Lane line extraction method, device and electronic device

Abstract

The present disclosure provides a method, device and electronic device for extracting lane lines, and relates to the fields of artificial intelligence technology such as high-precision maps, autonomous driving, vehicle networking, and smart cockpits. The specific scheme is: when performing lane line extraction, obtain multiple first lane line points corresponding to the target lane line to be extracted; cluster the multiple first lane line points according to the positions of the multiple first lane line points; cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the obtained first clustering results, and the second lane line points belong to the multiple first lane line points; and then extract the target lane line according to the multiple second clustering results obtained by the re-clustering. In this way, by re-clustering the first clustering results, the lane line points corresponding to the divergent lane lines can be effectively clustered into different second clustering results, so that the divergent lane lines can be accurately extracted according to the multiple second clustering results, thereby improving the accuracy of the extraction results.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to the field of artificial intelligence technologies such as high-precision maps, autonomous driving, vehicle networking, and smart cockpits, and specifically to a method, device, and electronic device for extracting lane lines. Background Art

[0002] Lane lines are an important component of high-precision maps and an essential element in autonomous driving scenarios. Usually, in autonomous driving scenarios, vehicles need to first obtain road images of the road they are on, extract lane lines from the road images, and then drive safely based on the extracted lane lines.

[0003] Therefore, how to accurately extract lane lines to improve the accuracy of extraction results is an urgent problem to be solved by those skilled in the art. Summary of the invention

[0004] The present disclosure provides a lane line extraction method, device and electronic device, which can accurately extract lane lines and improve the accuracy of extraction results.

[0005] According to a first aspect of the present disclosure, a method for extracting lane lines is provided, and the method for extracting lane lines may include:

[0006] Acquire multiple first lane line points corresponding to the target lane line to be extracted.

[0007] According to the positions of the multiple first lane line points, the multiple first lane line points are clustered to obtain multiple first clustering results; wherein the first clustering results include multiple second lane line points, and the second lane line points belong to the multiple first lane line points.

[0008] The plurality of second lane line points are clustered again according to the positions of the plurality of second lane line points included in the first clustering result to obtain a plurality of second clustering results.

[0009] The target lane line is extracted according to the multiple second clustering results.

[0010] According to a second aspect of the present disclosure, a lane line extraction device is provided, and the lane line extraction device may include:

[0011] The acquisition unit is used to acquire multiple first lane line points corresponding to the target lane line to be extracted.

[0012] The first clustering unit is used to cluster the multiple first lane line points according to their positions to obtain multiple first clustering results; wherein the first clustering results include multiple second lane line points, and the second lane line points belong to the multiple first lane line points.

[0013] The second clustering unit is used to cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the first clustering result to obtain multiple second clustering results.

[0014] An extraction unit is used to extract the target lane line according to the multiple second clustering results.

[0015] According to a third aspect of the present disclosure, an electronic device is provided, which may include:

[0016] at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the lane line extraction method described in the first aspect above.

[0019] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the lane line extraction method described in the first aspect above.

[0020] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising: a computer program, wherein the computer program is stored in a readable storage medium, and at least one processor of an electronic device can read the computer program from the readable storage medium, and the at least one processor executes the computer program so that the electronic device executes the lane line extraction method described in the first aspect.

[0021] According to the technical solution disclosed in the present invention, lane lines can be accurately extracted, thereby improving the accuracy of the extraction results.

[0022] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0024] Figure 1 is a schematic diagram of lane line extraction based on existing clustering results provided by an embodiment of the present disclosure;

[0025] Figure 2 is a schematic diagram of an application framework provided by an embodiment of the present disclosure;

[0026] Figure 3 is a flowchart of a lane line extraction method provided according to the first embodiment of the present disclosure;

[0027] Figure 4 is a schematic diagram of multiple second clustering results obtained by performing clustering twice provided by an embodiment of the present disclosure;

[0028] Figure 5 is a flow chart of a method for clustering multiple lane line points provided according to the second embodiment of the present disclosure;

[0029] Figure 6 is a schematic diagram of the distribution of three first clustering results satisfying the concentric circle form provided by an embodiment of the present disclosure;

[0030] Figure 7 is a schematic diagram of a clustering result of re-clustering a first clustering result provided by an embodiment of the present disclosure;

[0031] Figure 8 is a flowchart of a method for extracting a target lane line according to a plurality of second clustering results provided in accordance with a third embodiment of the present disclosure;

[0032] Fig. 9 is a schematic diagram of a connection situation provided by an embodiment of the present disclosure;

[0033] Fig.10 is a schematic diagram of an extracted target lane line provided by an embodiment of the present disclosure;

[0034] Fig.11 is a structural schematic diagram of a lane line extraction device provided according to a fourth embodiment of the present disclosure;

[0035] Fig.12 It is a schematic block diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0036] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0037] In the embodiments of the present disclosure, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the access relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist at the same time, and B exists alone, where A and B may be singular or plural. In the text description of the present disclosure, the character " / " generally indicates that the previous and next associated objects are in an "or" relationship. In addition, in the embodiments of the present disclosure, "first", "second", "third", "fourth", "fifth" and "sixth" are only used to distinguish the contents of different objects and have no other special meanings.

