Methods, apparatus, equipment and media for detecting lane line separation and merging points

Abstract

This application relates to a method, apparatus, device, and medium for detecting lane line separation and merging points. The method includes: after acquiring a lane line image, performing skeletonization processing on the lane lines in the lane line image to obtain a lane line skeleton diagram; identifying all branch nodes on the lane line skeleton diagram, and determining pseudo branch nodes among all branch nodes; and eliminating pseudo branch lane lines corresponding to the pseudo branch nodes in the lane line skeleton diagram to obtain the real branch nodes in the lane line image, i.e., obtaining the lane line separation and merging points. The solution provided by this application can accurately detect lane line separation and merging points.

Description

Technical Field

[0001] This application relates to the technical field of lane line image processing, and in particular to a method, apparatus, device and medium for detecting lane line separation and merging points. Background Technology

[0002] In autonomous driving technology, lane line data processing is a crucial component. Generally, in automated mapping processes, video of lanes on the road is first captured using camera equipment. Then, the actual coordinates of the lane lines are calculated by combining vehicle position information with lane pixel information from the video. This calculated information is then grouped or clustered into different lane lines. However, calculating actual coordinates from lane pixel information can lead to a mixture of data for different lane lines (e.g., main lane lines and their branch lane lines), resulting in inaccurate main lane line and branch lane line data. Summary of the Invention

[0003] To address or partially address the problems existing in related technologies, this application provides a method, apparatus, device, and medium for detecting lane line separation and merging points, which can accurately detect lane line separation and merging points to separate main lane line data and branch lane line data.

[0004] A first aspect of this application provides a method for detecting lane line separation and merging points, the method comprising:

[0005] After acquiring the lane line image, the lane lines on the lane line image are skeletonized to obtain the lane line skeleton map.

[0006] Identify all tree branch nodes on the lane line skeleton diagram, and determine pseudo tree branch nodes among all the tree branch nodes;

[0007] In the lane line skeleton diagram, the pseudo branch lane lines corresponding to the pseudo branch nodes are eliminated to obtain the real branch nodes in the lane line image, that is, to obtain the separation and merging points of the lane lines.

[0008] In one embodiment, after acquiring the lane line image and before skeletonizing the lane lines on the lane line image, the method further includes:

[0009] The lane lines in the lane line image are subjected to edge smoothing and / or hole filling.

[0010] In one embodiment, identifying pseudo-branch nodes among all said branch nodes includes:

[0011] If the number of pixels of the branch lane line corresponding to the branch node is greater than the preset number of pixels, and the hole filling of the branch lane line is abnormal, then the branch node is a pseudo branch node.

[0012] In one embodiment, identifying pseudo-branch nodes among all said branch nodes includes:

[0013] If the number of pixels of the branch lane line corresponding to the branch node is less than the preset number of pixels, then the branch node is a pseudo branch node.

[0014] In one embodiment, eliminating the pseudo-branch lane lines corresponding to the pseudo-tree branch nodes includes:

[0015] After determining the main lane line and pseudo branch lane line corresponding to the pseudo branch node based on the lane line skeleton diagram and the pseudo branch node, the pixel value of the pseudo branch lane line is set to zero.

[0016] In one embodiment, acquiring the lane line image includes:

[0017] Lane line images are obtained using camera equipment.

[0018] In one embodiment, after obtaining the lane line separation and merging points, the method further includes:

[0019] Mark the separation and merging points, and break the branch lane lines corresponding to the marked separation and merging points.

[0020] A second aspect of this application provides an apparatus for detecting lane line separation and merging points, the apparatus comprising:

[0021] An image processing module is configured to, after acquiring a lane line image, perform skeletonization processing on the lane lines in the lane line image to obtain a lane line skeleton map.

[0022] An image recognition module is configured to identify all tree branch nodes on the lane line skeleton map and identify pseudo tree branch nodes among all the tree branch nodes.

[0023] An image elimination module is configured to eliminate the pseudo-branch lane lines corresponding to the pseudo-tree branch nodes in the lane line skeleton image to obtain the real tree branch nodes in the lane line image, that is, to obtain the separation and merging points of the lane lines.

[0024] A third aspect of this application provides an electronic device, comprising:

[0025] Processor; and

[0026] A memory that stores executable code, which, when executed by the processor, causes the processor to perform the method described above.

