Rule-based pedestrian target trajectory prediction method, device, equipment and storage medium

Abstract

The present invention belongs to the field of autonomous driving technology, and specifically relates to a rule-based pedestrian target trajectory prediction method, device, equipment and storage medium. The rule-based pedestrian target trajectory prediction device includes a first sensor, a second sensor and a processor, and the processor is used to predict the movement trajectory of the pedestrian target. The prediction method includes: the first sensor collects the historical trajectory of the pedestrian target within a first preset time, and the second sensor collects the driving trajectory of the vehicle within the first preset time; according to a preset algorithm, the movement trend of the pedestrian toward the vehicle is analyzed, the crossing intention of the pedestrian target is judged, and a judgment result is obtained; according to the judgment result, the end point of the pedestrian's movement is predicted, and the predicted trajectory is generated in combination with the historical trajectory information. Its purpose is to collect the historical trajectory of the pedestrian target during the autonomous driving process, and combine it with the current position of the pedestrian target to predict the movement trajectory of the pedestrian target in a reliable manner, so as to provide reliable information for the subsequent operation of the car.

Description

Technical Field

[0001] The present invention belongs to the field of autonomous driving technology, and specifically relates to a rule-based pedestrian target trajectory prediction method, device, equipment and storage medium. Background Art

[0002] Autonomous driving in urban areas often encounters pedestrians, such as those crossing the road, waiting by the roadside, or walking along the curb. When approaching, vehicles are prone to missed or incorrect braking, impacting safety and comfort. Solving this problem relies on predicting the pedestrian's trajectory. For example, if a pedestrian is crossing the road, the predicted trajectory should be a continuous intersection of the road. Decision planning can reference the pedestrian's predicted position several seconds later, allowing for early deceleration and avoidance to ensure safety. Furthermore, after the pedestrian has crossed and the predicted position has significantly moved away, the vehicle can accelerate to optimize the autonomous driving experience.

[0003] Chinese patent CN110606019A discloses a pedestrian courtesy method, device, vehicle, and storage medium for autonomous driving. This patent's technical solution relies solely on the current speed and position of a pedestrian target, recursively extrapolating forward to derive a prediction point using a simple uniform motion model. However, its drawbacks include: first, the current speed of the pedestrian target is often inaccurate due to perception factors, significantly reducing the effectiveness of the prediction; second, it fails to fully utilize the target's historical time series information, predicting only from a single point forward, making stability difficult to guarantee. Summary of the Invention

[0004] The purpose of the present invention is to provide a rule-based pedestrian target trajectory prediction method, device, equipment and storage medium, which collects the historical trajectory of pedestrian targets during autonomous driving, and combines the current position of pedestrian targets to reliably predict the movement trajectory of pedestrian targets, providing reliable information for subsequent operations of the car.

[0005] In order to achieve the above technical objectives, the technical solutions adopted by the present invention are as follows:

[0006] In a first aspect, an embodiment of the present application provides a rule-based pedestrian target trajectory prediction method, which is applied to a rule-based pedestrian target trajectory prediction device, including a first sensor, a second sensor, and a processor, wherein the processor is configured to predict the movement trajectory of a pedestrian target. The prediction method includes:

[0007] The first sensor collects the historical trajectory of the pedestrian target within the first preset time, and the second sensor collects the driving trajectory of the vehicle within the first preset time;

[0008] Analyzing the movement trend of the pedestrian toward the vehicle according to a preset algorithm, determining the pedestrian's intention to cross the road, and obtaining a determination result;

[0009] The pedestrian's movement destination is predicted based on the judgment result, and a predicted trajectory is generated by combining the historical trajectory information.

[0010] In conjunction with the first aspect, in some optional implementations, the first sensor collects the historical trajectory of the pedestrian target within a first preset time, and the second sensor collects the driving trajectory of the vehicle within the first preset time, including:

[0011] A static array is preset to store the historical trajectory, and the historical trajectory point set in the static array is shifted at each moment, and the current moment position point is then loaded into the empty position;

[0012] The historical trajectory point set in the static array is subjected to coordinate transformation with the vehicle as a reference system.

