A vehicle intelligent driving guarantee method and system based on road feature fusion

Abstract

The present application discloses a vehicle intelligent driving assurance method based on road feature fusion, the method comprising: acquiring visual road information around the vehicle and determining the confidence of the visual road information, determining the positioning information of the vehicle and determining the positioning accuracy, determining the map road information around the vehicle according to the positioning information, performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result, determining the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the visual road information, the map road information and the road feature fusion result, so that various modes of intelligent driving are more flexible and intelligent, and the safety of intelligent driving is effectively guaranteed.

Description

Technical Field

[0001] The present application relates to the field of intelligent driving technology, and in particular to a vehicle intelligent driving guarantee method and system based on road feature fusion. Background Art

[0002] With the development of intelligent driving technology, more and more sensors are applied to vehicles. In order to improve the accuracy of perception results and reduce the impact of single sensor failure or misidentification, a variety of redundant sensing components are usually used to work together to perceive environmental information. Road feature information is one of the basic information for realizing intelligent driving. It is usually collected through cameras and high-precision maps to guide the intelligent driving of vehicles. The advantages of cameras are strong real-time perception and comprehensive available scenarios, but they are easily affected by weather, light, etc. and have a limited field of view. High-precision maps have high detection accuracy and long distance, but poor timeliness and rely on positioning accuracy.

[0003] In the prior art, when using road characteristic information obtained by cameras and high-precision maps to guide vehicle intelligent driving, the focus is usually on verifying the effectiveness of the road characteristic information obtained by cameras and high-precision maps, and selecting one of the road characteristic information as the basis for guiding vehicle intelligent driving based on the effectiveness. This selection method has a single judgment mechanism and is not very risk-resistant, and cannot guarantee the safety of vehicle intelligent driving under complex driving conditions. Summary of the invention

[0004] The main purpose of this application is to provide a vehicle intelligent driving assurance method and system based on road feature fusion, aiming to solve the technical problem in the prior art that the judgment mechanism for perceiving environmental information during vehicle intelligent driving is single and has low risk resistance.

[0005] In a first aspect, the present application provides a vehicle intelligent driving assurance method based on road feature fusion, the method comprising the following steps:

[0006] Acquiring visual road information around the vehicle and determining the confidence of the visual road information;

[0007] Determining the positioning information of the vehicle and determining the positioning accuracy;

[0008] Determining map road information around the vehicle according to the positioning information;

[0009] Performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result;

[0010] The intelligent driving level of the vehicle and the basis for intelligent driving path planning are determined according to the fusion result of the visual road information, the map road information and the road features.

[0011] In some embodiments, the determining the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the fusion result of the visual road information, the map road information and the road features includes:

[0012] Selecting different fusion strategies according to the visual road information and its confidence, the positioning accuracy and the map road information, and determining a fusion output result according to the consistency of the visual road information and the lane line cubic curve in the map road information, so as to determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning;

[0013] If the confidence of the visual road information is high, the positioning accuracy is high accuracy, and the consistency degree meets the consistency requirement, then the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level,

[0014] And use this as the optimal integration guarantee strategy;

[0015] If the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency degree meets the consistency requirement, and the visual lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a high confidence level;

[0016] If the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency meets the consistency requirement, and the map lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a medium confidence level;

[0017] If the confidence of the visual road information is medium and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; when the consistency requirement is not met, the visual lane line or the map lane line is determined as the basis for intelligent driving path planning according to the cubic curve parameters of the lane line, and the intelligent driving level is a low confidence level;

[0018] If the confidence of the visual road information is low and the positioning accuracy is high, then when the consistency meets the consistency requirement, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a high confidence level; when the consistency requirement is not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level;

[0019] If the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency level meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level, and when the consistency requirement is not met, the intelligent driving level is unavailable;

[0020] If the positioning information or the map road information is not available, the visual lane lines are used as the basis for intelligent driving path planning, and the confidence level of the visual road information is used as the confidence level of intelligent driving.

[0021] In some embodiments, acquiring visual road information around the vehicle and determining the confidence of the visual road information includes:

[0022] If the visual lane lines in the visual road information are complete, determining that the confidence level of the visual road information is high;

[0023] If the visual lane line missing area is smaller than a preset area, it is determined that the confidence of the visual road information is medium;

[0024] If the visual lane line missing area is greater than a preset area, it is determined that the confidence of the visual road information is low.

