Lane line fusion method and apparatus

Abstract

The application discloses a lane line fusion method and device. The method comprises the following steps: determining a first fitting function of a target lane line corresponding to a road where a vehicle is located according to positioning information of the vehicle in a first data collection period; determining a second fitting function of the target lane line according to image data of the road; obtaining a confidence degree corresponding to a lane line fusion function corresponding to the first fitting function, the second fitting function and a second data collection period; performing validity detection on the first fitting function and the second fitting function according to a preset check condition; and in the case that the first fitting function and the second fitting function pass the validity detection, performing weighted calculation according to the first fitting function, the second fitting function, the lane line fusion function corresponding to the second data collection period and the corresponding confidence degree, so as to obtain a target lane line fusion function corresponding to the first data collection period. In this way, the reliability of lane line parameters is improved.

Description

Technical Field

[0001] This application relates to the field of perception and positioning technology in autonomous driving technology, and in particular to a lane line fusion method and apparatus. Background Technology

[0002] Lane line recognition is crucial for the safety and comfort of autonomous vehicles. Currently, lane line parameter recognition methods mainly rely on computer vision detection (or camera detection) and high-precision map positioning. Lane line information based on a single signal source has poor noise resistance and low effective coverage. For example, in complex recognition conditions such as blurred lane lines, continuous curves, longitudinal speed bumps (herringbone lines), or inclement weather, lane line parameters may have significant errors or result in missed or false detections, making it difficult to guarantee the reliability of lane line parameters. Summary of the Invention

[0003] This application provides a lane line fusion method and apparatus that effectively improves the problems of poor noise resistance and limited coverage of traditional single-signal-source lane line parameter identification methods, thereby improving the accuracy and stability of lane line data information and ultimately enhancing the reliability of lane line parameters.

[0004] In a first aspect, embodiments of this application provide a lane line fusion method, the method comprising:

[0005] The system acquires the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period preceding the first data acquisition period.

[0006] Based on the location information, determine the first fitting function for the target lane line corresponding to the road where the vehicle is located, where the target lane line is any lane line among the lane lines on both sides of the road;

[0007] Based on the image data of the road, determine the second fitting function for the target lane line;

[0008] Obtain the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively.

[0009] Based on the preset verification conditions, the effectiveness of the first fitting function and the second fitting function are tested respectively.

[0010] If the first fitting function and the second fitting function pass the validity test, the target lane line fusion function corresponding to the first data acquisition period is obtained by weighted calculation based on the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, and the confidence scores corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, respectively.

[0011] In one possible implementation embodiment, if the first fitting function and the second fitting function pass the validity test respectively, a weighted calculation is performed based on the lane line fusion function corresponding to the first fitting function, the second fitting function, and the second data acquisition period, as well as the confidence levels corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function, and the second data acquisition period, to obtain the target lane line fusion function corresponding to the first data acquisition period, including:

[0012] If the first fitting function and the second fitting function pass the validity test respectively, the first fitting function and the second fitting function are subjected to preset smoothing processing according to the preset function smoothing algorithm to obtain the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function.

[0013] The target lane fusion function for the first data acquisition period is obtained by weighting the lane fusion functions corresponding to the first smoothing function, the second smoothing function, and the lane fusion function for the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane fusion function for the second data acquisition period.

[0014] In one possible implementation embodiment, according to a preset function smoothing algorithm, a preset smoothing process is performed on the first fitting function and the second fitting function respectively to obtain a first smoothing function corresponding to the first fitting function and a second smoothing function corresponding to the second fitting function, including:

[0015] The data acquisition cycle duration, vehicle operating parameters, and lane line fusion function coefficients corresponding to the second data acquisition cycle are obtained. The operating parameters include the current vehicle speed and yaw rate.

[0016] Based on the duration of the data collection cycle, the current vehicle speed, and the yaw rate, the coefficients of the lane line fusion function are adjusted to generate smoothing parameters.

[0017] Without changing lanes, based on smoothing parameters and a preset smoothing formula, the first fitting function and the second fitting function are subjected to preset smoothing processing respectively, thereby determining the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function. The preset smoothing formula is as follows:

[0018]

[0019]

[0020] Among them, M ′ Here, represents the coefficients of the first smoothing function, and M represents the coefficients of the first fitting function. Here, α is the smoothing parameter, C′ is the coefficient of the second smoothing function, and C is the coefficient of the second fitting function.

[0021] In one possible implementation, the first data acquisition period includes multiple sampling times, and the positioning information includes latitude and longitude information corresponding to each of the multiple sampling times;

[0022] Based on the location information, determine the first fitting function for the target lane line corresponding to the road where the vehicle is located, including:

[0023] Based on the preset coordinate transformation relationship, the coordinate information of each latitude and longitude information point in the vehicle coordinate system is determined. The preset coordinate transformation relationship is the transformation relationship between the ground coordinate system where the latitude and longitude information is located and the vehicle coordinate system.

[0024] The coordinate information corresponding to multiple coordinate points is fitted to generate the first fitting function for the target lane line.

[0025] In one possible implementation, the first fitting function corresponds to a first target confidence level; obtaining the first target confidence level corresponding to the first fitting function includes:

[0026] Obtain the location confidence level corresponding to the vehicle's positioning system, and the map confidence level of the map data configured within the positioning system; and...

[0027] Based on the coordinate information corresponding to multiple coordinate points, determine the projection distance between each coordinate point and the first fitting function;

[0028] The confidence level of the first fitting function is determined based on the preset confidence range corresponding to the maximum projection distance among multiple projection distances. The preset confidence range includes at least one preset numerical range, and each numerical range corresponds to a confidence level.

[0029] The confidence score of the first target is obtained by multiplying the map confidence score, the location confidence score, and the confidence score of the first fitting function.

[0030] In one possible implementation embodiment, the first fitting function is validated according to preset validation conditions, including:

[0031] The verification parameters of the first fitting function are obtained. The verification parameters of the first fitting function include the function coefficients of the third fitting function corresponding to the L data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each third fitting function. The third fitting function is determined based on the vehicle's positioning information.

[0032] If the verification parameters of the first fitting function meet the first preset validity judgment condition, the first fitting function passes the validity test.

[0033] The first presupposed validity determination criteria include at least one or more of the following determination criteria:

[0034] The coefficients of the L third fitting functions are all different;

[0035] The count indicators corresponding to the function coefficients of the L third fitted functions change sequentially;

[0036] The width of the target lane line determined based on the first fitting function is less than the preset lane line width;

[0037] The change in the coefficients of the third fitting function corresponding to the l-th data acquisition cycle compared to the coefficients of the third fitting function corresponding to the data acquisition cycles prior to the l-th data acquisition cycle is within a preset numerical range.

[0038] In one possible implementation embodiment, the method for determining a second fitting function for the target lane line based on image data of the road includes:

[0039] The image data of the road is identified and analyzed to generate an initial fitting function for the target lane line and the category information of the target lane line. The category information is associated with preset compensation parameters and the function movement direction.