[0027] 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.

[0028] The technical solution provided in this application may include the following beneficial effects:

[0029] The technical solution of this application performs skeletonization processing on the lane line image to ensure that all lane lines have a consistent line width, facilitating the identification of branch nodes formed by different lane lines. However, pseudo-branch nodes are also generated during skeletonization processing. Therefore, after identifying all branch nodes, it is necessary to filter out the pseudo-branch nodes and eliminate the pseudo-branch lane lines corresponding to them. The remaining branch nodes are the real branch nodes formed by different lane lines, and these real branch nodes are the separation and merging points. The technical solution of this application can accurately detect the separation and merging points of lane lines, thereby facilitating the subsequent segmentation of main lane line data and branch lane line data.

[0030] 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

[0031] 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.

[0032] Figure 1 This is a flowchart illustrating a method for detecting lane line separation and merging points according to an embodiment of this application;

[0033] Figure 2 This is a lane line image shown in an embodiment of this application;

[0034] Figure 3 Is with Figure 2 The corresponding lane line skeleton diagram;

[0035] Figure 4 yes Figure 3 Lane skeleton diagram after eliminating pseudo-tree branch nodes;

[0036] Figure 5 This is a schematic diagram of the main lane lines and branch lane lines after being interrupted, as shown in an embodiment of this application.

[0037] Figure 6 This is a schematic diagram of a structure for detecting lane line separation and merging points, as shown in an embodiment of this application;

[0038] Figure 7This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation

[0039] 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.

[0040] 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.

[0041] 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.

[0042] In autonomous driving technology, lane line data processing is a crucial component. Generally, in automated mapping processes, video of lanes on the road is first captured using camera equipment. Then, the actual coordinates of the lane lines are calculated by combining vehicle position information with lane pixel information from the video. This calculated information is then grouped or clustered into different lane lines. However, calculating actual coordinates from lane pixel information can lead to a mixture of data for different lane lines (e.g., main lane lines and their branch lane lines), resulting in inaccurate main lane line and branch lane line data.

[0043] To address the aforementioned issues, this application provides a method for detecting lane line separation and merging points, which can accurately detect lane line separation and merging points, thereby facilitating subsequent segmentation of main lane line data and branch lane line data.

[0044] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0045] like Figure 1 As shown in the figure, this application provides a method for detecting lane line separation and merging points, mainly including steps S101 to S103, which are described below:

[0046] Step S101: After acquiring the lane line image, the lane lines on the lane line image are skeletonized to obtain the lane line skeleton map.

[0047] In this embodiment, the line widths of the acquired lane line images are not entirely consistent. If the connection points of different lane lines are directly identified, it is easy to encounter abnormal connection point identification, which does not meet the accuracy requirements of high-precision maps. In order to more accurately identify connection points, this embodiment performs skeletonization processing on the acquired lane line images to obtain lane line skeleton diagrams with consistent lane line widths.

[0048] Lane line images are acquired through camera equipment. Generally, lane line images on roads are obtained by vehicle-mounted cameras traveling along the road. Examples of such cameras include binocular cameras installed on vehicles. These cameras capture lane lines in real time to provide basic data for subsequent image processing. Specifically, the binocular camera captures lane video, and then combines this data with vehicle position information and lane pixel information from the video to calculate the actual coordinates of the lane lines. This image representing the actual coordinates of the lane lines is defined as the lane line image.

[0049] To ensure more accurate image results in subsequent skeletonization processing, the method further includes the following steps after acquiring the lane line image and before skeletonizing the lane lines on the lane line image:

[0050] Step S101: Perform edge smoothing and / or hole filling on the lane lines in the lane line image.

[0051] In this embodiment, the lane line image finally acquired by the binocular camera is preprocessed. This preprocessing operation can be to smooth the edges of the lane lines in the lane line image, ensuring that the lane line skeleton in the resulting lane line skeleton map is sufficiently smooth. The preprocessing operation can also be to fill holes in the lane lines in the lane line image, or to perform edge smoothing and hole filling on the lane lines in the lane line image. Filling holes in the lane line image avoids the appearance of many unrealistic connection points after skeletonization, which would reduce the detection quality of lane line separation and merging points; and it also requires excluding large non-hole closed regions during hole filling. Lane line separation and merging points are the connection points of two connected different lane lines. After performing the above preprocessing on the lane lines in the lane line image, the skeleton map is directly generated, thus obtaining the lane line skeleton map.

