Lane line recognition method, device and electronic device based on roadside equipment

Abstract

This application relates to a lane line recognition method, apparatus, and electronic device for roadside equipment. The method includes: acquiring multiple first images taken by the roadside equipment at preset time intervals within a preset time period; performing lane line recognition on the multiple first images to obtain multiple second images; fusing the multiple second images to generate a third image; and calculating the pixel data of the lane lines in the third image to obtain the calculated lane line data in the third image. The solution provided by this application can improve the accuracy of lane line recognition.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to lane line recognition methods, devices and electronic devices based on roadside equipment. Background Technology

[0002] Lane recognition has become an indispensable technology for autonomous vehicles. Whether it is lane-level positioning technology, obstacle avoidance based on traffic rules, or local planning, all require high accuracy in lane recognition.

[0003] Currently, lane line recognition using visual detection is a common approach. Visual detection primarily involves first acquiring lane line images, then detecting lane line regions in the images based on image algorithms, and finally grouping the lane lines so that vehicles can automatically identify lane information while driving.

[0004] However, the quality of images acquired through visual detection is greatly affected by ambient light and background light, resulting in low accuracy in lane line recognition. Summary of the Invention

[0005] To address or partially address the problems existing in related technologies, this application provides a lane line recognition method, device, and electronic device based on roadside equipment, which can improve the accuracy of lane line recognition.

[0006] The first aspect of this application provides a lane line recognition method based on roadside equipment, the method comprising:

[0007] Acquire multiple first images taken by roadside equipment at preset time intervals within a preset time period;

[0008] Lane line recognition is performed on multiple first images to obtain multiple second images;

[0009] Multiple second images are merged to generate a third image;

[0010] The pixel data of the lane lines in the third image are calculated to obtain the calculated data of the lane lines in the third image.

[0011] A second aspect of this application provides a lane line recognition device based on roadside equipment, comprising:

[0012] The acquisition module is used to acquire multiple first images taken by the roadside equipment at preset time intervals within a preset time period;

[0013] The recognition module is used to perform lane line recognition on multiple first images to obtain multiple second images;

[0014] The fusion module is used to fuse multiple second images to generate a third image;

[0015] The calculation module is used to calculate the pixel data of the lane lines in the third image to obtain the calculated data of the lane lines in the third image.

[0016] A third aspect of this application provides an electronic device, comprising:

[0017] Processor; and

[0018] A memory that stores executable code, which, when executed by the processor, causes the processor to perform the method described above.

[0019] A fourth aspect of this application provides a computer-readable storage medium having executable code stored thereon, which, when executed by a processor of an electronic device, causes the processor to perform the method described above.

[0020] The technical solution provided in this application can include the following beneficial effects: The lane line recognition method in this application first obtains multiple second images by performing lane line recognition on multiple first images without complete lane lines captured by roadside equipment, then fuses the multiple second images into a third image with complete lane lines, and finally calculates the third image to obtain the calculated data. This can eliminate the disturbance in the quality of the images captured by the roadside equipment due to factors such as weather conditions and changes in road traffic flow, and improve the completeness and accuracy of lane line recognition.

[0021] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0022] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments thereof.

[0023] Figure 1 This is a schematic flowchart illustrating a lane line recognition method based on roadside equipment, as shown in an embodiment of this application.

[0024] Figure 2(a) is a second image at a first time interval within the same preset time period;

[0025] Figure 2(b) is a second image with a second time interval within the same preset time period;

[0026] Figure 2(c) is a second image with a third time interval within the same preset time period;

[0027] Figure 3 The third image is obtained by fusing Figure 2(a), Figure 2(b), and Figure 2(c);

[0028] Figure 4 This is a schematic diagram of the structure of a lane line recognition device based on roadside equipment shown in an embodiment of this application;

[0029] Figure 5 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation

[0030] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0031] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0032] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0033] Currently, lane line recognition using visual detection is a common approach. Visual detection primarily involves first acquiring lane line images, then detecting lane line regions in the images based on image algorithms, and finally grouping the lane lines so that vehicles can automatically identify lane information while driving.

