Pedestrian feature analysis method and device

Abstract

The present application relates to the field of vehicle-road collaboration and data statistics technology, and in particular to a method and device for analyzing pedestrian characteristics, wherein the method includes: using the roadside sensing device on the autonomous driving vehicle to collect the motion characteristic data of the pedestrian; extracting at least one motion feature from the motion characteristic data of the pedestrian; generating a characteristic data statistical graph and characteristic data statistical results based on at least one motion feature, and establishing an experience knowledge base based on the motion characteristic data and / or data statistical graphs and results. Thus, the technical problems in the related technology that it is difficult to obtain relatively complete basic data due to the fact that both the acquisition platform and the target object are in a dynamic environment, especially when too many sensing devices are installed on the vehicle, the data between different sensors cannot be synchronized, the stability of the system is reduced, and the degree of automation of the vehicle is affected, and the user experience is reduced.

Description

Technical Field

[0001] The present application relates to the field of vehicle-road collaboration and data statistics technology, and in particular to a pedestrian feature analysis method and device. Background Art

[0002] In related technologies, relevant images and data during the vehicle's driving process are obtained through on-board sensors, and a scene library is formed through labeling and training. After continuous training and learning, the autonomous driving decision-making ability is gradually improved. Labeling and training require a lot of time and effort. At the same time, the result of learning is the accumulation of experience in specific scenarios.

[0003] However, in related technologies, it is difficult to obtain relatively complete basic data because both the collection platform and the target object are in a dynamic environment. In particular, too many sensing devices are installed on the vehicle, making it impossible to synchronize data between different sensors, reducing the stability and security of the system, which in turn affects the degree of automation of the vehicle and reduces the user experience, which urgently needs to be solved. Summary of the invention

[0004] This application is based on the following problems and understandings made by the inventor:

[0005] When autonomous driving makes decisions about the next move, it also requires knowledge accumulated in the brain, just like driving a car. There are two ways to complete the knowledge base of autonomous driving: expert rule-based and AI (Artificial Intelligence). Learning the movement characteristics of pedestrians and vehicles is a relatively important part of the knowledge accumulation process.

[0006] At present, the more commonly used method for accumulating knowledge in the process of self-learning in autonomous driving is to obtain relevant images and data during the vehicle's driving process through on-board sensors, form a scene library through labeling and training, and then gradually improve the autonomous driving decision-making ability through continuous training and learning. Labeling and training require a lot of time and energy, and the result of learning is the accumulation of experience in specific scenarios.

[0007] At the same time, if Figure 1 As shown in the figure, the way in which vehicle-mounted sensors collect data has a limited range of perception, and it is difficult to describe the overall characteristics of traffic participants in the regional traffic environment. More importantly, they collect the vehicle's own operating status parameters and the vehicle's surrounding environment parameters. In addition, installing too many sensing devices on the vehicle poses a great challenge to the stable operation of the system and the synchronization between different sensor data. When vehicle-mounted sensors collect information about traffic participants, it is difficult to obtain relatively complete basic data because the collection platform and the target object are in a dynamic environment.

[0008] There are some problems and shortcomings in the method of collecting data through on-board sensors, but it is still a common and necessary method of data collection. In order to further enrich the knowledge base of autonomous driving and improve the level of decision-making, it is necessary to supplement data collection with roadside perception.

[0009] LiDAR is a key sensor for autonomous driving. The horizontal field of view of roadside laser and vehicle-mounted LiDAR is 360°, but the main difference is the vertical pitch angle. Commonly used LiDARs are 16-, 32-, and 64-line, with an effective range of 150 to 200 meters. The laser transmitter sends once, and the LiDAR rotates at high speed for one circle, generating multiple scan lines, which is called a frame of point cloud. Assume that the point cloud P of the tth frame is t , then P t ={p1, p2, ..., p n}, where n is the total number of point cloud data points, p i =(x i ,y i , z i ) is the 3D coordinate of the i-th data point in the laser radar coordinate system. Each data point in the point cloud contains multiple feature values, such as 3D spatial coordinates, reflection intensity, etc. The distance between points is determined by a certain measurement definition. Some commonly used clustering separation methods can be used to measure the distance between different points. For example, define a data point P in the point cloud i =(x i ,y i , z i ) and data point P j =(x j ,y j , z j ) i，j ,but By comparing the distances between different classes, different clustering cluster sets are generated.