[0052] Step S102: Identify all branch nodes on the lane line skeleton map and determine the pseudo branch nodes among all branch nodes.

[0053] In this embodiment, the connection point is formed by connecting two different lane lines. For example, one lane line is the main lane line, and the other lane line is the opposite branch lane line. The connection point is shaped like a tree branch, and therefore, the connection point is defined as a branch node. Specifically, the branch nodes involved in this embodiment are all star-shaped with three branches. The main lane line is determined by two branches on one line, and the remaining branch is the opposite branch lane line.

[0054] Furthermore, identifying pseudo-branch nodes among all branch nodes can include:

[0055] If the number of pixels of the branch lane line corresponding to the branch node is less than the preset number of pixels, then the branch node is a pseudo branch node.

[0056] In this embodiment, the lane line skeleton diagram will generate some small burrs during the generation process, which are also known as noise. Specifically, when the number of pixels of the branch lane line A2 corresponding to the branch node A1 is small, less than the preset number of pixels, the branch lane line A2 is considered a small burr, that is, the branch node A1 is a pseudo branch node.

[0057] In this embodiment, the number of pixels can characterize the length of the branch lane line corresponding to the branch node. Based on this, identifying pseudo-branch nodes among all branch nodes may also include:

[0058] If the length of the branch lane line corresponding to the branch node is less than the preset length, then the branch node is a pseudo-branch node.

[0059] Furthermore, identifying pseudo-branch nodes among all branch nodes may also include:

[0060] If the number of pixels of the branch lane line corresponding to the branch node is greater than the preset number of pixels, and the hole filling of the branch lane line is abnormal, then the branch node is a pseudo branch node.

[0061] In this embodiment, when the number of pixels of the branch lane line B2 corresponding to the branch node B1 is relatively long (the number of pixels of the branch lane line B2 is greater than or equal to the preset number of pixels), it does not mean that the branch node B1 is necessarily a real branch node. In the preprocessing of the lane line image, there may be some identification errors in the filling of holes on the lane line, that is, abnormal hole filling. The branch lane line B2 related to this phenomenon can also be regarded as noise. Then the branch node B1 is a pseudo branch node.

[0062] Step S103: In the lane line skeleton map, the pseudo branch lane lines corresponding to the pseudo branch nodes are eliminated to obtain the real branch nodes in the lane line image, that is, to obtain the lane line separation and merging points.

[0063] In the embodiments of this application, after the so-called noise is eliminated, the tree branch nodes related to the noise naturally disappear. That is to say, the pseudo tree branch nodes are eliminated, and the remaining tree branch nodes are the real tree branch nodes of the lane lines, that is, the separation and merging points of the lane lines are obtained.

[0064] Furthermore, the pseudo-branch lane lines corresponding to the pseudo-branch nodes are eliminated, specifically including:

[0065] After determining the main lane line and pseudo branch lane line corresponding to the pseudo branch node based on the lane line skeleton map and pseudo branch node, the pixel value of the pseudo branch lane line is set to zero.

[0066] In this embodiment, a pseudo-branch node C1 is found. Based on the lane lines on the lane line skeleton map, the main lane line C2 and the pseudo-branch lane line C3 that form the pseudo-branch node C1 are determined, along with the driving directions of the main lane line C2 and the pseudo-branch lane line C3. Then, the pixel values ​​of the pseudo-branch lane line C3 are zeroed out according to the determined directions. After zeroing out the pixel values ​​of the pseudo-branch lane line C3, it no longer exists; that is, the pseudo-branch lane line is erased, and correspondingly, the pseudo-branch node C1 is eliminated. The pseudo-branch lane line exists as noise. This embodiment directly identifies and judges the noise, which reduces the interference of noise on the detection results of separation and merging points compared to the clustering method used in related technologies.

[0067] like Figures 2-4 As shown, the images are, in order, the lane line image, the lane line skeleton diagram, and the lane line skeleton diagram after removing false branch nodes. Figure 3 As shown, there are some small burrs on the skeletonized lane lines, i.e., the circled parts; such as Figure 4 As shown, small burrs are eliminated to obtain the lane line skeleton diagram after eliminating pseudo-branch nodes. The small burrs in the circled part are eliminated.