[0013] In conjunction with the first aspect, in some optional implementations, performing coordinate transformation on the historical trajectory point set in the static array with the vehicle as the reference system includes:

[0014] The second sensor synchronously records the vehicle's operating information, calculates the displacement and rotation of the vehicle in each preset time step, determines the vehicle's position, and then progressively offsets the pedestrian's historical trajectory over time, iteratively calculating from the earliest moment to the current moment, thereby achieving coordinate transformation of the historical trajectory point set with the vehicle as the reference system;

[0015] Alternatively, the second sensor records the positioning information of the vehicle, and within each preset time step, determines the position of the vehicle based on the movement of the two positioning position points of the vehicle and the change in the heading angle, and performs coordinate transformation on the historical trajectory point set with the vehicle as the reference system.

[0016] In conjunction with the first aspect, in some optional implementations, analyzing the pedestrian's approaching trend toward the vehicle according to a preset algorithm, determining the pedestrian's intention to cross the road, and obtaining a determination result include:

[0017] Based on the historical trajectory of the pedestrian target, the amplitude of the pedestrian target's movement in a predetermined direction is recorded, and the number of times the pedestrian target approaches the direction in each preset time step is counted; when the number reaches a preset first constant value and the movement amplitude of the pedestrian target exceeds a preset first boundary, it is determined that the pedestrian target has a trend of moving toward the direction.

[0018] In conjunction with the first aspect, in some optional implementations, the method further includes:

[0019] On the basis of determining that the pedestrian target has a moving trend toward a preset own vehicle reference line, and the moving distance of the pedestrian target exceeds a preset second boundary, it is determined that the pedestrian target has an intention to cross the road.

[0020] In conjunction with the first aspect, in some optional implementations, predicting the pedestrian's movement destination based on the judgment result and generating a predicted trajectory in combination with the historical trajectory information includes:

[0021] When determining that a pedestrian intends to cross the road, determining, based on the pedestrian's movement trend, that the pedestrian's movement endpoint is located on the other side of the road after crossing the vehicle;

[0022] Alternatively, when it is determined that the pedestrian target has no intention of crossing the road but has a tendency to move forward, the pedestrian target's movement endpoint is determined to be located at a position in front of a side of the road that does not cross the vehicle.

[0023] In conjunction with the first aspect, in some optional implementations, the method further includes:

[0024] Selecting appropriate points in the historical trajectory point set, the end point, and the current position point of the pedestrian target, and generating a fitting curve using a general cubic curve fitting method;

[0025] A simulation point preset on the fitting curve starts from the current position of the pedestrian target, advances a preset step length along a tangent line at the current position on the fitting curve, arrives at a simulation end position, takes the ordinate at the simulation end position as a first abscissa, and determines the first end position on the fitting curve according to the first abscissa;

[0026] Taking the first terminal position as the starting point, the simulation point repeats the above steps several times, determines a predicted trajectory point set on the fitting curve, and connects the points in the predicted trajectory point set in sequence with a smooth curve to form a predicted trajectory of the pedestrian target.

[0027] In a second aspect, an embodiment of the present application further provides a rule-based pedestrian target trajectory prediction device, which is applied to a rule-based pedestrian target trajectory prediction device, including a first sensor, a second sensor, and a processor, wherein the processor is used to predict the movement trajectory of the pedestrian target, and the device includes:

[0028] Detection unit: used to collect and analyze the historical movement trajectory of pedestrian targets and the driving status information of the vehicle on the road;

[0029] Processing unit: used to determine the movement trend of pedestrian targets based on their historical trajectories, determine the end point of the pedestrian targets based on their movement trends, and generate the predicted trajectory of the pedestrian targets through fitting.