[0010] The present application provides a pedestrian feature analysis method and device to solve the technical problems in the related technology that it is difficult to obtain relatively complete basic data because the collection platform and the target object are in a dynamic environment, especially too many sensing devices are installed on the vehicle, which makes it impossible to synchronize data between different sensors, reduces the stability of the system, and further affects the degree of automation of the vehicle, reduces the user experience, and so on.

[0011] The first aspect of the present application provides a pedestrian feature analysis method, comprising the following steps: using a roadside sensing device on an autonomous driving vehicle to collect motion feature data of pedestrians; extracting at least one motion feature from the motion feature data of the pedestrians; generating feature data statistical graphs and feature data statistical results based on the at least one motion feature, and establishing an experience knowledge base based on the motion feature data and / or data statistical graphs and results.

[0012] Optionally, in one embodiment of the present application, extracting at least one motion feature from the motion feature data of the pedestrian includes: obtaining the target classification, longitude, latitude, speed, motion direction and time of the pedestrian from the motion feature data; obtaining at least one motion feature of the pedestrian's average speed, passing time, speed-density, flow-density, speed-flow, number of trajectory points and trajectory curve shape according to the target classification, longitude, latitude, speed, motion direction and time.

[0013] Optionally, in an embodiment of the present application, the method of the embodiment of the present application further includes: visually displaying the statistical results of the feature data, wherein the feature data statistics include trajectory playback data and statistical data display information.

[0014] Optionally, in one embodiment of the present application, after collecting the motion feature data of the pedestrian, it also includes: obtaining the actual category of each feature data; fusing all feature data according to the actual category to generate the motion feature data.

[0015] Optionally, in one embodiment of the present application, the use of roadside sensing equipment on an autonomous driving vehicle to collect motion feature data of pedestrians includes: obtaining initial motion feature data of the pedestrian; performing point cloud separation on the initial motion feature data to obtain point cloud data; and performing pedestrian clustering on the point cloud data using a preset clustering algorithm to obtain the motion feature data.

[0016] The second aspect of the present application provides a pedestrian feature analysis device, including: a collection module, used to use the roadside perception equipment on the autonomous driving vehicle to collect the motion feature data of the pedestrian; an extraction module, used to extract at least one motion feature from the motion feature data of the pedestrian; an analysis module, used to generate a feature data statistical graph and feature data statistical results based on the at least one motion feature, and establish an experience knowledge base based on the motion feature data and / or data statistical graphs and results.

[0017] Optionally, in one embodiment of the present application, the extraction module includes: a first acquisition unit, used to obtain the target classification, longitude, latitude, speed, movement direction and time of the pedestrian from the motion feature data; a second acquisition unit, used to obtain at least one motion feature of the pedestrian's average speed, passing time, speed-density, flow-density, speed-flow, trajectory points and trajectory curve shape according to the target classification, longitude, latitude, speed, movement direction and time.

[0018] Optionally, in an embodiment of the present application, the device of the embodiment of the present application further includes: a display module, which is used to visually display the characteristic data statistics, wherein the characteristic data statistics include trajectory playback data and statistical data display information.

[0019] Optionally, in one embodiment of the present application, the device of the embodiment of the present application also includes: an acquisition module, which is used to obtain the actual category of each feature data after collecting the motion feature data of the pedestrian; and a generation module, which is used to fuse all feature data according to the actual category after collecting the motion feature data of the pedestrian to generate the motion feature data.

[0020] Optionally, in one embodiment of the present application, the acquisition module includes: a third acquisition unit, used to acquire the initial motion feature data of the pedestrian; a processing unit, used to perform point cloud separation on the initial motion feature data to obtain point cloud data; and an analysis unit, used to perform pedestrian clustering on the point cloud data using a preset clustering algorithm to obtain the motion feature data.

[0021] A third aspect of the present application provides a vehicle, comprising: 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 feature analysis method as described in the above embodiment.

[0022] A fourth aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the pedestrian feature analysis method as described in the above embodiment.

