A collision behavior prediction method and device, electronic equipment and medium

Abstract

Embodiments of the present application disclose a collision behavior prediction method and device, electronic equipment and medium. The method comprises: predicting the movement direction and trajectory of a target according to the movement correlation information of the target; wherein the target comprises a first target and a second target; determining the blind area of the target at a prediction time according to the movement direction and trajectory of the target; wherein the prediction time is any time after the current time; determining a collision behavior prediction result according to the positional relationship between the blind area of the first target and the blind area of the second target at the prediction time, and the positional relationship between the blind area of the second target and the blind area of the first target at the prediction time; wherein the blind area of the first target is the blind area corresponding to the first target, and the blind area of the second target is the blind area corresponding to the second target. Through the above technical solution, the collision behavior of the second target and the first target can be accurately predicted, and the timeliness and accuracy of the collision behavior prediction are improved.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation technology, and in particular to a collision behavior prediction method, device, electronic device, and medium. Background Technology

[0002] With rapid economic development, the number and types of vehicles on the road are increasing daily. Blind spots often exist around moving vehicles, making it difficult for drivers to observe them in time. For example, large trucks, buses, and construction vehicles have large bodies that obstruct the driver's view, making it impossible to observe blind spots around the vehicle. This can lead to drivers being unaware of dangerous obstacles, pedestrians, or other vehicles within these blind spots, posing significant safety hazards to both vehicles and pedestrians.

[0003] The current solution is unable to accurately and promptly identify blind spots around the vehicle for timely alerts, resulting in poor timeliness and accuracy. Summary of the Invention

[0004] This application provides a collision behavior prediction method, apparatus, electronic device, and medium to improve the accuracy and timeliness of collision behavior prediction.

[0005] In one embodiment, this application provides a collision behavior prediction method, including:

[0006] Based on the motion association information of the target, the motion direction and trajectory of the target are predicted; wherein the target includes a first target and a second target;

[0007] Based on the target's direction of motion and trajectory, determine the target's blind zone at the predicted time; wherein, the predicted time is any time after the current time;

[0008] The collision behavior prediction result is determined based on the positional relationship between the blind zones of the first and second targets at the prediction time, and the positional relationship between the blind zones of the second and first targets at the prediction time; wherein, the blind zone of the first target is the blind zone corresponding to the first target, and the blind zone of the second target is the blind zone corresponding to the second target.

[0009] In one embodiment, this application provides a collision behavior prediction device, comprising:

[0010] An information prediction module is used to predict the motion direction and trajectory of the target based on the target's motion association information; wherein the target includes a first target and a second target;

[0011] The blind spot determination module is used to determine the blind spot of the target at a predicted time based on the target's direction of motion and trajectory; wherein the predicted time is any time after the current time.

[0012] The prediction result determination module is used to determine the collision behavior prediction result based on the positional relationship between the blind zones of the first target and the second target at the prediction time, and the positional relationship between the blind zones of the second target and the first target at the prediction time; wherein, the blind zone of the first target is the blind zone corresponding to the first target, and the blind zone of the second target is the blind zone corresponding to the second target.

[0013] In one embodiment, this application provides an electronic device, including:

[0014] One or more processors;

[0015] Storage device for storing one or more programs;

[0016] When the one or more programs are executed by the one or more processors, the one or more processors implement the collision behavior prediction method as described in any of the above embodiments.

[0017] In one embodiment, this application also provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the collision behavior prediction method as described in any of the above embodiments.