[0038] The technical solution provided by the embodiments of the present disclosure can be applied to artificial intelligence technology fields such as high-precision maps, autonomous driving, vehicle networking, and smart cockpits. Lane lines, as an important component of maps, are essential basic elements in autonomous driving scenarios. Accurately extracting lane lines from road images is crucial to the safety of autonomous driving vehicles.

[0039] In actual road scenarios, there is often a situation where a lane line is divided into two lane lines. For an example, see Figure 1 As shown, Figure 1 is a schematic diagram of lane line extraction based on existing clustering results provided by an embodiment of the present disclosure, wherein: Figure 1 The black solid dots in the figure represent multiple lane line points of the extracted lane line, and the dotted line represents the lane line obtained by clustering and fitting the lane line points based on the existing algorithm. When extracting lane lines, the multiple lane line points corresponding to the lane lines will be clustered first. During the clustering process, since such lane lines are connected, the clustering method will cluster the divergent lane line points into the same clustering result to obtain three clustering results. Among them, one clustering result includes the five closest lane line points in the upper left corner and the five closest lane line points in the lower left corner; one clustering result includes the five closest lane line points in the middle of the upper part of the dotted line and the five closest lane line points in the middle of the lower part of the dotted line; one clustering result includes the five closest lane line points on the right; and then the lane line points in these three clustering results are fitted to obtain a final lane line, that is, Figure 1 The dashed lane lines are shown.

[0040] Obviously, the lane line extracted by this method does not match the two lane lines separated in the actual road scene. Therefore, using the existing method to extract lane lines will lead to poor accuracy of the extraction results.

[0041] In order to improve the accuracy of the extraction results, in the embodiments of the present disclosure, it can be considered to cluster the lane lines to be extracted to obtain multiple clustering results, and then continue to cluster the multiple clustering results again. In this way, the lane line points corresponding to the divergent lane lines can be clustered into different clustering results through re-clustering, and then the lane lines can be extracted according to the clustering results obtained by the re-clustering. The divergent lane lines can be accurately extracted, thereby effectively improving the accuracy of the extraction results.

[0042] For examples, see Figure 2 As shown, Figure 2 This is a schematic diagram of an application framework provided by an embodiment of the present disclosure. The electronic device used to implement the technical solution can be a vehicle-mounted device or other electronic device, and can be specifically configured according to actual needs. Taking a vehicle-mounted device as an example, the vehicle-mounted device can include three modules, namely a concentric circle assignment module, a clustering module, and a lane line extraction module.

[0043] The concentric circle assignment module is mainly used to obtain multiple lane line points corresponding to the target lane line to be extracted, which can be recorded as the first lane line point for distinction. The lane line point refers to a point used to describe the lane line.

[0044] The clustering module is mainly used to cluster multiple first lane line points according to their positions to obtain multiple first clustering results; and to cluster multiple second lane line points included in the first clustering results again to obtain multiple second clustering results.

[0045] The lane line extraction module is mainly used to extract the target lane line according to the multiple second clustering results obtained by clustering again.

[0046] As described above, after clustering the multiple first lane line points to be extracted and obtaining multiple first clustering results, the multiple first clustering results are clustered again. In this way, the lane line points corresponding to the divergent lane lines can be clustered into different second clustering results through re-clustering, and then lane line extraction is performed based on the second clustering results obtained by re-clustering. The divergent lane lines can be accurately extracted, thereby effectively improving the accuracy of the extraction results.

[0047] The lane line extraction method provided by the present disclosure will be described in detail below through specific embodiments. It is understandable that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

[0048] Embodiment 1

[0049] Figure 3 is a flow chart of a lane line extraction method provided according to the first embodiment of the present disclosure. The lane line extraction method can be executed by software and / or hardware devices. For example, the hardware device can be a terminal or a server. For example, see Figure 3 As shown, the lane line extraction method may include:

[0050] S301. Acquire multiple first lane line points corresponding to the target lane line to be extracted.

[0051] For example, when obtaining multiple first lane line points corresponding to a target lane line, the multiple first lane line points corresponding to the target lane line can be directly extracted from a road image including the target lane line; the multiple first lane line points corresponding to the target lane line can also be searched and obtained from local storage; or the multiple first lane line points corresponding to the target lane line sent by other electronic devices can be received, etc. The specific settings can be made according to actual needs. Here, the embodiments of the present disclosure only take these three acquisition methods as examples for illustration, but it does not mean that the embodiments of the present disclosure are limited to this.

[0052] After obtaining a plurality of first lane line points, in order to accurately extract the target lane line based on the obtained plurality of first lane line points, the plurality of first lane line points may be clustered according to their positions to obtain a plurality of first clustering results, and the plurality of first clustering results may be clustered again, so that the lane line points corresponding to the divergent lane lines may be clustered into different second clustering results, that is, the following S302 is executed:

[0053] S302: Cluster the plurality of first lane line points according to their positions to obtain a plurality of first clustering results, wherein the first clustering results include a plurality of second lane line points, and the second lane line points belong to the plurality of first lane line points.

