Vehicle driving path determination method and system, vehicle and storage medium

Abstract

The invention discloses a vehicle driving path determination method and system, a vehicle and a storage medium, and the method comprises the steps: firstly collecting the image information during the driving of the vehicle; when the collected image information does not meet the preset definition requirement, on one hand, collecting point cloud information when the vehicle runs, on the other hand, determining the positioning position of the vehicle, and extracting environment characteristics of the positioning position; the point cloud information and the image information are synchronized to obtain actual environment characteristics of the position where the vehicle is located; and finally, carrying out information fusion on the environment characteristics and the actual environment characteristics, determining the actual positioning position and the environment characteristics of the vehicle, and determining a lane line and a driving path according to the actual positioning position and the environment characteristics of the vehicle. By synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then the point cloud information and the image information are fused with the environment characteristics of the positioning position, so that positioning and perceiving are more accurate, and the lane line and the driving path can be accurately determined.

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and in particular to a method, system, vehicle and storage medium for determining a vehicle driving path. Background Art

[0002] Currently, autonomous driving has become a new trend in vehicles. Users only need to enter the destination, and the intelligent driving system can automatically plan the route, drive automatically, avoid obstacles, etc., to take the user to the destination.

[0003] Current intelligent driving systems mostly utilize a multi-sensor plus high-precision map strategy for route planning. However, this strategy primarily relies on high-precision maps for positioning, and is unable to achieve the integration of perception and high-precision maps, or unify them in time and space, resulting in inaccurate perception and positioning. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention aims to provide a method, system, vehicle and storage medium for determining a vehicle driving path.

[0005] The present invention proposes a method for determining a vehicle driving path, comprising: collecting image information of the vehicle while driving; when the image information does not meet a preset clarity requirement, collecting point cloud information of the vehicle while driving, determining the positioning position of the vehicle, and extracting environmental features of the positioning position; synchronizing the point cloud information and the image information to obtain actual environmental features of the vehicle's location; fusing the environmental features with the actual environmental features to determine the actual positioning position of the vehicle and its environmental features, and determining the lane line and driving path based on the actual positioning position and its environmental features.

[0006] According to the method for determining the vehicle driving path of an embodiment of the present invention, image information of the vehicle while driving is first collected as a basis for determining whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, point cloud information of the vehicle while driving is collected on the one hand, and the vehicle's positioning position is determined on the other hand, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the vehicle's actual positioning position and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0007] In addition, the method for determining a vehicle driving path according to an embodiment of the present invention may also have the following additional technical features: Furthermore, the method for determining the vehicle driving path also includes: when the image information meets the clarity requirement, determining the lane line and driving path based on the image information; thereby helping to simplify the process and improve the vehicle information processing speed.

[0008] Furthermore, the determination of the positioning position of the vehicle includes: positioning the vehicle based on a preset first positioning system to determine the absolute positioning position and relative positioning position of the vehicle; correcting the relative positioning position based on a preset second positioning system to obtain a corrected relative positioning position; determining the positioning position of the vehicle based on the absolute positioning position and the corrected relative positioning position; thereby, by determining the absolute positioning position and relative positioning position of the vehicle through the first positioning system and correcting the relative positioning position through the second positioning system, the vehicle can be accurately positioned, which facilitates the subsequent extraction of environmental features of the positioning position, and further facilitates the accurate determination of the vehicle's driving path.

[0009] Furthermore, the extracting of the environmental features of the positioning position includes: fusing the positioning position with a pre-stored map to extract the environmental features of the positioning position, wherein the map includes different positions and their corresponding environmental features; thereby, the environmental features can be extracted based on the accurate positioning position, facilitating the subsequent determination of the vehicle's driving path based on the environmental features.

[0010] Furthermore, the synchronizing the point cloud information and the image information to obtain the actual environmental characteristics of the vehicle's location includes: synchronizing the original data of the point cloud information and the image information in time and space based on a preset network model to obtain the actual environmental characteristics of the vehicle's location; thereby synchronizing the original data of the point cloud information and the image information in time and space based on the preset network model, using the point cloud information to compensate for the problem of unclear image information, ensuring that the perceived surrounding environment is output more accurately, thereby facilitating the subsequent accurate determination of the vehicle's driving path.