[0025] In some embodiments, the feature fusion of the visual road information and the map road information to obtain a road feature fusion result also includes: when the visual road information has a medium confidence level or a low confidence level, the missing part of the visual lane line is predicted and completed, and then the consistency between the visual road information and the map road information is determined.

[0026] In some embodiments, determining the positioning information of the vehicle and determining the positioning accuracy includes:

[0027] Parameters of the positioning information sent by the positioning device are obtained, and the positioning accuracy is determined according to the parameters of the positioning information.

[0028] In some embodiments, the step of performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result further includes:

[0029] When the positioning accuracy is low, the current position information of the vehicle is calculated according to the track of the vehicle, and the map road information is determined according to the calculated position information.

[0030] In some embodiments, determining the consistency of cubic curve parameters of the lane lines in the visual road information and the map road information by feature fusion includes:

[0031] fusing and comparing the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information;

[0032] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information meets a preset value, determining that the degree of consistency meets the consistency requirement;

[0033] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information does not meet a preset value, determining that the degree of consistency does not meet the consistency requirement;

[0034] In some embodiments, the method further includes: determining a time difference between the positioning information update time and the feature fusion time;

[0035] determining an amount of movement of the vehicle within the time difference;

[0036] Determining the latest positioning information of the vehicle at the time of feature fusion according to the movement amount of the vehicle;

[0037] The latest map road information is determined according to the latest positioning information.

[0038] In some embodiments, it also includes:

[0039] During the vehicle driving process, when the intelligent driving path planning basis is converted from one of the visual lane line or the map lane line to another, the cubic curve parameters of the lane line are smoothed by a filtering algorithm.

[0040] In a second aspect, the present application also provides a vehicle intelligent driving assurance system based on road feature fusion, the system comprising:

[0041] A visual perception unit, which is used to obtain visual road information around the vehicle and determine the confidence of the visual road information;

[0042] A positioning unit, which is used to determine the positioning information of the vehicle and determine the positioning accuracy;

[0043] A high-precision map unit, which is used to determine map road information around the vehicle according to the positioning information;

[0044] A road feature fusion unit is used to perform feature fusion on the visual road information and the map road information to obtain a road feature fusion result, and determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning based on the visual road information, the map road information and the road feature fusion result.

[0045] In some embodiments, the road feature fusion unit is further used for:

[0046] Selecting different fusion strategies according to the visual road information and its confidence, the positioning accuracy and the map road information, and determining a fusion output result according to the consistency of the visual road information and the lane line cubic curve in the map road information, so as to determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning;

[0047] If the confidence level of the visual road information is high, the positioning accuracy is high accuracy, and the consistency level meets the consistency requirement, then the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level, and this is used as the optimal fusion guarantee strategy;

[0048] If the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency degree meets the consistency requirement, and the visual lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a high confidence level;

[0049] If the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency meets the consistency requirement, and the map lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a medium confidence level;

[0050] If the confidence of the visual road information is medium and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; when the consistency requirement is not met, the visual lane line or the map lane line is determined as the basis for intelligent driving path planning according to the cubic curve parameters of the lane line, and the intelligent driving level is a low confidence level;

[0051] If the confidence of the visual road information is low and the positioning accuracy is high, then when the consistency meets the consistency requirement, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a high confidence level; when the consistency requirement is not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level;

[0052] If the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency level meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level, and when the consistency requirement is not met, the intelligent driving level is unavailable;

[0053] If the positioning information or the map road information is not available, the visual lane lines are used as the basis for intelligent driving path planning.

[0054] In some embodiments, the visual perception unit is further used for:

[0055] If the visual lane lines in the visual road information are complete, determining that the confidence level of the visual road information is high;

[0056] If the visual lane line missing area is smaller than a preset area, it is determined that the confidence of the visual road information is medium;

[0057] If the visual lane line missing area is greater than a preset area, it is determined that the confidence of the visual road information is low.