[0034] However, the quality of images acquired through visual detection is greatly affected by ambient light and background light, resulting in low accuracy in lane line recognition.

[0035] To address the aforementioned issues, this application provides a lane line recognition method based on roadside equipment, which can improve the accuracy of lane line recognition.

[0036] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0037] Figure 1 This is a schematic flowchart illustrating a lane line recognition method based on roadside equipment, as shown in an embodiment of this application.

[0038] See Figure 1 The lane line recognition method based on roadside equipment in this application mainly includes steps S101 to S104, which are described below:

[0039] Step S101: Acquire multiple first images taken by the roadside equipment at preset time intervals within a preset time period. Acquiring multiple first images can be done periodically.

[0040] In this embodiment, the roadside equipment refers to devices such as cameras used to capture road conditions, which are installed on both sides of the road and set up continuously at intervals along the road.

[0041] Because the field of view of images captured by roadside equipment is consistent each time, the quality of these images can be affected by factors such as weather conditions and changes in traffic flow. For example, the images may be obstructed by obstacles such as vehicles or pedestrians. Therefore, the lane line images captured by the roadside equipment each time are inconsistent, and the obstructed portions of the lane lines in each image are not entirely the same. This application embodiment uses a method of periodically acquiring multiple first images captured by the roadside equipment at preset time intervals within a preset time period to obtain a complete lane line image. Specifically, for example, using a daily / weekly cycle, every other day / week, multiple first images captured by the roadside equipment at five-minute intervals are acquired within the time period from 10:00 AM to 3:00 PM. In this way, in some images, the obstructed portions of the lane lines may be unobstructed in other images, allowing the unobstructed portions of the lane lines in the multiple first images to be combined to form a complete lane line.

[0042] Step S102: Perform lane line recognition on multiple first images to obtain multiple second images.

[0043] The second image in this embodiment is a binary image of lane lines. Step S102 can be implemented by inputting multiple first images into a preset model to output multiple second images. The output multiple second images are shown in Figures 2(a), 2(b), and 2(c). Figures 2(a), 2(b), and 2(c) are three images taken at intervals of a first time, a second time, and a third time within the same preset time period. For example, three images taken at intervals of 5 minutes, 10 minutes, and 15 minutes within the time period from 10:00 AM to 3:00 PM.

[0044] The pixels at the lane line location are the foreground color (e.g., pixel value 255), while the pixels at other locations are the background color (e.g., pixel value 0).

[0045] The preset model in this application embodiment can be a trained U-Net network. When training the preset model, a certain number of lane line images need to be acquired through roadside equipment as original lane line image samples. The lane lines in the lane line image samples are then binarized using a computer, thereby converting the original lane line images into lane line binary maps. Finally, the U-Net network is trained based on the lane line image samples and the lane line binary maps to obtain the preset model.

[0046] It should be noted that image binarization is the process of setting the grayscale value of pixels in an image to 0 or 255, thus presenting the entire image with a clear black and white effect. This embodiment of the application simplifies lane line images and reduces data volume by performing image binarization processing on lane line images (e.g., setting the pixel value of lane line elements to 255 and the pixel value of other elements to 0), while also highlighting the outline of the lane lines for easier subsequent processing.

[0047] Step S103: Fuse multiple second images to generate a third image.

[0048] As mentioned above, due to the uniform field of view of roadside equipment, the image quality captured by the roadside equipment is disturbed by various factors. In order to obtain a complete lane line image, this embodiment of the application acquires multiple first images and then fuses the multiple first images to generate a third image with complete lane lines. The third image is as follows: Figure 3 As shown, Figure 3 The lane lines in the image are fused with lane line elements from Figures 2(a), 2(b), and 2(c), completing the occluded lane lines and displaying a complete lane line image. This achieves both completeness and accuracy in lane line recognition.