[0054] For example, according to the positions of the multiple first lane line points, the multiple first lane line points are clustered, usually based on a clustering algorithm, and the multiple first lane line points with a closer distance are divided into one clustering result, thereby obtaining multiple first clustering results. Among them, the multiple second lane line points included in the first clustering result are obtained by clustering the multiple first lane line points, belong to the multiple first lane line points, and each first lane line point can only be clustered into one first clustering result.

[0055] For example, in the embodiments of the present disclosure, the clustering algorithm used may be some existing clustering algorithms, such as the K-Means clustering algorithm, or a density-based clustering method, etc., which may be specifically set according to actual needs.

[0056] S303: Cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the first clustering result to obtain multiple second clustering results.

[0057] For example, when the multiple second lane line points are clustered again according to the positions of the multiple second lane line points included in the first clustering result, it is usually based on a clustering algorithm to divide the multiple second lane line points included in the first clustering result that are closer to each other into one clustering result, thereby obtaining multiple second clustering results.

[0058] The second clustering result includes multiple third lane line points, and the third lane line point belongs to multiple second lane line points, and each third lane line point can only be clustered into one second clustering result.

[0059] For example, the above Figure 1 The three clustering results shown in FIG. 1 can be recorded as three first clustering results. In the embodiment of the present disclosure, in order to accurately cluster the lane line points corresponding to the divergent lane lines into different clustering results, the multiple second lane line points included in the three first clustering results can be clustered again according to their positions to obtain five second clustering results. For example, see Figure 4 As shown, Figure 4 It is a schematic diagram of multiple second clustering results obtained by performing two clusterings provided by an embodiment of the present disclosure, wherein, among the five second clustering results, one second clustering result includes five lane line points that are closer to the upper left corner, one second clustering result includes five lane line points that are closer to the lower left corner, one second clustering result includes five lane line points that are closer to the middle of the upper part of the dotted line, and one second clustering result includes five lane line points that are closer to the middle of the lower part of the dotted line; one second clustering result includes five lane line points that are closer to the right. It can be seen that, through multiple second clustering results, the lane line points of divergent lane lines are clustered into different clustering results.

[0060] After obtaining multiple second clustering results, the target lane line can be extracted according to the multiple second clustering results, that is, the following S304 is executed:

[0061] S304: Extract target lane lines according to the multiple second clustering results.

[0062] It can be seen that in the disclosed embodiment, when performing lane line extraction, first obtain multiple first lane line points corresponding to the target lane line to be extracted; and cluster the multiple first lane line points according to the positions of the multiple first lane line points to obtain multiple first clustering results; then cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the first clustering results, and the second lane line points belong to the multiple first lane line points; and then extract the target lane line according to the multiple second clustering results obtained by the re-clustering. In this way, by re-clustering the first clustering results, the lane line points corresponding to the divergent lane lines can be effectively clustered into different second clustering results, so that the divergent lane lines can be accurately extracted according to the multiple second clustering results, thereby effectively improving the accuracy of the extraction results.

[0063] Based on the above Figure 3 In order to facilitate understanding of how to cluster the plurality of first lane line points to obtain the plurality of first clustering results in the above S302, and cluster the plurality of second lane line points included in the first clustering results again to obtain the plurality of second clustering results, the following will be described. Figure 5 The second embodiment shown is described in detail.

[0064] Embodiment 2

[0065] Figure 5 is a flow chart of a method for clustering multiple lane line points according to the second embodiment of the present disclosure, which method can also be executed by software and / or hardware devices. Figure 5 As shown, the method may include:

[0066] S501. Determine a center point from the plurality of first lane line points according to positions of the plurality of first lane line points.

[0067] The center point is determined based on the positions of multiple first lane line points.

[0068] For example, when determining the center point from multiple first lane line points based on the positions of multiple first lane line points, the second average position corresponding to the multiple first lane line points can be first determined based on the positions of the multiple first lane line points; and the first lane line point among the multiple first lane line points that is closest to the second average position is determined as the center point.

[0069] It can be understood that if there are multiple first lane line points that are closest to the second average position, any one of the multiple closest first lane line points can be determined as the center point.

[0070] For example, assuming that the number of the multiple first lane line points is 3, the 3 first lane line points include a first lane line point a, b and c respectively; wherein the position coordinates of the first lane line point a are (x1, y1), the position coordinates of the first lane line point b are (x2, y2), and the position coordinates of the first lane line point c are (x3, y3), then the coordinates of the second average position corresponding to the 3 first lane line points are ((x1+x2+x3) / 3, (y1+y2+y3) / 3), then the distance between the first lane line point a and the second average position, the distance between the first lane line point b and the second average position, and the distance between the first lane line point c and the second average position are determined respectively, and the first lane line point closest to the second average position is determined as the center point, so that the center point can be determined from the 3 first lane line points.

[0071] S502: Cluster the multiple first lane line points according to the positions and center points of the multiple first lane line points to obtain multiple first clustering results.

[0072] For example, when clustering multiple first lane line points according to their positions and center points, the distance between the position of each of the multiple first lane line points and the center point can be determined to cluster the multiple first lane line points. For example, a distance threshold s can be preset, and first lane line points with a distance less than one s can be clustered into one cluster, first lane line points with a distance equal to or equal to one s and less than 2 s can be clustered into one cluster, and first lane line points with a distance equal to or equal to 2 s and less than 3 s can be clustered into one cluster, and so on. Multiple first lane line points can be clustered according to the distance threshold s to obtain multiple first clustering results. The value of the distance threshold s can be set to 1 meter, 0.9 meters, 1.1 meters, etc., and can be set according to actual needs.