[0018] This application provides a collision behavior prediction method, apparatus, electronic device, and medium. The implementation involves predicting the motion direction and trajectory of a target based on its motion association information. The target includes a first target and a second target. Based on the target's motion direction and trajectory, the blind zone of the target at the prediction time is determined. The prediction time is any time after the current time. The collision behavior prediction result is determined based on the positional relationship between the blind zones of the first and second targets at the prediction time, and the positional relationship between the blind zones of the second and first targets at the prediction time. The first target blind zone is the blind zone corresponding to the first target, and the second target blind zone is the blind zone corresponding to the second target. This technical solution can accurately predict the collision behavior of the second and first targets, improving the timeliness and accuracy of collision behavior prediction. Attached Figure Description

[0019] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0020] Figure 1A flowchart illustrating a collision behavior prediction method provided in this application embodiment;

[0021] Figure 2 A schematic diagram of vehicle blind spots provided for embodiments of this application;

[0022] Figure 3 A flowchart of another collision behavior prediction method provided in the embodiments of this application;

[0023] Figure 4 A schematic diagram of blind zone intersection provided for an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of a pedestrian blind spot provided in an embodiment of this application;

[0025] Figure 6 At least two pedestrian blind spots are illustrated in the embodiments of this application;

[0026] Figure 7 This is a schematic diagram of the structure of a collision behavior prediction device provided in one embodiment of this application;

[0027] Figure 8 This is a schematic diagram of the structure of an electronic device provided in one embodiment of this application. Detailed Implementation

[0028] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are merely illustrative of the application and not intended to limit it. Furthermore, the embodiments and features described herein can be combined with each other unless otherwise specified. It should also be noted that, for ease of description, only the parts relevant to the present application are shown in the accompanying drawings, not the entire structure.

[0029] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the steps as sequential processes, many of the steps can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the steps can be rearranged. The process can be terminated when its operation is complete, but it may also have additional steps not included in the figures. The process may correspond to a method, function, procedure, subroutine, subroutine, etc. To better understand the embodiments of this application, related technologies are described below.

[0030] Figure 1This is a flowchart illustrating a collision behavior prediction method provided in an embodiment of this application. This embodiment is applicable to situations involving the prediction of collision behavior in traffic. Specifically, the collision behavior prediction method can be executed by a collision behavior prediction device, which can be implemented through software and / or hardware and integrated into an electronic device.

[0031] like Figure 1 As shown, the method specifically includes the following steps:

[0032] S110. Based on the motion association information of the target, predict the motion direction and trajectory of the target; wherein the target includes a first target and a second target.

[0033] The target can be a moving object on the road, including a primary target and a secondary target. For example, the primary target could be a vehicle, and the secondary target a pedestrian. Alternatively, the primary target could be a vehicle, and the secondary target could be another vehicle or another non-motorized vehicle. Motion-related information can be the target's own motion information and / or information about the road where the target is located, which can be obtained through vehicle reporting, road image acquisition devices, radar, positioning systems, etc. The direction of motion is the orientation of the target's movement, and the trajectory can determine the target's location at a future time. The direction of motion can be determined based on the target's speed, i.e., the direction of the speed. The direction of motion can also be determined based on the trajectory; for example, the direction of motion of a target at a trajectory point can be represented by the vector direction pointing from that trajectory point to the next trajectory point.

[0034] In this embodiment, if the target is a vehicle, the motion association information includes at least one of the following: the lane the vehicle is in, the vehicle speed, turn signal information, current time, and traffic light information at the intersection. If the target is a pedestrian, the motion association information includes at least one of the following: the pedestrian's location, the pedestrian's speed, current time, and traffic light information at the intersection. Predicting the target's motion direction and trajectory based on the motion association information includes: inputting the target's motion association information into a classification model to obtain the target's motion direction; and predicting the target's trajectory based on a learning algorithm.

