Methods, devices, vehicles, and storage media for predicting pedestrian intentions

Abstract

This application relates to the field of autonomous driving technology, and particularly to a method, device, vehicle, and storage medium for predicting pedestrian intentions. The method includes: collecting at least one action data point of at least one pedestrian in the current environment; obtaining at least one pedestrian group using the at least one action data point; calculating the center lateral velocity and boundary lateral velocity of the at least one pedestrian group based on the at least one action data point; obtaining an interaction expansion coefficient of the at least one pedestrian group from the center lateral velocity and boundary lateral velocity; and calculating the target lateral velocity of the at least one pedestrian group based on the interaction expansion coefficient, so as to predict the pedestrian intention of at least one pedestrian using the target lateral velocity. Embodiments of this application can aggregate multiple pedestrians to obtain a pedestrian group, and calculate the interaction expansion coefficient generated by the interaction of each pedestrian based on the motion characteristics of the pedestrian group, to obtain the predicted pedestrian intention and control the vehicle to avoid collisions, thereby improving the safety and reliability of vehicle driving.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a method, device, vehicle, and storage medium for predicting pedestrian intentions. Background Technology

[0002] As autonomous driving functions become more sophisticated, methods for predicting pedestrian movement on roads also increase. Accurately predicting pedestrian movement helps autonomous vehicles choose the correct route, avoiding collisions or accidental braking that could negatively impact the passenger experience. Among related technologies, pedestrian behavior prediction can predict a pedestrian's trajectory at a future time based on their movement patterns.

[0003] However, in related technologies, pedestrians have a high degree of freedom and are flexible in their movements. They are easily affected by sudden changes in speed due to the surrounding environment, which reduces the accuracy of predicting the future movement trajectory of pedestrians. In actual traffic scenarios, multiple pedestrians have behavioral needs such as obstacle avoidance and interaction, making it difficult to capture the overall characteristics of multiple pedestrians. This reduces the reliability of pedestrian trajectory prediction during autonomous driving and lowers the safety level of vehicle driving, which urgently needs to be addressed. Summary of the Invention

[0004] This application provides a method, device, vehicle, and storage medium for predicting pedestrian intentions, in order to solve the problems in related technologies, such as the high degree of freedom and flexible movement of pedestrians, which makes them susceptible to sudden changes in speed due to surrounding conditions, thus reducing the accuracy of predicting the future movement trajectory of pedestrians. Furthermore, in actual traffic scenarios, due to the need for obstacle avoidance and interaction among multiple pedestrians, it is difficult to capture the overall features of multiple pedestrians, which reduces the reliability of pedestrian trajectory prediction during autonomous driving and lowers the safety level of vehicle driving.

[0005] The first aspect of this application provides a method for predicting pedestrian intentions, comprising the following steps: collecting at least one action data of at least one pedestrian in the current environment, and using the at least one action data to obtain at least one pedestrian group;

[0006] The center lateral velocity and boundary lateral velocity of the at least one pedestrian group are calculated based on the at least one motion data, and the interaction expansion coefficient of the at least one pedestrian group is obtained from the center lateral velocity and the boundary lateral velocity.

[0007] The target lateral velocity of the at least one pedestrian group is calculated based on the interaction expansion coefficient, and the pedestrian intention of the at least one pedestrian is predicted using the target lateral velocity.

[0008] Optionally, in one embodiment of this application, obtaining at least one pedestrian group using the at least one action data includes: calculating the similarity of any pedestrian among all pedestrians based on the at least one action data, and extracting at least one pedestrian who meets a preset similarity condition from all pedestrians based on the similarity; and aggregating all pedestrians who meet the preset similarity condition to obtain the at least one pedestrian group.

[0009] Optionally, in one embodiment of this application, before predicting the pedestrian intention of the at least one pedestrian using the target lateral velocity, the method further includes: detecting whether the at least one pedestrian group meets a preset crossing condition; if the at least one pedestrian group does not meet the preset crossing condition, suppressing the interaction expansion coefficient until the at least one pedestrian group meets the preset crossing condition, and updating the target lateral velocity using the suppressed interaction expansion coefficient.