[0049] As an embodiment of this application, the step of fusing multiple second images to generate a third image can be implemented through steps S1031 to S1033, as explained below:

[0050] Step S1031: Select the target image with the most lane line pixel data from multiple second images.

[0051] The target image contains the most lane line pixels, indicating that the target image has the least occlusion and the best quality.

[0052] Step S1032: Retain the pixel data of the lane line segments in the remaining second image whose overlap with the target image is less than a preset threshold.

[0053] This application embodiment can determine the overlap between the remaining second image and the target image by comparing the appearance features of lane lines with a preset threshold. For example, the target image can roughly determine the number of lane lines, the length of each lane line, and the distance between adjacent lane lines. The corresponding appearance features of the lane lines are obtained from the remaining second image. Then, the remaining value of the appearance features of the target image and the appearance features of the remaining second image is compared with a preset threshold. If the value exceeds the threshold, it indicates that the corresponding lane line segment does not overlap with the target image, i.e., it does not appear in the target image, and the pixel data of that lane line segment is retained.

[0054] Step S1033: The pixel data of the retained lane line segments are added to the target image to generate a third image.

[0055] Finally, the pixel data of the retained lane line segments are added to the target image to generate a third image, which is an image containing the complete lane lines.

[0056] Step S104: Solve the pixel data of the lane lines in the third image to obtain the solved data of the lane lines in the third image.

[0057] Since the pixel data in the third image is in a pixel coordinate system and cannot represent the lane lines in a global coordinate system, it cannot be used to create high-definition maps. Therefore, the pixel data of the lane lines in the third image needs to be processed to obtain calculated data, which can then be used for high-definition map creation.

[0058] It should be noted that the solution results are the position, size, and shape of the lane lines on the actual lanes. They are generally identified using GPS coordinates and are used for marking on the map when creating high-precision maps.

[0059] As an embodiment of this application, the step of calculating the pixel data of the lane lines in the third image to obtain the calculated data of the lane lines in the third image can be implemented through steps S1041 to S1044, as explained below:

[0060] Step S1041: Extract pixel data of the target lane line in the preset confidence area in the third image.

[0061] Considering the characteristics of roadside equipment, distortion increases with distance, leading to larger errors. Therefore, this embodiment of the application sets a reliable region (e.g., a maximum distance of 100 meters from the roadside equipment, with a width of 15 meters on each side), and ignores data outside the reliable region during calculation to reduce calculation errors.

[0062] It should be noted that, in this embodiment, the trusted region refers to the region within which the calculated pixel data of the target lane line is relatively accurate and meets the accuracy requirements. The specific range of the trusted region can be flexibly set by those skilled in the art according to actual needs, and is not specifically limited here.

[0063] Step S1042: Combine the coordinates and extrinsic parameters of the roadside equipment to calculate the pixel data of the target lane line and obtain several feature points of the target lane line to be fitted.

[0064] In this embodiment, the first image is acquired by a roadside device, and the second image is obtained after image binarization. The third image, obtained by fusing the second image, contains not only the identified lane lines but also continuous positioning information, such as GPS data. For the third image, the pixel coordinates of the target lane lines in the third image can be calculated into coordinates in the camera coordinate system based on the camera's intrinsic and extrinsic parameters and the transformation matrix between the pixel coordinate system and the camera coordinate system. In other words, the pixels of the target lane lines in the pixel coordinate system are converted into feature points to be fitted in the camera coordinate system. These feature points to be fitted are discrete points corresponding to the target lane lines.

[0065] It should be noted that when defining the camera coordinate system, one of the coordinate axes, such as the horizontal or vertical axis, is perpendicular or nearly perpendicular to the direction of the lane line. Therefore, the various calculations performed in the camera coordinate system consume fewer resources, unlike in the world coordinate system where the angle between the direction of the lane line and the coordinate axis of the world coordinate system varies, making the various calculations performed in the world coordinate system very complex and consuming a lot of computational resources.