[0035] If the target is a vehicle, at least one of the following can be determined through methods such as vehicle reporting, image capture, positioning system location, and radar location: the vehicle's lane, speed, turn signal information, current time, and traffic light information at the intersection. If the target is a pedestrian, at least one of the following can be determined through methods such as image capture, positioning system location, and radar location: the pedestrian's location, speed, current time, and traffic light information at the intersection. The target's motion association information is input into the classification model to determine the probability of the target's movement direction, thereby determining the target's movement direction, which includes left turn, straight ahead, right turn, and U-turn. Real-time data on target movement on roads in four directions is collected, including target speed and position information, resulting in Vs1, Vs2, Vs3, Vs4, Vl1, Vl2, Vl3, Vl4, Vr1, Vr2, Vr3, Vr4, Vt1, Vt2, Vt3, Vt4, where s represents going straight, l represents turning left, r represents turning right, and t represents making a U-turn. For each road, the relative position and speed of a target entering that road with targets on other roads are determined as environmental variables s = [Dr14R, Dr13S, Dr12L, Vr14R, Vr13S, Vr12L], where Dr represents the relative position relationship, Vr represents the relative speed, 1 indicates the currently analyzed road, 2 indicates a road from which a left or right turn is made, 3 indicates a road entered straight from that road, and 4 indicates the remaining roads. For example, road network information can be used to determine the number of lanes, speed limits, turning distances, and other information for each road, facilitating subsequent prediction of the target's movement direction and trajectory. External factors affecting road traffic conditions, such as weather, traffic conditions, and road conditions, can be obtained through target reporting or proactive queries. The influencing factors of each factor can then be determined, and the actual maximum speed limit of the road can be calculated. Where Vmax is the theoretical maximum speed limit of the road, and φ k Let be the influencing factor of the k-th influencing factor, and n be the number of influencing factors. Based on the above information and a learning algorithm, such as Q-learning, the motion trajectory of the target is predicted according to its motion association information. A reward function is used to fit the current traffic flow speed; the closer the traffic flow speed is to V′max, the larger the reward value, thus obtaining the target's motion trajectory.

[0036] S120. Determine the blind zone of the target at the predicted time based on the target's direction of motion and trajectory; wherein the predicted time is any time after the current time.

[0037] For example, at road intersections, oncoming traffic frequently passes each other, allowing for the prediction of blind spots for targets at intersections. After determining the target's trajectory, its position at the prediction time can be determined. Based on the target's direction of movement and position, its blind spot can be identified. The target's blind spot can be caused by occlusion by other targets, it can be a blind spot behind the target, or it can be a blind spot determined by the target type, such as... Figure 2 As shown, areas A, B, and C are semi-blind spots. Due to the high height of the driver's cab, the driver can only observe objects at the same height as the windows, making it difficult to observe objects below the windows; therefore, they are semi-blind spots. Area D is a complete blind spot, where the driver has difficulty observing objects. Based on the target's direction and trajectory of movement, the specific area of ​​the target's blind spot on the actual road at the predicted time can be determined, facilitating subsequent analysis.

[0038] S130. Based on the positional relationship between the blind zones of the first target and the second target at the prediction time, and the positional relationship between the blind zones of the second target and the first target at the prediction time, determine the collision behavior prediction result; wherein, the blind zone of the first target is the blind zone corresponding to the first target, and the blind zone of the second target is the blind zone corresponding to the second target.

[0039] The positions of the first and second targets can be determined based on their movement trajectories. The blind spots of the first and second targets can be represented using a Cartesian coordinate system centered at the road intersection, thus defining their respective areas. Alternatively, the areas of the first and second target blind spots can be represented using latitude and longitude ranges.

[0040] For example, when predicting whether a collision is possible between a first target and a second target, the positional relationship between their blind spots and the blind spots of the first and second targets can be used to determine this possibility. For instance, if the first target is within the blind spot of the second target, there is a possibility of it colliding with the first target relative to the second target. If the second target is also within the blind spot of the first target, and the first target cannot detect its presence in time or avoid it, there is a possibility of it colliding with the second target; in this case, the probability of a collision between the first and second targets is highest. If the second target is not within the blind spot of the first target, meaning the first target can see the second target, it can avoid the second target, reducing the probability of a collision. If neither the first nor the second target is within the blind spot of the second target, the probability of a collision between the first and second targets is lowest. The probability of a collision can be represented by preset data; for example, 10 represents the highest probability of a collision, 8 represents a decrease in the probability relative to the highest probability, and 1 represents the lowest probability. Based on the above analysis results, the predicted collision behavior is determined.

