A method, device and equipment for determining risky driving behavior and a storage medium

Abstract

The present disclosure provides a risk driving behavior determination method and device, equipment and storage medium, relates to the technical field of computers, and particularly to the technical field of intelligent transportation. The specific implementation scheme is: determining a behavior feature by using a driving image; in the case that the behavior feature is a dangerous behavior feature, determining at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule; the adjacent behavior feature is a behavior feature determined by using an adjacent frame driving image; and determining a risk driving behavior by using the dangerous behavior feature and the adjacent behavior feature.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computer, and particularly relates to the technical field of intelligent transportation, deep learning, etc. BACKGROUND

[0002] Statistics show that the bad driving habits of drivers are one of the main reasons for frequent traffic accidents. Correcting bad driving habits is of great significance for reducing traffic accidents and reducing the driving risk of users. The existing driving risk prompting technology mainly gives a log report based on driving mileage, driving time length and driving speed and other related information. The log report has certain reference value, but cannot comprehensively and timely propose correction measures for the bad driving habits of drivers.

[0003] Therefore, how to generate relatively comprehensive driving suggestions based on driving information in real time becomes a problem to be solved. SUMMARY

[0004] The present disclosure provides a risk driving behavior determination method and device, equipment and storage medium.

[0005] According to an aspect of the present disclosure, a risk driving behavior determination method is provided, comprising:

[0006] determining a behavior feature by using driving images;

[0007] in a case where the behavior feature is a dangerous behavior feature, determining at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule; the adjacent behavior feature is a behavior feature determined by using adjacent frame driving images;

[0008] determining a risk driving behavior by using the dangerous behavior feature and the adjacent behavior feature.

[0009] According to another aspect of the present disclosure, a risk driving behavior determination device is provided, comprising:

[0010] a behavior feature determination module configured to determine a behavior feature by using driving images;

[0011] an adjacent behavior feature determination module configured to, in a case where the behavior feature is a dangerous behavior feature, determine at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule; the adjacent behavior feature is a behavior feature determined by using adjacent frame driving images;

[0012] a risk driving behavior determination module configured to determine a risk driving behavior by using the dangerous behavior feature and the adjacent behavior feature.

[0013] According to another aspect of the present disclosure, an electronic device is provided, comprising:

[0014] at least one processor; and

[0015] a memory in communication with the at least one processor; wherein

[0016] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method in any embodiment of the present disclosure.

[0017] According to another aspect of the present disclosure, a non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the method in any embodiment of the present disclosure is provided.

[0018] According to another aspect of the present disclosure, a computer program product comprising a computer program which, when executed by a processor, implements the method in any embodiment of the present disclosure is provided.

[0019] According to the technology of the present disclosure, by determining the behavior feature by using the driving image, obtaining the dangerous behavior feature and the adjacent behavior feature, and further determining the risky driving behavior. In this way, the user can be effectively guided to correct the bad driving habits in the driving process, and the safety hidden danger is reduced.

[0020] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0021] The accompanying drawings are used to better understand the present scheme, and do not constitute a limitation on the present disclosure. Among them:

[0022] Figure 1 is a flowchart of a method for determining a risky driving behavior according to the present disclosure;

[0023] Figure 2 is a flowchart of a method for determining a behavior feature according to the present disclosure;

[0024] Figure 3 is a flowchart of a method for executing a risky driving behavior according to the present disclosure;

[0025] Figure 4 is a flowchart of a method for determining a driving image according to the present disclosure;

[0026] Figure 5 is a structural diagram of a device for determining a risky driving behavior according to the present disclosure;

[0027] Figure 6 is a structural diagram of a behavior feature determination module according to the present disclosure;

[0028] Figure 7is a structural diagram of a risk driving behavior execution module according to the present disclosure;

[0029] Figure 8 is a block diagram of an electronic device for implementing a risk driving behavior determination method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0030] Exemplary embodiments of the present disclosure are described herein with reference to the accompanying drawings, which are included to provide a thorough understanding of embodiments of the present disclosure, and are taken to illustrate the preferred embodiments of the present disclosure, alone, and should not be taken to be the only embodiments of the present disclosure. Accordingly, those of ordinary skill in the art will recognize that there are numerous variations and modifications of this description that will come to mind to those skilled in the art, but are not described herein for the sake of brevity and so as not to unnecessarily obscure the application of the present disclosure. Those of ordinary skill in the art will recognize or be able to ascertain using no more than routine experimentation, materials and instruments known to them, the equivalents of such specific embodiments.

