prompt operations, such as displaying a flashing warning pattern on the display screen and emitting an alarm sound of a specific frequency from the horn to alert the crossing dynamic target to potential dangers.

[0111] As an example, when a vehicle is turning right at an intersection, it detects an electric bicycle with a speed of 18 km / h approaching quickly from the left side of the vehicle, and its direction of travel is obviously deviated from the vehicle's forward path. Based on the speed and direction of the electric bicycle, the query rule base determines that the first prompt strategy is "flashing the blue light of the left side display at a frequency of 3 times per second, and playing the warning voice of 'Please pay attention, the vehicle is turning right' in a loop"; at the same time, it is calculated through sensor data that the electric bicycle is about 8 meters behind the left side of the vehicle, which is a side position. Based on this, the left side display screen and the left side speaker of the vehicle are selected as prompt devices, and then control instructions are sent to them. The left side display screen quickly starts flashing blue light at a high frequency, and the left side speaker plays a warning voice synchronously, reminding the electric bicycle rider to pay attention to avoid and avoid collision accidents.

[0112] The embodiment of the present application compares the calculated target collision risk value with a pre-set first risk threshold (which can be dynamically adjusted according to factors such as road environment and traffic flow). If the target collision risk value is greater than or equal to the first risk threshold, it indicates that there is not only the possibility of a direct collision between the candidate object and the crossing dynamic target, but also a secondary collision due to the change in the motion trajectory of the vehicle after being hit. The comprehensive risk has reached a level that requires active intervention, so the candidate object is judged as a "target object with a collision risk with a crossing dynamic target"; conversely, if the target collision risk value is lower than the first risk threshold, it means that the collision risk between the two is within a controllable range, and subsequent safety prompts or avoidance operations are not triggered for the time being. Through this quantitative comparison mechanism, the vehicle can quickly and accurately screen out high-risk targets, provide a clear basis for real-time decision-making of autonomous driving, and ensure timely response to potential dangers in complex traffic scenarios.

[0113] In addition, by monitoring the target's movement in real time and matching it with a corresponding warning strategy, the system can flexibly adjust the intensity and method of warnings based on the target's speed, direction, and other characteristics, ensuring that the warning measures are appropriate to the level of risk. The system also detects the relative position of the crossing dynamic target and the vehicle, selecting the appropriate warning device accordingly. This allows for precise and targeted delivery of warning information, allowing the warning device to maximize its effectiveness for targets in different directions, avoiding wasted warning resources and ineffective interference. For example, targets in front of the vehicle are alerted using the front-facing display and front horn, while targets to the side are alerted using side devices, effectively improving the pertinence and effectiveness of warnings.

[0114] As a complete example, in a downhill mountain road scenario, a vehicle uses a wide-angle camera mounted on the rear to continuously monitor the video stream in the direction opposite to the vehicle's direction of travel, providing real-time monitoring of the mountain and road conditions behind it. Based on the video stream, a computer vision algorithm is used to identify and track objects in the image. A preset speed of 1m / s is set. If a rock rolling down the mountain is detected exceeding this speed, it is identified as a candidate object.

[0115] Subsequently, the second motion data of the rolling stone is obtained, including information such as the current position, speed, acceleration, and motion trajectory of the rolling stone. At the same time, the third motion data of the vehicle is obtained, such as the vehicle's driving speed, braking status, lane position, etc. It is known that the crossing dynamic target is a pedestrian crossing the road. Based on the second motion data of the rolling stone and the first motion data of the pedestrian, the dynamic model is used to predict the first collision risk value between the rolling stone and the pedestrian; based on the second motion data of the rolling stone and the third motion data of the vehicle, the second collision risk value between the rolling stone and the vehicle is predicted. If the second collision risk value is greater than the set second risk threshold, the motion state of the vehicle after being hit by the rolling stone is further simulated to obtain the fourth motion data. Based on the fourth motion data and the first motion data of the pedestrian, the third collision risk value between the vehicle and the pedestrian after being hit is predicted. The target collision risk value is calculated by combining the first collision risk value and the third collision risk value.

[0116] Next, the calculated target collision risk value is compared with a first risk threshold. If the target collision risk value is greater than or equal to the first risk threshold, the rolling stone is determined to be a target object with a collision risk with a pedestrian crossing the road.