[0023] The embodiment of the present application can use the roadside sensing equipment on the autonomous driving vehicle to collect the motion feature data of pedestrians, thereby obtaining a large sample size of pedestrian motion feature data in a natural state, and extracting at least one motion feature from the pedestrian motion feature data, thereby generating a feature data statistical graph and feature data statistical results, thereby establishing an experience knowledge base, and then providing a rich experience knowledge base for intelligent decision-making of autonomous driving, effectively increasing the stability of the system, thereby improving the degree of automation of the vehicle, and effectively meeting the user's usage needs. As a result, the technical problems in the related technology that it is difficult to obtain relatively complete basic data due to the fact that both the acquisition platform and the target object are in a dynamic environment, especially too many sensing devices are installed on the vehicle, so that the data between different sensors cannot be synchronized, reducing the stability of the system, thereby affecting the degree of automation of the vehicle, and reducing the user's usage experience, etc.

[0024] 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 the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] 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:

[0026] Figure 1 It is a schematic diagram of a vehicle-side data collection method in the related art;

[0027] Figure 2 A flowchart of a pedestrian feature analysis method provided according to an embodiment of the present application;

[0028] Figure 3 This is a functional schematic diagram of a pedestrian feature analysis method according to a specific embodiment of the present application;

[0029] Figure 4 This is a flowchart of a pedestrian feature analysis method according to a specific embodiment of the present application;

[0030] Figure 5 is a schematic structural diagram of a pedestrian feature analysis device according to an embodiment of the present application;

[0031] Figure 6 It is a schematic diagram of the structure of a vehicle provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0033] The following describes the pedestrian feature analysis method and device of the embodiment of the present application with reference to the accompanying drawings. In view of the above-mentioned related technology mentioned in the background technology center, it is difficult to obtain relatively complete basic data because the acquisition platform and the target object are in a dynamic environment, especially too many sensing devices are installed on the vehicle, so that the data between different sensors cannot be synchronized, the stability of the system is reduced, and the degree of automation of the vehicle is affected, and the user experience is reduced. The present application provides a pedestrian feature analysis method, in which the roadside sensing device on the autonomous driving vehicle can be used to collect the motion feature data of the pedestrian, so as to obtain a large sample volume and the pedestrian motion feature data in a natural state, and extract at least one motion feature from the motion feature data of the pedestrian, and then generate a feature data statistical graph and feature data statistical results, so as to establish an experience knowledge base, and then provide a rich experience knowledge base for the intelligent decision-making of the autonomous driving, effectively increase the stability of the system, and then improve the degree of automation of the vehicle, and effectively meet the user's use needs. Thus, the technical problems such as the difficulty in obtaining relatively complete basic data due to the dynamic environment of the acquisition platform and the target object, especially too many sensing devices are installed on the vehicle, so that the data between different sensors cannot be synchronized, the stability of the system is reduced, and the degree of automation of the vehicle is affected, and the user experience is reduced are solved.

[0034] Specifically, Figure 2 A flowchart of a pedestrian feature analysis method provided in an embodiment of the present application.

[0035] like Figure 2 As shown, the pedestrian feature analysis method includes the following steps:

[0036] In step S201, the motion characteristic data of pedestrians are collected using the roadside perception equipment on the autonomous driving vehicle.

[0037] It is understandable that the embodiments of the present application can utilize roadside sensing equipment on autonomous driving vehicles to collect motion feature data of pedestrians. For example, the vehicle can utilize roadside sensing equipment such as lidar and cameras to collect motion status data of pedestrians around the vehicle, and then combine the synchronized data of radar and video to perform the analysis of the following steps, thereby improving the user's driving safety and avoiding traffic hazards.

[0038] It should be noted that the embodiment of the present application synchronizes the time and space between the laser radar and the camera. In terms of time synchronization, the camera shooting is triggered synchronously according to the frame period of the laser radar, so as to achieve complete frame synchronization between the radar and the camera and achieve millisecond-level synchronization. In terms of space, the coordinates of the laser radar and the camera need to be fused. The data obtained by the laser radar can be converted from the laser radar coordinate system to the world coordinate system, and then from the world coordinate system to the pixel coordinate system of the camera. For example, assuming that the coordinates of a point in the world coordinate system are P W =(x W ,y W ,z W ) T , the coordinate in the camera coordinate system is P c =(x c ,y c ,z c ),but T=[t x t y t z ] T ; Let the coordinate of a point in the camera coordinate be P c =(x c ,y c ,z c ), the pixel coordinates can be expressed as:

[0039]

[0040] Among them, k is the camera internal parameter matrix, c x , c y is the offset of the camera optical axis in the pixel coordinate system, f x 、f y is the normalized focal length on the u-axis and v-axis.