[0073] Combined with the above description, after clustering multiple first lane line points by distance threshold s, the distribution of multiple first clustering results satisfies the form of concentric circles, and the center of the concentric circle is the center point. Figure 1 The lane line point shown can be recorded as the first lane line point, and the distance threshold s is used to Figure 1 After clustering the multiple first lane line points shown in FIG. 1 , the distribution of the first clustering results satisfies the form of concentric circles. For an example, see Figure 6 As shown, Figure 6This is a schematic diagram of the distribution of three first clustering results that satisfy the concentric circle form provided by an embodiment of the present disclosure. It can be seen that among the three first clustering results, the first clustering result with the closest distance is in the central circle area of ​​the concentric circles, the first clustering result with the second closest distance is in the annular area formed by the central circle and its adjacent circles, and the first clustering result with the farthest distance is in the annular area formed by the second circle and the third circle from the inside to the outside.

[0074] In order to accurately extract the divergent lane lines, on the basis of obtaining multiple first clustering results that satisfy the concentric circle form, the multiple first clustering results can be further clustered again. In this way, through re-clustering, the lane line points corresponding to the divergent lane lines can be clustered into different second clustering results, and then lane line extraction is performed based on the second clustering results obtained by re-clustering, so that the divergent lane lines can be accurately extracted.

[0075] For example, before re-clustering multiple first clustering results, it can be determined whether the number of second lane line points included in the first clustering result in the central circle area of ​​the concentric circles is greater than or equal to a preset number threshold; if the number of second lane line points included in the first clustering result in the central circle area is less than the number threshold, it means that there is an error in the selection of the center point, and the center point can be re-determined, and the multiple second lane line points corresponding to the target lane line to be extracted can be clustered again to obtain a new first clustering result; on the contrary, if the number of second lane line points included in the first clustering result in the central circle area is greater than or equal to the number threshold, the multiple first clustering results can be directly clustered again.

[0076] For example, when re-clustering multiple first clustering results, in order to reduce the amount of data that needs to be re-clustered, considering that the distance of the second lane line points included in the first clustering result in the central circle area is relatively close, it is not necessary to re-cluster the second lane line points included in the first clustering result in the central circle area; and only the second lane line points included in the first clustering result in the annular area formed by two adjacent concentric circles are re-clustered, which can not only achieve re-clustering, but also reduce the amount of data for re-clustering. It can be understood that if the reduction of the amount of data that needs to be re-clustered is not considered, the second lane line points included in the first clustering result in the central circle area can also be re-clustered, and the second clustering result obtained by re-clustering is still itself, that is, the same as its first clustering result.

[0077] For example, in the disclosed embodiment, taking the case of re-clustering only the second lane line points included in the first clustering result in the annular area formed by two adjacent concentric circles, when re-clustering them, it is necessary to first determine the first clustering result in the annular area formed by two adjacent concentric circles from multiple first clustering results based on the positions of the multiple second lane line points included in the first clustering result, that is, execute the following S503.

[0078] S503: Determine, from the plurality of first clustering results, a third clustering result located in an annular region formed by two adjacent concentric circles according to the positions of the plurality of second lane line points included in the first clustering results.

[0079] Among them, the distances between the multiple second lane line points included in the third clustering result and the center point are greater than the radius of the center circle of the concentric circles.

[0080] It can be understood that in order to distinguish the first clustering result in the annular area formed by two adjacent concentric circles from the multiple first clustering results obtained by the first clustering of the lane line points corresponding to the target lane line, in the embodiment of the present disclosure, the first clustering result in the annular area formed by two adjacent concentric circles can be recorded as the third clustering result, which is a subset of the multiple first clustering results.

[0081] For example, when determining a third clustering result in a ring area formed by two adjacent concentric circles from multiple first clustering results, the first clustering result closest to the center point can be first determined from the multiple first clustering results, and the other first clustering results among the multiple first clustering results except the first clustering result closest to the center point can be determined as the third clustering results, thereby determining the third clustering result.

[0082] For example, when determining the distance between the first clustering result and the center point, at least two possible implementation methods may be included:

[0083] In a possible implementation method, the distance between each second lane line point in the first clustering result and the center point can be calculated based on the position of each second lane line point in the first clustering result to obtain multiple distances; then the average of the multiple distances is calculated, and the average is determined as the distance between the first clustering result and the center point.

[0084] In another possible implementation, the average position corresponding to the first clustering result can be calculated based on the position of each second lane line point in the first clustering result, and the distance between the average position and the center point can be determined as the distance between the first clustering result and the center point.

[0085] It can be understood that in this possible implementation method, the method of calculating the average position corresponding to the first clustering result based on the position of each second lane line point in the first clustering result is similar to the method of determining the second average position mentioned above. Please refer to the relevant description of determining the second average position above, and the embodiments of the present disclosure will not be repeated here.