[0011] Furthermore, the network model includes a BEV+Transformer model; thereby, based on the BEV+Transformer model, the original data of the point cloud information and the image information are synchronized in time and space, and the point cloud information is used to compensate for the problem of unclear image information, ensuring that the output of the perceived surrounding environment is more accurate, thereby facilitating the subsequent accurate determination of the vehicle's driving path.

[0012] Furthermore, when the brightness and / or contrast of the lane lines in the image information do not meet the preset conditions, it is determined that the image information does not meet the preset clarity requirements; thereby, by setting the preset conditions, it is convenient to judge whether the image information meets the preset clarity requirements based on whether the brightness and / or contrast of the lane lines in the image information meet the preset conditions, which serves as the basis for whether to switch the path determination method, and helps to select an appropriate path determination method.

[0013] In response to the above-mentioned problems, the present invention also proposes a vehicle driving path determination system, comprising: an acquisition module for acquiring image information of the vehicle when it is driving; a processing module for acquiring point cloud information of the vehicle when it is driving, and determining the positioning position of the vehicle and extracting environmental features of the positioning position when the image information does not meet the preset clarity requirements; a synchronization module for synchronizing the point cloud information and the image information to obtain the actual environmental features of the vehicle; a determination module for fusing the environmental features with the actual environmental features to determine the actual positioning position of the vehicle and its environmental features, and determining the lane line and driving path based on the actual positioning position and its environmental features.

[0014] According to an embodiment of the present invention, a vehicle driving path determination system executes the vehicle driving path determination method of the above embodiment, firstly, by collecting image information of the vehicle while driving, as the basis for determining whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle while driving is collected, and on the other hand, the vehicle's positioning position is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0015] In response to the above-mentioned problems, the present invention also proposes a vehicle, comprising: a vehicle driving path determination system as described in the above-mentioned second aspect embodiment of the present invention, or the vehicle comprises: a processor, a memory, and a vehicle driving path determination program stored in the memory and executable on the processor, wherein the vehicle driving path determination program, when executed by the processor, implements the vehicle driving path determination method as described in the above-mentioned first aspect embodiment of the present invention.

[0016] According to an embodiment of the present invention, a vehicle is provided with the vehicle driving path determination system of the above embodiment, and executes the vehicle driving path determination method of the above embodiment, firstly, by collecting image information when the vehicle is driving, as a basis for judging whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle is collected when it is driving, and on the other hand, the positioning position of the vehicle is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined according to the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0017] In response to the above-mentioned problems, the present invention also proposes a computer-readable storage medium, on which a program for determining a vehicle driving path is stored. When the program for determining a vehicle driving path is executed by a processor, the method for determining a vehicle driving path as described in the first aspect of the present invention is implemented.

[0018] According to the computer-readable storage medium of an embodiment of the present invention, when the vehicle driving path determination program stored thereon is executed by a processor, the vehicle driving path determination method of the above embodiment is executed, firstly, image information of the vehicle when driving is collected as the basis for judging whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle when driving is collected, and on the other hand, the positioning position of the vehicle is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0019] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments with reference to the accompanying drawings, in which: Figure 1 is a schematic diagram of a drive system of a hybrid vehicle according to one embodiment of the present invention; Figure 2 is a flow chart of a method for determining a vehicle driving path according to another embodiment of the present invention; Figure 3 is a flow chart of a method for determining a vehicle driving path according to a specific embodiment of the present invention; Figure 4 4 is a structural block diagram of a system for determining a vehicle driving path according to an embodiment of the present invention.

[0021] Reference numerals: 100 - vehicle driving path determination system; 110 - acquisition module; 120 - processing module; 130 - synchronization module; 140 - determination module. DETAILED DESCRIPTION

[0022] The embodiments of the present invention will be described in detail below. The embodiments described with reference to the accompanying drawings are exemplary. The embodiments of the present invention will be described in detail below.