[0031] As shown in Figure 1 the present disclosure relates to a risk driving behavior determination method, which can include the following steps:

[0032] S101: determining a behavior feature from a driving image;

[0033] S102: in a case where the behavior feature is a dangerous behavior feature, determining at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule; the adjacent behavior feature is a behavior feature in an adjacent frame image;

[0034] S103: determining a risk driving behavior using the dangerous behavior feature and the adjacent behavior feature.

[0035] The present embodiment can be applied to a computer device, and specifically can include a vehicle-mounted driving computer or a server in communication connection with a vehicle.

[0036] The driving image can be an image acquired in real time based on a vehicle-mounted camera, wherein the driving video can include lane lines, nearby vehicles, pedestrians, and traffic lights, and the like.

[0037] The behavior feature can include a normal driving feature and a dangerous behavior feature. The dangerous behavior feature can be a behavior feature determined based on the driving image and having a certain risk, such as line pressing, driving into a non-motor vehicle lane, running a red light, and being too close to a pedestrian, and the like, which are not exhaustively listed herein.

[0038] Determining the behavior feature from the driving image can be determining a behavior state feature based on a positional relationship between the target vehicle and road information.

[0039] In the case that the behavior feature is a dangerous behavior feature, at least one adjacent behavior feature adjacent to the dangerous behavior feature is determined based on a preset rule. In a specific implementation, at least one adjacent frame image is first determined based on the preset rule, and then the behavior features in the adjacent frame image are taken as the adjacent behavior features.

[0040] The adjacent frame image can be one or more driving images having an adjacent relationship with the current frame image in the driving video. The adjacent behavior feature can be a behavior feature determined based on the adjacent frame image.

[0041] The preset rule includes at least one of a selection direction rule and a selection interval rule. The selection direction rule can be forward selection or backward selection based on the current frame image, which is not limited here.

[0042] The selection interval rule can be to select the adjacent frame video image based on a preset interval time length, where the preset interval time length can be 1s, 2s, 3s, etc., which is not limited here. The selection interval rule can also be to select the adjacent frame video image based on a preset interval image number starting from the current frame video image, where the interval image number can be 1, 3, 5, etc., which is not limited here.

[0043] With the dangerous behavior feature and the adjacent behavior feature, multiple behavior features in the time dimension can be obtained, and the risk driving behavior can be determined based on the multiple behavior features in the time dimension.

[0044] Through the above process, the risk driving behavior can be determined based on the behavior feature, thereby guiding the user to correct the bad driving habits in the driving process and reduce the safety hazards.

[0045] As shown in FIG. 1, Figure 2 a method for determining a behavior feature includes:

[0046] S201: inputting a driving image into a semantic segmentation model to obtain a first feature; the first feature is used to represent semantic information in the driving image;

[0047] S202: inputting the driving image into a depth recognition model to obtain a second feature; the second feature is used to represent depth information of the driving image;

[0048] S203: obtaining a behavior feature by using the first feature and the second feature.

[0049] The semantic segmentation model can be a statistical semantic segmentation model, a geometric semantic segmentation model, or a neural network-based semantic segmentation model, which is not limited here.

[0050] Preferably, a neural network-based semantic segmentation model can automatically learn the features of driving images, thereby achieving end-to-end feature extraction. The neural network-based semantic segmentation model can include multiple convolutional layers, deconvolutional layers, and pooling layers. After inputting the 2D driving image captured by the vehicle camera into the convolutional neural network, a series of feature maps corresponding to the driving image are obtained through multiple convolution and pooling processes. Then, the feature map obtained from the last convolutional layer is upsampled using deconvolutional and pooling layers to obtain the first feature. This ensures that the upsampled first feature has the same size as the original driving image, thus preserving the same 2D positional information as the original image while predicting each pixel value. The first feature output by the semantic segmentation model can be an image feature containing 2D semantic information.