[0066] Step S1043: Cluster several feature points to be fitted on the same target lane line into a group.

[0067] Since there are multiple lane lines, and the feature points to be fitted obtained above are discrete points, these discrete points need to be clustered into groups in order to complete the subsequent fitting of straight lines or curves.

[0068] Step S1044: Fit several feature points to be fitted within the same group to obtain the solution data of the target lane line.

[0069] As mentioned earlier, the feature points to be fitted are discrete points and cannot be used as points of the target lane line in the map coordinate system. Therefore, curve fitting can be used to fit the feature points of the target lane line in the camera coordinate system to obtain the solution data. The solution data is then used as the data of the target lane line in the Earth coordinate system, ultimately obtaining the position, size, and shape of the target lane line on the actual road. The solution data is then used to create a high-precision map.

[0070] From the above Figure 1 As can be seen from the example of the lane line recognition method based on roadside equipment, the lane line recognition method of this application embodiment first obtains multiple second images by performing lane line recognition on multiple first images without complete lane lines captured by the roadside equipment, then fuses the multiple second images into a third image with complete lane lines, and finally processes the third image to obtain the calculated data. This can eliminate the disturbance to the image quality caused by factors such as weather conditions and changes in road traffic flow caused by the roadside equipment, and improve the completeness and accuracy of lane line recognition.

[0071] Currently, lane line data is typically obtained through data collection vehicles, which are costly and have long collection cycles, usually several months. During this long period, lane lines may have changed, but it's impossible to collect lane line data again using a collection vehicle. Therefore, this application's embodiment utilizes lane line images captured by roadside equipment to assist in detecting changes to lane lines on the road to be inspected, significantly reducing costs.

[0072] Specifically, the lane marking changes on the road to be inspected are detected through the following steps:

[0073] Step S105: If there is no existing lane line data within the preset road area, the solution data for the current period is confirmed in order to create a high-precision map.

[0074] If there is no existing lane line data within the preset road area, it means that the current period's calculated data is the latest data within the preset road area, and the data collection vehicle has not yet collected lane line data within this preset road area. Therefore, the current period's calculated data can be directly confirmed, and after confirmation, it can be used to create a high-precision map.

[0075] It should be noted that the verification of the solution data can be carried out in different ways. For example, algorithms can be used to determine the technical indicators of the solution results, or manual intervention can be used to verify the results by comparing them with images, videos, point clouds, etc. Those skilled in the art can set the parameters flexibly according to actual needs, and no specific limitations are made here.

[0076] Step S106: If existing lane line data exists within the preset road range, determine whether the lane lines within the preset road range have changed based on the existing lane line data or the current period's calculation data.

[0077] If existing lane line data exists within the preset road area, it means that the lane line data within the preset road area could have been collected by the data collection vehicle before. In this embodiment, the current period's solution data and existing lane line data can be used to detect changes in the lane lines within the preset area, so as to update or correct the lane line data in a timely manner.

[0078] As an embodiment of this application, the step of determining whether the lane lines within the preset road area have changed can be implemented through steps S1061 to S1063, as explained below:

[0079] Step S1061: Based on the existing lane line data, find multiple solution data for the current period within the neighborhood of the existing lane line data.

[0080] During the search process, two scenarios may occur:

[0081] The first scenario involves existing lane line data within the preset road area, where only existing lane line data exists, with no calculated data for the current period. In other words, the preset road area contains only old data, with no new data. This means the lane lines within the preset road area may have been erased, or they may have failed to be recognized by the roadside equipment due to obstruction or other reasons. This is the first scenario requiring confirmation, and the specific situation needs to be verified to update the lane line information within the preset road area in real time.

[0082] The second scenario involves existing lane line data within the preset road area, where multiple calculated data points exist for the current period, indicating the presence of new data. In this case, the following steps are used to determine lane line changes:

[0083] Step S1062: If there is a solution data in the current cycle that completely matches the existing lane line data, then it is determined that the lane lines within the preset road range have not changed.