[0040] Based on the compensation parameters associated with the target lane line category information and the function movement direction, the initial fitting function is moved to obtain the second fitting function.

[0041] In one possible implementation, the second fitting function corresponds to a second target confidence level; the second target confidence level is determined based on the sharpness of the road image data.

[0042] In one possible implementation embodiment, the second fitting function is validated according to preset verification conditions, including:

[0043] The verification parameters of the second fitting function are obtained. The verification parameters of the second fitting function include the function coefficients of the fourth fitting function corresponding to the M data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each fourth fitting function. The fourth fitting function is determined based on the vehicle's positioning information.

[0044] If the verification parameters of the second fitting function meet the second preset validity judgment condition, the second fitting function passes the validity test.

[0045] The second pre-set validity determination criteria include at least one or more of the following determination criteria:

[0046] The coefficients of the M fourth fitting functions are all different;

[0047] The count indicators corresponding to the function coefficients of the M fourth fitting functions change sequentially;

[0048] The width of the target lane line determined based on the second fitting function is less than the preset lane line width;

[0049] The change in the coefficients of the fourth fitting function corresponding to the m-th data acquisition cycle compared to the coefficients of the fourth fitting function corresponding to the data acquisition cycles prior to the m-th data acquisition cycle is within a preset range.

[0050] Secondly, embodiments of this application provide a lane line fusion device, comprising:

[0051] The acquisition module is used to acquire the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period before the first data acquisition period.

[0052] The processing module is used to determine a first fitting function for the target lane line corresponding to the road where the vehicle is located based on the positioning information, wherein the target lane line is any lane line among the lane lines on both sides of the road; and to determine a second fitting function for the target lane line based on the image data of the road.

[0053] The acquisition module is also used to acquire the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively.

[0054] The processing module is also used to perform validity checks on the first fitting function and the second fitting function respectively according to preset verification conditions;

[0055] The processing module is further configured to, when the first fitting function and the second fitting function pass the validity test respectively, perform a weighted calculation based on the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, and the confidence levels corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period respectively, to obtain the target lane line fusion function corresponding to the first data acquisition period.

[0056] Thirdly, embodiments of this application provide an electronic device, the device comprising:

[0057] Processor and memory storing computer program instructions;

[0058] The processor executes computer program instructions to implement any of the above-mentioned lane line fusion methods.

[0059] Fourthly, embodiments of this application provide a computer storage medium storing computer program instructions, which, when executed by a processor, implement the lane fusion method described above.

[0060] Fifthly, embodiments of this application provide a computer program product, characterized in that, when the instructions in the computer program product are executed by the processor of an electronic device, the electronic device is able to execute any of the lane fusion methods described above.

[0061] The lane fusion method, apparatus, device, and computer storage medium of this application embodiment acquire vehicle positioning information, image data of the road where the vehicle is located during the first data acquisition period, and lane fusion function corresponding to a second data acquisition period, wherein the second data acquisition period is the data acquisition period prior to the first data acquisition period. Based on the positioning information, a first fitting function for the target lane line corresponding to the road where the vehicle is located is determined, wherein the target lane line is any lane line on both sides of the road. Based on the road image data, a second fitting function for the target lane line is determined. The confidence levels corresponding to the first fitting function, the second fitting function, and the lane fusion function corresponding to the second data acquisition period are obtained respectively. According to preset verification conditions, the first fitting function and the second fitting function are respectively tested for validity. If the first fitting function and the second fitting function pass the validity test, a weighted calculation is performed based on the first fitting function, the second fitting function, and the lane fusion function corresponding to the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane fusion function corresponding to the second data acquisition period, to obtain the target lane fusion function corresponding to the first data acquisition period. By performing a weighted calculation based on the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, the target lane line fusion function corresponding to the first data acquisition period effectively improves the problems of poor noise resistance and limited coverage of the traditional single signal source lane line parameter identification method, improves the accuracy and stability of lane line data information, and thus improves the reliability of lane line parameters. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 This is a schematic flowchart of a lane line fusion method provided in one embodiment of this application;

[0064] Figure 2 This is a schematic flowchart of a lane line fusion method provided in another embodiment of this application;

[0065] Figure 3 This is a schematic flowchart of a lane line fusion method provided in another embodiment of this application;

[0066] Figure 4 This is a schematic diagram of the first fitting function involved in another embodiment of this application;

[0067] Figure 5 This is a schematic flowchart of a lane line fusion method provided in another embodiment of this application;

[0068] Figure 6 This is a schematic diagram of the validity detection process in another embodiment of this application;

[0069] Figure 7 This is an example diagram of the representation position of the first fitting function in another embodiment of this application;

[0070] Figure 8 This is an example diagram of the representation position of the second fitting function according to another embodiment of this application;

[0071] Figure 9 This is a schematic diagram of the validity detection process in another embodiment of this application;

[0072] Figure 10 This is a schematic diagram of the lane line fusion device provided in another embodiment of this application;

[0073] Figure 11 This is a schematic diagram of the structure of an electronic device provided in another embodiment of this application. Detailed Implementation

[0074] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0075] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0076] Lane line recognition is crucial for the safety and comfort of autonomous vehicles. Currently, lane line parameter recognition methods mainly rely on computer vision detection (or camera detection) and high-precision map positioning. Lane line information based on a single signal source has poor noise resistance and low effective coverage. For example, in complex recognition conditions such as blurred lane lines, continuous curves, longitudinal speed bumps (herringbone lines), or inclement weather, lane line parameters may have significant errors or result in missed or false detections, making it difficult to guarantee the reliability of lane line parameters.

[0077] To address the problems of the prior art, embodiments of this application provide a lane line fusion method, apparatus, device, and computer storage medium. The lane line fusion method provided in this application embodiment will be described first below.

[0078] Figure 1 A schematic flowchart of a lane line fusion method provided in one embodiment of this application is shown.

[0079] like Figure 1 As shown, the lane line fusion method provided in this application includes S110 to S160.

[0080] S110. Obtain the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period prior to the first data acquisition period.

[0081] Here, the location information is stored in the high-precision map and is obtained through a high-precision positioning system. The image data is acquired through a camera, which can be a roadside camera or a vehicle-side camera. The lane line fusion function corresponding to the second data acquisition cycle is the lane line fusion function corresponding to the data acquisition cycle before the first data acquisition cycle, for example, the lane line fusion function corresponding to the previous data acquisition cycle.

[0082] In some embodiments, the positioning information is discontinuous positioning information. The positioning information may be the positioning information of a vehicle, and the positioning information of the lane line can be determined based on the positioning information of the vehicle.

[0083] In some embodiments, the location information may also be the location information of the lane lines.

[0084] S120. Based on the positioning information, determine the first fitting function of the target lane line corresponding to the road where the vehicle is located, where the target lane line is any lane line among the lane lines on both sides of the road.

[0085] In some embodiments, the target lane line corresponding to the road where the vehicle is located is fitted based on discontinuous positioning information to obtain a first fitting function. The fitting method may include various methods, and the first fitting function is any function that can be fitted based on the positioning information.