[0068] To verify this, separation and merging points on lane line images acquired on different roads were detected, as shown in Table 1. Table 1 shows the actual number of separation and merging points, the number of detected points, and the detection accuracy on the lane line images.

[0069] Table 1. Actual number of separation and merging points, detection number of detection points, and detection accuracy on lane line images.

[0070]

[0071] As shown in Table 1, the technical solution of this application can accurately detect the separation and merging points on the road.

[0072] In summary, the technical solution of this application performs skeletonization processing on the lane line image to make all lane lines have the same line width, which facilitates the identification of tree branch nodes formed by different lane lines. In the skeletonization process, pseudo tree branch nodes will also be generated. After identifying all tree branch nodes, the pseudo tree branch nodes are further filtered and the pseudo branch lane lines corresponding to the pseudo tree branch nodes are eliminated. Then the remaining tree branch nodes are the real tree branch nodes formed by different lane lines, and these real tree branch nodes are the separation and merging points.

[0073] like Figure 5 As shown, preferably, after obtaining the lane line separation and merging points, the method further includes: marking the separation and merging points, and breaking the branch lane lines corresponding to the marked separation and merging points, so as to perform data segmentation of the main lane lines and branch lane lines in the lane lines.

[0074] In this embodiment, after obtaining the lane line separation and merging points, in order to accurately segment two connected different lane line data in the later stage, the obtained separation and merging points are marked to facilitate the breaking of branch lane lines, thereby realizing the segmentation of lane line data according to the markings.

[0075] Corresponding to the aforementioned application function implementation method embodiments, this application also provides a device, electronic device, and corresponding embodiments for detecting lane line separation and merging points.

[0076] Figure 6 This is a schematic diagram illustrating the structure of a device for detecting lane line separation and merging points according to an embodiment of this application. The device includes:

[0077] The image processing module 601 is configured to perform skeletonization processing on the lane lines in the lane line image after acquiring the lane line image, so as to obtain a lane line skeleton map.

[0078] Image recognition module 602 is configured to recognize all tree branch nodes on the lane line skeleton map and identify pseudo tree branch nodes among all tree branch nodes.

[0079] The image elimination module 603 is configured to eliminate the pseudo-branch lane lines corresponding to the pseudo-tree branch nodes in the lane line skeleton map to obtain the real tree branch nodes in the lane line image, that is, to obtain the lane line separation and merging points.

[0080] The technical solution of this application uses an image processing module 601 to perform skeletonization processing on the lane line image, so that all lane lines have the same line width, which facilitates the image recognition module 602 to identify the branch nodes formed by different lane lines. In the skeletonization process, pseudo branch nodes will also be generated. After identifying all branch nodes, the pseudo branch nodes are filtered, and the pseudo branch lane lines corresponding to the pseudo branch nodes are eliminated by the image elimination module 603. Then the remaining branch nodes are the real branch nodes formed by different lane lines, and these real branch nodes are the separation and merging points.

[0081] Lane line images are acquired through camera equipment. Generally, lane line images on roads are obtained by vehicle-mounted cameras traveling along the road. Examples of such cameras include binocular cameras installed on vehicles. These cameras capture lane lines in real time to provide basic data for subsequent image processing. Specifically, the binocular camera captures lane video, and then combines this data with vehicle position information and lane pixel information from the video to calculate the actual coordinates of the lane lines. This image representing the actual coordinates of the lane lines is defined as the lane line image.

[0082] It should be noted that the line widths of the acquired lane line images are not entirely consistent. If the connection points of different lane lines are directly identified, it is easy to encounter abnormal connection point identification, which does not meet the accuracy requirements of high-precision maps. In order to more accurately identify connection points, this embodiment of the application performs skeletonization processing on the acquired lane line images to obtain lane line skeleton maps with consistent lane line widths.

[0083] To ensure more accurate image results in subsequent skeletonization processing, the device also includes:

[0084] The image preprocessing module is configured to perform edge smoothing and / or hole filling on the lane lines in the lane line image after acquiring the lane line image and before skeletonizing the lane lines in the lane line image.