[0030] In a third aspect, an embodiment of the present application also provides a rule-based pedestrian target trajectory prediction device, comprising a first sensor, a second sensor, a processor and a memory, wherein the first sensor and the processor are electrically connected, the second sensor and the processor are electrically connected, the processor is used to predict the movement trajectory of the pedestrian target, and a computer program is stored in the memory. When the computer program is executed by the processor, the rule-based pedestrian target trajectory prediction device performs the above method.

[0031] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program runs on a computer, the computer executes the above method.

[0032] The invention adopting the above technical solution has the following advantages:

[0033] The first sensor senses and collects the historical trajectory information of pedestrian targets on both sides of the road where the autonomous driving vehicle is traveling. The second sensor collects the driving information of the vehicle, and uses the driving vehicle as a reference system to convert and form a historical trajectory point set of the pedestrian target; the pedestrian's movement end point is determined from the pedestrian's historical trajectory, and the pedestrian's historical trajectory information, current position and movement end point are fitted using a general cubic curve fitting method. According to the idea of ​​integration, the predicted trajectory of the pedestrian target is determined on the fitting curve, which is more reliable and the operations made by the vehicle based on the predicted curve are more accurate, avoiding misoperation. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The present invention can be further illustrated by the non-limiting examples given in the accompanying drawings;

[0035] Figure 1 A block diagram of a rule-based pedestrian target trajectory prediction device provided in an embodiment of the present application.

[0036] Figure 2 A flowchart of a rule-based pedestrian target trajectory prediction method provided in an embodiment of the present application.

[0037] Figure 3 A block diagram of a rule-based pedestrian target trajectory prediction device provided in an embodiment of the present application.

[0038] The main component symbols are described as follows:

[0039] 10. Rule-based pedestrian target trajectory prediction device; 11. First sensor; 12. Second sensor; 13. Processor; 200. Rule-based pedestrian target trajectory prediction device; 210. Detection unit; 220. Processing unit. DETAILED DESCRIPTION

[0040] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that in the drawings or descriptions, similar or identical parts are numbered the same. Implementations not shown or described in the drawings are forms known to those of ordinary skill in the art. In addition, directional terms mentioned in the embodiments, such as "upper," "lower," "top," "bottom," "left," "right," "front," and "back," are merely references to the directions in the drawings and are not intended to limit the scope of protection of the present invention.

[0041] like Figure 1 As shown, an embodiment of the present application provides a rule-based pedestrian target trajectory prediction device 10, which may include a first sensor 11, a second sensor 12, a processor 13 and a memory, wherein the first sensor 11 and the processor 13 are electrically connected, and the second sensor 12 and the processor 13 are electrically connected, and the processor 13 is used to predict the movement trajectory of the pedestrian target.

[0042] The first sensor 11 may include, but is not limited to, image sensors and radar sensors, and is used to collect the historical movement trajectory and movement status of pedestrian targets. The second sensor 12 may include, for example, an ultrasonic sensor and a speed sensor, and is used to collect and record the vehicle's driving information.

[0043] In this embodiment, when the autonomous driving vehicle is driving on the road, the first sensor 11 collects historical trajectory information of pedestrian targets on the road, and the second sensor 12 collects driving information of the vehicle. The historical trajectory information of the pedestrian target is converted into a historical trajectory point set in the processor 13 with the vehicle as the reference system. The terminal position of the pedestrian target is determined according to the moving state of the pedestrian target collected by the first sensor 11. A general cubic fitting method is used to form a fitting curve for the historical trajectory point set, the terminal position and the current position of the pedestrian target. A number of predicted trajectory points are determined on the fitting curve by "converting curves into straight lines". The predicted trajectory points are connected in sequence with a smooth curve to obtain a predicted trajectory of the pedestrian target.

[0044] The memory stores a computer program, and when the computer program is executed by the processor 13 , the rule-based pedestrian target trajectory prediction device 10 is enabled to perform corresponding steps in the following rule-based pedestrian target trajectory prediction method.