[0086] As an example, a cubic polynomial can be used for fitting, with the first fitting function being:

[0087] y = m0 + m1x + m2x 2 +m3x 3 (1)

[0088] Where M = [m0, m1, m2, m3] are the coefficients of the first fitting function.

[0089] In some embodiments, the lane lines have width, and a first fitting function determined based on the positioning information characterizes the center position of the target lane line.

[0090] S130. Based on the image data of the road, determine the second fitting function for the target lane line.

[0091] In some embodiments, the lane lines have a width, and the second fitting function determined based on the road image data characterizes the inner edge curve of the target lane line, or it can be the curve after the inner edge curve of the target lane line is moved towards the center position of the lane line.

[0092] In some embodiments, a second fitting function is obtained by fitting the target lane line corresponding to the road where the vehicle is located based on the image data of the road. The fitting method may include a variety of methods, and the second fitting function is any function that can be fitted based on the positioning information.

[0093] As an example, a cubic polynomial can be used for fitting, and the second fitting function is:

[0094] y = c0 + c1x + c2x 2 +c3x 3 (2)

[0095] Where C = [c0, c1, c2, c3] are the coefficients of the second fitting function.

[0096] S140. Obtain the confidence levels of the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively.

[0097] In some embodiments, there are errors between the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period and the actual situation. The first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period each have corresponding confidence levels. The confidence levels can be obtained in different ways. The higher the confidence level, the higher the reliability of the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, and the more accurate the function expression.

[0098] In some embodiments, at least one factor affects the confidence level, and the confidence level can be set to [0,1], where 0 represents completely untrustworthy and 1 represents completely trustworthy.

[0099] S150. According to the preset verification conditions, the effectiveness of the first fitting function and the second fitting function are tested respectively.

[0100] Here, the preset verification conditions are pre-defined detection conditions, and each detection condition includes at least one condition.

[0101] In some embodiments, the validity of the first fitting function and the second fitting function are tested according to preset verification conditions to detect whether there are any abnormalities in the first fitting function and the second fitting function.

[0102] S160. If the first fitting function and the second fitting function pass the validity test respectively, the target lane line fusion function corresponding to the first data acquisition period is obtained by weighted calculation based on the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, and the confidence levels corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period respectively.

[0103] In some embodiments, the weights for the first and second fitting functions can be set as a weighting function with respect to confidence. Here, the weighting function may also include at least one of the following as independent variables: lane line type (e.g., solid line, dashed line, longitudinal speed bump, double line, etc.), driving mode (e.g., manual driving, human-machine co-driving, fully automated driving, etc.), and driving area (e.g., highway, urban expressway, urban road, closed park, etc.).

[0104] As an example, the weight functions for the first and second fitting functions are as follows:

[0105] β M =f(γ) M ,LineType,DriveMode,DriveZone) (3)

[0106] β C =f(γ) C ,LineType,DriveMode,DriveZone) (4)

[0107] Where, γ M γ C , respectively, represent the confidence levels of the first and second fitted functions, LineType represents the lane line type, DriveMode represents the driving mode, and DriveZone represents the driving zone.

[0108] In some embodiments, the lane line fusion function corresponding to the second data acquisition period can be fused to the target lane line fusion function corresponding to the first data acquisition period. The weight of the lane line fusion function corresponding to the second data acquisition period can be set to a constant (for example, the weight of the lane line fusion function corresponding to the second data acquisition period is 0.5). It can also be adjusted according to at least one of the weights of the first fitting function and the second fitting function, lane line type, driving mode, and driving area.

[0109] As an example, the weight function of the lane line fusion function corresponding to the second data acquisition cycle is:

[0110] β P =f(β) M ,β C,LineType,DriveMode,DriveZone) (5)

[0111] Where, β M ,β C represents the weighting function of the first and second fitting functions, LineType represents the lane line type, DriveMode represents the driving mode, and DriveZone represents the driving zone.

[0112] In some embodiments, the weights corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period are normalized, and then weighted calculation is performed.

[0113] As an example, the weight β corresponding to the first fitting function M The weights β corresponding to the second fitting function C The weight β of the lane line fusion function corresponding to the second data acquisition period P After normalization, we obtain β′ M ,β′ C and β′ p The first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period are fused together. The coefficients of the target lane line fusion function corresponding to the first data acquisition period are:

[0114] P fusion =β′ M M+β′ C C+β′ P P fusion_las (6)

[0115] Where, β′ M ,β′ C and β′ p P represents the weights of the normalized first fitting function, second fitting function, and lane line fusion function corresponding to the second data acquisition period, respectively. fusion_las These are the coefficients of the lane line fusion function corresponding to the second data acquisition cycle.

[0116] Here, P fusion =[p 0_fusion ,p 1_fusion ,p 2_fusion ,p 3_fusion [ ] represents the coefficients of the target lane line fusion function corresponding to the first data acquisition period. The target lane line fusion function corresponding to the first data acquisition period is:

[0117] y = p 0_fusion +p 1_fusionx+p 2_fusion x 2 +p 3_fusion x 3 (7)

[0118] In some embodiments, if the first fitting function fails the validity test, the confidence level corresponding to the first fitting function is set to low confidence (e.g., the confidence level corresponding to the first fitting function is 0), and if the second fitting function fails the validity test, the confidence level corresponding to the second fitting function is set to low confidence (e.g., the confidence level corresponding to the second fitting function is 0).

[0119] In some embodiments, the confidence level of the target lane line fusion function corresponding to the first data acquisition period can be calculated by the weights corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively, and at least one of the lane line type (e.g., solid line, dashed line, longitudinal speed bump, double line, etc.), driving mode (e.g., manual driving, human-machine co-driving, fully automated driving, etc.) and driving area (e.g., highway, urban expressway, urban road, closed park, etc.), without specific limitations.

[0120] As an example, the confidence level of the target lane line fusion function corresponding to the first data acquisition period is:

[0121] γ P =f(β) M ,β C ,β P ,LineType,DriveMode,DriveZone) (8)

[0122] The weights corresponding to the first fitting function are β. M The weights corresponding to the second fitting function are β. C The weight of the lane line fusion function corresponding to the second data acquisition period is β. P LineType is the lane line type, DriveMode is the driving mode, and DriveZone is the driving zone.

[0123] In this way, by performing a weighted calculation based on the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, the target lane line fusion function corresponding to the first data acquisition period effectively improves the problems of poor noise resistance and limited coverage of the traditional single signal source lane line parameter identification method, improves the accuracy and stability of lane line data information, and thus improves the reliability of lane line parameters.

[0124] In some embodiments, such as Figure 2 As shown, S160 may specifically include S210 to S220.

[0125] S210. If the first fitting function and the second fitting function pass the validity test respectively, according to the preset function smoothing algorithm, the first fitting function and the second fitting function are subjected to preset smoothing processing to obtain the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function.

[0126] In some embodiments, the first fitting function and the second fitting function are denoised and subjected to preset smoothing processing. Here, the preset function smoothing algorithm can be designed based on the error of the positioning information and image data.