[0023] The existing intelligent driving solutions still use the multi-sensor plus high-precision map strategy based on the traditional 2D+CNN algorithm. Even with the support of high-precision maps, the perception effect is not ideal, especially for the high-level intelligent driving L3 level, which is obviously not sufficient in terms of computing power and perception. At the positioning level, it mainly relies on high-precision maps, and cannot achieve the integration of perception and high-precision maps, and cannot achieve temporal unification from the time and space dimensions, and cannot accurately perceive and locate.

[0024] In view of the above problems, the present invention provides a method, system, vehicle and storage medium for determining a vehicle driving path, which are described below with reference to Figures 1-4 A method, system, vehicle, and storage medium for determining a vehicle driving path according to an embodiment of the present invention are described. In the following embodiments of the present invention, the vehicle may be any one of a fuel vehicle, an electric vehicle, and a hybrid vehicle.

[0025] Figure 1 FIG. 1 is a flow chart of a method for determining a vehicle driving path according to an embodiment of the present invention. Figure 1 As shown, a method for determining a vehicle driving path according to an embodiment of the present invention includes the following steps: Step S1: collecting image information of the vehicle while it is traveling.

[0026] In a specific embodiment, image information of a vehicle while it is traveling may be collected by an image acquisition device such as a vehicle-mounted camera, etc. Specifically, the image information is a 2D image.

[0027] Specifically, by collecting image information while the vehicle is traveling, as a basis for determining whether to switch the path determination method, it is helpful to select an appropriate path determination method.

[0028] Step S2: When the image information does not meet the preset clarity requirement, the point cloud information of the vehicle while driving is collected, the positioning position of the vehicle is determined, and the environmental features of the positioning position are extracted.

[0029] In a specific embodiment, when the collected image information does not meet the preset clarity requirements, point cloud information of the vehicle while in motion is collected, and the vehicle's location is determined, and the environmental characteristics of the location are extracted. Specifically, when the collected image information is dim or has low contrast, it may be considered that it does not meet the preset clarity requirements. The preset clarity requirements can be set by the operator and stored in a relevant vehicle controller, such as a vehicle controller. Point cloud information of the vehicle while in motion can be collected using point cloud collection equipment, such as an on-board laser radar. The vehicle's location can be determined by combining it with an on-board high-precision map, and the environmental characteristics of the location can be extracted.

[0030] Specifically, when the collected image information does not meet the preset clarity requirements, on the one hand, the point cloud information of the vehicle is collected while it is driving, and on the other hand, the vehicle's positioning position is determined and the environmental characteristics of the positioning position are extracted; this facilitates the subsequent accurate determination of the vehicle's driving path based on the point cloud information and environmental characteristics.

[0031] Step S3: Synchronize the point cloud information and image information to obtain the actual environmental characteristics of the vehicle's location.

[0032] In a specific embodiment, an algorithm can be used to first derive 3D data from image information. The point cloud information and the 3D image data can then be synchronized in time and space to capture the actual environmental characteristics of the vehicle's location, ensuring a more accurate output of the perceived surrounding environment. Specifically, the synchronization of point cloud information and image information can be achieved using a BEV (Bird's Eye View) + Transformer algorithm, including CNN (Convolutional Neural Networks), DNN (Deep Neural Networks), and RNN (Recurrent Neural Networks).

[0033] Specifically, point cloud information and image information are synchronized to obtain the actual environmental characteristics of the vehicle's location, so as to use point cloud information to make up for the unclear image information and ensure that the perception of the surrounding environment is output more accurately, thereby facilitating the subsequent accurate determination of the vehicle's driving path.

[0034] Step S4: Fusing the environmental features and the actual environmental features to determine the actual positioning position of the vehicle and its environmental features, and determining the lane line and driving path based on the actual positioning position and its environmental features.

[0035] Specifically, the environmental characteristics and actual environmental characteristics are integrated to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and driving path are determined based on the actual positioning position and its environmental characteristics. In this way, by synchronizing point cloud information and image information, the surrounding environment can be perceived in real time, and then integrated with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and driving path can be accurately determined.