[0117] Once a rockfall is identified as a target object posing a collision risk with a pedestrian crossing the road, the system monitors its movement in real time. As the rock continues to roll, its speed accelerates from an initial 2m / s to 3.5m / s, and its trajectory gradually shifts toward the center of the road. Based on the rock's increasing speed and approaching the vehicle's path, the system queries a pre-defined rule library and obtains the corresponding first warning strategy: "The external display screen flashes red five times per second, and the external speakers play a looped message at 90dB, with a two-second interval between announcements."

[0118] At the same time, sensor data detected a pedestrian crossing the road 8 meters in front of the vehicle, directly in the direction of travel. Based on this relative position, the vehicle's forward-facing display and two speakers on the front of the vehicle were activated as pedestrian safety warning devices. Subsequently, the vehicle, following the first warning strategy, sent a control command to the forward-facing display, causing it to flash red rapidly. Simultaneously, the speakers on the front of the vehicle played a warning voice at a volume of 90 decibels in a loop, reminding the pedestrian to avoid rolling rocks and vehicles, thereby minimizing the risk of a potential collision.

[0119] In the embodiment of the present application, after performing a safety prompt operation on a crossing dynamic target, a dynamic target prompt method is also provided. Figure 5 FIG. 1 is a flow chart of a method for prompting a dynamic target according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:

[0120] Step S301 , monitoring the motion change data after crossing the dynamic target based on the safety prompt operation.

[0121] In this embodiment, after a safety alert is activated for a crossing dynamic target (e.g., a flashing display screen, honking horn, or voice notification), onboard sensors (e.g., lidar, camera, and millimeter-wave radar) are continuously used to collect data on the crossing dynamic target in real time. The lidar emits a laser beam to obtain the target's three-dimensional position, velocity, and acceleration changes; the camera uses computer vision algorithms and target detection and tracking technology to identify the target's displacement and posture changes within the video frame; and the millimeter-wave radar accurately measures the target's radial velocity and range fluctuations.

[0122] These data are fused and processed to calculate the motion change data such as the position offset, speed change rate, and motion direction change angle of the crossing dynamic target after the safety prompt operation.

[0123] Step S302: If the motion change data is lower than the preset data, the object type of the target object is obtained.

[0124] In the embodiment of the present application, the speed change rate threshold, the motion direction change angle threshold, etc. (for example, the position offset must be greater than 1 meter, the speed change rate must exceed 0.5m / s 2 , the angle of change in motion direction is greater than 15°). If the motion change data are all lower than the corresponding preset data, it indicates that the safety prompt operation has not effectively prompted the target to change its motion state, and there is a high risk of collision.

[0125] At this time, the image data collected by the camera is used to classify the target object through a deep learning image recognition model (such as YOLO, ResNet) to identify whether it is a pedestrian, bicycle, motorcycle or other type. At the same time, the data of lidar and millimeter wave radar are combined to assist in judgment and obtain the accurate object type.

[0126] Step S303: Acquire a second prompt strategy corresponding to the object type.

[0127] In the embodiment of the present application, after determining the type of the target object, a pre-established prompt strategy database is queried. This database stores a variety of prompt strategies corresponding to different object types. Each strategy includes parameters such as prompt mode (such as light flashing at different frequencies, horn beeping at different rhythms, and voice broadcasting with different content), prompt intensity (such as voice volume, light brightness level), and prompt duration.

[0128] For example, if the target object is a pedestrian, the corresponding secondary warning strategy might be a high-frequency flashing red light and a loud, looping voice prompt of "Please quickly leave the danger zone." If it is a bicycle, a specific rhythmic horn sound combined with a text prompt might be used. The matching secondary warning strategy is retrieved from the database based on the object type.

[0129] Step S304: performing an avoidance prompt operation on the target object according to the second prompt strategy.

[0130] In this embodiment, the acquired second prompt strategy is converted into specific control instructions and sent to the corresponding off-board device execution unit. For the off-board display, it is controlled to display a warning pattern or text at the frequency set by the strategy; for the off-board speaker, its ringing rhythm and volume are adjusted; and for the voice player, the corresponding voice content is played.

[0131] At the same time, the vehicle's automatic driving system plans and executes the corresponding avoidance path based on the second prompt strategy, combined with the current relative position, speed and other information between the vehicle and the target object. By controlling the vehicle's steering, braking and acceleration systems, it performs deceleration, steering and other operations while ensuring safety, actively avoids the target object, reduces the risk of collision, and ensures driving safety and the safety of traffic participants.