[0127] S220. Based on the first smoothing function, the second smoothing function, the lane line fusion function corresponding to the second data acquisition period, and the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, a weighted calculation is performed to obtain the target lane line fusion function corresponding to the first data acquisition period.

[0128] Wherein, the first smoothing function is the first fitting function after pre-smoothing processing, the second smoothing function is the second fitting function after pre-smoothing processing, the confidence level of the first fitting function is the confidence level of the first smoothing function, and the confidence level of the second fitting function is the confidence level of the second smoothing function.

[0129] In this way, the first and second fitting functions are pre-smoothed, filtering out the error noise of the first and second fitting functions, thereby improving the accuracy of the target lane line fusion function corresponding to the first data acquisition cycle.

[0130] In some embodiments, such as Figure 3 As shown, the above-mentioned S210 may specifically include S310 to S330.

[0131] S310. Obtain the duration of the data acquisition cycle, the vehicle's operating parameters, and the coefficients of the lane line fusion function corresponding to the second data acquisition cycle. The operating parameters include the current vehicle's operating speed and yaw rate.

[0132] In some embodiments, location information and image data are acquired. There is a time delay in the acquisition process, which needs to be compensated for. Compensation can be achieved in various ways.

[0133] In some embodiments, the duration of the data acquisition cycle is obtained, and the coefficients of the lane line fusion function corresponding to the second data acquisition cycle are obtained based on the duration of the cycle. As an example, the coefficients of the lane line fusion function corresponding to the second data acquisition cycle are P.fusion_last =[p 0_last ,p 1_last ,p 2_last ,p 3_last The current vehicle speed is v, the yaw rate is ω, and the cycle duration is T.

[0134] S320. Based on the duration of the data acquisition cycle, the current vehicle speed, and the yaw rate, adjust the coefficients of the lane line fusion function to generate smoothing parameters.

[0135] As an example, the coefficients of the lane fusion function are adjusted based on the duration of the data acquisition cycle, the current vehicle speed, and the yaw rate to generate smoothing parameters. in, The calculation formula (9-12) is as follows:

[0136]

[0137]

[0138]

[0139]

[0140] Where, p 0_last ,p 1_last ,p 2_last ,p 3_last ω represents the coefficient of the lane line fusion function corresponding to the second data acquisition cycle, v represents the current vehicle speed, ω represents the yaw rate, and T represents the cycle duration.

[0141] S330. When the vehicle does not change lanes, according to the smoothing parameters and the preset smoothing formula, the first fitting function and the second fitting function are subjected to preset smoothing processing respectively, and the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function are determined. The preset smoothing formula is formula (13-14):

[0142]

[0143]

[0144] Among them, M ′ Here, represents the coefficients of the first smoothing function, and M represents the coefficients of the first fitting function. Here, α is the smoothing parameter, C′ is the coefficient of the second smoothing function, and C is the coefficient of the second fitting function.

[0145] In some embodiments, when the first fitting function and the second fitting function pass the validity detection and the vehicle does not change lanes, the coefficients of the first fitting function and the second fitting function are low-pass filtered (for example, the coefficient of the low-pass filter can be set to α = 0.2). According to the smoothing parameter and the preset smoothing formula, the first fitting function and the second fitting function are respectively subjected to preset smoothing processing to determine the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function.

[0146] In some embodiments, when the first smoothing function and the second smoothing function pass the validity detection and the vehicle changes lanes, the coefficients of the first smoothing function and the second smoothing function retain their original values, for example, M′=M, C′=C.

[0147] In some embodiments, if the first fitting function and the second fitting function fail the validity test, the coefficients of the first smoothing function and the second smoothing function are set to default invalid values ​​(the default invalid values ​​can be set to the default invalid values ​​agreed upon by the high-precision map protocol, the default invalid values ​​agreed upon by the camera communication protocol, or values ​​that do not coincide with the actual range of parameter values, such as M). default =

[0148] [-99,-99,-99,-99],C default = [-99, -99, -99, -99]), at this time the coefficients of the first smoothing function and the second smoothing function are M′ = M default C′=C default .

[0149] In this way, filtering is applied to both the first and second fitting functions to ensure their smoothness.

[0150] Based on this, in some embodiments, the first data acquisition period includes multiple sampling times, and the positioning information includes latitude and longitude information corresponding to the multiple sampling times respectively;

[0151] Based on the location information, determine the first fitting function for the target lane line corresponding to the road where the vehicle is located, including:

[0152] Based on the preset coordinate transformation relationship, the coordinate information of each latitude and longitude information point in the vehicle coordinate system is determined. The preset coordinate transformation relationship is the transformation relationship between the ground coordinate system where the latitude and longitude information is located and the vehicle coordinate system.

[0153] The coordinate information corresponding to multiple coordinate points is fitted to generate the first fitting function for the target lane line.

[0154] In some embodiments, according to a preset coordinate transformation relationship, each latitude and longitude information is projected into the vehicle coordinate system to determine the coordinate information of the coordinate points in the vehicle coordinate system. The coordinate information corresponding to multiple coordinate points is fitted to generate a first fitting function for the target lane line. The vehicle coordinate system is determined by the latitude and longitude information and heading angle of the vehicle output by the high-precision positioning system.

[0155] As an example, Figure 4 This is a schematic diagram of the first fitted function, as shown below. Figure 4 As shown, within the vehicle coordinate system xoy, the coordinate information of each latitude and longitude information in the vehicle coordinate system is determined according to the preset coordinate transformation relationship. The coordinate information corresponding to multiple coordinate points is fitted to generate the first fitting function of the target lane line.

[0156] This allows for a more accurate fit to lane lines.

[0157] In some embodiments, such as Figure 5 As shown, the first fitting function corresponds to the first target confidence level; obtaining the first target confidence level corresponding to the first fitting function includes S510 to S540.

[0158] S510: Obtain the location confidence level of the positioning system in the vehicle and the map confidence level of the map data configured in the positioning system.

[0159] In some embodiments, the positioning confidence is high-precision positioning confidence. The more accurate the high-precision positioning result, the smaller the projection error when the first fitting function is projected onto the vehicle coordinate system, and the higher the positioning confidence. Conversely, the lower the confidence, the lower the confidence. The accuracy of the high-precision positioning result can be obtained through various means, such as determining the accuracy of the high-precision positioning result by at least one of the state level output by the high-precision positioning system, the variance of the positioning result output by the high-precision positioning system, and the state level of the GPS signal.

[0160] In some embodiments, the higher the accuracy and the faster the update frequency of the map data configured in the positioning system, the higher the map confidence level; conversely, the lower the map confidence level, the lower the accuracy and the faster the update frequency. This can be determined by the reference error value of the map data configured in the positioning system and the map update frequency, which is typically updated monthly, weekly, or daily.

[0161] In some embodiments, the location confidence is calculated by the positioning system in the vehicle itself and the confidence score is directly output.