[0045] like Figure 2 As shown, the present application also provides a rule-based pedestrian target trajectory prediction method. The rule-based pedestrian target trajectory prediction method may include the following steps:

[0046] Step 110: The first sensor 11 collects the historical trajectory of the pedestrian target within the first preset time, and the second sensor 12 collects the driving trajectory of the vehicle within the first preset time;

[0047] Step 120: Analyze the movement trend of the pedestrian toward the vehicle according to a preset algorithm, determine the pedestrian's intention to cross the road, and obtain a determination result;

[0048] Step 130: Predict the pedestrian's movement destination based on the judgment result, and generate a predicted trajectory in combination with the historical trajectory information.

[0049] The historical trajectory of the pedestrian target is acquired through step 110 and converted into a historical trajectory point set with the vehicle as the reference frame. In step 120, the pedestrian target's crossing intention is analyzed through a preset algorithm, and the pedestrian target's movement end point is determined in step 130. According to the pedestrian target's movement end point, historical trajectory points and current position, a general cubic fitting is used to obtain a fitting curve. On the fitting curve, a number of predicted trajectory points are determined by "converting the curve into a straight line". The predicted trajectory points are sequentially connected with a smooth curve to obtain the predicted trajectory of the pedestrian target.

[0050] As an optional embodiment, the first sensor 11 collects the historical trajectory of the pedestrian target within the first preset time, and the second sensor 12 collects the driving trajectory of the vehicle within the first preset time, including:

[0051] A static array is preset to store the historical trajectory, and the historical trajectory point set in the static array is shifted at each moment, and the current moment position point is then loaded into the empty position;

[0052] The historical trajectory point set in the static array is subjected to coordinate transformation with the vehicle as a reference system.

[0053] In this embodiment, when a pedestrian is present on the roadside, the first sensor 11 collects the pedestrian's movement trajectory at each moment and stores it in a static array. After the current moment's trajectory is collected, the previous moment's trajectory point is shifted in the static array, and the current moment's point is inserted into the empty position, forming a historical trajectory point set for the pedestrian. The second sensor 12 collects the vehicle's driving information.

[0054] As an optional implementation, the historical trajectory point set in the static array is subjected to coordinate transformation with the vehicle as a reference system, including:

[0055] The second sensor 12 synchronously records the vehicle's operating information, calculates the displacement and rotation of the vehicle in each preset time step, determines the vehicle's position, and then progressively offsets the pedestrian's historical trajectory over time, iteratively calculating from the earliest moment to the current moment, thereby achieving coordinate transformation of the historical trajectory point set with the vehicle as the reference system;

[0056] Alternatively, the second sensor 12 records the positioning information of the vehicle, and within each preset time step, determines the position of the vehicle based on the movement of the two positioning position points of the vehicle and the change in the heading angle, and transforms the coordinates of the historical trajectory point set with the vehicle as the reference system.

[0057] In this embodiment, the historical trajectory point set collected by the first sensor 11 is converted into the historical trajectory point set of the pedestrian target with the vehicle as the reference frame. The second sensor 12 records the displacement and rotation of the vehicle at each preset time step to determine the vehicle position. The coordinates are then converted based on the position of the pedestrian target at the corresponding time on the historical trajectory. This conversion is performed in chronological order to obtain the historical trajectory point set of the pedestrian target with the vehicle as the reference frame.

[0058] The method for determining the position of the vehicle may also be: the second sensor 12 records the displacement and heading angle change of the vehicle in each preset time step, and determines the position of the vehicle in sequence.

[0059] As an optional implementation, according to a preset algorithm, the pedestrian's approaching trend toward the vehicle is analyzed, and the pedestrian's intention to cross the road is determined, and a determination result is obtained, including:

[0060] Based on the historical trajectory of the pedestrian target, the amplitude of the pedestrian target's movement in a predetermined direction is recorded, and the number of times the pedestrian target approaches the direction in each preset time step is counted; when the number reaches a preset first constant value and the movement amplitude of the pedestrian target exceeds a preset first boundary, it is determined that the pedestrian target has a trend of moving toward the direction.