[0041] In summary, the embodiments of the present application can effectively solve the problems of insufficient perception dimensions and lack of information when using a single sensor of a laser radar or camera, and can collect complete pedestrian feature data.

[0042] Optionally, in one embodiment of the present application, motion feature data of pedestrians are collected using roadside sensing equipment on an autonomous driving vehicle, including: obtaining initial motion feature data of pedestrians; performing point cloud separation on the initial motion feature data to obtain point cloud data; and performing pedestrian clustering on the point cloud data using a preset clustering algorithm to obtain motion feature data.

[0043] As a possible implementation method, the embodiments of the present application can utilize roadside sensing equipment on autonomous driving vehicles, such as lidar and cameras, to obtain initial motion feature data of pedestrians, and perform point cloud data separation, i.e., point cloud slicing, to obtain point cloud data, and utilize a preset clustering algorithm to cluster pedestrians in the point cloud data, such as using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm, which can effectively improve the accuracy of clustering and thereby obtain motion feature data.

[0044] Specifically, point cloud data separation, i.e., point cloud slicing, is performed because the point cloud data sensed by the laser radar has the property of uneven density, and the farther the distance from the sensor, the lower the density of the point cloud data and the more noise. In this embodiment of the application, the point cloud data is divided into different parts according to the distance, and the steps of point cloud slicing are as follows:

[0045] Step S1: Determine the farthest distance and the shortest distance of the slice;

[0046] Step S2: Determine the interval parameter λ, and determine the slice interval and number N;

[0047] Step S3: determine whether there are enough point cloud data in the slice, that is, whether the number of point cloud elements is greater than the threshold Thresh. If it is less than Thresh, merge it with the next slice and repeat this process N times;

[0048] Step S4: Determine whether the last slice contains enough point cloud data. If the number of elements is less than Thresh, the slice is discarded as noise.

[0049] It should be noted that the threshold is set by those skilled in the art according to actual conditions and is not specifically limited here.

[0050] Among them, the embodiment of the present application can use a preset clustering algorithm to cluster pedestrians on point cloud data, such as using the DBSCAN algorithm. Compared with other clustering algorithms, the advantage of the DBSCAN algorithm is that the number of targets in the data does not need to be determined in advance. The clustering radius and the minimum number of elements of the algorithm are globally uniquely determined, and are often suitable for data with uniform density. When encountering uneven data such as lidar, the clustering effect is relatively poor. By preprocessing the point cloud data, the density is balanced in the height dimension, and then DBSCAN is used for clustering, thereby effectively improving the clustering accuracy.

[0051] Furthermore, in one embodiment of the present application, after collecting the motion feature data of the pedestrian, it also includes: obtaining the actual category of each feature data; fusing all the feature data according to the actual category to generate motion feature data.

[0052] It can be understood that the embodiments of the present application can obtain the actual category of each feature data, such as the camera obtaining dynamic images of pedestrians, the lidar obtaining pedestrian motion feature images, etc., so as to fuse all the collected feature data according to the actual category, and fuse and synchronize the data to generate motion feature data. Therefore, when both the acquisition platform and the target object are in a dynamic environment, relatively complete motion feature data can also be obtained, thereby improving the stability of the system.

[0053] In step S202, at least one motion feature is extracted from the motion feature data of the pedestrian.

[0054] Specifically, the embodiments of the present application can obtain a large sample volume of pedestrian motion feature data in a natural state through roadside equipment, such as the specific motion feature data of the following steps, and through a multi-dimensional analysis of the motion feature data and the overall environment around the vehicle, it can effectively describe the motion characteristics of pedestrians, thereby extracting at least one motion feature based on the pedestrian's motion feature data to achieve trajectory tracking of a single target. Then, the roadside equipment will send the collected data to the background server of the following steps for storage through a dedicated network or wireless network in the form of TCP (Transmission Control Protocol) or UDP (User Data Protocol) communication, thereby achieving data synchronization, effectively improving the stability of the system, and effectively improving driving safety.

[0055] Among them, in one embodiment of the present application, at least one motion feature is extracted from the motion feature data of the pedestrian, including: obtaining the target classification, longitude, latitude, speed, motion direction and time of the pedestrian from the motion feature data; obtaining at least one motion feature of the pedestrian's average speed, passing time, speed-density, flow-density, speed-flow, number of trajectory points and trajectory curve shape according to the target classification, longitude, latitude, speed, motion direction and time.