[0132] In this embodiment, a dynamic target prompting device is also provided. The device is used to implement the above-mentioned embodiments and preferred embodiments. The details already described will not be repeated here. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0133] This embodiment provides a dynamic target prompting device, such as Figure 6 As shown, including:

[0134] An acquisition module 601 is configured to acquire a target sequence in a driving environment of a vehicle, wherein the target sequence includes a dynamic target and first motion data of the dynamic target;

[0135] An analysis module 602 is configured to detect the current driving direction of the vehicle and analyze the first motion data to determine whether there is a dynamic target crossing the driving direction in the target sequence;

[0136] A detection module 603 is configured to detect, based on the direction opposite to the traveling direction, whether there is a target object with a collision risk with the crossing dynamic target if there is a crossing dynamic target with a tendency to intersect the traveling direction;

[0137] The sending module 604 is configured to perform a safety prompt operation on the crossing dynamic target if there is a target object that has a collision risk with the crossing dynamic target.

[0138] In an embodiment of the present application, the analysis module 602 is used to predict the motion trajectory of the dynamic target using the first motion data; extract the trajectory direction corresponding to the motion trajectory; compare the trajectory direction with the driving direction of the vehicle to determine whether there is a crossing dynamic target in the target sequence that has a cross-trend with the driving direction.

[0139] In an embodiment of the present application, the device also includes: a trigger module for obtaining point cloud data of the vehicle's driving environment and a road section image of the vehicle's current driving section; using the road section image and point cloud data to analyze the environmental attributes of the driving environment; querying the road section label of the vehicle's road section; based on the environmental attributes and the road section label, analyzing whether the current driving section has a crossing attribute; if the current driving section has a crossing attribute, executing the step of detecting the vehicle's driving direction.

[0140] In an embodiment of the present application, the detection module 603 is used to detect a video stream in the opposite direction of the driving direction; based on the video stream, detect whether there is a candidate object with a driving speed greater than a preset speed in the opposite direction of the vehicle's driving direction; if there is at least one candidate object with a driving speed greater than the preset speed, predict the target collision risk value between the candidate object and the crossing dynamic target; based on the comparison result of the target collision risk value and the first risk threshold, determine whether there is a target object with a collision risk with the crossing dynamic target; wherein, the candidate object with a target collision risk value greater than or equal to the first risk threshold is determined as a target object with a collision risk with the crossing dynamic target.

[0141] In an embodiment of the present application, the detection module 603 is used to obtain the second motion data of the candidate object and the third motion data of the vehicle; based on the second motion data and the first motion data, predict the first collision risk value between the candidate object and the crossing dynamic target; based on the second motion data and the third motion data, predict the second collision risk value between the candidate object and the vehicle; if the second collision risk value is greater than the second risk threshold, predict the fourth motion data of the vehicle after being collided with the candidate object; based on the fourth motion data and the first motion data, predict the third collision risk value of the vehicle with the crossing dynamic target after being collided with the candidate object; based on the first collision risk value and the third collision risk value, calculate the target collision risk value.

[0142] In an embodiment of the present application, the sending module 604 is used to monitor the movement of the target object in real time and obtain a first prompt strategy corresponding to the movement; detect the relative position relationship between the crossing dynamic target and the vehicle; determine the prompt device in the vehicle for providing safety prompts for the crossing dynamic target based on the relative position relationship; and control the prompt device to perform safety prompt operations on the crossing dynamic target according to the first prompt strategy.

[0143] In an embodiment of the present application, the device also includes: a prompt module for monitoring the motion change data of a crossing dynamic target after a safety prompt operation; if the motion change data is lower than the preset data, obtaining the object type of the target object; obtaining a second prompt strategy corresponding to the object type; and performing an avoidance prompt operation on the target object according to the second prompt strategy.

[0144] See also Figure 7 , Figure 7 is a structural diagram of an electronic device provided by an optional embodiment of the present invention, such as Figure 7As shown, the electronic device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system).

[0145] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0146] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0147] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created based on the use of an electronic device presented by a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0148] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0149] The electronic device further includes a communication interface 30 for the electronic device to communicate with other devices or a communication network.