[0162] S520. Based on the coordinate information corresponding to multiple coordinate points, determine the projection distance between each coordinate point and the first fitting function.

[0163] S530. Determine the confidence level of the first fitting function based on the preset confidence range corresponding to the maximum projection distance among multiple projection distances, wherein the preset confidence range includes at least one preset numerical range, and each numerical range corresponds to a confidence level.

[0164] In some embodiments, the closer the first fitting function is to the coordinate point, the higher the confidence level of the first fitting function; conversely, the further away the first fitting function is from the coordinate point, the lower the confidence level of the first fitting function.

[0165] As an example, each coordinate point can determine a projection distance for the first fitting function. Based on the preset confidence range corresponding to the maximum projection distance among multiple projection distances (for example, the preset confidence range is 0.5m), the confidence of the first fitting function is determined. If the maximum projection distance exceeds the preset confidence range, the confidence of the first fitting function is 0. The formula for calculating the confidence of the first fitting function (15) is as follows:

[0166]

[0167] Where Dmax is the maximum projection distance and D0 is the preset confidence range.

[0168] S540. Calculate the product of map confidence, location confidence, and confidence of the first fitting function to obtain the first target confidence corresponding to the first fitting function.

[0169] In some embodiments, the first target confidence level can be determined in various ways. For example, the first target confidence level can be determined by the product of map confidence level, location confidence level, and the confidence level of the first fitting function. As an example, the first target confidence level corresponding to the first fitting function is:

[0170] γ M =γ M1 γ M2 γ M3 (16)

[0171] Where, γ M1 γ represents the confidence level of the first fitted function. M2， γ M3 These are location confidence and map confidence, respectively.

[0172] In this way, the accuracy of the first pseudofunction can be determined by calculating the confidence level.

[0173] Based on this, in some embodiments, the validity of the first fitting function is tested according to preset verification conditions, including:

[0174] The verification parameters of the first fitting function are obtained. The verification parameters of the first fitting function include the function coefficients of the third fitting function corresponding to the L data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each third fitting function. The third fitting function is determined based on the vehicle's positioning information.

[0175] If the verification parameters of the first fitting function meet the first preset validity judgment condition, the first fitting function passes the validity test.

[0176] The first presupposed validity determination criteria include at least one or more of the following determination criteria:

[0177] The coefficients of the L third fitting functions are all different;

[0178] The count indicators corresponding to the function coefficients of the L third fitted functions change sequentially;

[0179] The width of the target lane line determined based on the first fitting function is less than the preset lane line width;

[0180] The change in the coefficients of the third fitting function corresponding to the l-th data acquisition cycle compared to the coefficients of the third fitting function corresponding to the data acquisition cycles prior to the l-th data acquisition cycle is within a preset numerical range.

[0181] The counter is incremented by 1 for each output of the location information.

[0182] In some embodiments, as the vehicle moves in its lane, the function coefficients corresponding to the third fitting function will change accordingly. If the function coefficients corresponding to the third fitting function remain unchanged, the first fitting function is abnormal and fails the validity test.

[0183] In some embodiments, if the width of the target lane line determined by the first fitting function is less than the normal lane width or the lane width provided by the high-precision map, the first fitting function is abnormal and fails the validity test.

[0184] In some embodiments, it is detected whether the coefficients of the first fitting function do not fluctuate or jitter. That is, the change in the coefficients of the third fitting function corresponding to the l-th data acquisition period and the coefficients of the third fitting function corresponding to the data acquisition periods before the l-th data acquisition period are within a preset value range. If the change is not within the preset value range, the first fitting function is considered to have fluctuates or jitter and fails the validity test.

[0185] In some embodiments, if the first fitting function fails the validity test, the confidence level corresponding to the first fitting function is changed to low confidence (e.g., the confidence level corresponding to the first fitting function is 0).

[0186] As an example, Figure 6 This is a flowchart illustrating the validity testing process, such as... Figure 6 As shown, the validity of the first fitting function is tested. The parameter is the coefficient of the first fitting function. Whether the parameter is updated is to determine whether the function coefficients corresponding to the L third fitting functions are all different, and whether the count indicators corresponding to the function coefficients of the L third fitting functions change sequentially. Whether the parameter is within the valid value range is to determine whether the width of the target lane line determined based on the first fitting function is less than the preset lane line width. Whether the parameter does not jump or jitter is to determine whether the change in the function coefficient of the third fitting function corresponding to the l-th data acquisition cycle and the function coefficient of the third fitting function corresponding to the data acquisition cycle before the l-th data acquisition cycle is within the preset value range.

[0187] In this way, the effectiveness can be checked, and any anomalies that occur during the fitting process can be detected in a timely manner.

[0188] Based on this, in some embodiments, the above-mentioned S130 may specifically include:

[0189] The image data of the road is identified and analyzed to generate an initial fitting function for the target lane line and the category information of the target lane line. The category information is associated with preset compensation parameters and the function movement direction.

[0190] Based on the compensation parameters associated with the target lane line category information and the function movement direction, the initial fitting function is moved to obtain the second fitting function.

[0191] In some embodiments, lane lines have width. Based on road image data, an initial fitting function is determined to characterize the inner edge curve of the target lane line. Furthermore, different initial fitting functions are determined for different types of target lane lines. Preset compensation parameters and function movement directions are associated with lane line information of different categories, for example, such as... Figure 7 and Figure 8 The images show different types of lane markings. Figure 7 The image shows the actual location represented by the initial fitting function, the first fitting function, and the second fitting function for lane lines of non-longitudinal speed bump type. Figure 8 The image shows the actual location represented by the initial fitting function, the first fitting function, and the second fitting function for the longitudinal speed bump type lane line. The actual location represented by the second fitting function includes the actual locations of longitudinal speed bumps that are identified and those that are not identified based on the image data.

[0192] As an example, road image data is analyzed. If the actual target lane line is classified as a longitudinal speed bump type lane line, and a non-longitudinal speed bump type lane line is identified, the compensation parameter is 1 / 2 lane width. The initial fitting function is shifted 1 / 2 lane width towards the lane center to obtain the second fitting function. Similarly, if the actual target lane line is classified as a longitudinal speed bump type lane line, and a longitudinal speed bump type lane line is identified, the compensation parameter is 1 / 2 longitudinal speed bump bandwidth. The initial fitting function is shifted 1 / 2 longitudinal speed bump bandwidth towards the lane center to obtain the second fitting function. Here, the lane width and longitudinal speed bump bandwidth can be obtained from a high-precision map or empirical values.

[0193] In this way, both the compensated second fitting function and the first fitting function are used to characterize the center position of the lane line.

[0194] Based on this, in some embodiments, the second fitting function corresponds to the second target confidence level; the second target confidence level is determined based on the clarity of the road image data.

[0195] In some embodiments, road image data is susceptible to weather, lighting, obstacle occlusion, and the original lane line clarity, and the generated target lane line category information may contain errors, which may affect the accuracy of the final second fitting function. The second target confidence is determined based on the clarity of the road image data.