[0041] This application provides a collision behavior prediction method, apparatus, electronic device, and medium. The implementation involves predicting the motion direction and trajectory of a target based on its motion association information. The target includes a first target and a second target. Based on the target's motion direction and trajectory, the blind zone of the target at the prediction time is determined. The prediction time is any time after the current time. The collision behavior prediction result is determined based on the positional relationship between the blind zones of the first and second targets at the prediction time, and the positional relationship between the blind zones of the second and first targets at the prediction time. The first target blind zone is the blind zone corresponding to the first target, and the second target blind zone is the blind zone corresponding to the second target. This technical solution can accurately predict the collision behavior of the second and first targets, improving the timeliness and accuracy of collision behavior prediction.

[0042] Figure 3 This is a flowchart illustrating another collision behavior prediction method provided in this application embodiment. This embodiment is an optimization based on the above embodiments. It should be noted that technical details not described in detail in this embodiment can be found in any of the above embodiments.

[0043] Specifically, such as Figure 3 As shown, the method specifically includes the following steps:

[0044] S210. Based on the motion association information of the target, predict the motion direction and trajectory of the target; wherein the target includes a first target and a second target.

[0045] S220. Based on the target's direction of motion and trajectory, determine the target's blind zone at the predicted time; wherein the predicted time is any time after the current time.

[0046] In this embodiment of the application, if the target is a vehicle, the blind zone of the target at the predicted time is determined based on the target's direction of motion and trajectory, including: determining the overlap level of the blind zone based on the overlap of the blind zones of at least two targets; wherein, the more overlapping blind zones there are, the higher the overlap level of the overlapping area of ​​the blind zone; determining the danger level of the blind zone based on the collision results of historical collision behaviors in the blind zone; and determining the thermal level of the blind zone based on the overlap level and the danger level.

[0047] For example, at the prediction time, each vehicle located at the road intersection is identified, and the blind spot of each vehicle is determined. The blind spots of each vehicle are intersected to determine the overlapping area of ​​the blind spots. The more overlapping blind spots there are, the higher the overlap level of that area is determined. Figure 4 As shown, area A has the most overlapping blind spots among multiple vehicles, therefore its overlap level is the highest, represented by 100. Area B has fewer overlapping blind spots than area A, and its overlap level is lower, represented by 80. Area C has fewer overlapping blind spots than area B, and its overlap level is lower, represented by 60. Area D has fewer overlapping blind spots than area C, or there are no overlapping blind spots, represented by 40.

[0048] Historical collision events at this intersection are acquired, and the hazard level of the blind spot is determined based on the collision results. For example, the vehicle type, speed, direction of movement, and collision location of each historical collision are identified to pinpoint the corresponding blind spot. The hazard level is then determined based on the injury details in the collision results. As shown in Table 1, each row represents a historical collision event, and the data corresponding to the first and second injury scenarios represent their probabilities of occurrence. We can set the additive coefficient for the first injury scenario to α, the additive coefficient for the second injury scenario to β, the probability of the first injury scenario to d, and the probability of the second injury scenario to h. Then, the hazard level of this historical collision event is d*α + h*β, and this hazard level is used as the hazard level of the blind spot where the historical collision event occurred. The overlap level of the blind spot is then superimposed with the hazard level to obtain the thermal level of the blind spot.

[0049] Table 1

[0050]

[0051]

[0052] In this embodiment of the application, if the first target is a vehicle and the second target is a pedestrian, then the blind zone of the target at the prediction time is determined according to the direction and trajectory of the target's movement, including: if there are at least two second targets, then the overlap level of the blind zones of each second target at the prediction time is determined according to the overlap of the blind zones of the at least two second targets.