[0058] In some embodiments, the road feature fusion unit is further used for:

[0059] When the visual road information has a medium confidence level or a low confidence level, the missing portion of the visual lane line is predicted and completed, and then the consistency between the visual road information and the map road information is determined.

[0060] In some embodiments, the positioning unit is further used for:

[0061] Parameters of the positioning information sent by the positioning device are obtained, and the positioning accuracy is determined according to the parameters of the positioning information.

[0062] In some embodiments, the road feature fusion unit is further used for:

[0063] When the positioning accuracy is low, the current position information of the vehicle is calculated according to the track of the vehicle, and the map road information is determined according to the calculated position information.

[0064] In some embodiments, the road feature fusion unit is further used for:

[0065] fusing and comparing the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information;

[0066] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information meets a preset value, determining that the degree of consistency meets the consistency requirement;

[0067] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information does not meet a preset value, it is determined that the degree of consistency does not meet the consistency requirement.

[0068] In some embodiments, the positioning unit is further used to: determine the time difference between the positioning information update time and the feature fusion time, determine the movement amount of the vehicle within the time difference, and determine the latest positioning information of the vehicle at the feature fusion time according to the movement amount of the vehicle;

[0069] The high-precision map unit is also used to determine the latest map road information based on the latest positioning information.

[0070] In some embodiments, the road feature fusion unit is further used for:

[0071] During the vehicle driving process, when the intelligent driving path planning basis is converted from one of the visual lane line or the map lane line to another, the cubic curve parameters of the lane line are smoothed by a filtering algorithm.

[0072] The present application provides a vehicle intelligent driving assurance method based on road feature fusion, which obtains visual road information around the vehicle and the confidence of the visual road information, determines the positioning accuracy according to the positioning information of the vehicle, obtains the map road information around the vehicle according to the positioning information, and fuses the road features with the visual information and the map road information to obtain reliable road features. And according to the confidence of the visual road information and the map road information, as well as the consistency of the lane lines in the road feature fusion results, different state combinations are performed to determine the different path planning bases of the vehicle during intelligent driving, and different self-intelligent driving levels make various modes of autonomous driving more flexible and intelligent, effectively ensuring the safety of autonomous driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0074] Figure 1 A flowchart of a vehicle intelligent driving assurance method based on road feature fusion provided in an embodiment of the present application;

[0075] Figure 2 A schematic block diagram of a vehicle intelligent driving assurance system based on road feature fusion provided in an embodiment of the present application;

[0076] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0078] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0079] The embodiments of the present application provide a vehicle intelligent driving assurance method and system based on road feature fusion.

[0080] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0081] Please refer to Figure 1 and Figure 2 , Figure 1 A flowchart of a vehicle intelligent driving assurance method based on road feature fusion provided in an embodiment of the present application is provided. Figure 2 A schematic flow chart of a vehicle intelligent driving assurance method based on road feature fusion provided in an embodiment of the application. The vehicle intelligent driving assurance method based on road feature fusion in the present application is specifically described below in conjunction with the accompanying drawings.

[0082] like Figure 1 As shown, the method includes steps S1 to S2.

[0083] Step S1: Acquire visual road information around a vehicle and determine the confidence level of the visual road information.

[0084] Exemplary, such as Figure 2As shown, the visual road information around the vehicle is obtained through the visual perception unit, which includes a high-definition intelligent front-view wide-angle camera and a side-view camera. The camera in the visual perception unit is arranged outside the vehicle, and the image in front of the vehicle and the side of the vehicle is collected by recording video or pictures. The image includes road information, and the road information obtained by the visual perception unit is named visual road information. The visual perception unit can separate lane lines, road edges, traffic road signs TSR, signboards, road scenes (ramp, tunnel, merge, divergence), etc. from the visual road information.

[0085] It is worth noting that the vehicle uses lane lines as the main basis for intelligent driving. The vehicle plans its driving path according to the lane lines, adjusts its own position and heading, etc. to ensure that the vehicle drives safely in the lane. In this application, the confidence of the visual road information is determined by the completeness of the visual lane lines in the visual road information.