[0086] After determining the third clustering result in the annular area formed by two adjacent concentric circles from the plurality of first clustering results, only the plurality of second lane line points included in the third clustering result may be clustered again, that is, executing the following S504:

[0087] S504. Cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the third clustering result to obtain multiple second clustering results.

[0088] For example, when clustering the multiple lane line points according to the positions of the multiple second lane line points included in the third clustering result, it is usually based on a clustering algorithm to divide the multiple lane line points included in the third clustering result that are closer to each other into one clustering result, thereby obtaining multiple second clustering results.

[0089] For example, in the embodiments of the present disclosure, the clustering algorithm used may be some existing clustering algorithms, such as the K-Means clustering algorithm, or a density-based clustering method, etc., which may be specifically set according to actual needs.

[0090] Example, example, combined with the above Figure 6 The three first clustering results shown in FIG. 1 and FIG. 2 are assumed to be clustered again only for the second lane line points included in the first clustering result in the annular area formed by two adjacent concentric circles. Figure 6 The first clustering result in the annular area formed by the central circle and its adjacent circles, that is, the first clustering result with the second closest distance; and the first clustering result in the annular area formed by the second and third circles from the inside to the outside of the concentric circles, that is, the first clustering result with the farthest distance, are clustered again. For an example, see Figure 7 As shown, Figure 7 It is a schematic diagram of a clustering result of re-clustering the first clustering result provided by an embodiment of the present disclosure, which can obtain two second clustering results corresponding to the first clustering result with the second closest distance, and two second clustering results corresponding to the first clustering result with the farthest distance.

[0091] It can be seen that in the disclosed embodiment, when clustering multiple lane line points, the center point can be determined from the multiple first lane line points according to the positions of the multiple first lane line points; and the multiple first lane line points can be clustered according to the positions and the center point of the multiple first lane line points to obtain multiple first clustering results; and from the multiple first clustering results, a third clustering result in the annular area formed by two adjacent concentric circles is determined; and then the multiple second lane line points are clustered again according to the positions of the multiple second lane line points included in the third clustering result to obtain multiple second clustering results. In this way, through re-clustering, the lane line points corresponding to the divergent lane lines can be clustered into different second clustering results, so that when the lane line is extracted later according to the second clustering results obtained by re-clustering, the divergent lane lines can be accurately extracted.

[0092] After obtaining multiple second clustering results, the target lane line can be extracted according to the multiple second clustering results, that is, the above S203 is executed. In order to facilitate understanding of how to extract the target lane line according to the multiple second clustering results in the above S203, the following will be described. Figure 8 The third embodiment shown is described in detail.

[0093] Embodiment 3

[0094] Figure 8 FIG. 1 is a flow chart of a method for extracting a target lane line according to a plurality of second clustering results provided in accordance with the third embodiment of the present disclosure. The method can also be executed by software and / or hardware devices. For example, see Figure 8 As shown, the method may include:

[0095] S801. Determine a target lane line point from the plurality of third lane line points according to positions of the plurality of third lane line points included in the second clustering result; wherein the third lane line point belongs to the plurality of second lane line points.

[0096] For example, when determining the target lane line point from multiple third lane line points based on the positions of multiple third lane line points included in the second clustering result, the first average position can be first determined based on the positions of the multiple third lane line points included in the second clustering result; and the target lane line point can be determined from the multiple third lane line points included in the second clustering result based on the first average position.

[0097] It can be understood that in this possible implementation method, the method for determining the first average position based on the positions of multiple third lane line points included in the second clustering result is similar to the method for determining the second average position mentioned above. Please refer to the relevant description of determining the second average position above, and the embodiments of the present disclosure will not be repeated here.

[0098] For example, when determining the target lane line point from multiple third lane line points based on the first average position, the third lane line point closest to the first average position can be first determined from the multiple third lane line points; and the third lane line point closest to the first average position can be determined as the target lane line point.

[0099] It can be understood that if the second clustering result includes multiple third lane line points, and the number of third lane line points closest to the first average position is multiple, any third lane line point among the multiple closest third lane line points can be determined as the target lane line point.

[0100] After determining the target lane line point from the plurality of third lane line points included in the second clustering result, the target lane line point represents the second clustering result, and then connecting the target lane line points corresponding to the plurality of second clustering results, the target lane line can be extracted. For example, when connecting the target lane line points corresponding to the plurality of second clustering results, the following S802 can be performed:

[0101] S802. Starting from the center point, the target lane line points corresponding to the second clustering results are connected in sequence in the order of concentric circles from the inside to the outside to determine multiple initial lane lines; wherein, only one target lane line point in each annular area is in the initial lane line.

[0102] For example, starting from the center point, when connecting the target lane line points corresponding to the second clustering results in sequence in concentric circles from the inside to the outside, a Hungarian algorithm, for example, can be used to determine all possible connection schemes for the target lane line points corresponding to the second clustering results in two adjacent annular areas. In this way, multiple initial lane lines can be obtained based on all possible connection schemes.

[0103] For example, you can combine Figure 7 The four second clustering results shown can respectively determine the target lane line points corresponding to the four second clustering results. Assuming that the four target lane line points are target lane line point 1, target lane line point 2, target lane line point 3, and target lane line point 4, when all possible connection schemes of the target lane line points corresponding to the second clustering results in two adjacent annular areas are determined by using, for example, the Hungarian algorithm, two possible connection schemes can be obtained.