[0010] Optionally, in one embodiment of this application, predicting the pedestrian intention of the at least one pedestrian using the target lateral velocity includes: identifying the actual position of the at least one pedestrian group; and generating the pedestrian intention based on the actual position and the target lateral velocity.

[0011] Optionally, in one embodiment of this application, generating the pedestrian intention based on the actual location and the target lateral velocity includes: calculating the road travel time of the at least one pedestrian group based on the target lateral velocity when the actual location is in a preset pedestrian crossing area; constructing a first target obstacle based on the road travel time and the actual pedestrian crossing; and generating the pedestrian intention from the first target obstacle.

[0012] Optionally, in one embodiment of this application, generating the pedestrian intention based on the actual position and the target lateral velocity includes: when the actual position is in a preset non-pedestrian crossing area, determining whether the target lateral velocity is greater than or equal to a preset threshold; if the target lateral velocity is greater than or equal to the preset threshold, constructing a second target obstacle based on the target lateral velocity, and generating the pedestrian intention from the second target obstacle; otherwise, determining that the pedestrian intention is no intention to cross.

[0013] A second aspect of this application provides a pedestrian intention prediction device, comprising: a data acquisition module for acquiring at least one action data of at least one pedestrian in the current environment, and obtaining at least one pedestrian group using the at least one action data; a calculation module for calculating the center lateral velocity and boundary lateral velocity of the at least one pedestrian group based on the at least one action data, and obtaining the interaction expansion coefficient of the at least one pedestrian group from the center lateral velocity and the boundary lateral velocity; and a prediction module for calculating the target lateral velocity of the at least one pedestrian group based on the interaction expansion coefficient, so as to predict the pedestrian intention of the at least one pedestrian using the target lateral velocity.

[0014] Optionally, in one embodiment of this application, the acquisition module includes: an extraction unit, configured to calculate the similarity of any pedestrian among all pedestrians based on the at least one action data, and extract at least one pedestrian who meets a preset similarity condition from all pedestrians based on the similarity; and an aggregation unit, configured to aggregate all pedestrians that meet the preset similarity condition to obtain the at least one pedestrian group.

[0015] Optionally, in one embodiment of this application, it further includes: a detection module, configured to detect whether the at least one pedestrian group meets a preset crossing condition before predicting the pedestrian intention of the at least one pedestrian using the target lateral velocity; and a suppression module, configured to suppress the interaction expansion coefficient when it is detected that the at least one pedestrian group does not meet the preset crossing condition, until the at least one pedestrian group meets the preset crossing condition, and update the target lateral velocity using the suppressed interaction expansion coefficient.

[0016] Optionally, in one embodiment of this application, the prediction module includes: an identification unit for identifying the actual position of the at least one pedestrian group; and a generation unit for generating the pedestrian intention based on the actual position and the target lateral velocity.

[0017] Optionally, in one embodiment of this application, the generation unit is specifically used to: calculate the road travel time of the at least one pedestrian group based on the target lateral speed when the actual location is in a preset pedestrian crossing area; construct a first target obstacle based on the road travel time and the actual pedestrian crossing; and generate the pedestrian intention from the first target obstacle.

[0018] Optionally, in one embodiment of this application, the generation unit is specifically used to: determine whether the target lateral speed is greater than or equal to a preset threshold when the actual location is in a preset non-pedestrian crossing area; if the target lateral speed is greater than or equal to the preset threshold, construct a second target obstacle based on the target lateral speed, and generate the pedestrian intention from the second target obstacle; otherwise, determine that the pedestrian intention is no intention to cross.

[0019] A third aspect of this application provides a vehicle including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the pedestrian intention prediction method as described in the above embodiments.

[0020] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the pedestrian intention prediction method described above.

[0021] A fifth aspect of this application provides a computer program that, when executed, implements the pedestrian intention prediction method described above.