[0084] The above description refers to the existence of existing lane line data within a preset road area, and the existence of current-period solution data that perfectly matches the existing lane line data within the domain of the preset road area. In other words, new data that perfectly matches the old data exists around the old data. This indicates that the new data is consistent with the old data, and the old data has not changed; that is, the lane lines within the preset road area have not changed.

[0085] The matching mentioned in the embodiments of this application refers to the fact that the calculated data within the preset road range and the existing lane line data within the preset road range meet a predetermined threshold in terms of position, length, angle, etc. (e.g., position or length within 0.5 meters, angle within 5 degrees, etc., which are set according to actual needs and are not specifically limited here).

[0086] Step S1063: If there is no solution data in the multiple solution data of the current period that completely matches the existing lane line data, then it is determined that the lane lines within the preset road range have changed.

[0087] The above description refers to the existence of existing lane line data within a preset road area, and the absence of any current-period solution data within that preset road area that perfectly matches the existing lane line data. In other words, there is no new data that perfectly matches the old data in its vicinity. This indicates that the new data is inconsistent with the old data, and the old data has changed, meaning the lane lines within the preset road area have changed.

[0088] As another embodiment of this application, the step of determining whether the lane lines within the preset road range have changed can be implemented through steps S1061' to S1063', as explained below:

[0089] Step S1061': Based on the solution data of the current period, find multiple existing lane line data in the neighborhood of the solution data of the current period;

[0090] During the search process, two scenarios may occur:

[0091] The first scenario involves a situation where, within the current period's calculated data area of ​​the preset road range, only the current period's calculated data is available, with no existing lane line data. In other words, within the preset road range, all data is new, with no old data. Therefore, the lane lines within the preset road range could be newly added, or they could be the result of misidentification by the roadside equipment. This constitutes the second scenario requiring confirmation. In such cases, it is necessary to verify the specific situation to update the lane line information within the preset road range in real time.

[0092] The second scenario involves multiple individuals having existing lane line data within the current cycle's calculated data area of ​​a preset road zone, i.e., existing old data. In this case, the following steps are used to determine lane line changes:

[0093] Step S1062': If there is existing lane line data among multiple existing lane line data that completely matches the solution data of the current period, then it is determined that the lane lines within the preset road range have not changed.

[0094] The above description refers to the existence of solution data for the current period within the preset road area, and the existence of existing lane line data that completely matches the solution data for the current period within the domain of the preset road area. In other words, there is old data that completely matches the new data. This indicates that the new data is consistent with the old data, the old data has not changed, and that is, the lane lines within the preset road area have not changed.

[0095] Step S1063: If there is no existing lane line data among the multiple existing lane line data that completely matches the solution data of the current period, then it is determined that the lane lines within the preset road range have changed.

[0096] The above description refers to the existence of solution data for the current period within the preset road area, and the absence of existing lane line data that perfectly matches the solution data for the current period within the preset road area. In other words, there is no old data that perfectly matches the new data in its vicinity. This indicates that the new data is inconsistent with the old data; the new data has changed, meaning the lane lines within the preset road area have changed.

[0097] As one embodiment of this application, the existing lane line data is lane line data in a geospatial database or historical periodic calculation data.

[0098] The existing lane line data is lane line data within the geospatial database. In other words, the existing lane line data can be data collected by a data collection vehicle and stored in the geospatial database.

[0099] The existing lane line data is calculated from historical periods, meaning it could be data from the previous period or even the period before that. Comparing the historical data with the current data detects lane line changes and avoids errors caused by poor accuracy of external parameters from roadside equipment.

[0100] As an embodiment of this application, if the existing lane line data is lane line data in a geospatial database, and it is determined that the lane lines within the preset road range have not changed, then the external parameters of the roadside equipment are calibrated using the lane line data in the geospatial database.