[0104] Among them, one possible connection scheme can be recorded as connection scheme 1: the center point is connected to the target lane line point 1 and the target lane line point 2 respectively, the target lane line point 1 is then connected to the target lane line point 3, and the target lane line point 2 is then connected to the target lane line point 4. Another possible connection scheme can be recorded as connection scheme 2: the center point is connected to the target lane line point 1 and the target lane line point 2 respectively, the target lane line point 1 is then connected to the target lane line point 4, and the target lane line point 2 is then connected to the target lane line point 3.

[0105] After connection scheme 1 and connection scheme 2 are determined respectively, an initial lane line can be obtained based on connection scheme 1, which can be recorded as initial lane line 1; another initial lane line can be obtained based on connection scheme 2, which can be recorded as initial lane line 2, thereby obtaining two initial lane lines.

[0106] It can be understood that when connecting to obtain multiple initial lane lines, the above example is only used to illustrate the connection starting from the center point and connecting the target lane line points corresponding to the second clustering result in sequence in the order of concentric circles from the inside to the outside. The target lane line points and the center point corresponding to the second clustering result can also be connected in sequence in the order of concentric circles from the outside to the inside. The specific setting can be made according to actual needs.

[0107] For example, when connecting the target lane line points, there are usually five cases, see Fig. 9 As shown, Fig. 9 is a schematic diagram of a connection situation provided by an embodiment of the present disclosure. The five situations may include a situation where there is no successor, that is, the target lane line point has no subsequent target lane line point that can be connected. Fig. 9 9a in the figure; in the case of no preceding drive, that is, there is no target lane line point connected to the target lane line point, see Fig. 9 9b in ; One-to-one situation, that is, one target lane line point is connected to one target lane line point, that is, the target lane line points are connected one by one, see Fig. 9 9c in the figure; for the bifurcation case, that is, a target lane line point bifurcates, corresponding to connecting two target lane line points, see Fig. 9 9d in the figure; for the merged case, that is, two target lane line points are connected to one target lane line point, see Fig. 9 After determining multiple initial lane lines, the target lane line can be determined from the multiple initial lane lines, that is, the following S803 is executed:

[0108] S803: Determine a target lane line from multiple initial lane lines.

[0109] For example, when determining a target lane line from multiple initial lane lines, the lengths of the multiple initial lane lines may be determined respectively first; and the initial lane line with the shortest length among the multiple initial lane lines may be determined as the target lane line.

[0110] For example, it can also be combined with Figure 7 The four second clustering results shown in FIG. 4 are assumed to correspond to four target lane line points, namely, target lane line point 1, target lane line point 2, target lane line point 3, and target lane line point 4, and the determined initial lane line includes initial lane line 1 and initial lane line 2. Obviously, the length of initial lane line 1 is less than the length of initial lane line 2, and initial lane line 1 can be determined as the target lane line finally extracted. For example, see Fig.10 As shown, Fig.10 It is a schematic diagram of an extracted target lane line provided by an embodiment of the present disclosure, where the target lane line is a divergent lane line.

[0111] It can be seen that in the disclosed embodiment, when performing lane line extraction, the target lane line point is first determined from the multiple third lane line points included in the second clustering result obtained by re-clustering; and starting from the center point, the target lane line points corresponding to the second clustering result are sequentially connected in the order of concentric circles from the inside to the outside to determine multiple initial lane lines; and then the target lane line is determined from the multiple initial lane lines. In view of the re-clustering of the first clustering result, the lane line points corresponding to the divergent lane lines can be effectively clustered into different second clustering results, so that the divergent lane lines can be accurately extracted based on the target lane line points corresponding to the second clustering result, thereby effectively improving the accuracy of the extraction result.

[0112] Embodiment 4

[0113] Fig.11 is a schematic diagram of a lane line extraction device 110 according to the fourth embodiment of the present disclosure, for example, see Fig.11 As shown, the lane line extraction device 110 may include:

[0114] The acquisition unit 1101 is used to acquire a plurality of first lane line points corresponding to the target lane line to be extracted.

[0115] The first clustering unit 1102 is used to cluster the multiple first lane line points according to their positions to obtain multiple first clustering results; wherein the first clustering results include multiple second lane line points, and the second lane line points belong to the multiple first lane line points.

[0116] The second clustering unit 1103 is used to cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the first clustering result to obtain multiple second clustering results.

[0117] The extraction unit 1104 is used to extract the target lane line according to the multiple second clustering results.

[0118] Optionally, the first clustering unit 1102 includes a first clustering module and a second clustering module.

[0119] The first clustering module is used to determine a center point from the plurality of first lane line points according to positions of the plurality of first lane line points.

[0120] The second clustering module is used to cluster the multiple first lane line points according to the positions and center points of the multiple first lane line points to obtain multiple first clustering results.

[0121] Optionally, the distribution of the multiple first clustering results satisfies the form of concentric circles, and the centers of the concentric circles are the center points; wherein the second clustering unit 1103 includes a third clustering module and a fourth clustering module.