[0086] Furthermore, if the visual lane line in the visual road information is complete, the confidence of the visual road information is determined to be high; if the visual lane line is lost and the lost area is smaller than the preset area, this small area loss may be caused by wear, occlusion and loss of the lane line itself. In this case, the confidence of the visual road information is determined to be medium; if the visual lane line is lost and the lost area is larger than the preset area, the loss of the lane line in this case may be because the visual perception unit acquires the visual lane line due to weather and light, resulting in blurred footage and short-term loss of the lane line. In this case, the confidence of the visual road information is determined to be low.

[0087] Step S2: Determine the positioning information of the vehicle and determine the positioning accuracy.

[0088] Exemplarily, the positioning information of the vehicle is determined by a positioning unit, and the positioning information includes the specific location of the vehicle. The positioning unit includes a global navigation satellite system GNSS, an inertial measurement unit IMU and an odometer for combined positioning to ensure the accuracy of positioning.

[0089] Furthermore, when positioning the vehicle, the positioning unit assumes that the positioning information is accurate to the lane range as the highest accuracy. If the positioning information is accurate to the lane, the positioning information is determined to be high-precision; if the positioning information cannot be accurate to the lane due to signal fluctuations, the positioning information is determined to be low-precision; if the signal is lost or the positioning unit fails, the positioning information is considered unavailable. The lane range only describes the positioning accuracy and is not limited to the lane.

[0090] Step S3: determining map road information around the vehicle based on the positioning information.

[0091] Specifically, after determining the positioning information, the high-precision map unit searches the high-definition map according to the positioning information to determine the map road information at the current location of the vehicle, wherein the map road information can also be output in the form of video or image. The map road information also includes: lane lines, road edges, traffic road signs TSR, signboards, road scenes (ramp, tunnel, merge, divergence) and other information.

[0092] It is worth mentioning that since the lane lines in the map road information output by the high-precision map unit are shape point coordinates, the map lane line fitting module in the road feature fusion unit uses a cubic fitting curve to fit the lane lines. In order to improve the stability of the lane lines during the fitting process, the Kalman filtering algorithm is used to smooth the fitted lane lines.

[0093] In some embodiments, after obtaining visual road information and map road information, the features in these two types of information are matched and verified by the map data verification module in the road feature fusion unit, and the positioning information is corrected according to the matching and verification results to make the positioning information more accurate.

[0094] As a preferred implementation, after acquiring the map road information, the lane lines, traffic road signs TSR and scenes in the map in the map road information are parsed into a function-defined interface form through the data management and parsing module in the road feature fusion unit, so that the format of the map road information is consistent with the format of the visual road information for vehicle use. The data management and parsing module will also perform ROI data filtering to retain only the required data. The filtered data and retained data can be set according to the vehicle's intelligent driving needs. When the vehicle's positioning information is updated to indicate that the vehicle has traveled to a new location, the data management and parsing module will also delete the expired map road information based on the latest positioning information, and update the visual road information according to the change in the vehicle's position.

[0095] It is worth noting that the positioning information determined by the positioning unit and the map road information of the high-precision map are both positioning information of the 84 coordinate system, while the visual road information is a road data description under the vehicle coordinate system. Therefore, it is necessary to convert the points output by the positioning unit and the improved map unit into the vehicle coordinate system through the coordinate conversion module to unify the coordinate system. The specific conversion has a public method and will not be repeated here.

[0096] Step S4: performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result.

[0097] Specifically, the road feature fusion result includes the consistency degree of the cubic curve parameters of the lane lines in the visual road information and the map road information. The acquisition of the consistency degree includes fusing and comparing the cubic curve parameters of the lane lines in the visual road information and the cubic curve parameters of the lane lines in the map road information. If the consistency degree of the cubic curve parameters of the lane lines in the visual road information and the cubic curve parameters of the lane lines in the map road information meets the preset value, it is determined that the consistency degree meets the consistency requirement. If the consistency degree of the cubic curve parameters of the lane lines in the visual road information and the cubic curve parameters of the lane lines in the map road information does not meet the preset value, it is determined that the consistency degree does not meet the consistency requirement.

[0098] Preferably, when the visual road information has a medium confidence level or a low confidence level, and the lane line is missing, the lane line tracking module in the road feature fusion unit predicts and completes the missing part of the visual lane line based on the existing lane lines and historical lane widths, and then determines the consistency between the visual road information and the map road information.