[0196] In some embodiments, the confidence level of the second fitting function can be determined by the detection quality of the camera. The better the detection quality, the higher the confidence level. For example, if the detection quality provided by the camera is high, then the confidence level γ of the second fitting function is higher. C =1, if the detection quality is medium, then γ C =0.7, if the detection quality is low, then γ C =0.4.

[0197] In this way, a second fitting function corresponding to the second target confidence level can be set for the image data based on the effects of weather, lighting, obstacle occlusion, and the original lane line clarity.

[0198] Based on this, in some embodiments, the second fitting function is validated according to preset verification conditions, including:

[0199] The verification parameters of the second fitting function are obtained. The verification parameters of the second fitting function include the function coefficients of the fourth fitting function corresponding to the M data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each fourth fitting function. The fourth fitting function is determined based on the vehicle's positioning information.

[0200] If the verification parameters of the second fitting function meet the second preset validity judgment condition, the second fitting function passes the validity test.

[0201] The second pre-set validity determination criteria include at least one or more of the following determination criteria:

[0202] The coefficients of the M fourth fitting functions are all different;

[0203] The count indicators corresponding to the function coefficients of the M fourth fitting functions change sequentially;

[0204] The width of the target lane line determined based on the second fitting function is less than the preset lane line width;

[0205] The change in the coefficients of the fourth fitting function corresponding to the m-th data acquisition cycle compared to the coefficients of the fourth fitting function corresponding to the data acquisition cycles prior to the m-th data acquisition cycle is within a preset range.

[0206] Here, M and L can be the same or different.

[0207] The counter is determined by the number of times the camera outputs image data; the counter increments by 1 for each output.

[0208] In some embodiments, as the vehicle moves in its lane, the function coefficients corresponding to the fourth fitting function will change accordingly. If the function coefficients corresponding to the fourth fitting function remain unchanged, the second fitting function is abnormal and fails the validity test.

[0209] In some embodiments, if the width of the target lane line determined by the second fitting function is less than the normal lane width or the lane width provided by the high-precision map, the second fitting function is abnormal and fails the validity test.

[0210] In some embodiments, it is detected whether the coefficients of the second fitting function do not fluctuate or jitter. That is, the change in the coefficients of the fourth fitting function corresponding to the m-th data acquisition period and the coefficients of the fourth fitting function corresponding to the data acquisition periods before the m-th data acquisition period is within a preset value range. If the change is not within the preset value range, the second fitting function is considered to have fluctuates or jitter and fails the validity test.

[0211] In some embodiments, if the second fitting function fails the validity test, the confidence level corresponding to the second fitting function is changed to a low confidence level (e.g., the confidence level corresponding to the second fitting function is 0).

[0212] As an example, Figure 8 Here is a flowchart illustrating the validity check process, such as... Figure 8 As shown, the validity of the second fitting function is tested. The parameters are the coefficients of the second fitting function. Whether the parameters are updated is determined by whether the function coefficients corresponding to the M fourth fitting functions are all different, and whether the count indicators corresponding to the function coefficients of the M fourth fitting functions change sequentially. Whether the parameters are within the valid value range is determined by whether the width of the target lane line determined based on the second fitting function is less than the preset lane line width. Whether the parameters do not jump or jitter is determined by whether the change in the function coefficient of the fourth fitting function corresponding to the m-th data acquisition cycle and the function coefficient of the fourth fitting function corresponding to the data acquisition cycle before the m-th data acquisition cycle is within the preset value range.

[0213] In this way, the effectiveness can be checked, and any anomalies that occur during the fitting process can be detected in a timely manner.

[0214] Based on the lane line fusion method provided in the above embodiments, this application also provides specific implementations of the lane line fusion device. Please refer to the following embodiments.

[0215] First see Figure 10 The lane line fusion device 900 provided in this application embodiment includes:

[0216] The acquisition module 1010 is used to acquire the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period before the first data acquisition period.

[0217] The processing module 1020 is used to determine a first fitting function for the target lane line corresponding to the road where the vehicle is located based on the positioning information, wherein the target lane line is any lane line among the lane lines on both sides of the road; and to determine a second fitting function for the target lane line based on the image data of the road.

[0218] The acquisition module 1010 is also used to acquire the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively.

[0219] The processing module 1020 is also used to perform validity checks on the first fitting function and the second fitting function respectively according to preset verification conditions;

[0220] The processing module 1020 is further configured to, when the first fitting function and the second fitting function pass the validity test respectively, perform a weighted calculation based on the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, and the confidence levels corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period respectively, to obtain the target lane line fusion function corresponding to the first data acquisition period.

[0221] Based on this, in some embodiments, the processing module 1020 may include:

[0222] The processing submodule is used to perform preset smoothing processing on the first fitting function and the second fitting function respectively according to the preset function smoothing algorithm, when the first fitting function and the second fitting function pass the validity detection respectively, to obtain the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function.

[0223] The calculation submodule is used to perform weighted calculations based on the first smoothing function, the second smoothing function, the lane line fusion function corresponding to the second data acquisition period, and the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively, to obtain the target lane line fusion function corresponding to the first data acquisition period.

[0224] Based on this, in some embodiments, the processing submodule may include:

[0225] The acquisition unit is used to acquire the duration of the data acquisition cycle, the vehicle's operating parameters, and the coefficients of the lane fusion function corresponding to the second data acquisition cycle. The operating parameters include the current vehicle's operating speed and yaw rate.

[0226] The adjustment unit is used to adjust the coefficients of the lane line fusion function based on the duration of the data acquisition cycle, the current vehicle speed, and the yaw rate, thereby generating smoothing parameters.

[0227] The processing unit is configured to, when the vehicle has not changed lanes, perform preset smoothing processing on a first fitting function and a second fitting function according to smoothing parameters and a preset smoothing formula, respectively, to determine the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function, wherein the preset smoothing formula is:

[0228]

[0229]

[0230] Among them, M ′ Here, represents the coefficients of the first smoothing function, and M represents the coefficients of the first fitting function. Here, α is the smoothing parameter, C′ is the coefficient of the second smoothing function, and C is the coefficient of the second fitting function.

[0231] Based on this, in some embodiments, the first data acquisition period includes multiple sampling times, and the positioning information includes latitude and longitude information corresponding to the multiple sampling times respectively;

[0232] Processing module 1020 may include:

[0233] The determination submodule is used to determine the coordinate information of each latitude and longitude information point in the vehicle coordinate system according to the preset coordinate transformation relationship. The preset coordinate transformation relationship is the transformation relationship between the ground coordinate system where the latitude and longitude information is located and the vehicle coordinate system.

[0234] The fitting submodule is used to fit the coordinate information corresponding to multiple coordinate points to generate the first fitting function for the target lane line.

[0235] Based on this, in some embodiments, the first fitting function corresponds to the first target confidence level; the acquisition module 1010 may include:

[0236] The acquisition submodule is used to acquire the location confidence level corresponding to the positioning system in the vehicle, the map confidence level of the map data configured within the positioning system, and,

[0237] The determination submodule is used to determine the projection distance between each coordinate point and the first fitting function based on the coordinate information corresponding to multiple coordinate points.