[0022] This application's embodiments can aggregate multiple pedestrians to form a pedestrian group, and calculate the interaction expansion coefficient generated by the interaction of each pedestrian based on the movement characteristics of the pedestrian group. This yields a prediction of the pedestrian's intentions, which is then used to control the vehicle to avoid obstacles, thereby improving the safety and reliability of vehicle driving. This solves the problems in related technologies, such as the high degree of freedom and flexible movement of pedestrians, which makes them susceptible to sudden speed changes due to surrounding conditions, leading to decreased accuracy in predicting future pedestrian trajectories. Furthermore, in real-world traffic scenarios, the need for obstacle avoidance and interaction among multiple pedestrians makes it difficult to capture the overall characteristics of all pedestrians, reducing the reliability of pedestrian trajectory prediction during autonomous driving and lowering the safety level of vehicle driving.

[0023] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. Attached Figure Description

[0024] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0025] Figure 1 A flowchart illustrating a method for predicting pedestrian intent according to an embodiment of this application;

[0026] Figure 2 This is a schematic diagram of pedestrian clustering in one embodiment of this application;

[0027] Figure 3 This is a schematic diagram of a pedestrian group crossing a crosswalk according to an embodiment of this application;

[0028] Figure 4 This is a schematic diagram of a pedestrian group crossing a road according to an embodiment of this application;

[0029] Figure 5 This is a schematic diagram of the structure of a pedestrian intention prediction device according to an embodiment of this application;

[0030] Figure 6 This is a structural schematic diagram of a vehicle according to an embodiment of this application. Detailed Implementation

[0031] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0032] The following description, with reference to the accompanying drawings, outlines a method, apparatus, vehicle, and storage medium for predicting pedestrian intentions according to embodiments of this application. Addressing the issues raised in the background section regarding pedestrians' high degrees of freedom and flexible movement, which make them susceptible to sudden speed changes due to surrounding conditions, the accuracy of future trajectory predictions decreases. Furthermore, in real-world traffic scenarios, the need for obstacle avoidance and interaction among multiple pedestrians makes it difficult to capture the overall characteristics of all pedestrians, reducing the reliability of pedestrian trajectory prediction during autonomous driving and lowering vehicle safety. This application provides a method for predicting pedestrian intentions. This method aggregates multiple pedestrians to form a pedestrian group and calculates the interaction expansion coefficient based on the group's movement characteristics to obtain a predicted pedestrian intention. This prediction allows the vehicle to be controlled to avoid obstacles, thereby improving driving safety and reliability. This solves the problems in related technologies, such as the high degree of freedom and flexible movement of pedestrians, which makes them susceptible to sudden changes in speed due to surrounding conditions, thus reducing the accuracy of pedestrian trajectory prediction. Furthermore, in real traffic scenarios, the need for obstacle avoidance and interaction among multiple pedestrians makes it difficult to capture the overall characteristics of multiple pedestrians, which reduces the reliability of pedestrian trajectory prediction during autonomous driving and lowers the safety level of vehicle driving.

[0033] Specifically, Figure 1 This is a flowchart illustrating a method for predicting pedestrian intent provided in an embodiment of this application.

[0034] like Figure 1As shown, the method for predicting pedestrian intent includes the following steps:

[0035] In step S101, at least one action data of at least one pedestrian in the current environment is collected, and at least one pedestrian group is obtained using the at least one action data.

[0036] It is understood that, in the embodiments of this application, all pedestrian information in the environment can be extracted, including the current and historical position, displacement, speed, orientation, etc. of pedestrians, as at least one action data of at least one pedestrian in the current environment, so as to obtain a pedestrian group based on at least one action data of pedestrians in the current environment.

[0037] In particular, since there are many pedestrians in vehicle scenarios and their movement routes are consistent, predicting pedestrians individually can easily lead to insufficient computing resources. Pedestrians have a high degree of freedom and are prone to sudden changes in speed due to the surrounding situation. Therefore, pedestrian groups can be used to cluster and merge pedestrians in the current environment.