[0036] Therefore, according to the method for determining the vehicle driving path of an embodiment of the present invention, image information of the vehicle while driving is first collected as the basis for determining whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, point cloud information of the vehicle while driving is collected on the one hand, and the vehicle's positioning position is determined on the other hand, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the vehicle's actual positioning position and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0037] Figure 2 FIG. 1 is a flow chart of a method for determining a vehicle driving path according to another embodiment of the present invention. Figure 2 As shown, in one embodiment of the present invention, the method for determining a vehicle driving path further includes step S5: when the image information meets the clarity requirement, determining the lane line and the driving path according to the image information.

[0038] In a specific embodiment, when the captured image information meets preset clarity requirements, lane lines and driving paths can be directly determined based on the image information. This avoids operations such as point cloud information collection, positioning location determination, and unnecessary information synchronization and fusion. This prevents resource waste, simplifies the process, and improves vehicle information processing speed. Specifically, for example, when the captured image information has high brightness and high contrast, it can be considered to meet the preset clarity requirements. The preset clarity requirements can be set by the operator and stored in a relevant vehicle controller, such as the vehicle controller.

[0039] Specifically, according to the method for determining the vehicle driving path of an embodiment of the present invention, when the collected image information meets the preset clarity requirements, the lane lines and driving path can be directly determined based on the image information, which helps to simplify the process and improve the vehicle information processing speed.

[0040] In one embodiment of the present invention, step S2 determines the positioning position of the vehicle, including: positioning the vehicle based on a preset first positioning system to determine the absolute positioning position and relative positioning position of the vehicle; correcting the relative positioning position based on a preset second positioning system to obtain a corrected relative positioning position; and determining the positioning position of the vehicle based on the absolute positioning position and the corrected relative positioning position.

[0041] In a specific embodiment, the vehicle is first positioned based on a preset first positioning system to determine the vehicle's absolute and relative positions. The relative position is then corrected based on a preset second positioning system to obtain a corrected relative position. Finally, the vehicle's position is determined based on the absolute and corrected relative positions. Specifically, the first positioning system may be, for example, a GNSS (Global Navigation Satellite System) + IMU (Inertial Measurement Unit), and the second positioning system may be, for example, a PTK (Positioning, Tracking, and Navigation) positioning system.

[0042] In a specific embodiment, the GNSS+IMU combined navigation system is a navigation technology that combines the Global Navigation Satellite System (GNSS) and the Inertial Measurement Unit (IMU). GNSS determines geographic location by receiving satellite signals, while the IMU provides precise position and attitude information in a short period of time by measuring acceleration and angular velocity. The combination of the two can provide continuous, high-precision positioning and navigation services. GNSS determines geographic location by receiving satellite signals, providing long-distance and high-precision positioning information. However, GNSS performs poorly in signal-blocked or weak signal environments. The IMU, by measuring acceleration and angular velocity, can provide precise position and attitude information in a short period of time, but errors accumulate over long periods of time. Combining the two, it is possible to rely on GNSS for positioning when the GNSS signal is good, and rely on the IMU for position estimation when the signal is poor, thereby achieving continuous, high-precision navigation. GNSS+IMU combined navigation can be used in self-driving cars. GNSS+IMU combined navigation can provide highly precise position and attitude information, helping vehicles accurately navigate and avoid obstacles in various environments.

[0043] In a specific embodiment, PTK is a revolutionary GPS application. Unlike traditional static, fast static, and dynamic measurement methods, it achieves real-time centimeter-level positioning accuracy without the need for complex post-calculation. Its core principle is to utilize carrier phase dynamic real-time differential technology. The base station and rover exchange data in real time via a wireless communication network. The user receiver combines the received satellite signals with the base station signals to calculate precise coordinate increments, thereby achieving centimeter-level three-dimensional positioning. The RTK workflow involves setting up a base station, which can transmit received satellite signals to the rover via a data link at a known or unknown location. After receiving these signals, the rover compares them with its own satellite signals in real time to calculate precise positioning information. Even with large station spacing, such as 30 kilometers, RTK can ensure planar accuracy of 1-2 centimeters and even provide centimeter-level three-dimensional positioning. In RTK operation mode, this real-time feedback and data sharing capability makes the measurement process more efficient, greatly improving measurement accuracy and real-time performance, and is of great value in application scenarios requiring high-precision positioning, such as surveying and mapping, agriculture, and engineering.