[0099] Preferably, when the positioning accuracy is low, the vehicle's current high-precision position information is calculated based on the vehicle's track through feature fusion, and the map road information is determined based on the calculated high-precision position information.

[0100] As a preferred implementation, since there is a delay in the updating of the positioning information by the positioning unit and there is a phenomenon of different frequencies when the road feature fusion unit performs fusion, in order to ensure the accuracy of the map data of the road feature fusion unit in the vehicle coordinate system when the fusion is performed, it is necessary to determine the time difference between the positioning information update time and the feature fusion time according to the timestamp, and determine the vehicle's movement amount according to the time difference, and then determine the latest positioning information of the vehicle at the feature fusion time according to the vehicle's movement amount, and finally determine the latest map road information according to the latest positioning information.

[0101] Step S5: determining the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the fusion result of the visual road information, the map road information and the road features.

[0102] Specifically, different fusion strategies are selected according to the visual road information and its confidence, the positioning accuracy and the map road information, and the fusion output result is determined according to the consistency of the visual road information and the lane line cubic curve in the map road information to determine the vehicle's intelligent driving level and the basis for intelligent driving path planning. Specifically, the intelligent driving level confidence determination module in the system determines the intelligent driving level confidence and the basis for intelligent driving path planning according to the confidence of the visual road information, the positioning accuracy of the map road information and the consistency between the visual lane line and the map lane line, which can be divided into the following situations:

[0103] The first type is that if the confidence of the visual road information is high, the positioning accuracy is high precision, and the consistency meets the consistency requirements, the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level, and this is used as the optimal fusion guarantee strategy. The preset distance in the embodiment of this application is set to 80 meters.

[0104] The second type is that if the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency level meets the consistency requirements. When the consistency requirements are not met, the visual lane lines are used as the basis for intelligent driving path planning. The intelligent driving level is a high confidence level.

[0105] The third type is that if the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency level meets the consistency requirements. When the consistency requirements are not met, the map lane lines are used as the basis for intelligent driving path planning. The intelligent driving level is a medium confidence level.

[0106] The fourth type, if the confidence of the visual road information is medium and the positioning accuracy is low, the visual lane line is used as the basis for intelligent driving path planning when the consistency degree meets the consistency requirements, and the intelligent driving level is a medium confidence level. When the consistency requirements are not met, the visual lane line and the map lane line are fitted into a new lane line. The cubic curve parameters C0 and C1 of the new lane line trust the visual lane line more, and the cubic curve parameters C2 and C3 trust the map lane line more. Then, the updated cubic curve parameters are output in combination with smoothing processing.

[0107] The fifth type is that if the confidence of the visual road information is low and the positioning accuracy is high, the map lane lines are used as the basis for intelligent driving path planning when the consistency meets the consistency requirements, and the intelligent driving level is a high confidence level. When the consistency requirements are not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level.

[0108] The sixth type is that if the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency level meets the consistency requirements, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level. When the consistency requirements are not met, the intelligent driving level is unavailable.

[0109] Seventh, if the positioning unit fails or the high-precision map unit fails, resulting in the loss of positioning information and map road information, the visual lane lines in the visual lane information are used as the basis for path planning for intelligent driving, and the confidence level of the vehicle's intelligent driving is determined by the confidence level of the visual road information.

[0110] It is worth noting that the high confidence level, medium confidence level and low confidence level in the vehicle's intelligent driving level represent the degree of intelligent driving of the vehicle. For example, the intelligent driving under the high confidence level can be set to L4 intelligent driving, the intelligent driving under the medium confidence level can be set to L3 intelligent driving, and the intelligent driving under the low confidence level can be set to L2 intelligent driving. It can also be set according to specific needs. In short, the higher the vehicle's intelligent driving confidence level, the higher the vehicle's automatic performance and the less manual intervention. The lower the vehicle's intelligent driving confidence level, the more the intelligent driving function will be gradually downgraded. In the case that the vehicle's intelligent driving level is unavailable, all intelligent driving functions will be restricted, and the driver will perform manual driving throughout the process.

[0111] In some embodiments, when the intelligent driving path planning basis is converted from one of the visual lane lines or the map lane lines to another, the road feature fusion unit predicts, updates and smoothes the cubic curve parameters of the lane lines through the Kalman filter algorithm to ensure a stable transition during the conversion.