[0038] Optionally, in one embodiment of this application, obtaining at least one pedestrian group using at least one action data includes: calculating the similarity of any pedestrian among all pedestrians based on at least one action data, and extracting at least one pedestrian who meets a preset similarity condition from all pedestrians based on the similarity; and aggregating all pedestrians who meet the preset similarity condition to obtain at least one pedestrian group.

[0039] It should be noted that the preset similar conditions can be set by those skilled in the art according to the actual situation, and no specific limitations are made here.

[0040] In practice, a pedestrian with distinctive features can be randomly selected from all pedestrians in the current scene. This pedestrian's information is then compared with the similarity of other pedestrians within a certain range. If similar, the pedestrian is added to the pedestrian group. The similarity formula is:

[0041]

[0042] in, As weight, and Given pedestrian motion data, including position, displacement, speed, and orientation, the above formula can achieve a weighted summation of feature similarity between two pedestrians at a given time point. By calculating the similarity of N pedestrians, M pedestrian clusters can be obtained. Regularizing the boundaries of these clusters results in regular rectangular pedestrian clusters, such as... Figure 2 The diagram shown is a schematic representation of pedestrian clustering in one embodiment of this application, where ① is a pedestrian crossing and ② is a pedestrian cluster formed by pedestrian clustering.

[0043] In step S102, the center lateral velocity and boundary lateral velocity of at least one pedestrian group are calculated based on at least one motion data, and the interaction expansion coefficient of at least one pedestrian group is obtained from the center lateral velocity and boundary lateral velocity.

[0044] It is understood that, in the embodiments of this application, the motion data of all pedestrians within the pedestrian group, mainly the relative position of each pedestrian, can be used. , direction angle Information such as these is used to calculate the traveler's status information via a neural network:

[0045]

[0046] in, for MLP (Multilayer Perceptron) network, Max pooling is used. The resulting state information... Including the center velocity of the pedestrian group Given the direction h, the lateral velocity of the pedestrian group relative to the center of the road can be calculated. .

[0047] Furthermore, the method for calculating the lateral velocity of the pedestrian group at its boundary is as follows: calculate the lateral distance l of each pedestrian at the boundary of the pedestrian group relative to the center line of the crosswalk.

[0048] Then, calculate the lateral velocity of the pedestrians at the boundary based on the lateral distance between the current time and the historical time. :

[0049]

[0050] in, The horizontal distance. This represents the time difference between the current moment and a historical moment. The interaction expansion coefficient C is calculated as follows:

[0051]

[0052] That is, the lateral velocity of the center of the pedestrian group Lateral velocity of pedestrians at the boundary of the pedestrian group By making comparisons, we can obtain the interaction expansion coefficient C of the pedestrian group on each edge. The interaction expansion coefficient of the pedestrian group can reflect the interaction relationship of features within the pedestrian group.

[0053] In step S103, the target lateral velocity of at least one pedestrian group is calculated based on the interaction expansion coefficient, so as to predict the pedestrian intention of at least one pedestrian using the target lateral velocity.

[0054] It is understood that, in the embodiments of this application, the lateral velocity of the center of the pedestrian group can be... Using the interaction expansion coefficient C, calculate the overall target lateral velocity of the pedestrian group relative to the crosswalk. :

[0055]

[0056] Based on the obtained target lateral velocity, the vehicle's autonomous driving system can predict the pedestrian intentions of at least one pedestrian in the pedestrian group according to the actual traffic scene where the pedestrian is located, in order to help the vehicle's autonomous driving system avoid or slow down.

[0057] Optionally, in one embodiment of this application, before predicting the pedestrian intention of at least one pedestrian using the target lateral velocity, the method further includes: detecting whether at least one pedestrian group meets a preset crossing condition; if at least one pedestrian group does not meet the preset crossing condition, suppressing the interaction expansion coefficient until at least one pedestrian group meets the preset crossing condition, and updating the target lateral velocity using the suppressed interaction expansion coefficient.

[0058] It should be noted that the preset crossing conditions can be set by those skilled in the art according to the actual situation, and no specific limitations are made here.