[0051] Semantic segmentation models can be trained using adversarial training methods. Specifically, the adversarial network is first pre-trained by combining traditional multi-class cross-entropy loss with the adversarial network, and then the adversarial loss between the network and labeled data is used to fine-tune the parameters of the segmentation network.

[0052] The deep recognition model can be either a supervised learning model or a semi-supervised learning model; this is not a limitation. For example, it could be a Deep Belief Network (DBN) or a Deep Boltzmann Machine (DBM). The driving image is input into the deep recognition model to obtain a second feature; this second feature represents the depth information of the driving image.

[0053] Deep recognition models can be trained using training and validation sets, where the validation set can be a dataset carrying depth labels. Determining the accuracy of the training and validation sets facilitates the adjustment of the deep recognition model's parameters.

[0054] Behavioral features are obtained by using the first and second features. Specifically, the first and second features can be fused to obtain image features containing deep semantic information.

[0055] Through the above process, 2D semantic information and depth information can be generated using semantic segmentation models and depth recognition models, respectively. This will greatly improve the speed of semantic recognition, enabling real-time driving alerts for risky driving behaviors.

[0056] like Figure 3 As shown, in one embodiment, step S103 includes the following sub-steps:

[0057] S301: Construct continuous behavioral characteristics using dangerous behavioral characteristics and adjacent behavioral characteristics;

[0058] S302: In the case that the continuous behavior feature does not conform to the preset traffic behavior rule, the driving behavior corresponding to the continuous behavior feature is regarded as a risky driving behavior.

[0059] The continuous behavior feature is used to represent the driving behavior feature in a period of time. The implementation manner of constructing the continuous behavior feature can be based on the time sequence corresponding to the behavior feature, and the dangerous behavior feature and the adjacent behavior feature are spliced.

[0060] For example, in the case that the dangerous behavior feature is line pressing, the continuous behavior feature in a period of time based on the dangerous behavior feature can be normal lane changing, continuous lane changing, long-time line pressing, etc. For example, in the case that the dangerous behavior feature is line pressing, and other adjacent behavior features are all normal driving states in the lane, the corresponding continuous behavior feature is normal lane changing. In the case that the dangerous behavior feature is line pressing, and other adjacent behavior features are all line pressing, the corresponding continuous behavior feature is long-time line pressing. In the case that the dangerous behavior feature and the adjacent behavior feature are line pressing, normal driving, line pressing, and normal driving in turn, the corresponding continuous behavior feature is continuous lane changing.

[0061] The dangerous behavior feature can also include driving into a non-motor vehicle lane, at which time the corresponding continuous behavior feature can include temporary parking, long-time driving in a non-motor vehicle lane, etc., which are not exhaustively listed here.

[0062] In the case that the continuous behavior feature does not conform to the preset traffic behavior rule, the driving behavior corresponding to the continuous behavior feature is regarded as a risky driving behavior. The preset traffic behavior rule can include: prohibition of continuous lane changing, prohibition of long-time line pressing, etc., which are not exhaustively listed here.

[0063] In the case that the continuous behavior feature does not conform to the preset traffic behavior rule, the corresponding driving behavior is regarded as a risky driving behavior.

[0064] Through the above process, the risky driving behavior can be determined based on the continuous behavior feature, and then the user can be guided to correct the bad driving habits in the driving process and reduce the safety hazards.

[0065] In an embodiment, the method further includes: generating driving prompt information based on the risky driving behavior; and displaying the driving prompt information.

[0066] The driving prompt information can be in the form of text, sound, etc., which is not limited here.

[0067] In an embodiment, the displaying of the driving prompt information includes:

[0068] obtaining a preset information display mode;

[0069] rendering the driving prompt information based on the information display mode.

[0070] The preset information display mode can be set according to the form of the prompt information. For example, the driving prompt information in the form of text can be displayed through the vehicle-mounted screen, such as highlighting, dynamic prompting, etc. The driving prompt information in the form of sound can be voice broadcasting or prompt sound, etc., which are not limited herein.

[0071] As shown in the embodiment, Figure 4 The driving image determination method comprises the following steps:

[0072] S401: Obtain a driving video.

[0073] S402: Obtain a driving image based on the driving video and a preset image frame extraction rule.