[0056] In some embodiments, the embodiments of the present application can establish a database in advance on the background server, that is, a system for storing collected data, so as to receive data sent from the roadside equipment collection end and store it in a data table. The data table mainly contains basic data, that is, obtaining the target classification, longitude, latitude, speed, movement direction and time of pedestrians from the motion feature data, so as to obtain the walking direction or speed of pedestrians around the vehicle, and then use the basic data in the database, that is, according to the target classification, longitude, latitude, speed, movement direction and time, obtain the average speed of pedestrians, passing time, speed-density, flow-density, speed-flow, trajectory points and at least one motion feature of the trajectory curve shape, so as to obtain relatively complete basic data, thereby providing a rich experience knowledge base for intelligent decision-making of autonomous driving, improving driving safety performance, and avoiding the occurrence of traffic accidents.

[0057] In step S203, a feature data statistical graph and a feature data statistical result are generated based on at least one motion feature, and an experience knowledge base is established based on the motion feature data and / or the data statistical graph and result.

[0058] In the actual implementation process, Figure 3 As shown, the embodiment of the present application can collect the motion state data of pedestrians through roadside collection equipment, such as lidar and cameras, and integrate and synchronize the motion state data of pedestrians, so that the roadside controller sends the collected data to the background server for storage through a dedicated network or wireless network in a TCP or UDP communication mode. The background server establishes a database in advance, receives the data sent from the roadside end, and stores it in a data table, and then uses the basic data in the database to carry out pedestrian feature analysis and statistics, thereby generating feature data statistical graphics and feature data statistical results based on at least one motion feature, and establishing an experience knowledge base based on the motion feature data and / or data statistical graphics and results.

[0059] In summary, this embodiment can provide a rich experience knowledge base for intelligent decision-making in autonomous driving, and can also provide support for intelligent traffic operation supervision, effectively improving driving safety, greatly improving the degree of vehicle automation, avoiding traffic accidents, and improving user experience.

[0060] Optionally, in an embodiment of the present application, the method of the embodiment of the present application further includes: visually displaying characteristic data statistics, wherein the characteristic data statistics include trajectory playback data and statistical data display information.

[0061] It can be understood that the embodiments of the present application can collect the motion data characteristics of pedestrians through roadside laser radar and other equipment to establish a database, and through the extraction and statistics of basic data, form feature statistical results and output feature statistical graphics based on a large sample size and multi-dimensional analysis, and then summarize the motion characteristics of pedestrians. And output it to the autonomous driving as an experience knowledge base, and at the same time visualize the feature data statistical results, for example: trajectory playback data, statistical data display information, so as to query the results and intuitively observe the results according to the required statistical dimensions, thereby effectively increasing the stability of the system, improving the intelligence level of the vehicle, and effectively meeting the user's usage needs.

[0062] like Figure 4 As shown, the working principle of the embodiment of the present application is described in detail below with a specific embodiment.

[0063] Step S401: The roadside collection equipment laser radar and camera collect feature data of pedestrians.

[0064] Step S402: Fusing and synchronizing the pedestrian feature data collected by the road test equipment.

[0065] Step S403: Determine whether the single ID target data is continuous and complete. If it is complete, execute step S404; otherwise, execute step S401 to re-collect the pedestrian's feature data.

[0066] Step S404: Send the data to the backend server via TCP / UDP.

[0067] Step S405: store data, including ID, speed, longitude, latitude, movement direction, etc.

[0068] Step S406: Count the dimensions such as speed, trajectory, and traffic density.

[0069] Step S407: the display terminal outputs statistical graphics, trajectory playback or query results.

[0070] According to the pedestrian feature analysis method proposed in the embodiment of the present application, the roadside sensing device on the autonomous driving vehicle can be used to collect the motion feature data of pedestrians, thereby obtaining a large sample size of pedestrian motion feature data in a natural state, and extracting at least one motion feature from the pedestrian motion feature data, thereby generating feature data statistical graphs and feature data statistical results, thereby establishing an experience knowledge base, and then providing a rich experience knowledge base for intelligent decision-making of autonomous driving, effectively increasing the stability of the system, thereby improving the degree of automation of the vehicle, and effectively meeting the user's usage needs. As a result, the technical problems in the related technology that it is difficult to obtain relatively complete basic data due to the fact that both the acquisition platform and the target object are in a dynamic environment, especially too many sensing devices are installed on the vehicle, so that the data between different sensors cannot be synchronized, reducing the stability of the system, thereby affecting the degree of automation of the vehicle, and reducing the user's usage experience, etc.