[0059] In actual execution, at least one pedestrian group that meets the preset crossing conditions can be a pedestrian group with the intention to cross. The target lateral velocity calculated in the above steps is used to determine whether the pedestrian has the intention to cross. If at least one pedestrian group does not meet the preset crossing conditions, the interaction expansion coefficient of the pedestrian group is suppressed until the target lateral velocity calculated by the suppressed interaction expansion coefficient meets the preset crossing conditions.

[0060] Optionally, in one embodiment of this application, predicting the pedestrian intention of at least one pedestrian using a target lateral velocity includes: identifying the actual location of at least one pedestrian group; and generating the pedestrian intention based on the actual location and the target lateral velocity.

[0061] In actual execution, the pedestrian intentions of at least one pedestrian group can be obtained by identifying the actual location of the pedestrian group on the current road and the target lateral speed.

[0062] Optionally, in one embodiment of this application, generating pedestrian intent based on actual position and target lateral velocity includes: when the actual position is in a preset non-pedestrian crossing area, determining whether the target lateral velocity is greater than or equal to a preset threshold; if the target lateral velocity is greater than or equal to the preset threshold, constructing a second target obstacle based on the target lateral velocity, and generating pedestrian intent from the second target obstacle; otherwise, determining that the pedestrian intent is no crossing intent.

[0063] It should be noted that the preset non-pedestrian crossing area and preset threshold can be set by those skilled in the art according to the actual situation, and no specific limitation is made here.

[0064] In practice, the preset non-pedestrian crossing area can be a road surface area where pedestrians have no possibility of crossing the road, such as when a group of pedestrians is beside the road. If the target's lateral velocity reaches a preset threshold when the actual location is within the preset non-pedestrian crossing area... Then according to speed Generate pedestrian wall ,in To accommodate the location information of the pedestrian wall crossing the road, where t is the duration of the pedestrian wall, the pedestrian wall... As a secondary obstacle, the pedestrian's intention is conveyed to the planning and control module, which then decelerates and avoids the obstacle based on this information. If the target's lateral speed is less than a preset threshold... If the pedestrian does not meet the conditions to interfere with the passage of vehicles, then the pedestrian's intention is considered to be no intention to cross.

[0065] The pedestrian intent prediction method proposed in this application can aggregate multiple pedestrians to form a pedestrian group, and calculate the interaction expansion coefficient caused by the interaction of each pedestrian based on the movement characteristics of the pedestrian group. This yields the predicted pedestrian intent, which is then used to control the vehicle to avoid obstacles, thereby improving the safety and reliability of vehicle driving. This solves the problems in related technologies, such as the high degree of freedom and flexible movement of pedestrians, which makes them susceptible to sudden speed changes due to surrounding conditions, leading to decreased accuracy in predicting future pedestrian trajectories. Furthermore, in real-world traffic scenarios, the need for obstacle avoidance and interaction among multiple pedestrians makes it difficult to capture the overall characteristics of all pedestrians, reducing the reliability of pedestrian trajectory prediction during autonomous driving and lowering the safety level of vehicle driving.

[0066] Next, with reference to the accompanying drawings, a pedestrian intention prediction device according to an embodiment of this application is described.

[0067] Figure 5 This is a schematic diagram of the pedestrian intention prediction device according to an embodiment of this application.

[0068] like Figure 5 As shown, the pedestrian intention prediction device 10 includes: a data acquisition module 100, a calculation module 200, and a prediction module 300.

[0069] The acquisition module 100 is used to acquire at least one action data of at least one pedestrian in the current environment, and to obtain at least one pedestrian group using the at least one action data.

[0070] The calculation module 200 is used to calculate the center lateral velocity and the boundary lateral velocity of at least one pedestrian group based on at least one motion data, and to obtain the interaction expansion coefficient of at least one pedestrian group from the center lateral velocity and the boundary lateral velocity.

[0071] The prediction module 300 is used to calculate the target lateral velocity of at least one pedestrian group based on the interaction expansion coefficient, so as to predict the pedestrian intention of at least one pedestrian using the target lateral velocity.

[0072] Optionally, in one embodiment of this application, the acquisition module 100 includes an extraction unit and an aggregation unit.