[0074] When the driving video is obtained, it can be a video collected in real time based on a vehicle-mounted camera. Then, the obtained video can be segmented into a plurality of frames of images according to the preset image frame extraction rule, so as to facilitate the extraction of the behavior feature from the plurality of frames of images. The preset image frame extraction rule can be a time interval rule or a frequency rule. For example, a driving image is extracted every 5s based on the driving video. The interval time can be set according to requirements, which is not limited herein. The frequency rule can be that a fixed number of driving images are extracted in a unit of time, for example, 20 or 30 driving images are extracted per minute, and the specific number is not limited.

[0075] As shown in the embodiment, Figure 5 The present disclosure relates to a risk driving behavior determination device, which can comprise:

[0076] The behavior feature determination module 501 is configured to determine a behavior feature by using the driving image.

[0077] The adjacent behavior feature determination module 502 is configured to determine at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule when the behavior feature is a dangerous behavior feature. The adjacent behavior feature is a behavior feature in an adjacent frame of image.

[0078] The risk driving behavior determination module 503 is configured to determine a risk driving behavior by using the dangerous behavior feature and the adjacent behavior feature.

[0079] As shown in the embodiment, Figure 6 In an embodiment, the behavior feature determination module 501 comprises:

[0080] The first feature determination sub-module 601 is configured to input the driving image into a semantic segmentation model to obtain a first feature. The first feature is used to represent semantic information in the driving image.

[0081] The second feature determination sub-module 602 is configured to input the driving image into a depth recognition model to obtain a second feature, where the second feature is used to represent depth information of the driving image.

[0082] The behavior feature determination sub-module 603 is configured to obtain a behavior feature by using the first feature and the second feature.

[0083] As shown in FIG. 6, in an embodiment, the risk driving behavior determination module 503 includes: Figure 7

[0084] The continuous behavior feature construction sub-module 701 is configured to construct a continuous behavior feature by using the dangerous behavior feature and the adjacent behavior feature.

[0085] The risk driving behavior execution sub-module 702 is configured to, in a case where the continuous behavior feature does not conform to a preset traffic behavior rule, regard a driving behavior corresponding to the continuous behavior feature as a risk driving behavior.

[0086] In an embodiment, the method further includes:

[0087] The prompt information generation sub-module is configured to generate driving prompt information based on the risk driving behavior.

[0088] The display sub-module is configured to display the driving prompt information.

[0089] In an embodiment, the display sub-module includes:

[0090] The display sub-module is configured to obtain a preset information display mode.

[0091] The rendering sub-module is configured to render the driving prompt information based on the information display mode.

[0092] In an embodiment, the method for determining a driving image includes:

[0093] Obtaining a driving video.

[0094] Obtaining a driving image based on the driving video and a preset extraction rule.

[0095] In the technical solution of the present disclosure, the acquisition, storage and application of user personal information comply with relevant laws and regulations and do not violate public order and good customs.

[0096] According to the embodiments of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium and a computer program product.

[0097] Figure 8 ​A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0098] like Figure 8 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0099] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0100] The computing unit 801 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs various methods and processes described above, such as the method of determining risky driving behavior. For example, in some embodiments, the method of determining risky driving behavior can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded onto the RAM 803 and executed by the computing unit 801, one or more steps of the method of determining risky driving behavior described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the method of determining risky driving behavior by any other suitable means, such as by means of firmware.

[0101] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0102] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces a means for implementing the functions / acts specified in the flowcharts and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0103] In the context of this disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0104] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0105] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0106] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0107] It should be understood that the various forms of flow shown above can be used to reorder, add, or remove steps. For example, the steps described in the present disclosure can be performed in parallel, in series, or in a different order, without limitation herein, so long as the desired results of the technology described in the present disclosure are achieved.