[0122] A third clustering module is used to determine, from multiple first clustering results, a third clustering result in an annular area formed by two adjacent concentric circles based on the positions of multiple second lane line points included in the first clustering results; wherein the distance between the multiple second lane line points included in the third clustering result and the center point is greater than the radius of the center circle of the concentric circles.

[0123] The fourth clustering module is used to cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the third clustering result to obtain multiple second clustering results.

[0124] Optionally, the extraction unit 1104 includes a first extraction module, a second extraction module and a third extraction module.

[0125] The first extraction module is used to determine the target lane line point from the multiple third lane line points according to the positions of the multiple third lane line points included in the second clustering result; wherein the third lane line point belongs to the multiple second lane line points.

[0126] The second extraction module is used to connect the target lane line points corresponding to the second clustering results in sequence from the center point to the outside in concentric circles to determine multiple initial lane lines; wherein only one target lane line point in each annular area is in the initial lane line.

[0127] The third extraction module is used to determine a target lane line from a plurality of initial lane lines.

[0128] Optionally, the third extraction module includes a first extraction submodule and a second extraction submodule.

[0129] The first extraction submodule is used to determine the lengths of multiple initial lane lines respectively.

[0130] The second extraction submodule is used to determine the shortest initial lane line among the multiple initial lane lines as the target lane line.

[0131] Optionally, the first extraction module includes a first extraction submodule and a second extraction submodule.

[0132] The first extraction submodule is used to determine a first average position corresponding to the second clustering result according to the positions of the plurality of third lane line points included in the second clustering result.

[0133] The second extraction submodule is used to determine the target lane line point from the plurality of third lane line points according to the first average position.

[0134] Optionally, the second extraction submodule is specifically used to determine, from multiple third lane line points, a third lane line point that is closest to the first average position; and determine the third lane line point that is closest to the first average position as the target lane line point.

[0135] Optionally, the first clustering module includes a first clustering submodule and a second clustering submodule.

[0136] The first clustering submodule is used to determine a second average position corresponding to the plurality of first lane line points according to the positions of the plurality of first lane line points.

[0137] The second clustering submodule is used to determine, among the plurality of first lane line points, a first lane line point that is closest to the second average position as a center point.

[0138] The lane line extraction device 110 provided in the embodiment of the present disclosure can execute the technical solution of the lane line extraction method shown in any of the above embodiments, and its implementation principle and beneficial effects are similar to the implementation principle and beneficial effects of the lane line extraction method. Please refer to the implementation principle and beneficial effects of the lane line extraction method, and no further details will be given here.

[0139] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0140] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.

[0141] Fig.121 is a schematic block diagram of an electronic device 120 provided by an embodiment of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0142] like Figure 5 As shown, the device 120 includes a computing unit 1201, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1202 or a computer program loaded from a storage unit 1208 into a random access memory (RAM) 1203. In the RAM 1203, various programs and data required for the operation of the device 120 can also be stored. The computing unit 1201, the ROM 1202, and the RAM 1203 are connected to each other via a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0143] A number of components in the device 120 are connected to the I / O interface 1205, including: an input unit 1206, such as a keyboard, a mouse, etc.; an output unit 1207, such as various types of displays, speakers, etc.; a storage unit 1208, such as a disk, an optical disk, etc.; and a communication unit 1209, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1209 allows the device 120 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0144] The computing unit 1201 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1201 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1201 performs the various methods and processes described above, such as the lane line extraction method. For example, in some embodiments, the lane line extraction method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 1208. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 120 via the ROM 1202 and / or the communication unit 1209. When the computer program is loaded into the RAM 1203 and executed by the computing unit 1201, one or more steps of the lane line extraction method described above may be performed. Alternatively, in other embodiments, the computing unit 1201 may be configured to execute the lane line extraction method in any other appropriate manner (eg, by means of firmware).

[0145] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0146] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0147] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0148] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0149] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0150] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0151] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0152] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for extracting lane lines, comprising: Acquire multiple first lane line points corresponding to the target lane line to be extracted; Clustering the plurality of first lane line points according to the positions of the plurality of first lane line points to obtain a plurality of first clustering results; wherein the first clustering results include a plurality of second lane line points, and the second lane line points belong to the plurality of first lane line points; the distribution of the plurality of first clustering results satisfies the form of concentric circles, and the center of the concentric circles is a center point, and the center point is determined based on the positions of the plurality of first lane line points; According to the positions of the plurality of second lane line points included in the first clustering results, determining a third clustering result located in an annular area formed by two adjacent circles of the concentric circles from the plurality of first clustering results; wherein the distance between the plurality of second lane line points included in the third clustering result and the center point is greater than the radius of the center circle of the concentric circles; Clustering the plurality of second lane line points again according to the positions of the plurality of second lane line points included in the third clustering result to obtain a plurality of second clustering results; The target lane line is extracted according to the multiple second clustering results.

2. The method according to claim 1, wherein: The clustering of the plurality of first lane line points according to the positions of the plurality of first lane line points to obtain a plurality of first clustering results includes: determining a center point from among the plurality of first lane line points according to positions of the plurality of first lane line points; The multiple first lane line points are clustered according to the positions of the multiple first lane line points and the center point to obtain multiple first clustering results.