[0071] Next, the pedestrian feature analysis device proposed according to the embodiment of the present application is described with reference to the accompanying drawings.

[0072] Figure 5 It is a block diagram of a pedestrian feature analysis device according to an embodiment of the present application.

[0073] like Figure 5 As shown, the pedestrian feature analysis device 10 includes: a collection module 100 , an extraction module 200 and an analysis module 300 .

[0074] Specifically, the collection module 100 is used to collect motion characteristic data of pedestrians using the roadside perception equipment on the autonomous driving vehicle.

[0075] The extraction module 200 is used to extract at least one motion feature from the motion feature data of the pedestrian.

[0076] The analysis module 300 is used to generate a characteristic data statistical graph and a characteristic data statistical result based on at least one motion feature, and to establish an experience knowledge base based on the motion characteristic data and / or the data statistical graph and result.

[0077] Optionally, in one embodiment of the present application, the extraction module 200 includes: a first acquisition unit and a second acquisition unit.

[0078] Among them, the first acquisition unit is used to obtain the pedestrian's target classification, longitude, latitude, speed, movement direction and time from the motion feature data.

[0079] The second acquisition unit is used to acquire at least one motion feature of the pedestrian's average speed, passing time, speed-density, flow-density, speed-flow, trajectory points and trajectory curve shape according to the target classification, longitude, latitude, speed, movement direction and time.

[0080] Optionally, in one embodiment of the present application, the device 10 of the embodiment of the present application further includes: a display module.

[0081] Among them, the display module is used to visually display the characteristic data statistics results, wherein the characteristic data statistics include trajectory playback data and statistical data display information.

[0082] Optionally, in one embodiment of the present application, the device 10 of the embodiment of the present application further includes: an acquisition module and a generation module.

[0083] The acquisition module is used to acquire the actual category of each feature data after collecting the motion feature data of the pedestrian.

[0084] The generation module is used to fuse all feature data according to actual categories after collecting the motion feature data of pedestrians to generate motion feature data.

[0085] Optionally, in one embodiment of the present application, the acquisition module 100 includes: a third acquisition unit, a processing unit and an analysis unit.

[0086] The third acquisition unit is used to acquire initial motion feature data of the pedestrian.

[0087] The processing unit is used to perform point cloud separation on the initial motion feature data to obtain point cloud data.

[0088] The analysis unit is used to cluster pedestrians on the point cloud data using a preset clustering algorithm to obtain motion feature data.

[0089] It should be noted that the above explanation of the embodiment of the pedestrian feature analysis method is also applicable to the pedestrian feature analysis device of this embodiment, and will not be repeated here.

[0090] According to the pedestrian feature analysis device proposed in the embodiment of the present application, the roadside sensing device on the autonomous driving vehicle can be used to collect the motion feature data of pedestrians, thereby obtaining a large sample volume of pedestrian motion feature data in a natural state, and extracting at least one motion feature from the pedestrian motion feature data, thereby generating a feature data statistical graph and feature data statistical results, thereby establishing an experience knowledge base, and then providing a rich experience knowledge base for intelligent decision-making of autonomous driving, effectively increasing the stability of the system, thereby improving the degree of automation of the vehicle, and effectively meeting the user's usage needs. As a result, the technical problems in the related technology that it is difficult to obtain relatively complete basic data due to the fact that both the acquisition platform and the target object are in a dynamic environment, especially too many sensing devices are installed on the vehicle, so that the data between different sensors cannot be synchronized, reducing the stability of the system, thereby affecting the degree of automation of the vehicle, and reducing the user's usage experience, etc.

[0091] Figure 6 A schematic diagram of the structure of a vehicle provided in an embodiment of the present application. The vehicle may include:

[0092] A memory 601 , a processor 602 , and a computer program stored in the memory 601 and executable on the processor 602 .

[0093] When the processor 602 executes the program, the pedestrian feature analysis method provided in the above embodiment is implemented.

[0094] Furthermore, the vehicle also includes:

[0095] The communication interface 603 is used for communication between the memory 601 and the processor 602 .

[0096] The memory 601 is used to store computer programs that can be executed on the processor 602 .