[0053] For example, such as Figure 5 As shown, Figure 5 The fan-shaped area in the diagram represents the pedestrian's field of vision, and the area within this field of vision obstructed by large vehicles is the pedestrian blind spot. If there are at least two pedestrians, such as... Figure 6 As shown, pedestrian 2 can provide pedestrian 1 with a certain field of vision, and pedestrian 2 will also provide pedestrian 1 with a certain warning based on their own field of vision. Therefore, if there are at least two second targets, the blind spots of each of the at least two second targets are superimposed, and the overlap level of the blind spots of each second target is determined according to the overlap. For example, the overlap level of the shared blind spot of pedestrian 1 and pedestrian 2 is the highest, represented by 100. The overlap level of the area belonging to the blind spot of pedestrian 1 but not to the blind spot of pedestrian 2 is the second lowest, represented by 80.

[0054] In this embodiment of the application, determining the blind zone of the target at the predicted time based on the target's direction of motion and trajectory includes: determining the initial level of the second target blind zone based on the number of historical collision behaviors in the second target blind zone; determining the weighted level of the second target blind zone based on the type of the first target located in the second target blind zone; determining the danger level of the second target blind zone based on the initial level and the weighted level; and determining the thermal level of the second target blind zone based on the overlap level and danger level of the second target blind zone.

[0055] The information on the number of historical collisions in the second target blind spot can be the proportion of historical collisions that occurred in that area, such as the ratio of the number of historical collisions involving a single pedestrian in their own blind spot to the total number of historical collisions that occurred in that area. The higher this ratio, the higher the initial rating. For example, as shown in Table 2, if the vehicle type affecting a pedestrian's vision in their own blind spot is a large truck, the weighted rating is higher; if the vehicle type is a car, the weighted rating is lower. The initial rating and the weighted rating are then combined, for example, by multiplying the initial rating by the weighted rating, to obtain the hazard level of the second target blind spot.

[0056] Table 2

[0057] Blind spot type Initial level Vehicle types in blind spots Weighted levels Hazard level Blind spots 50 large trucks 50 2500 Blind spots 50 Car 30 1500 Common blind spots among companions 50 large trucks 20 1000 Blind spots that can be seen by companions 30 large trucks 20 600 Blind spots that can be seen by companions 30 motorcycle 10 300 … … … … …

[0058] S230. If, at the prediction time, the first target is located in the region with the highest thermal level in the blind zone of the second target, and the second target to which the blind zone of the second target belongs is located in the region with the highest thermal level in the blind zone of the first target corresponding to the first target, then the collision behavior prediction result is determined to be that a collision will occur.

[0059] For example, if the first target is located in the area with the highest thermal intensity in the blind zone of the second target, it is determined that the first target is more likely to collide with the second target. If the second target is also located in the area with the highest thermal intensity in the blind zone of the first target, it is determined that the first target is most likely to collide with the second target, and the collision behavior prediction result is determined to be that a collision will occur.

[0060] S240. Determine the collision time based on the predicted time and the preset avoidance time.

[0061] The preset evacuation time can be determined based on the actual situation, such as the pedestrian's reaction time.

[0062] S250. Generate early warning information and send the early warning information to the first target and / or the second target to remind the first target and / or the second target that a collision is predicted to occur at the time of the collision.

[0063] For example, a warning message can be sent to the first target and / or the second target to remind them that a collision may occur at the moment of collision, and to remind them to take evasive action.

[0064] In this embodiment, if the first target is located in the area with the highest thermal intensity within the blind zone of the second target, and the second target is also located in the area with the highest thermal intensity within the blind zone of the first target, then warning messages can be sent to both the first and second targets respectively. If the first target is located in the area with the highest thermal intensity within the blind zone of the second target, but the second target is not located in the area with the highest thermal intensity within the blind zone of the first target, or if the second target is not located within the blind zone of the first target, then only a warning message can be sent to the second target.