[0101] Because the lane line data in the geospatial database is highly accurate, when it is confirmed that there have been no changes in the lane lines within the preset road area, the extrinsic parameters of the roadside equipment can be calibrated using the lane line data in the geospatial database to improve the calculation accuracy.

[0102] Corresponding to the aforementioned application function implementation method embodiments, this application also provides a lane line recognition device, electronic device, and corresponding embodiments based on roadside equipment.

[0103] Figure 4 This is a schematic diagram of the structure of a lane line recognition device based on roadside equipment, as shown in an embodiment of this application.

[0104] See Figure 4 The lane line recognition device based on roadside equipment according to an embodiment of this application mainly includes an acquisition module 401, a recognition module 402, a fusion module 403 and a calculation module 404.

[0105] The acquisition module 401 is used to acquire multiple first images taken by the roadside equipment at preset time intervals within a preset time period;

[0106] The recognition module 402 is used to perform lane line recognition on multiple first images to obtain multiple second images;

[0107] The fusion module 403 is used to fuse multiple second images to generate a third image;

[0108] The calculation module 404 is used to calculate the pixel data of the lane lines in the third image to obtain the calculated data of the lane lines in the third image.

[0109] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0110] From the above Figure 4 As can be seen from the example of the lane line recognition device based on roadside equipment, the lane line recognition device of this application first obtains multiple second images by performing lane line recognition on multiple first images that do not have complete lane lines captured by the roadside equipment, then fuses the multiple second images into a third image with complete lane lines, and finally processes the third image to obtain the processed data. This can eliminate the disturbance to the quality of the images captured by the roadside equipment due to factors such as weather conditions and changes in road traffic flow, and improve the completeness and accuracy of lane line recognition.

[0111] Figure 5 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application.

[0112] See Figure 5 The electronic device 500 includes a memory 510 and a processor 520.

[0113] The processor 520 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0114] Memory 510 may include various types of storage units, such as system memory, read-only memory (ROM), and permanent storage devices. ROM may store static data or instructions required by the processor 520 or other modules of the computer. Permanent storage devices may be read-write storage devices. Permanent storage devices may be non-volatile storage devices that retain stored instructions and data even when the computer is powered off. In some embodiments, permanent storage devices use mass storage devices (e.g., magnetic or optical disks, flash memory) as permanent storage devices. In other embodiments, permanent storage devices may be removable storage devices (e.g., floppy disks, optical drives). System memory may be a read-write storage device or a volatile read-write storage device, such as dynamic random access memory. System memory may store some or all of the instructions and data required by the processor during operation. Furthermore, memory 510 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (e.g., DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), and disks and / or optical disks may also be used. In some embodiments, memory 510 may include a removable storage device that is readable and / or writable, such as a laser disc (CD), a read-only digital multifunction optical disc (e.g., DVD-ROM, dual-layer DVD-ROM), a read-only Blu-ray disc, an ultra-high-density optical disc, a flash memory card (e.g., SD card, mini SD card, Micro-SD card, etc.), a magnetic floppy disk, etc. Computer-readable storage media do not contain carrier waves or transient electronic signals transmitted wirelessly or via wired connections.

[0115] The memory 510 stores executable code, which, when processed by the processor 520, can cause the processor 520 to execute part or all of the methods described above.

[0116] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0117] Alternatively, this application may be implemented as a computer-readable storage medium (or a non-transitory machine-readable storage medium or a machine-readable storage medium) storing executable code (or computer program or computer instruction code) thereon, which, when executed by a processor of an electronic device (or server, etc.), causes the processor to perform part or all of the steps of the methods described above according to this application.