[0044] Specifically, according to the method for determining the vehicle driving path of an embodiment of the present invention, when determining the positioning position of the vehicle, the vehicle is first positioned based on a preset first positioning system to determine the absolute positioning position and relative positioning position of the vehicle; then, based on a preset second positioning system, the relative positioning position is corrected to obtain the corrected relative positioning position; finally, the positioning position of the vehicle is determined based on the absolute positioning position and the corrected relative positioning position; thereby, by determining the absolute positioning position and relative positioning position of the vehicle through the first positioning system and correcting the relative positioning position through the second positioning system, the vehicle can be accurately positioned, which facilitates the subsequent extraction of environmental features of the positioning position, and further facilitates the accurate determination of the vehicle's driving path.

[0045] In one embodiment of the present invention, step S2 extracts environmental features of the positioning location, including: fusing information of the positioning location with a pre-stored map to extract the environmental features of the positioning location, wherein the map includes different locations and their corresponding environmental features.

[0046] In a specific embodiment, the positioning location is integrated with a pre-stored map to extract the environmental characteristics of the positioning location. Specifically, the pre-stored map is, for example, a vehicle-mounted high-precision map that includes different locations and their corresponding environmental characteristics. The environmental characteristics can be mapped based on the positioning location.

[0047] Specifically, according to the method for determining the vehicle driving path of an embodiment of the present invention, the positioning position is integrated with the pre-stored map information to extract the environmental characteristics of the positioning position; thereby, the environmental characteristics can be extracted based on the accurate positioning position, which facilitates the subsequent determination of the vehicle's driving path based on the environmental characteristics.

[0048] In one embodiment of the present invention, step S3 synchronizes the point cloud information and the image information to obtain the actual environmental characteristics of the vehicle's location, including: based on a preset network model, synchronizing the original data of the point cloud information and the image information in time and space to obtain the actual environmental characteristics of the vehicle's location.

[0049] In a specific embodiment, based on a preset network model, the point cloud information and image information are synchronized in time and space to obtain the actual environmental characteristics of the vehicle's location. Specifically, the preset network model is, for example, BEV+Transformer.

[0050] Specifically, according to the method for determining the vehicle driving path of an embodiment of the present invention, based on a preset network model, the original data of point cloud information and image information are synchronized in time and space to obtain the actual environmental characteristics of the vehicle's location; in this way, based on the preset network model, the original data of point cloud information and image information are synchronized in time and space, and the point cloud information is used to compensate for the problem of unclear image information, ensuring that the perception of the surrounding environment is output more accurately, thereby facilitating the subsequent accurate determination of the vehicle's driving path.

[0051] In one embodiment of the present invention, the network model includes a BEV+Transformer model.

[0052] In a specific embodiment, BEV+Transformer is an autonomous driving perception technology that combines a bird's-eye view (BEV) and a Transformer model, aiming to enhance the autonomous driving system's ability to understand and predict complex traffic scenarios. It achieves high-precision perception of the environment and decision support by fusing multi-sensor data into a unified bird's-eye view and using the Transformer's self-attention mechanism to process global information. Its core lies in integrating traditional multi-sensor data (such as cameras, lidar, etc.) into a bird's-eye view through geometric projection and coordinate transformation, and processing this information in combination with the Transformer model. It not only provides a global perspective, but also can capture different objects in the environment. The BEV can understand the spatial relationship between objects, thereby enhancing the system's perception ability; its technical advantages are: 1. Global perspective and spatial understanding: The BEV perspective provides a global view of the vehicle's surroundings, and the Transformer's self-attention mechanism can effectively capture the spatial relationship between objects, enabling the system to more comprehensively understand complex traffic scenarios; 2. Multimodal data fusion: BEV+Transformer can uniformly process data from different sensors, such as visual images, point clouds, etc., improving the accuracy and robustness of perception; 3. Efficient prediction ability: By combining BEV information and Transformer models, the system can more accurately predict the movement trajectories of vehicles, pedestrians, etc., and provide support for autonomous driving decisions.