[0085] In this embodiment, the lane line image finally acquired by the binocular camera is preprocessed. This preprocessing operation can be to smooth the edges of the lane lines in the lane line image, ensuring that the lane line skeleton in the resulting lane line skeleton map is sufficiently smooth. The preprocessing operation can also be to fill holes in the lane lines in the lane line image, or to perform edge smoothing and hole filling on the lane lines in the lane line image. Filling holes in the lane line image avoids the appearance of many unrealistic connection points after skeletonization, which would reduce the detection quality of lane line separation and merging points; and it also requires excluding large non-hole closed regions during hole filling. Lane line separation and merging points are the connection points of two connected different lane lines. After performing the above preprocessing on the lane lines in the lane line image, the skeleton map is directly generated, thus obtaining the lane line skeleton map.

[0086] Furthermore, the image recognition module 602 includes a pseudo-tree branch node determination unit.

[0087] The pseudo-branch node determination unit is configured to determine that if the number of pixels of the branch lane line corresponding to the branch node is lower than the preset number of pixels, the branch node is determined to be a pseudo-branch node.

[0088] In this embodiment, the connection point is formed by connecting two different lane lines. For example, one lane line is the main lane line, and the other lane line is a corresponding branch lane line. The connection point is shaped like a tree branch, and therefore, it is defined as a branch node. In this embodiment, the branch nodes are all star-shaped with three branches. Two branches on one line define the main lane line, and the remaining branch is the corresponding branch lane line. During the generation of the lane line skeleton diagram, some small burrs are generated, which are essentially noise. When the number of pixels in the branch lane line A2 corresponding to branch node A1 is small, less than a preset number of pixels, then branch lane line A2 is considered a burr; that is, branch node A1 is a pseudo-branch node.

[0089] In this embodiment, the number of pixels can characterize the length of the branch lane line corresponding to the branch node. Based on this, identifying pseudo-branch nodes among all branch nodes may also include:

[0090] If the length of the branch lane line corresponding to the branch node is less than the preset length, then the branch node is a pseudo-branch node.

[0091] The pseudo-branch node determination unit is further configured to determine that if the number of pixels of the branch lane line corresponding to the branch node is greater than a preset number of pixels, and the hole filling of the branch lane line is abnormal, then the branch node is determined to be a pseudo-branch node.

[0092] In this embodiment, when the number of pixels of the branch lane line B2 corresponding to the branch node B1 is relatively long (the number of pixels of the branch lane line B2 is greater than or equal to the preset number of pixels), it does not mean that the branch node B1 is necessarily a real branch node. In the preprocessing of the lane line image, there may be some identification errors in the filling of holes on the lane line, that is, abnormal hole filling. The branch lane line B2 related to this phenomenon can also be regarded as noise. Then the branch node B1 is a pseudo branch node.

[0093] Furthermore, the image removal module 603 includes a pseudo-branch lane line removal unit.

[0094] The pseudo-branch lane line elimination unit is configured to set the pixel value of the pseudo-branch lane line to zero after determining the main lane line and pseudo-branch lane line corresponding to the pseudo-branch node based on the lane line skeleton map and the pseudo-branch node.

[0095] In this embodiment, after eliminating the so-called noise, the tree branch nodes related to the noise naturally disappear, meaning the pseudo tree branch nodes are eliminated, and the remaining tree branch nodes are the true tree branch nodes of the lane lines, thus obtaining the lane line separation and merging points. A pseudo tree branch node C1 is found. Based on the lane lines on the lane line skeleton diagram, the main lane line C2 and the pseudo branch lane line C3 that form the pseudo tree branch node C1 are determined, along with the driving directions of the main lane line C2 and the pseudo branch lane line C3. Then, the pixel values ​​of the pseudo branch lane line C3 are zeroed out according to the determined directions. After zeroing out the pixel values ​​of the pseudo branch lane line C3, the pseudo branch lane line C3 no longer exists, and correspondingly, the pseudo tree branch node C1 is eliminated. Here, the pseudo branch lane line exists as noise. This embodiment directly identifies and judges the noise, which reduces the interference of noise on the separation and merging point detection results compared to the clustering method used in related technologies.

[0096] Preferably, the device further includes a separation and merging point marking module.