[0112] It is worth noting that since the HD map unit does not output speed limit signs, when determining the vehicle's intelligent driving speed, the TSR / scenario fusion module combines the road speed limit information in the HD map, the speed limit information detected by the visual perception unit, and the navigation map speed limit information to select a safe speed. For traffic road signs, TSR needs to be displayed based on the visual road information and map road information. For scenes in the intelligent driving process, in fast-driving scenes such as highways and cities, the scene results of the HD map are used as the basis for intelligent driving, and on other roads, the results of the time perception unit are used as the basis for intelligent driving.

[0113] This implementation provides a vehicle intelligent driving guarantee method based on road feature fusion, which fuses visual road information and map road information, and determines the path planning basis for vehicle intelligent driving based on the fusion results, avoiding driving hazards caused by functional failure or information failure of a certain unit module, and making the path planning of intelligent driving more reliable. In addition, the level of intelligent driving is determined based on the fusion results, and the intelligent driving function is gradually reduced according to the situation, providing a clear intervention boundary between intelligent driving and human-access driving, and improving the safety of intelligent driving.

[0114] Please refer to Figure 2 , Figure 2A flowchart of a vehicle intelligent driving assurance system based on road feature fusion is provided for an embodiment of the present application.

[0115] like Figure 2 As shown, the system includes: a visual perception unit, a positioning unit, a high-precision map unit and a road feature fusion unit.

[0116] A visual perception unit, which is used to obtain visual road information around the vehicle and determine the confidence of the visual road information;

[0117] A positioning unit, which is used to determine the positioning information of the vehicle and determine the positioning accuracy;

[0118] A high-precision map unit, which is used to determine map road information around the vehicle according to the positioning information;

[0119] A road feature fusion unit is used to perform feature fusion on the visual road information and the map road information to obtain a road feature fusion result, and determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning based on the visual road information, the map road information and the road feature fusion result.

[0120] In some embodiments, the road feature fusion unit is further used for:

[0121] Selecting different fusion strategies according to the visual road information and its confidence, the positioning accuracy and the map road information, and determining a fusion output result according to the consistency of the visual road information and the lane line cubic curve in the map road information, so as to determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning;

[0122] If the confidence level of the visual road information is high, the positioning accuracy is high accuracy, and the consistency level meets the consistency requirement, then the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level, and this is used as the optimal fusion guarantee strategy;

[0123] If the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency degree meets the consistency requirement, and the visual lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a high confidence level;

[0124] If the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency meets the consistency requirement, and the map lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a medium confidence level;

[0125] If the confidence of the visual road information is medium and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; when the consistency requirement is not met, the visual lane line or the map lane line is determined as the basis for intelligent driving path planning according to the cubic curve parameters of the lane line, and the intelligent driving level is a low confidence level;

[0126] If the confidence of the visual road information is low and the positioning accuracy is high, then when the consistency meets the consistency requirement, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a high confidence level; when the consistency requirement is not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level;

[0127] If the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency level meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level, and when the consistency requirement is not met, the intelligent driving level is unavailable;

[0128] If the positioning information or the map road information is not available, the visual lane lines are used as the basis for intelligent driving path planning, and the confidence level of the visual road information is used as the automatic driving level.

[0129] In some embodiments, the visual perception unit is further used for:

[0130] If the visual lane lines in the visual road information are complete, determining that the confidence level of the visual road information is high;

[0131] If the visual lane line missing area is smaller than a preset area, it is determined that the confidence of the visual road information is medium;

[0132] If the visual lane line missing area is greater than a preset area, it is determined that the confidence of the visual road information is low.

[0133] In some embodiments, the road feature fusion unit is further used for:

[0134] When the visual road information has a medium confidence level or a low confidence level, the missing portion of the visual lane line is predicted and completed, and then the consistency between the visual road information and the map road information is determined.

[0135] In some embodiments, the positioning unit is further used for:

[0136] Parameters of the positioning information sent by the positioning device are obtained, and the positioning accuracy is determined according to the parameters of the positioning information.

[0137] In some embodiments, the road feature fusion unit is further used for:

[0138] When the positioning accuracy is low, the current position information of the vehicle is calculated according to the track of the vehicle, and the map road information is determined according to the calculated position information.