[0065] This application provides a collision behavior prediction method. If, at the prediction time, the first target is located in the area with the highest thermal intensity within the blind zone of the second target, and the second target to which the blind zone belongs is located in the area with the highest thermal intensity within the blind zone of the first target corresponding to the first target, then the collision behavior prediction result is determined to be a collision. Based on the prediction time and a preset avoidance time, the collision time is determined; a warning message is generated and sent to the first target and / or the second target to remind them that a collision is predicted to occur at the predicted collision time. This method can predict the occurrence of collision behavior in a timely and accurate manner, and by generating a timely warning message, remind the first and second targets to avoid collisions.

[0066] Figure 7 This is a schematic diagram of a collision behavior prediction device provided in one embodiment of this application. The collision behavior prediction device provided in this embodiment includes:

[0067] The information prediction module 310 is used to predict the motion direction and trajectory of the target based on the motion association information of the target; wherein the target includes a first target and a second target;

[0068] The blind spot determination module 320 is used to determine the blind spot of the target at a predicted time based on the target's direction of motion and trajectory; wherein the predicted time is any time after the current time.

[0069] The prediction result determination module 330 is used to determine the collision behavior prediction result based on the positional relationship between the blind zones of the first target and the second target at the prediction time, and the positional relationship between the blind zones of the second target and the first target at the prediction time; wherein, the first target blind zone is the blind zone corresponding to the first target, and the second target blind zone is the blind zone corresponding to the second target.

[0070] In this embodiment of the application, if the target is a vehicle, the blind spot determination module 320 includes:

[0071] The first overlap level determination unit is used to determine the overlap level of the blind zone based on the overlap of the blind zones of at least two targets; wherein, the more overlapping blind zones there are, the higher the overlap level of the overlapping area of ​​the blind zone.

[0072] The first hazard level determination unit is used to determine the hazard level of the blind zone based on the collision results of historical collision behaviors in the blind zone;

[0073] The first thermal level determination unit is used to determine the thermal level of the blind zone based on the overlap level and the danger level.

[0074] In this embodiment of the application, if the first target is a vehicle and the second target is a pedestrian, the blind spot determination module 320 includes:

[0075] The second overlap level determination unit is used to determine the overlap level of the second target blind zone of each second target at the prediction time based on the overlap of the second target blind zones of the at least two second targets if there are at least two second targets.

[0076] In this embodiment of the application, the blind spot determination module 320 includes:

[0077] An initial level determination unit is used to determine the initial level of the second target blind zone based on the quantity information of historical collision behaviors in the second target blind zone;

[0078] The weighting level determination unit is used to determine the weighting level of the second target blind zone based on the type of the first target located in the second target blind zone;

[0079] The second hazard level determination unit is used to determine the hazard level of the second target blind zone based on the initial level and the weighted level.

[0080] The second thermal level determination unit is used to determine the thermal level of the second target blind zone based on the overlap level and danger level of the second target blind zone.

[0081] In this embodiment of the application, the prediction result determination module 330 is specifically used for:

[0082] If, at the prediction time, the first target is located in the region with the highest thermal intensity in the blind zone of the second target, and the second target to which the blind zone of the second target belongs is located in the region with the highest thermal intensity in the blind zone of the first target corresponding to the first target, then the collision behavior prediction result is determined to be that a collision will occur.

[0083] In this embodiment of the application, the device further includes:

[0084] The collision time determination module is used to determine the collision time based on the predicted time and the preset avoidance time.

[0085] The warning information generation module is used to generate warning information and send the warning information to the first target and / or the second target to remind the first target and / or the second target that a collision is predicted to occur at the time of the collision.