[0238] The determination submodule is also used to determine the confidence level of the first fitting function based on the preset confidence range corresponding to the maximum projection distance among multiple projection distances, wherein the preset confidence range includes at least one preset numerical range, and each numerical range corresponds to a confidence level;

[0239] The calculation submodule is used to calculate the product of map confidence, location confidence, and the confidence of the first fitting function to obtain the first target confidence corresponding to the first fitting function.

[0240] Based on this, in some embodiments, the processing module 1020 may include:

[0241] The acquisition submodule is used to acquire the verification parameters of the first fitting function. The verification parameters of the first fitting function include the function coefficients of the third fitting function corresponding to the L data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each third fitting function. The third fitting function is determined based on the vehicle's positioning information.

[0242] The judgment submodule is used to ensure that the first fitting function passes the validity test if the verification parameters of the first fitting function meet the first preset validity judgment condition.

[0243] The first presupposed validity determination criteria include at least one or more of the following determination criteria:

[0244] The coefficients of the L third fitting functions are all different;

[0245] The count indicators corresponding to the function coefficients of the L third fitted functions change sequentially;

[0246] The width of the target lane line determined based on the first fitting function is less than the preset lane line width;

[0247] The change in the coefficients of the third fitting function corresponding to the l-th data acquisition cycle compared to the coefficients of the third fitting function corresponding to the data acquisition cycles prior to the l-th data acquisition cycle is within a preset numerical range.

[0248] Based on this, in some embodiments, the processing module 1020 may include:

[0249] The generation submodule is used to identify and analyze road image data, generate an initial fitting function for the target lane line and the category information of the target lane line, wherein the category information is associated with preset compensation parameters and function movement direction;

[0250] The processing submodule is also used to move the initial fitting function based on the compensation parameters associated with the target lane line category information and the function movement direction to obtain the second fitting function.

[0251] Based on this, in some embodiments, the second fitting function corresponds to the second target confidence level; the second target confidence level is determined based on the clarity of the road image data.

[0252] Based on this, in some embodiments, the processing module 1020 may include:

[0253] The acquisition submodule is also used to acquire the verification parameters of the second fitting function. The verification parameters of the second fitting function include the function coefficients of the fourth fitting function corresponding to the M data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficients of each fourth fitting function. The fourth fitting function is determined based on the vehicle's positioning information.

[0254] The judgment submodule is also used to ensure that the second fitting function passes the validity test when the verification parameters of the second fitting function meet the second preset validity judgment condition;

[0255] The second pre-set validity determination criteria include at least one or more of the following determination criteria:

[0256] The coefficients of the M fourth fitting functions are all different;

[0257] The count indicators corresponding to the function coefficients of the M fourth fitting functions change sequentially;

[0258] The width of the target lane line determined based on the second fitting function is less than the preset lane line width;

[0259] The change in the coefficients of the fourth fitting function corresponding to the m-th data acquisition cycle compared to the coefficients of the fourth fitting function corresponding to the data acquisition cycles prior to the m-th data acquisition cycle is within a preset range.

[0260] The various modules of the lane line fusion device provided in this application embodiment can achieve Figures 1 to 8 It provides the functionality for each step of the lane line fusion method and achieves the corresponding technical effects. For the sake of brevity, it will not be elaborated here.

[0261] Based on the same inventive concept, embodiments of this application also provide an electronic device.

[0262] Figure 11 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0263] An electronic device may include a processor 1101 and a memory 1102 storing computer program instructions.

[0264] Specifically, the processor 1101 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0265] Memory 1102 may include mass storage for data or instructions. For example, and not limitingly, memory 1102 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 1102 may include removable or non-removable (or fixed) media. Where appropriate, memory 1102 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 1102 is non-volatile solid-state memory.

[0266] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0267] The processor 1101 implements any of the lane line fusion methods in the above embodiments by reading and executing computer program instructions stored in the memory 1102.

[0268] In one example, the electronic device may also include a communication interface 1103 and a bus 1110. For example, Figure 11 As shown, the processor 1101, memory 1102, and communication interface 1103 are connected through bus 1110 and complete communication with each other.

[0269] The communication interface 1103 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0270] Bus 1110 includes hardware, software, or both, that couples components of an electronic device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Linear Predictive Coding (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (Peripheral Component Interconnect-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VESA Local Bus, VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 1110 may include one or more buses. Although specific buses are described and illustrated in the embodiments of this application, this application contemplates any suitable bus or interconnection. The electronic device can perform the lane fusion method in the embodiments of the present invention, thereby achieving... Figures 1 to 8 The lane line fusion method is described.

[0271] Furthermore, in conjunction with the lane line fusion methods in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the lane line fusion methods in the above embodiments.

[0272] This application also provides a computer program product in which the instructions, when executed by a processor of an electronic device, cause the electronic device to perform various processes implementing any of the lane line fusion method embodiments described above.

[0273] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0274] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, read-only memory (ROM), flash memory, erasable read-only memory (EROM), floppy disks, compact disc read-only memory (CD-ROM), optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0275] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0276] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0277] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A lane line fusion method, characterized in that, include: The system acquires the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period preceding the first data acquisition period. Based on the positioning information, a first fitting function is determined for the target lane line corresponding to the road where the vehicle is located, wherein the target lane line is any one of the lane lines on both sides of the road; Based on the image data of the road, a second fitting function for the target lane line is determined; Obtain the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively; According to the preset verification conditions, the effectiveness of the first fitting function and the second fitting function are tested respectively. If the first fitting function and the second fitting function pass the validity detection, a weighted calculation is performed based on the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, and the confidence levels corresponding to the lane line fusion function corresponding to the first fitting function, the second fitting function and the second data acquisition period, respectively, to obtain the target lane line fusion function corresponding to the first data acquisition period. Wherein, the first fitting function corresponds to the first target confidence level; obtaining the first target confidence level corresponding to the first fitting function includes: Based on the projection error when the first fitting function is projected onto the vehicle coordinate system, the positioning confidence of the positioning system in the vehicle is determined. Based on the reference error value and map update frequency of the map data, the map confidence level of the map data configured in the positioning system is determined. Based on the coordinate information corresponding to the coordinate points of the multiple positioning information in the vehicle coordinate system, the projection distance between each coordinate point and the first fitting function is determined; The confidence level of the first fitting function is determined based on the preset confidence range corresponding to the maximum projection distance among the multiple projection distances, wherein the preset confidence range includes at least one preset numerical range, and each numerical range corresponds to a confidence level; The confidence of the map, the confidence of the location, and the confidence of the first fitting function are calculated to obtain the first target confidence corresponding to the first fitting function. Based on preset verification conditions, the first fitted function is subjected to validity testing, including: Obtain the verification parameters of the first fitting function. The verification parameters of the first fitting function include the function coefficients of the third fitting function corresponding to the L data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each third fitting function. The third fitting function is determined based on the vehicle's positioning information. If the verification parameters of the first fitting function meet the first preset validity determination condition, the first fitting function passes the validity detection.