[0053] Specifically, according to the method for determining the vehicle driving path of an embodiment of the present invention, based on the BEV+Transformer model, the original data of point cloud information and image information are synchronized in time and space to obtain the actual environmental characteristics of the vehicle's location; thereby, based on the BEV+Transformer model, the original data of point cloud information and image information are synchronized in time and space, and the point cloud information is used to compensate for the problem of unclear image information, ensuring that the output of the perceived surrounding environment is more accurate, thereby facilitating the subsequent accurate determination of the vehicle's driving path.

[0054] In one embodiment of the present invention, when the brightness and / or contrast of the lane lines in the image information do not meet a preset condition, it is determined that the image information does not meet a preset clarity requirement.

[0055] In a specific embodiment, when the brightness and / or contrast of the lane lines in the image information do not meet preset conditions, the image information is determined to not meet preset clarity requirements. Specifically, whether the image information meets the preset conditions can be determined by setting a brightness threshold and / or a contrast threshold. For example, when the brightness and contrast of the lane lines in the image information meet the brightness threshold and the contrast threshold, respectively, the image information can be determined to meet the preset clarity requirements. When the brightness of the lane lines in the image information does not meet the brightness threshold, or the contrast of the lane lines in the image information does not meet the contrast threshold, the image information can be determined to not meet the preset clarity requirements. The preset conditions can be set by an operator and stored in a relevant controller of the vehicle, such as a vehicle controller.

[0056] Specifically, according to the method for determining a vehicle driving path in an embodiment of the present invention, when the brightness and / or contrast of the lane lines in the image information do not meet the preset conditions, it is determined that the image information does not meet the preset clarity requirements; thereby, by setting the preset conditions, it is convenient to judge whether the image information meets the preset clarity requirements based on whether the brightness and / or contrast of the lane lines in the image information meet the preset conditions, which serves as the basis for whether to switch the path determination method, thereby facilitating the selection of an appropriate path determination method.

[0057] Figure 3 FIG. 1 is a flow chart of a method for determining a vehicle driving path according to a specific embodiment of the present invention. Figure 3 As shown, in this specific embodiment, the method for determining the vehicle driving path is explained from the positioning level, perception level, and fusion and matching level.

[0058] In this specific embodiment, Figure 3 The judgment conditions are: lane lines in heavy rain and fog / lane lines are unclear, and the shadow of the guardrail forms the lane line.

[0059] In this specific embodiment, the positioning layer includes: positioning the longitude and latitude through GNSS (GPS / Beidou), making a "position estimate" for the relative position and absolute position of the vehicle, and then "correcting" the relative position through PTK positioning to obtain the "position prediction value" of the current vehicle. The position prediction value is integrated with the high-precision map through the algorithm to comprehensively obtain the "environmental characteristics of the predicted position" around the vehicle.

[0060] In this specific embodiment, the perception layer includes: using the camera's planar data (through CNN, DNN, RNN) algorithm to obtain 3D data, the 3D point cloud data of the lidar, the data of the lidar and camera through the BEV+Transformer algorithm, synchronizing the raw data in time and space in the BEV space, and outputting more accurate perception of the surrounding environment.

[0061] In this specific embodiment, the fusion and matching level includes: performing path planning through the fusion and matching of the positioning level and the perception level, outputting the current position and the surrounding perception environment, and achieving real-time path planning and outputting lane lines.

[0062] In this specific embodiment, it can be seen that the method for determining the vehicle driving path provided by this specific embodiment integrates multiple cameras and lidar in the BEV 3D space through the BEV fusion route, perceives the surrounding environment in real time, and achieves more accurate positioning and perception with the support of GNSS+IMU+RTK, which is more conducive to end-to-end L3 driving. It can also drive normally without high-precision maps, and can avoid the disadvantages when high-precision maps are not working. Specifically, it has the following advantages: 1. In heavy rain and fog, the data collected by the camera has poor ability to recognize lane lines. The 3D point cloud data of the lidar is used to compensate, and point cloud cutting and point cloud compensation are used to recognize lane lines on the BEV to ensure normal driving and performance. 2. Under the shadow of the guardrail, lane line recognition is prone to misidentification. The BEV+Trangformer perception algorithm is combined with the lidar point cloud to compensate, and the time and space dimensions are unified in the BEV space to effectively filter out non-true values.