[0086] In this embodiment of the application, if the target is a vehicle, the motion association information includes at least one of the following: the lane the vehicle is in, the vehicle speed, turn signal information, current time, and traffic light information at the intersection. If the target is a pedestrian, the motion association information includes at least one of the following: the pedestrian's location, the pedestrian's speed, current time, and traffic light information at the intersection.

[0087] Information prediction module 310 includes:

[0088] A motion direction determination unit is used to input the motion association information of the target into a classification model to obtain the motion direction of the target;

[0089] The motion trajectory determination unit is used to predict the motion trajectory of the target based on the Q-learning algorithm and the motion association information of the target.

[0090] The collision behavior prediction device provided in this application embodiment can be used to execute the collision behavior prediction method provided in any of the above embodiments, and has corresponding functions and beneficial effects.

[0091] Figure 8 This is a schematic diagram of an electronic device provided according to one embodiment of the present application. The electronic device 50 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 50 may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, user equipment, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present application described and / or claimed herein.

[0092] like Figure 8 As shown, the electronic device 50 includes at least one processor 510 and a memory, such as a read-only memory (ROM) 520 or a random access memory (RAM) 530, communicatively connected to the at least one processor 510. The memory stores computer programs executable by the at least one processor. The processor 510 can perform various appropriate actions and processes based on the computer program stored in the ROM 520 or loaded into the RAM 530 from storage unit 580. The RAM 530 may also store various programs and data required for the operation of the electronic device 50. The processor 510, ROM 520, and RAM 530 are interconnected via a bus 540. An input / output (I / O) interface 550 is also connected to the bus 540.

[0093] Multiple components in electronic device 50 are connected to I / O interface 550, including: input unit 560, such as keyboard, mouse, etc.; output unit 570, such as various types of monitors, speakers, etc.; storage unit 580, such as disk, optical disk, etc.; and communication unit 590, such as network card, modem, wireless transceiver, etc. Communication unit 590 allows electronic device 50 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks and wireless networks.

[0094] Processor 510 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 510 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 510 performs the various methods and processes described above, such as collision behavior prediction methods.

[0095] In some embodiments, the collision behavior prediction method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 580. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 50 via ROM 520 and / or communication unit 590. When the computer program is loaded into RAM 530 and executed by processor 510, one or more steps of the method described above may be performed. Alternatively, in other embodiments, processor 510 may be configured to perform the collision behavior prediction method by any other suitable means (e.g., by means of firmware).

[0096] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0097] Computer programs used to implement the methods of this application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0098] In the context of this application, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0099] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device 50, which includes: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device 50. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0100] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0101] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0102] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this application can be achieved, and this is not limited herein.

[0103] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A collision behavior prediction method, characterized in that, The method includes: Based on the motion association information of the target, the motion direction and trajectory of the target are predicted; wherein the target includes a first target and a second target; Based on the target's direction of motion and trajectory, determine the target's blind zone at the predicted time; wherein, the predicted time is any time after the current time; The collision behavior prediction result is determined based on the positional relationship between the blind zones of the first and second targets at the prediction time, and the positional relationship between the blind zones of the second and first targets at the prediction time; wherein, the blind zone of the first target is the blind zone corresponding to the first target, and the blind zone of the second target is the blind zone corresponding to the second target. If the target is a vehicle, then based on the target's direction of motion and trajectory, the blind zone of the target at the predicted time is determined, including: The overlap level of the blind zones is determined based on the overlap of the blind zones of at least two targets; wherein, the more overlapping blind zones there are, the higher the overlap level of the overlapping areas of the blind zones. The danger level of the blind zone is determined based on the collision results of historical collision behaviors in the blind zone; The thermal level of the blind zone is determined based on the overlap level and the hazard level. If the first target is a vehicle and the second target is a pedestrian, then based on the direction and trajectory of the target's movement, the blind zone of the target at the predicted time is determined, including: If there are at least two second targets, then the overlap level of the second target blind zone of each second target at the prediction time is determined based on the overlap of the second target blind zones of the at least two second targets. Based on the target's direction of motion and trajectory, determine the target's blind zone at the predicted time, including: The initial level of the second target blind zone is determined based on the number of historical collision behaviors in the second target blind zone; The weighting level of the second target blind zone is determined based on the type of the first target located in the second target blind zone; The danger level of the second target blind zone is determined based on the initial level and the weighted level. The thermal level of the second target blind zone is determined based on the overlap level and hazard level of the second target blind zone.