[0118] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A roadside device-based lane line recognition method, characterized by, The method comprises: acquiring a plurality of first images taken by a roadside device at preset time intervals within a preset time period; performing lane line identification on the plurality of first images to obtain a plurality of second images; the second images are lane line binary images; fusing the plurality of second images to generate a third image; which comprises: selecting a target image with the most pixel data of lane lines from the plurality of second images; retaining pixel data of lane line segments in the remaining second images that have less than a preset threshold degree of coincidence with the target image; supplementing the retained pixel data of lane line segments to the target image to generate a third image; solving the pixel data of lane lines in the third image to obtain the solving data of lane lines in the third image; which comprises: extracting pixel data of a target lane line within a preset trusted area in the third image; combining the coordinates and external parameters of the roadside device to solve the pixel data of the target lane line to obtain a plurality of feature points to be fitted of the target lane line; clustering the plurality of feature points to be fitted on the same target lane line into a group; fitting the plurality of feature points to be fitted in the same group to obtain the solving data of the target lane line. 2.The roadside device based lane line identification method of claim 1, wherein, The method further comprises: if there is no existing lane line data within a preset road range, confirming the solving data of the current period to make a high-precision map; if there is existing lane line data within a preset road range, judging whether the lane lines within the preset road range have changed based on the existing lane line data or the solving data of the current period. 3.The roadside device based lane line identification method of claim 2, wherein, The judgment of whether the lane lines within the preset road range have changed comprises: based on the existing lane line data, finding a plurality of solving data of the current period within the neighborhood of the existing lane line data; if there is solving data that completely matches the existing lane line data in the plurality of solving data of the current period, it is judged that the lane lines within the preset road range have not changed; if there is no solving data that completely matches the existing lane line data in the plurality of solving data of the current period, it is judged that the lane lines within the preset road range have changed. 4.The roadside device based lane line identification method of claim 3, wherein, if there is no solving data of the current period within the neighborhood of the existing lane line data, it is judged that the lane lines within the preset road range are in a first to-be-confirmed situation. 5.The roadside device based lane line identification method of claim 2, wherein, The judgment of whether the lane lines within the preset road range have changed comprises: based on the solving data of the current period, finding a plurality of existing lane line data within the neighborhood of the solving data of the current period; if there is existing lane line data that completely matches the solving data of the current period in the plurality of existing lane line data, it is judged that the lane lines within the preset road range have not changed; if there is no existing lane line data that completely matches the solving data of the current period in the plurality of existing lane line data, it is judged that the lane lines within the preset road range have changed. 6.The roadside device based lane line identification method of claim 5, wherein, If the existing lane line data does not exist in the calculation data neighborhood of the current cycle, it is determined that the lane line in the preset road range appears a second to-be-confirmed situation. 7.A lane line recognition apparatus based on a roadside device, characterized by, Comprise: An acquisition module is configured to acquire a plurality of first images captured by a roadside device every interval of a preset time within a preset time period; An identification module is configured to perform lane line identification on the plurality of first images to obtain a plurality of second images; The second image is a lane line binary image; A fusion module is configured to fuse the plurality of second images to generate a third image; It comprises: Filtering a target image with the most pixel data of lane lines from the plurality of second images; Reserving pixel data of lane line segments in the remaining second images that have a coincidence degree less than a preset threshold with the target image; Supplementing the pixel data of the reserved lane line segments to the target image to generate a third image; A calculation module is configured to calculate pixel data of lane lines in the third image to obtain calculation data of lane lines in the third image; it comprises: Extracting pixel data of a target lane line in a preset trusted area in the third image; Combining coordinates and external parameters of the roadside device to calculate the pixel data of the target lane line to obtain a plurality of to-be-fitted feature points of the target lane line; Clustering the plurality of to-be-fitted feature points on the same target lane line into a group; Fitting the plurality of to-be-fitted feature points in the same group to obtain the calculation data of the target lane line.

8. An electronic device, comprising: Comprise: A processor; And A memory having executable code stored thereon, when the executable code is executed by the processor, the processor executes the method of any one of claims 1-6.

9. A computer-readable storage medium having executable code stored thereon, when the executable code is executed by a processor of an electronic device, the processor executes the method of any one of claims 1-6.

Images