[0097] The separation and merging point marking module is configured to mark the separation and merging points after obtaining the lane line separation and merging points, and break the branch lane lines corresponding to the marked separation and merging points to perform data segmentation of the main lane lines and branch lane lines in the lane lines.

[0098] In this embodiment, after obtaining the lane line separation and merging points, in order to accurately segment two connected different lane line data in the later stage, the obtained separation and merging points are marked to facilitate the breaking of branch lane lines, thereby realizing the segmentation of lane line data according to the markings.

[0099] Figure 7 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application.

[0100] See Figure 7 The electronic device 700 includes a memory 701 and a processor 702.

[0101] The processor 702 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.

[0102] Memory 701 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 processor 702 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 701 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 701 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, a 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.

[0103] The memory 701 stores executable code, which, when processed by the processor 702, can cause the processor 702 to execute part or all of the methods described above.

[0104] 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.

[0105] 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.

[0106] 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 method for detecting lane line separation and merging points, characterized in that, The method includes: After acquiring the lane line image, the lane lines on the lane line image are skeletonized to obtain the lane line skeleton map. Identify all branch nodes on the lane line skeleton map, and determine pseudo branch nodes among all branch nodes, including: determining the pseudo branch nodes based on the number of pixels of the branch lane lines, wherein the branch nodes are used to indicate the main lane line determined by two branches connected to the branch node and located on the same straight line, and the branch lane lines determined by the remaining branches connected to the branch node, wherein the lane lines on the lane line image are subjected to edge smoothing and / or hole filling, and if the number of pixels of the branch lane line corresponding to the branch node is greater than a preset number of pixels, and the hole filling of the branch lane line is abnormal, then the branch node is a pseudo branch node; In the lane line skeleton diagram, the pseudo branch lane lines corresponding to the pseudo branch nodes are eliminated to obtain the real branch nodes in the lane line image, that is, to obtain the separation and merging points of the lane lines.

2. The method for detecting lane line separation and merging points as described in claim 1, characterized in that, Identifying pseudo-branch nodes among all the stated branch nodes includes: If the number of pixels of the branch lane line corresponding to the branch node is less than the preset number of pixels, then the branch node is a pseudo branch node.

3. The method for detecting lane line separation and merging points as described in claim 1 or 2, characterized in that, Eliminating the pseudo-branch lane lines corresponding to the pseudo-tree branch nodes includes: After determining the main lane line and pseudo branch lane line corresponding to the pseudo branch node based on the lane line skeleton diagram and the pseudo branch node, the pixel value of the pseudo branch lane line is set to zero.

4. The method for detecting lane line separation and merging points as described in claim 1, characterized in that, The acquisition of the lane line image includes: Lane line images are obtained using camera equipment.

5. The method for detecting lane line separation and merging points as described in claim 1, characterized in that, After obtaining the lane line separation and merging points, the method further includes: Mark the separation and merging points, and break the branch lane lines corresponding to the marked separation and merging points.

6. A device for detecting lane line separation and merging points, characterized in that, The device includes: An image processing module is configured to, after acquiring a lane line image, perform skeletonization processing on the lane lines in the lane line image to obtain a lane line skeleton map. An image recognition module is configured to recognize all branch nodes on the lane line skeleton map and identify pseudo branch nodes among all branch nodes, including: determining the pseudo branch node based on the number of pixels of the branch lane line, wherein the branch node is used to indicate the main lane line determined by two branches connected to the branch node and located on the same straight line, and the branch lane line determined by the remaining branches connected to the branch node, wherein the lane lines on the lane line image are subjected to edge smoothing and / or hole filling, and if the number of pixels of the branch lane line corresponding to the branch node is greater than a preset number of pixels and the hole filling of the branch lane line is abnormal, then the branch node is a pseudo branch node; An image elimination module is configured to eliminate the pseudo-branch lane lines corresponding to the pseudo-tree branch nodes in the lane line skeleton image to obtain the real tree branch nodes in the lane line image, that is, to obtain the separation and merging points of the lane lines.

7. An electronic device, characterized in that, include: processor; as well as A memory having executable code stored thereon, which, when executed by the processor, causes the processor to perform the method as described in any one of claims 1 to 5.

8. 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 as described in any one of claims 1 to 5.

Images