[0139] In some embodiments, the road feature fusion unit is further used for:

[0140] fusing and comparing the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information;

[0141] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information meets a preset value, determining that the degree of consistency meets the consistency requirement;

[0142] If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information does not meet a preset value, it is determined that the degree of consistency does not meet the consistency requirement.

[0143] In some embodiments, the positioning unit is further used to: determine the time difference between the positioning information update time and the feature fusion time, determine the movement amount of the vehicle within the time difference, and determine the latest positioning information of the vehicle at the feature fusion time according to the movement amount of the vehicle;

[0144] The high-precision map unit is also used to determine the latest map road information based on the latest positioning information.

[0145] In some embodiments, the road feature fusion unit is further used for:

[0146] During the vehicle driving process, when the intelligent driving path planning basis is converted from one of the visual lane line or the map lane line to another, the cubic curve parameters of the lane line are smoothed by a filtering algorithm.

[0147] It should be noted that, in the embodiment of the present invention, the various functional modules or functional units of the vehicle intelligent driving security system based on road feature fusion can exchange information through electrical connection, where "electrical connection" is used to describe the signal connection between two components, such as control signals and feedback signals, and the electrical power connection between two components. In addition, the "connection" involved here can be a wired connection or a wireless connection, and the "electrical connection" involved can be a direct electrical connection between two components or an indirect electrical connection through other components.

[0148] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and conciseness of description, the specific working process of the system and each unit and module described above can refer to the corresponding process in the aforementioned embodiment of the vehicle intelligent driving assurance method based on road feature fusion, and will not be repeated here.

[0149] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0150] The serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific implementation mode of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A vehicle intelligent driving assurance method based on road feature fusion, characterized in that: include: Acquiring visual road information around the vehicle and determining the confidence of the visual road information; Determining the positioning information of the vehicle and determining the positioning accuracy; Determining map road information around the vehicle according to the positioning information; Performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result; Determining the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the fusion result of the visual road information, the map road information and the road features; Wherein, determining the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the fusion result of the visual road information, the map road information and the road features includes: Selecting different fusion strategies according to the visual road information and its confidence, the positioning accuracy and the map road information, and determining a fusion output result according to the consistency of the visual road information and the lane line cubic curve in the map road information, so as to determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning; If the confidence of the visual road information is high, the positioning accuracy is high accuracy and the consistency meets the consistency requirements, then the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level, and this is used as the optimal fusion guarantee strategy; If the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency degree meets the consistency requirement, and the visual lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a high confidence level; If the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency meets the consistency requirement, and the map lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a medium confidence level; If the confidence of the visual road information is medium and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; when the consistency requirement is not met, the visual lane line and the map lane line are fitted into a new lane line as the basis for intelligent driving path planning, and the intelligent driving level is a low confidence level; If the confidence of the visual road information is low and the positioning accuracy is high, then when the consistency meets the consistency requirement, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a high confidence level; when the consistency requirement is not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; If the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level. When the consistency requirement is not met, the intelligent driving level is unavailable; If the positioning information or the map road information is not available, the visual lane lines are used as the basis for intelligent driving path planning, and the confidence level of the visual road information is used as the automatic driving level.

2. The vehicle intelligent driving guarantee method based on road feature fusion according to claim 1 is characterized in that: The acquiring visual road information around the vehicle and determining the confidence of the visual road information includes: If the visual lane lines in the visual road information are complete, determining that the confidence level of the visual road information is high; If the visual lane line missing area is smaller than a preset area, it is determined that the confidence of the visual road information is medium; If the visual lane line missing area is greater than a preset area, it is determined that the confidence level of the visual road information is low.

3. The vehicle intelligent driving assurance method based on road feature fusion according to claim 2 is characterized in that: The performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result also includes: When the visual road information has a medium confidence level or a low confidence level, the missing portion of the visual lane line is predicted and completed, and then the consistency between the visual road information and the map road information is determined.

4. The vehicle intelligent driving assurance method based on road feature fusion according to claim 1 is characterized in that: The determining the positioning information of the vehicle and determining the positioning accuracy includes: Parameters of positioning information sent by a positioning device are obtained, and the positioning accuracy is determined according to the parameters of the positioning information.