2. The method according to claim 1, characterized in that, When the first fitting function and the second fitting function pass the validity detection, a weighted calculation is performed based on the lane line fusion function corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, to obtain the target lane line fusion function corresponding to the first data acquisition period, including: If the first fitting function and the second fitting function pass the validity detection, the first fitting function and the second fitting function are respectively subjected to preset smoothing processing according to the preset function smoothing algorithm to obtain the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function; The target lane fusion function corresponding to the first data acquisition period is obtained by weighting the lane fusion functions corresponding to the first smoothing function, the second smoothing function, and the lane fusion function corresponding to the second data acquisition period, as well as the confidence levels corresponding to the first fitting function, the second fitting function, and the lane fusion function corresponding to the second data acquisition period.

3. The method according to claim 2, characterized in that, The step of performing preset smoothing processing on the first fitting function and the second fitting function according to a preset function smoothing algorithm to obtain a first smoothing function corresponding to the first fitting function and a second smoothing function corresponding to the second fitting function includes: The system acquires the duration of the data acquisition cycle, the vehicle's operating parameters, and the coefficients of the lane fusion function corresponding to the second data acquisition cycle. The operating parameters include the current vehicle speed and yaw rate. Based on the duration of the data acquisition cycle, the current vehicle speed, and the yaw rate, the coefficients of the lane line fusion function are adjusted to generate smoothing parameters. When the vehicle does not change lanes, based on the smoothing parameters and the preset smoothing formula, the first fitting function and the second fitting function are respectively subjected to preset smoothing processing to determine the first smoothing function corresponding to the first fitting function and the second smoothing function corresponding to the second fitting function, wherein the preset smoothing formula is: in, Let M be the coefficient of the first smoothing function, and let M be the coefficient of the first fitting function. Here, α is the smoothing parameter, and α is the preset filtering coefficient. is the coefficient of the second smoothing function, and C is the coefficient of the second fitting function.

4. The method according to claim 1, characterized in that, The first data acquisition period includes multiple sampling times, and the positioning information includes latitude and longitude information corresponding to each of the multiple sampling times; The step of determining the first fitting function for the target lane line corresponding to the road where the vehicle is located based on the positioning information includes: Based on a preset coordinate transformation relationship, the coordinate information of each latitude and longitude information point in the vehicle coordinate system is determined, wherein the preset coordinate transformation relationship is the transformation relationship between the ground coordinate system where the latitude and longitude information is located and the vehicle coordinate system; The coordinate information corresponding to the multiple coordinate points is fitted to generate the first fitting function of the target lane line.

5. The method according to claim 1, characterized in that, The first preset validity determination criteria include at least one or more of the following determination criteria: The coefficients of the L third fitting functions are all different; The count indicators corresponding to the function coefficients of the L third fitting functions change sequentially; The width of the target lane line determined based on the first fitting function is less than the preset lane line width; No. l The coefficients of the third fitting function corresponding to the data acquisition cycle are related to the first data acquisition cycle. l The change in the coefficients of the third fitting function corresponding to the data acquisition period prior to the previous data acquisition period is within the preset numerical range.

6. The method according to claim 1, characterized in that, The step of determining the second fitting function for the target lane line based on the image data of the road includes: The image data of the road is identified and analyzed to generate an initial fitting function for the target lane line and category information of the target lane line, wherein the category information is associated with preset compensation parameters and function movement direction; Based on the compensation parameters associated with the category information of the target lane line and the function movement direction, the initial fitting function is moved to obtain the second fitting function.

7. The method according to claim 6, characterized in that, The second fitting function corresponds to the second target confidence level; the second target confidence level is determined based on the clarity of the image data of the road.

8. The method according to claim 7, characterized in that, Based on preset verification conditions, the second fitted function is subjected to validity testing, including: Obtain the verification parameters of the second fitting function. The verification parameters of the second fitting function include the function coefficients of the fourth fitting function corresponding to the M data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each fourth fitting function. The fourth fitting function is determined based on the vehicle's positioning information. If the verification parameters of the second fitting function meet the second preset validity determination condition, the second fitting function passes the validity detection. The second preset validity determination criteria include at least one or more of the following determination criteria: The coefficients of the M fourth fitting functions are all different; The count indicators corresponding to the function coefficients of the M fourth fitting functions change sequentially; The width of the target lane line determined based on the second fitting function is less than the preset lane line width; The change in the coefficients of the fourth fitting function corresponding to the m-th data acquisition cycle compared to the coefficients of the fourth fitting function corresponding to the data acquisition cycles prior to the m-th data acquisition cycle is within a preset range.

9. A lane line fusion device, characterized in that, The device includes: The acquisition module is used to acquire the vehicle's location information within the first data acquisition period, the image data of the road where the vehicle is located within the first data acquisition period, and the lane line fusion function corresponding to the second data acquisition period, wherein the second data acquisition period is the data acquisition period before the first data acquisition period; The processing module is configured to determine a first fitting function for the target lane line corresponding to the road where the vehicle is located based on the positioning information, wherein the target lane line is any lane line among the lane lines on both sides of the road; and to determine a second fitting function for the target lane line based on the image data of the road. The acquisition module is further configured to acquire the confidence levels corresponding to the first fitting function, the second fitting function, and the lane line fusion function corresponding to the second data acquisition period, respectively. The processing module is further configured to perform validity checks on the first fitting function and the second fitting function respectively according to preset verification conditions; The processing module is further configured to perform a weighted calculation based on the first fitting function, the second fitting function and the lane line fusion function corresponding to the second data acquisition period, and the confidence levels corresponding to the first fitting function, the second fitting function and the lane line fusion function corresponding to the second data acquisition period, respectively, to obtain the target lane line fusion function corresponding to the first data acquisition period, when the first fitting function and the second fitting function pass the validity detection respectively. The acquisition module is further configured to: determine the location confidence of the positioning system in the vehicle based on the projection error when the first fitting function is projected onto the vehicle coordinate system; determine the map confidence of the map data configured in the positioning system based on the reference error value and map update frequency of the map data; determine the projection distance between each coordinate point and the first fitting function according to the coordinate information corresponding to the coordinate points of the multiple positioning information in the vehicle coordinate system; determine the confidence of the first fitting function according to the preset confidence range corresponding to the largest projection distance among the multiple projection distances, wherein the preset confidence range includes at least one preset numerical range, and each numerical range corresponds to a confidence level; and calculate the product of the map confidence, the location confidence, and the confidence of the first fitting function to obtain the first target confidence of the first fitting function. The processing module is further configured to obtain the verification parameters of the first fitting function. The verification parameters of the first fitting function include the function coefficients of the third fitting function corresponding to the L data acquisition cycles before the first data acquisition cycle and the count identifier corresponding to the function coefficient of each third fitting function. The third fitting function is determined based on the vehicle's positioning information. If the verification parameters of the first fitting function meet the first preset validity judgment condition, the first fitting function passes the validity detection.

Images