[0108] The specific implementation described above does not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for determining a risky driving behavior, comprising: determining a behavior feature of a target vehicle by using a driving image of the target vehicle, wherein the driving image is an image of an outside of the target vehicle obtained in real time based on a vehicle-mounted camera, and the behavior feature is a behavior state feature determined based on a positional relationship between the target vehicle and road information; in a case where the behavior feature is a dangerous behavior feature, determining at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule, wherein the adjacent behavior feature is a behavior feature determined by using an adjacent frame driving image; determining a risky driving behavior of the target vehicle by using the dangerous behavior feature and the adjacent behavior feature; wherein the determining of the risky driving behavior by using the dangerous behavior feature and the adjacent behavior feature comprises: constructing a continuous behavior feature by using the dangerous behavior feature and the adjacent behavior feature, wherein the continuous behavior feature comprises one or more of normal lane changing, continuous lane changing, long-time lane pressing, temporary parking, and long-time driving on a non-motor vehicle lane; in a case where the continuous behavior feature does not conform to a preset traffic behavior rule, regarding a driving behavior corresponding to the continuous behavior feature as a risky driving behavior. 2.The method of claim 1, wherein the determining of the behavior feature comprises: inputting the driving image into a semantic segmentation model to obtain a first feature, wherein the first feature is used to represent semantic information in the driving image; inputting the driving image into a depth recognition model to obtain a second feature, wherein the second feature is used to represent depth information of the driving image; obtaining the behavior feature by using the first feature and the second feature. 3.The method of claim 1, further comprising: generating driving prompt information based on the risky driving behavior; and displaying the driving prompt information. 4.The method of claim 3, wherein the displaying of the driving prompt information comprises: obtaining a preset information display mode; and rendering the driving prompt information based on the information display mode. 5.The method of any one of claims 1-4, wherein the determining of the driving image comprises: obtaining a driving video; and obtaining the driving image based on the driving video and a preset image frame extraction rule. 6.An apparatus for determining a risky driving behavior, comprising: a behavior feature determination module configured to determine a behavior feature of a target vehicle by using a driving image of the target vehicle, wherein the driving image is an image of an outside of the target vehicle obtained in real time based on a vehicle-mounted camera, and the behavior feature is a behavior state feature determined based on a positional relationship between the target vehicle and road information; an adjacent behavior feature determination module configured to, in a case where the behavior feature is a dangerous behavior feature, determine at least one adjacent behavior feature adjacent to the dangerous behavior feature based on a preset rule, wherein the adjacent behavior feature is a behavior feature determined by using an adjacent frame driving image; a risky driving behavior determination module configured to determine a risky driving behavior of the target vehicle by using the dangerous behavior feature and the adjacent behavior feature. The risk driving behavior determination module comprises: a continuous behavior feature construction submodule, configured to construct a continuous behavior feature by using the dangerous behavior feature and the adjacent behavior feature; the continuous behavior feature comprises one or more of normal lane changing, continuous lane changing, long-time lane pressing, temporary parking, and long-time driving on a non-motor vehicle lane; a risk driving behavior execution submodule, configured to, in a case where the continuous behavior feature does not conform to a preset traffic behavior rule, take a driving behavior corresponding to the continuous behavior feature as a risk driving behavior.

7. The apparatus of claim 6, wherein the behavior feature determination module comprises: a first feature determination submodule, configured to input the driving image into a semantic segmentation model to obtain a first feature; the first feature is used to represent semantic information in the driving image; a second feature determination submodule, configured to input the driving image into a depth recognition model to obtain a second feature; the second feature is used to represent depth information of the driving image; a behavior feature determination submodule, configured to obtain the behavior feature by using the first feature and the second feature.

8. The apparatus of claim 6, wherein the risk driving behavior determination module comprises: a continuous behavior feature construction submodule, configured to construct a continuous behavior feature by using the dangerous behavior feature and the adjacent behavior feature; a risk driving behavior execution submodule, configured to, in a case where the continuous behavior feature does not conform to a preset traffic behavior rule, take a driving behavior corresponding to the continuous behavior feature as a risk driving behavior.

9. The apparatus of claim 6, further comprising: a prompt information generation submodule, configured to generate driving prompt information based on the risk driving behavior; a display submodule, configured to display the driving prompt information.

10. The apparatus of claim 9, wherein the display submodule comprises: a display submodule, configured to obtain a preset information display mode; a rendering submodule, configured to render the driving prompt information based on the information display mode.

11. The apparatus of any one of claims 6-10, wherein the driving image determination method comprises: obtaining a driving video; obtaining the driving image based on the driving video and a preset image frame extraction rule.

12. An electronic device, comprising: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

13. A non-transitory computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to make the computer execute the method of any one of claims 1-5.

14. A computer program product comprising a computer program which, when executed by a processor, implements the method of any one of claims 1-5.

Images