[0097] The memory 601 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0098] If the memory 601, the processor 602 and the communication interface 603 are implemented independently, the communication interface 603, the memory 601 and the processor 602 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a 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 only one type of bus.

[0099] Optionally, in a specific implementation, if the memory 601, the processor 602 and the communication interface 603 are integrated on a chip, the memory 601, the processor 602 and the communication interface 603 can communicate with each other through an internal interface.

[0100] The processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0101] This embodiment also provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the above pedestrian feature analysis method is implemented.

[0102] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0103] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0104] 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 more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0105] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or N wirings (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0106] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiment, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0107] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0108] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0109] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A pedestrian feature analysis method, characterized in that: The following steps are involved: The motion feature data of pedestrians are collected by using the roadside sensing device on the autonomous driving vehicle, wherein the step of collecting the motion feature data of pedestrians by using the roadside sensing device on the autonomous driving vehicle comprises: obtaining initial motion feature data of the pedestrians, performing point cloud separation on the initial motion feature data to obtain point cloud data, and performing pedestrian clustering on the point cloud data by using a preset clustering algorithm to obtain motion feature data, wherein the step of separating the point cloud comprises: Step S1: Determine the farthest distance and the shortest distance of the slice; Step S2: Determine the interval parameter λ, and determine the slice interval and number N; Step S3: determining whether there are enough point cloud data in the slice, that is, whether the number of point cloud elements is greater than a threshold Thresh; if less than Thresh, merging with the next slice, and repeating this process N times; Step S4: determining whether the last slice contains enough point cloud data; if the number of elements is less than the Thresh, discarding the slice as noise; Extracting at least one motion feature from the motion feature data of the pedestrian, wherein the extracting the at least one motion feature from the motion feature data of the pedestrian comprises: acquiring the target classification, longitude, latitude, speed, motion direction and time of the pedestrian from the motion feature data, and acquiring at least one motion feature of the average speed, passing time, speed-density, flow-density, speed-flow, number of trajectory points and trajectory curve shape of the pedestrian according to the target classification, longitude, latitude, speed, motion direction and time; and Generate a feature data statistical graph and feature data statistical results based on the at least one motion feature, establish an experience knowledge base based on the motion feature data, data statistical graph and results, and visualize the feature data statistical results, wherein the feature data statistics include trajectory playback data and statistical data display information.

2. The method according to claim 1, characterized in that After collecting the motion characteristic data of the pedestrian, the method further includes: Get the actual category of each feature data; All feature data are fused according to the actual category to generate the motion feature data.

3. A pedestrian feature analysis device, characterized in that: include: A collection module, used for collecting motion feature data of pedestrians by using a roadside sensing device on an autonomous driving vehicle, wherein the collecting of motion feature data of pedestrians by using a roadside sensing device on an autonomous driving vehicle comprises: obtaining initial motion feature data of the pedestrian, performing point cloud separation on the initial motion feature data to obtain point cloud data, and performing pedestrian clustering on the point cloud data by using a preset clustering algorithm to obtain motion feature data, wherein the step of point cloud separation comprises: Step S1: Determine the farthest distance and the shortest distance of the slice; Step S2: Determine the interval parameter λ, and determine the slice interval and number N; Step S3: determining whether there are enough point cloud data in the slice, that is, whether the number of point cloud elements is greater than a threshold Thresh; if less than Thresh, merging with the next slice, and repeating this process N times; Step S4: determining whether the last slice contains enough point cloud data; if the number of elements is less than the Thresh, discarding the slice as noise; an extraction module, configured to extract at least one motion feature from the motion feature data of the pedestrian, wherein the extracting the at least one motion feature from the motion feature data of the pedestrian comprises: acquiring a target classification, longitude, latitude, speed, motion direction and time of the pedestrian from the motion feature data, and acquiring at least one motion feature of an average speed, passing time, speed-density, flow-density, speed-flow, number of trajectory points and trajectory curve shape of the pedestrian according to the target classification, longitude, latitude, speed, motion direction and time; and An analysis module is used to generate a feature data statistical graph and feature data statistical results based on the at least one motion feature, and to establish an experience knowledge base based on the motion feature data, data statistical graph and results, and to visually display the feature data statistical results, wherein the feature data statistics include trajectory playback data and statistical data display information.

4. A vehicle, characterized in that: include: 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 feature analysis method according to any one of claims 1 to 2.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the pedestrian feature analysis method as described in any one of claims 1-2.

Images