[0150] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0151] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for prompting a dynamic target, characterized in that: The method comprises: Acquire a target sequence in a driving environment of the vehicle, wherein the target sequence includes a dynamic target and first motion data of the dynamic target; detecting the current driving direction of the vehicle and analyzing the first motion data to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction; If there is a crossing dynamic target that has a tendency to intersect the driving direction, detecting whether there is a target object that has a collision risk with the crossing dynamic target based on the opposite direction of the driving direction; If there is a target object that has a collision risk with the crossing dynamic target, a safety prompt operation is performed on the crossing dynamic target.

2. The method according to claim 1, characterized in that The analyzing the first motion data to determine whether there is a crossing dynamic target having a trend crossing the driving direction in the target sequence includes: Predicting a motion trajectory of the dynamic target using the first motion data; Extracting the trajectory direction corresponding to the motion trajectory; The trajectory direction and the driving direction of the vehicle are compared to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction.

3. The method according to claim 1, characterized in that Before detecting the current driving direction of the vehicle, the method further includes: Acquire point cloud data of the driving environment of the vehicle and a road section image of the road section on which the vehicle is currently traveling; Analyzing environmental attributes of the driving environment using the road segment image and the point cloud data; Querying the road section label of the road section where the vehicle is located; Analyzing whether the current driving section has a crossing attribute based on the environmental attributes and the section label; If the current driving section has a crossing attribute, the step of detecting the vehicle's driving direction is executed.

4. The method according to claim 1, wherein The detecting, based on the direction opposite to the driving direction, whether there is a target object with a collision risk with the crossing dynamic target includes: Detecting a video stream in a direction opposite to the driving direction; Based on the video stream, detecting whether there is a candidate object in the opposite direction of the vehicle's travel direction and having a travel speed greater than a preset speed; If there is at least one candidate object whose travel speed is greater than a preset speed, predicting a target collision risk value between the candidate object and the crossing dynamic target; Based on the comparison result of the target collision risk value and the first risk threshold, it is determined whether there is a target object with a collision risk with the crossing dynamic target; wherein, the candidate object whose target collision risk value is greater than or equal to the first risk threshold is determined as the target object with a collision risk with the crossing dynamic target.

5. The method according to claim 4, characterized in that The predicting of a target collision risk value between the candidate object and the traversing dynamic target includes: Acquiring second motion data of the candidate object and third motion data of the vehicle; predicting a first collision risk value between the candidate object and the traversing dynamic target based on the second motion data and the first motion data; predicting a second collision risk value between the candidate object and the vehicle based on the second motion data and the third motion data; If the second collision risk value is greater than a second risk threshold, predicting fourth motion data of the vehicle after being collided with the candidate object; predicting, based on the fourth motion data and the first motion data, a third collision risk value of the vehicle with the traversing dynamic target after being collided with the candidate object; A target collision risk value is calculated based on the first collision risk value and the third collision risk value.

6. The method according to claim 1, characterized in that The performing of a safety prompt operation on the traversing dynamic target includes: monitoring the movement of the target object in real time and obtaining a first prompt strategy corresponding to the movement; detecting a relative positional relationship between the crossing dynamic target and the vehicle; Determining, based on the relative positional relationship, a prompting device in the vehicle for providing a safety prompt to the crossing dynamic object; The prompt device is controlled to perform a safety prompt operation on the crossing dynamic target according to the first prompt strategy.

7. The method according to claim 1, characterized in that After performing a safety prompt operation on the traversing dynamic object, the method further includes: Monitoring the motion change data of the crossing dynamic target after the safety prompt operation; If the motion change data is lower than a preset data, obtaining the object type of the target object; Obtaining a second prompt strategy corresponding to the object type; Perform an avoidance prompt operation on the target object according to the second prompt strategy.

8. A dynamic target prompting device, characterized in that: The device comprises: An acquisition module, configured to acquire a target sequence in a driving environment in which the vehicle is located, wherein the target sequence includes a dynamic target and first motion data of the dynamic target; an analysis module, configured to detect the current driving direction of the vehicle and analyze the first motion data to determine whether there is a crossing dynamic target in the target sequence that has a trend crossing the driving direction; a detection module configured to, if there is a crossing dynamic target having a tendency to intersect the driving direction, detect, based on a direction opposite to the driving direction, whether there is a target object with a collision risk with the crossing dynamic target; The sending module is configured to perform a safety prompt operation on the crossing dynamic target if there is a target object that has a collision risk with the crossing dynamic target.

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images