[0073] The extraction unit is used to calculate the similarity of any pedestrian among all pedestrians based on at least one action data, and to extract at least one pedestrian that meets a preset similarity condition from all pedestrians based on the similarity.

[0074] The aggregation unit is used to aggregate all pedestrians that meet the preset similarity conditions to obtain at least one pedestrian group.

[0075] Optionally, in one embodiment of this application, the device 10 further includes a detection module and a suppression module.

[0076] The detection module is used to detect whether at least one group of pedestrians meets the preset crossing conditions before predicting the pedestrian intention of at least one pedestrian using the target lateral velocity.

[0077] The suppression module is used to suppress the interaction expansion coefficient when at least one pedestrian group does not meet the preset crossing conditions, until at least one pedestrian group meets the preset crossing conditions, and then update the target lateral velocity using the suppressed interaction expansion coefficient.

[0078] Optionally, in one embodiment of this application, the prediction module 300 includes an identification unit and a generation unit.

[0079] The identification unit is used to identify the actual location of at least one group of pedestrians.

[0080] The generation unit is used to generate pedestrian intent based on the actual position and the target's lateral velocity.

[0081] Optionally, in one embodiment of this application, the generation unit is specifically used to: calculate the road travel time of at least one pedestrian group based on the target lateral speed when the actual location is in a preset pedestrian crossing area; construct a first target obstacle based on the road travel time and the actual pedestrian crossing; and generate pedestrian intention from the first target obstacle.

[0082] Optionally, in one embodiment of this application, the generation unit is specifically used to: determine whether the target lateral speed is greater than or equal to a preset threshold when the actual location is in a preset non-pedestrian crossing area; if the target lateral speed is greater than or equal to the preset threshold, construct a second target obstacle based on the target lateral speed, and generate a pedestrian intention from the second target obstacle; otherwise, determine that the pedestrian intention is no intention to cross.

[0083] It should be noted that the explanation of the aforementioned method embodiment for predicting pedestrian intentions also applies to the pedestrian intention prediction device of this embodiment, and will not be repeated here.

[0084] The pedestrian intention prediction device proposed in this application can aggregate multiple pedestrians to form a pedestrian group, and calculate the interaction expansion coefficient caused by the interaction of each pedestrian based on the movement characteristics of the pedestrian group, thereby obtaining the predicted pedestrian intention and controlling the vehicle to avoid obstacles, thus improving the safety and reliability of vehicle driving. This solves the problems in related technologies, such as the high degree of freedom and flexible movement of pedestrians, which makes them susceptible to sudden speed changes due to surrounding conditions, leading to a decrease in the accuracy of pedestrian future trajectory prediction; and the difficulty in capturing the overall characteristics of multiple pedestrians in actual traffic scenarios due to obstacle avoidance and interaction needs, which reduces the reliability of pedestrian trajectory prediction during autonomous driving and lowers the safety level of vehicle driving.

[0085] Figure 6 A schematic diagram of the structure of a vehicle provided in an embodiment of this application. The vehicle may include:

[0086] The memory 601, the processor 602, and the computer program stored on the memory 601 and capable of running on the processor 602.

[0087] When the processor 602 executes the program, it implements the pedestrian intention prediction method provided in the above embodiments.

[0088] Furthermore, the vehicle also includes:

[0089] Communication interface 603 is used for communication between memory 601 and processor 602.

[0090] The memory 601 is used to store computer programs that can run on the processor 602.

[0091] The memory 601 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0092] If the memory 601, processor 602, and communication interface 603 are implemented independently, then the communication interface 603, memory 601, and processor 602 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0093] Optionally, in a specific implementation, if the memory 601, processor 602, and communication interface 603 are integrated on a single chip, then the memory 601, processor 602, and communication interface 603 can communicate with each other through an internal interface.

[0094] The processor 602 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0095] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method for predicting pedestrian intentions.

[0096] This embodiment also provides a computer program that, when executed, implements the above-described method for predicting pedestrian intentions.