5. The vehicle intelligent driving assurance method based on road feature fusion according to claim 4 is characterized in that: The performing feature fusion on the visual road information and the map road information to obtain a road feature fusion result also includes: When the positioning accuracy is low, the current position information of the vehicle is calculated according to the track of the vehicle, and the map road information is determined according to the calculated position information.

6. The vehicle intelligent driving guarantee method based on road feature fusion according to claim 1 is characterized in that: Determining the consistency of cubic curve parameters of the lane line in the visual road information and the map road information through feature fusion includes: fusing and comparing the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information; If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information meets a preset value, determining that the degree of consistency meets the consistency requirement; If the degree of consistency between the cubic curve parameters of the lane line in the visual road information and the cubic curve parameters of the lane line in the map road information does not meet a preset value, it is determined that the degree of consistency does not meet the consistency requirement.

7. The vehicle intelligent driving guarantee method based on road feature fusion according to claim 1 is characterized in that: Also includes: Determine the time difference between the positioning information update time and the feature fusion time; determining an amount of movement of the vehicle within the time difference; Determining the latest positioning information of the vehicle at the time of feature fusion according to the movement amount of the vehicle; The latest map road information is determined according to the latest positioning information.

8. The vehicle intelligent driving assurance method based on road feature fusion according to claim 1 is characterized in that: Also includes: During the vehicle driving process, when the intelligent driving path planning basis is converted from one of the visual lane line or the map lane line to another, the cubic curve parameters of the lane line are smoothed by a filtering algorithm.

9. A vehicle intelligent driving guarantee system based on road feature fusion, characterized in that: include: A visual perception unit, which is used to obtain visual road information around the vehicle and determine the confidence of the visual road information; A positioning unit, which is used to determine the positioning information of the vehicle and determine the positioning accuracy; A high-precision map unit, which is used to determine map road information around the vehicle according to the positioning information; a road feature fusion unit, which performs feature fusion on the visual road information and the map road information to obtain a road feature fusion result, and determines the intelligent driving level of the vehicle and the basis for intelligent driving path planning according to the visual road information, the map road information and the road feature fusion result; Wherein, the road feature fusion unit is also used for: Selecting different fusion strategies according to the visual road information and its confidence, the positioning accuracy and the map road information, and determining a fusion output result according to the consistency of the visual road information and the lane line cubic curve in the map road information, so as to determine the intelligent driving level of the vehicle and the basis for intelligent driving path planning; If the confidence of the visual road information is high, the positioning accuracy is high accuracy and the consistency meets the consistency requirements, then the visual lane lines in the visual road information are selected as the basis for intelligent driving path planning within the preset distance, and the map lane lines in the map road information are selected as the basis for intelligent driving path planning outside the preset distance, and the intelligent driving level is determined to be a high confidence level, and this is used as the optimal fusion guarantee strategy; If the confidence of the visual road information is high and the positioning accuracy is low, the optimal fusion guarantee strategy is adopted when the consistency degree meets the consistency requirement, and the visual lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a high confidence level; If the confidence of the visual road information is medium and the positioning accuracy is high, the optimal fusion guarantee strategy is adopted when the consistency meets the consistency requirement, and the map lane line is used as the basis for intelligent driving path planning when the consistency requirement is not met, and the intelligent driving level is a medium confidence level; If the confidence of the visual road information is medium and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; when the consistency requirement is not met, the visual lane line and the map lane line are fitted into a new lane line as the basis for intelligent driving path planning, and the intelligent driving level is a low confidence level; If the confidence of the visual road information is low and the positioning accuracy is high, then when the consistency meets the consistency requirement, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a high confidence level; when the consistency requirement is not met, the map lane lines are used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level; If the confidence of the visual road information is low and the positioning accuracy is low, then when the consistency meets the consistency requirement, the visual lane line is used as the basis for intelligent driving path planning, and the intelligent driving level is a medium confidence level. When the consistency requirement is not met, the intelligent driving level is unavailable; If the positioning information or the map road information is not available, the visual lane lines are used as the basis for intelligent driving path planning, and the confidence level of the visual road information is used as the automatic driving level.

Images