[0061] In this embodiment, based on the pedestrian's historical trajectory, if the distance of the pedestrian in a predetermined direction changes within each preset time step (for example, if the distance between the pedestrian and the vehicle decreases within a preset time step), the pedestrian approaches the vehicle once. The number of times the pedestrian approaches the predetermined direction is recorded. To avoid misjudgments, the pedestrian's movement amplitude in the predetermined direction is also recorded, and a first threshold is preset based on the movement amplitude. Only when the number of times the pedestrian approaches the predetermined direction reaches a preset first value and the movement amplitude exceeds the first preset threshold is the pedestrian considered to have a movement trend in the predetermined direction.

[0062] Among them, the calibration object of the pedestrian target movement amplitude can be the swing amplitude of the pedestrian's arm, the forward leaning angle of the pedestrian's body, or the movement distance of the pedestrian in a predetermined direction within a preset time step.

[0063] As an optional implementation, the method further includes:

[0064] On the basis of determining that the pedestrian target has a moving trend toward a preset own vehicle reference line, and the moving distance of the pedestrian target exceeds a preset second boundary, it is determined that the pedestrian target has an intention to cross the road.

[0065] As an optional implementation, based on the judgment result, predicting the pedestrian's movement destination, and combining the historical trajectory information to generate a predicted trajectory, includes:

[0066] When determining that a pedestrian intends to cross the road, determining, based on the pedestrian's movement trend, that the pedestrian's movement endpoint is located on the other side of the road after crossing the vehicle;

[0067] Alternatively, when it is determined that the pedestrian target has no intention of crossing the road but has a tendency to move forward, the pedestrian target's movement endpoint is determined to be located at a position in front of a side of the road that does not cross the vehicle.

[0068] The vehicle reference line can be the center line of a lane on the road, or can be the vehicle trajectory generated based on the vehicle's trajectory. The preset second boundary is a pipe boundary formed by expanding a certain distance to both sides of the vehicle's trajectory as the center.

[0069] In this embodiment, if a pedestrian is determined to be moving toward the preset vehicle reference line, the pedestrian's movement distance is compared to see if it exceeds a preset second threshold. If so, the pedestrian is determined to have intended to cross the road. Because the determination of the first threshold has a certain degree of uncertainty, setting a second threshold enhances the robustness of the decision.

[0070] As an optional implementation, the method further includes:

[0071] Selecting appropriate points in the historical trajectory point set, the end point, and the current position point of the pedestrian target, and generating a fitting curve using a general cubic curve fitting method;

[0072] A simulation point preset on the fitting curve starts from the current position of the pedestrian target, advances a preset step length along a tangent line at the current position on the fitting curve, arrives at a simulation end position, takes the ordinate at the simulation end position as a first abscissa, and determines the first end position on the fitting curve according to the first abscissa;

[0073] Taking the first terminal position as the starting point, the simulation point repeats the above steps several times, determines a predicted trajectory point set on the fitting curve, and connects the points in the predicted trajectory point set in sequence with a smooth curve to form a predicted trajectory of the pedestrian target.

[0074] In this embodiment, to enhance the focus's directivity and ensure the fitted curve retains a certain amount of inertia from the historical trajectory, appropriate points are selected from the pedestrian's historical trajectory point set for encryption. Then, through a weighted calculation, these points are fitted to the pedestrian's current location and endpoint to create a fitted curve. On the fitted curve, simulated points are used using the aforementioned "straightening" method to determine a predicted trajectory point set. This set of points is then connected using a smooth curve to obtain the pedestrian's predicted trajectory.

[0075] like Figure 3 As shown, the present application also provides a rule-based pedestrian target trajectory prediction device 200, which includes at least one software function module that can be stored in the form of software or firmware in a storage module or fixed in the operating system (OS) of the rule-based pedestrian target trajectory prediction device 10. The processor 13 is used to execute the executable modules stored in the storage module, such as the software function modules and computer programs included in the rule-based pedestrian target trajectory prediction device 200.