[0063] In summary, according to the method for determining the vehicle driving path of an embodiment of the present invention, image information of the vehicle when it is driving is first collected as the basis for determining whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle when it is driving is collected, and on the other hand, the positioning position of the vehicle is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's location; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0064] A further embodiment of the present invention also discloses a system for determining a vehicle's travel path. Figure 4 FIG. 1 is a structural block diagram of a vehicle driving path determination system according to an embodiment of the present invention. Figure 4 As shown, the vehicle driving path determination system 100 includes: a collection module 110 , a processing module 120 , a synchronization module 130 and a determination module 140 .

[0065] Specifically, the acquisition module 110 is used to acquire image information when the vehicle is traveling.

[0066] The processing module 120 is used to collect point cloud information of the vehicle while it is traveling, determine the vehicle's location, and extract environmental features of the location when the image information does not meet the preset clarity requirements.

[0067] The synchronization module 130 is used to synchronize the point cloud information and the image information to obtain the actual environmental characteristics of the vehicle's location.

[0068] The determination module 140 is used to fuse the environmental characteristics and the actual environmental characteristics to determine the actual positioning position of the vehicle and its environmental characteristics, and determine the lane line and driving path based on the actual positioning position and its environmental characteristics.

[0069] In one embodiment of the present invention, the determination module 140 is further configured to determine the lane line and the driving path based on the image information when the image information meets the clarity requirement.

[0070] In one embodiment of the present invention, the processing module 120 determines the positioning position of the vehicle, including: positioning the vehicle based on a preset first positioning system to determine the absolute positioning position and relative positioning position of the vehicle; correcting the relative positioning position based on a preset second positioning system to obtain a corrected relative positioning position; and determining the positioning position of the vehicle based on the absolute positioning position and the corrected relative positioning position.

[0071] In one embodiment of the present invention, the processing module 120 extracts the environmental features of the positioning location, including: fusing the positioning location with a pre-stored map to extract the environmental features of the positioning location, wherein the map includes different locations and their corresponding environmental features.

[0072] In one embodiment of the present invention, the synchronization module 130 synchronizes the point cloud information and the image information to obtain the actual environmental characteristics of the vehicle's location, including: based on a preset network model, synchronizing the original data of the point cloud information and the image information in time and space to obtain the actual environmental characteristics of the vehicle's location.

[0073] In one embodiment of the present invention, the network model includes a BEV+Transformer model.

[0074] In one embodiment of the present invention, when the brightness and / or contrast of the lane line in the image information does not meet a preset condition, it is determined that the image information does not meet a preset clarity requirement.

[0075] According to an embodiment of the present invention, the vehicle driving path determination system 100 executes the vehicle driving path determination method of the above embodiment, firstly, by collecting image information of the vehicle while driving, as the basis for determining whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle while driving is collected, and on the other hand, the vehicle's positioning position is determined and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0076] A further embodiment of the present invention also discloses a vehicle.

[0077] In some embodiments, the vehicle includes: a vehicle driving path determination system 100 as described in any of the above embodiments of the present invention.

[0078] In other embodiments, the vehicle includes: a processor, a memory, and a vehicle driving path determination program stored in the memory and executable on the processor. When the vehicle driving path determination program is executed by the processor, the vehicle driving path determination method described in any of the above embodiments of the present invention is implemented.

[0079] In a specific embodiment, the vehicle may be any one of a fuel vehicle, an electric vehicle, or a hybrid vehicle.

[0080] According to an embodiment of the present invention, a vehicle is provided with the vehicle driving path determination system 100 in the above embodiment to execute the vehicle driving path determination method in the above embodiment, firstly, by collecting image information when the vehicle is driving, as a basis for judging whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle is collected when it is driving, and on the other hand, the positioning position of the vehicle is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0081] A further embodiment of the present invention also discloses a computer-readable storage medium, on which a program for determining a vehicle driving path is stored. When the program for determining a vehicle driving path is executed by a processor, the method for determining a vehicle driving path as described in any of the above embodiments of the present invention is implemented.