[0097] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0098] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Thus, a feature specified as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this application, "N" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0099] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing a custom logical function or process step, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed in a different order than shown or discussed, including performing functions in a substantially simultaneous manner or in a reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application pertain.

[0100] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0101] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0102] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0103] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0104] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for predicting pedestrian intentions, characterized in that, Includes the following steps: Collect at least one action data of at least one pedestrian in the current environment, and use the at least one action data to obtain at least one pedestrian group; The center lateral velocity and boundary lateral velocity of the at least one pedestrian group are calculated based on the at least one motion data, and the interaction expansion coefficient of the at least one pedestrian group is obtained from the center lateral velocity and the boundary lateral velocity. The target lateral velocity of the at least one pedestrian group is calculated based on the interactive expansion coefficient, and the pedestrian intention of the at least one pedestrian is predicted using the target lateral velocity; The step of calculating the center lateral velocity and boundary lateral velocity of the at least one pedestrian group based on the at least one motion data, and obtaining the interaction expansion coefficient of the at least one pedestrian group from the center lateral velocity and the boundary lateral velocity, includes: Calculate the lateral distance of the boundary pedestrians of the at least one pedestrian group relative to the center line of the crosswalk; calculate the boundary lateral velocity of the boundary pedestrians based on the lateral distances at the current time and historical time; compare the center lateral velocity with the boundary lateral velocity to obtain the interaction expansion coefficient of the at least one pedestrian group; The method of predicting the pedestrian intention of at least one pedestrian using the target lateral velocity includes: Identify the actual position of the at least one pedestrian group; generate the pedestrian intention based on the actual position and the target lateral velocity; The step of generating the pedestrian intent based on the actual position and the target lateral velocity includes: If the actual location is in a preset non-pedestrian crossing area, determine whether the target lateral speed is greater than or equal to a preset threshold; if the target lateral speed is greater than or equal to the preset threshold, construct a second target obstacle based on the target lateral speed, and generate the pedestrian intention from the second target obstacle; otherwise, determine that the pedestrian intention is no intention to cross.

2. The method according to claim 1, characterized in that, The process of obtaining at least one pedestrian group using the at least one motion data includes: Calculate the similarity of any pedestrian among all pedestrians based on the at least one action data, and extract at least one pedestrian who meets the preset similarity condition from all pedestrians based on the similarity. All pedestrians that meet the preset similarity conditions are aggregated to obtain at least one pedestrian group.

3. The method according to claim 1, characterized in that, Before predicting the pedestrian intention of the at least one pedestrian using the target lateral velocity, the method further includes: Detect whether the at least one group of pedestrians meets the preset crossing conditions; If at least one pedestrian group is detected to not meet the preset crossing condition, the interaction expansion coefficient is suppressed until the at least one pedestrian group meets the preset crossing condition, and the target lateral velocity is updated using the suppressed interaction expansion coefficient.

4. A device for predicting pedestrian intentions, characterized in that, A method for predicting pedestrian intent as described in any one of claims 1-3, comprising: The acquisition module is used to acquire at least one action data of at least one pedestrian in the current environment, and to obtain at least one pedestrian group using the at least one action data; The calculation module is used to calculate the center lateral velocity and the boundary lateral velocity of the at least one pedestrian group based on the at least one motion data, and to obtain the interaction expansion coefficient of the at least one pedestrian group from the center lateral velocity and the boundary lateral velocity; A prediction module is used to calculate the target lateral velocity of the at least one pedestrian group based on the interaction expansion coefficient, so as to predict the pedestrian intention of the at least one pedestrian using the target lateral velocity.

5. The apparatus according to claim 4, characterized in that, The acquisition module includes: The extraction unit is used to calculate the similarity of any pedestrian among all pedestrians based on the at least one action data, and extract at least one pedestrian who meets a preset similarity condition from all pedestrians based on the similarity. The aggregation unit is used to aggregate all pedestrians that meet the preset similarity conditions to obtain at least one pedestrian group.

6. A vehicle, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the method for predicting pedestrian intent as described in any one of claims 1-3.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the method for predicting pedestrian intent as described in any one of claims 1-3.

Images