[0076] The rule-based pedestrian trajectory prediction device 200 includes a selection detection unit 210 and a processing unit 220. The functions of each unit may be as follows:

[0077] Processing unit 220: used to collect and analyze the historical movement trajectory of pedestrian targets and the driving status information of the vehicle on the road;

[0078] Processing unit 220: used to determine the movement trend of the pedestrian target based on the historical trajectory of the pedestrian target, determine the movement end point of the pedestrian target based on the movement trend of the pedestrian target, and fit and generate the predicted trajectory of the pedestrian target.

[0079] In this embodiment, the storage module may be, but is not limited to, a random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, etc. In this embodiment, the storage module may be used to store a static array acquired by the first sensor 11, a preset time step, a vehicle reference line, a first boundary, and a second boundary. Of course, the storage module may also be used to store a program, which the processor executes upon receiving an execution instruction.

[0080] It is understandable that Figure 1The structure of the rule-based pedestrian target trajectory prediction device 10 shown in FIG is only a structural diagram. The rule-based pedestrian target trajectory prediction device 10 may also include Figure 1 More components shown. Figure 1 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0081] It should be noted that technical personnel in the relevant field can clearly understand that for the convenience and conciseness of description, the specific working processes of the rule-based pedestrian target trajectory prediction device 10 and the rule-based pedestrian target trajectory prediction device 200 described above can refer to the corresponding processes of each step in the aforementioned method, and will not be elaborated here.

[0082] The present application also provides a computer-readable storage medium that stores a computer program, which, when executed on a computer, causes the computer to execute the rule-based pedestrian trajectory prediction method described in the above embodiment.

[0083] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented through hardware, or can be implemented with the help of software plus a necessary general hardware platform. Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a braking device, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0084] In summary, the embodiments of the present application provide a rule-based pedestrian trajectory prediction method, apparatus, rule-based pedestrian trajectory prediction device 10, and storage medium. In this solution, when a pedestrian is present on the road during autonomous driving, a first sensor 11 collects the pedestrian's historical trajectory information and derives the pedestrian's movement trend. A processor 13 determines the pedestrian's destination based on the pedestrian's movement trend. A fitting curve is derived from the pedestrian's historical trajectory, current position, and destination. The predicted pedestrian trajectory is then further determined based on the fitting curve.

[0085] In the embodiments provided in the present application, it should be understood that the disclosed devices, systems and methods can also be implemented in other ways. The device, system and method embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of code, and a part of the module, program segment or code includes one or more executable instructions for implementing the specified logical function. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, the functional modules in the various embodiments of the present application can be integrated together to form an independent part, or each module can exist alone, or two or more modules can be integrated to form an independent part.

[0086] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A rule-based pedestrian trajectory prediction method, characterized by: A rule-based pedestrian trajectory prediction device includes a first sensor, a second sensor, and a processor. The processor is used to predict the movement trajectory of a pedestrian. The prediction method includes: The first sensor collects the historical trajectory of the pedestrian target within the first preset time, and the second sensor collects the driving trajectory of the vehicle within the first preset time; Analyzing the movement trend of the pedestrian toward the vehicle according to a preset algorithm, determining the pedestrian's intention to cross the road, and obtaining a determination result; Predicting the pedestrian's destination based on the judgment result, and generating a predicted trajectory by combining the historical trajectory information; Analyzing the pedestrian's approaching trend toward the vehicle, determining the pedestrian's intention to cross the road, and obtaining a determination result, including: Based on the historical trajectory of the pedestrian target, the amplitude of the pedestrian target's movement in a predetermined direction is recorded, and the number of times the pedestrian target approaches the predetermined direction in each preset time step is counted; when the number reaches a preset first constant value and the movement amplitude of the pedestrian target exceeds a preset first boundary, it is determined that the pedestrian target has a movement trend approaching the predetermined direction, and the calibration object of the movement amplitude is the swing amplitude of the pedestrian's arm, the forward lean angle of the pedestrian's body, or the distance the pedestrian moves in the predetermined direction in a preset time step; The method further comprises: On the basis of determining that the pedestrian target has a moving trend toward a preset own vehicle reference line, the moving distance of the pedestrian target exceeds a preset second boundary, and it is determined that the pedestrian target has an intention to cross the road.