[0082] According to the computer-readable storage medium of an embodiment of the present invention, when the vehicle driving path determination program stored thereon is executed by a processor, the vehicle driving path determination method of the above embodiment is executed, firstly, image information of the vehicle when driving is collected as the basis for judging whether to switch the path determination method; when the collected image information does not meet the preset clarity requirements, on the one hand, point cloud information of the vehicle when driving is collected, and on the other hand, the positioning position of the vehicle is determined, and the environmental characteristics of the positioning position are extracted; then the point cloud information and the image information are synchronized to obtain the actual environmental characteristics of the vehicle's position; finally, the environmental characteristics and the actual environmental characteristics are fused to determine the actual positioning position of the vehicle and its environmental characteristics, and the lane line and the driving path are determined based on the actual positioning position and its environmental characteristics; in this way, by synchronizing the point cloud information and the image information, the surrounding environment can be perceived in real time, and then fused with the environmental characteristics of the positioning position to ensure more accurate positioning and perception, so that the lane line and the driving path can be accurately determined.

[0083] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "example," "specific example," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example.

[0084] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. A method for determining a vehicle driving path, characterized in that: include: collecting image information of the vehicle while it is traveling; When the image information does not meet the preset clarity requirement, collecting point cloud information of the vehicle while it is traveling, determining the location of the vehicle, and extracting environmental features of the location; Synchronizing the point cloud information and the image information to obtain actual environmental characteristics of the vehicle; The environmental features and the actual environmental features are fused to determine the actual positioning position of the vehicle and its environmental features, and the lane line and driving path are determined based on the actual positioning position and its environmental features.

2. The method for determining a vehicle driving path according to claim 1, wherein: Also includes: When the image information meets the clarity requirement, the lane line and the driving path are determined according to the image information.

3. The method for determining a vehicle driving path according to claim 1, wherein: Determining the positioning position of the vehicle includes: Positioning the vehicle based on a preset first positioning system to determine the absolute positioning position and relative positioning position of the vehicle; Based on a preset second positioning system, the relative positioning position is corrected to obtain a corrected relative positioning position; The positioning position of the vehicle is determined based on the absolute positioning position and the corrected relative positioning position.

4. The method for determining a vehicle driving path according to claim 1, wherein: The extracting the environmental features of the positioning location includes: The positioning location is fused with information from a pre-stored map to extract environmental features of the positioning location, wherein the map includes different locations and their corresponding environmental features.

5. The method for determining a vehicle driving path according to claim 1, wherein: The step of synchronizing the point cloud information and the image information to obtain actual environmental characteristics of the vehicle's location includes: Based on a preset network model, the point cloud information and the image information are synchronized with each other in time and space to obtain the actual environmental characteristics of the vehicle's location.

6. The method for determining a vehicle driving path according to claim 5, wherein: The network model includes a BEV+Transformer model.

7. The method for determining a vehicle driving path according to claim 1, wherein: When the brightness and / or contrast of the lane lines in the image information do not meet a preset condition, it is determined that the image information does not meet a preset clarity requirement.

8. A vehicle travel path determination system, characterized in that: include: An acquisition module, configured to acquire image information of the vehicle while it is traveling; a processing module, configured to, when the image information does not meet a preset clarity requirement, collect point cloud information of the vehicle while it is traveling, determine the location of the vehicle, and extract environmental features of the location; A synchronization module, configured to synchronize the point cloud information with the image information to obtain actual environmental characteristics of the vehicle's location; The determination module is used to fuse the environmental characteristics and the actual environmental characteristics to determine the actual positioning position of the vehicle and its environmental characteristics, and determine the lane line and driving path based on the actual positioning position and its environmental characteristics.

9. A vehicle, characterized in that: include: The vehicle travel path determination system according to claim 8; or, A processor, a memory, and a vehicle driving path determination program stored in the memory and executable on the processor, wherein the vehicle driving path determination program, when executed by the processor, implements the vehicle driving path determination method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a vehicle driving path determination program, and when the vehicle driving path determination program is executed by a processor, the vehicle driving path determination method according to any one of claims 1 to 7 is implemented.

Images