2. The method according to claim 1, characterized in that, Based on the positional relationship between the blind zones of the first and second targets at the prediction time, and the positional relationship between the blind zones of the second and first targets at the prediction time, the collision behavior prediction result is determined, including: If, at the prediction time, the first target is located in the region with the highest thermal intensity in the blind zone of the second target, and the second target to which the blind zone of the second target belongs is located in the region with the highest thermal intensity in the blind zone of the first target corresponding to the first target, then the collision behavior prediction result is determined to be that a collision will occur.

3. The method according to claim 2, characterized in that, After determining that a collision will occur based on the collision behavior prediction, the method further includes: The collision time is determined based on the predicted time and the preset avoidance time; A warning message is generated and sent to the first target and / or the second target to remind the first target and / or the second target that a collision is predicted to occur at the time of the collision.

4. The method according to claim 1, characterized in that, If the target is a vehicle, the motion association information includes at least one of the following: the lane the vehicle is in, the vehicle speed, turn signal information, current time, and traffic light information at the intersection. If the target is a pedestrian, the motion association information includes at least one of the following: the pedestrian's location, the pedestrian's speed, current time, and traffic light information at the intersection. Based on the target's motion association information, predict the target's motion direction and trajectory, including: The motion association information of the target is input into the classification model to obtain the motion direction of the target; Based on the learning algorithm, the motion trajectory of the target is predicted according to the motion association information of the target.

5. A collision behavior prediction device, characterized in that, The device includes: An information prediction module is used to predict the motion direction and trajectory of the target based on the target's motion association information; wherein the target includes a first target and a second target; The blind spot determination module is used to determine the blind spot of the target at a predicted time based on the target's direction of motion and trajectory; wherein the predicted time is any time after the current time. The prediction result determination module is used to determine the collision behavior prediction result based on the positional relationship between the blind zones of the first target and the second target at the prediction time, and the positional relationship between the blind zones of the second target and the first target at the prediction time; wherein, the first target blind zone is the blind zone corresponding to the first target, and the second target blind zone is the blind zone corresponding to the second target; If the target is a vehicle, the blind spot determination module includes: The first overlap level determination unit is used to determine the overlap level of the blind zone based on the overlap of the blind zones of at least two targets; wherein, the more overlapping blind zones there are, the higher the overlap level of the overlapping area of ​​the blind zone. The first hazard level determination unit is used to determine the hazard level of the blind zone based on the collision results of historical collision behaviors in the blind zone; The first thermal level determination unit is used to determine the thermal level of the blind zone based on the overlap level and the danger level. If the first target is a vehicle and the second target is a pedestrian, the blind spot determination module includes: The second overlap level determination unit is used to determine the overlap level of the second target blind zone of each second target at the prediction time based on the overlap of the second target blind zones of the at least two second targets if there are at least two second targets. The blind spot determination module includes: An initial level determination unit is used to determine the initial level of the second target blind zone based on the quantity information of historical collision behaviors in the second target blind zone; The weighting level determination unit is used to determine the weighting level of the second target blind zone based on the type of the first target located in the second target blind zone; The second hazard level determination unit is used to determine the hazard level of the second target blind zone based on the initial level and the weighted level. The second thermal level determination unit is used to determine the thermal level of the second target blind zone based on the overlap level and danger level of the second target blind zone.

6. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the collision behavior prediction method as described in any one of claims 1-4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the collision behavior prediction method as described in any one of claims 1-4.

Images