2. The method according to claim 1, wherein: The first sensor collects the historical trajectory of the pedestrian target within the first preset time, and the second sensor collects the driving trajectory of the vehicle within the first preset time, including: A static array is preset to store the historical trajectory, and the historical trajectory point set in the static array is shifted at each moment, and the current moment position point is then loaded into the empty position; The historical trajectory point set in the static array is subjected to coordinate transformation with the vehicle as a reference system.

3. The method according to claim 2, wherein: The historical trajectory point set in the static array is subjected to coordinate transformation with the vehicle as a reference system, including: The second sensor synchronously records the vehicle's operating information, calculates the displacement and rotation of the vehicle in each preset time step, determines the vehicle's position, and then progressively offsets the pedestrian's historical trajectory over time, iteratively calculating from the earliest moment to the current moment, thereby achieving coordinate transformation of the historical trajectory point set with the vehicle as the reference system; Alternatively, the second sensor records the positioning information of the vehicle, and within each preset time step, determines the position of the vehicle based on the movement of the two positioning position points of the vehicle and the change in the heading angle, and performs coordinate transformation on the historical trajectory point set with the vehicle as the reference system.

4. The method according to claim 1, wherein: Predicting the pedestrian's destination based on the judgment result and generating a predicted trajectory by combining the historical trajectory information includes: When determining that the pedestrian intends to cross the road, determining, based on the pedestrian's movement trend, that the pedestrian's movement endpoint is located on the other side of the road after crossing the vehicle; Alternatively, when it is determined that the pedestrian target has no intention of crossing the road but has a tendency to move forward, the pedestrian target's movement endpoint is determined to be located at a position in front of a side of the road that does not cross the vehicle.

5. The method according to claim 4, characterized in that: The method further comprises: Selecting appropriate points in the historical trajectory point set, the end point, and the current position point of the pedestrian target, and generating a fitting curve using a general cubic curve fitting method; A simulation point preset on the fitting curve starts from the current position of the pedestrian target, advances a preset step length along a tangent line at the current position on the fitting curve, arrives at a simulation end position, takes the ordinate at the simulation end position as a first abscissa, and determines the first end position on the fitting curve according to the first abscissa; Taking the first terminal position as the starting point, the simulation point repeats the above steps several times, determines a predicted trajectory point set on the fitting curve, and connects the points in the predicted trajectory point set in sequence with a smooth curve to form a predicted trajectory of the pedestrian target.

6. A rule-based pedestrian trajectory prediction device, characterized by: A rule-based pedestrian trajectory prediction device includes a first sensor, a second sensor, and a processor. The processor is used to predict the movement trajectory of the pedestrian target. The device includes: Detection unit: used to collect and analyze the historical movement trajectory of pedestrian targets and the driving status information of the vehicle on the road; Processing unit: used to determine the movement trend of pedestrian targets based on their historical trajectories, determine the end point of the pedestrian targets based on their movement trends, and generate the predicted trajectory of the pedestrian targets through fitting.

7. A rule-based pedestrian trajectory prediction device, characterized by: The device includes a first sensor, a second sensor, a processor, and a memory. The first sensor and the processor are electrically connected, the second sensor and the processor are electrically connected, the processor is used to predict the movement trajectory of a pedestrian target, and a computer program is stored in the memory. When the computer program is executed by the processor, the rule-based pedestrian target trajectory prediction device performs the method described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is run on a computer, the computer is caused to execute the method according to any one of claims 1 to 5.

Images