3. The method according to claim 1, wherein: The step of extracting the target lane line according to the plurality of second clustering results comprises: Determine a target lane line point from the plurality of third lane line points according to the positions of the plurality of third lane line points included in the second clustering result; wherein the third lane line point belongs to the plurality of second lane line points; Starting from the center point, in the order of the concentric circles from the inside to the outside, the target lane line points corresponding to the second clustering results are connected in sequence to determine a plurality of initial lane lines; wherein only one target lane line point in each annular area is located in the initial lane line; The target lane line is determined from the multiple initial lane lines.

4. The method according to claim 3, wherein: The step of determining the target lane line from the multiple initial lane lines comprises: Determining the lengths of the multiple initial lane lines respectively; An initial lane line with the shortest length among the multiple initial lane lines is determined as the target lane line.

5. The method according to claim 3, wherein: The determining the target lane line point from the plurality of third lane line points according to the positions of the plurality of third lane line points included in the second clustering result comprises: Determining a first average position corresponding to the second clustering result according to the positions of the plurality of third lane line points included in the second clustering result; The target lane line point is determined from the plurality of third lane line points according to the first average position.

6. The method according to claim 5, wherein: The step of determining the target lane line point from the plurality of third lane line points according to the first average position comprises: Determine, from the plurality of third lane line points, a third lane line point that is closest to the first average position; The third lane line point that is closest to the first average position is determined as the target lane line point.

7. The method according to any one of claims 2 to 6, wherein: The determining the center point from the plurality of first lane line points according to the positions of the plurality of first lane line points comprises: Determining, based on the positions of the plurality of first lane line points, a second average position corresponding to the plurality of first lane line points; A first lane line point among the multiple first lane line points that is closest to the second average position is determined as the center point.

8. A lane line extraction device, comprising: An acquisition unit, used to acquire a plurality of first lane line points corresponding to a target lane line to be extracted; A first clustering unit is used to cluster the plurality of first lane line points according to the positions of the plurality of first lane line points to obtain a plurality of first clustering results; wherein the first clustering results include a plurality of second lane line points, and the second lane line points belong to the plurality of first lane line points; the distribution of the plurality of first clustering results satisfies the form of concentric circles, and the center of the concentric circles is a center point, and the center point is determined based on the positions of the plurality of first lane line points; a second clustering unit, configured to cluster the plurality of second lane line points again according to the positions of the plurality of second lane line points included in the first clustering result, to obtain a plurality of second clustering results; An extraction unit, configured to extract the target lane line according to the plurality of second clustering results; Wherein, the second clustering unit includes a third clustering module and a fourth clustering module; The third clustering module is used to determine, from the plurality of first clustering results, a third clustering result located in an annular area formed by two adjacent circles of the concentric circles according to the positions of the plurality of second lane line points included in the first clustering result; wherein the distance between the plurality of second lane line points included in the third clustering result and the center point is greater than the radius of the center circle of the concentric circles; The fourth clustering module is used to cluster the multiple second lane line points again according to the positions of the multiple second lane line points included in the third clustering result to obtain the multiple second clustering results.

9. The device according to claim 8, wherein: The first clustering unit includes a first clustering module and a second clustering module; The first clustering module is used to determine a center point from the plurality of first lane line points according to positions of the plurality of first lane line points; The second clustering module is used to cluster the multiple first lane line points according to the positions of the multiple first lane line points and the center point to obtain multiple first clustering results.

10. The device according to claim 8, wherein: The extraction unit includes a first extraction module, a second extraction module and a third extraction module; The first extraction module is used to determine a target lane line point from the plurality of third lane line points according to positions of the plurality of third lane line points included in the second clustering result; wherein the third lane line point belongs to the plurality of second lane line points; The second extraction module is used to connect the target lane line points corresponding to the second clustering results in sequence from the center point to the outside of the concentric circles to determine multiple initial lane lines; wherein only one target lane line point in each annular area is located in the initial lane line; The third extraction module is used to determine the target lane line from the multiple initial lane lines.

11. The device according to claim 10, wherein: The third extraction module includes a first extraction submodule and a second extraction submodule; The first extraction submodule is used to respectively determine the lengths of the multiple initial lane lines; The second extraction submodule is used to determine the initial lane line with the shortest length among the multiple initial lane lines as the target lane line.

12. The device according to claim 10, wherein: The first extraction module includes a first extraction submodule and a second extraction submodule; The first extraction submodule is used to determine a first average position corresponding to the second clustering result according to the positions of the plurality of third lane line points included in the second clustering result; The second extraction submodule is used to determine the target lane line point from the plurality of third lane line points according to the first average position.

13. The device according to claim 12, wherein: The second extraction submodule is specifically used to determine the third lane line point closest to the first average position from the multiple third lane line points; and determine the third lane line point closest to the first average position as the target lane line point.

14. The device according to any one of claims 8 to 13, wherein: The first clustering module includes a first clustering submodule and a second clustering submodule; The first clustering submodule is used to determine a second average position corresponding to the plurality of first lane line points according to the positions of the plurality of first lane line points; The second clustering submodule is used to determine the first lane line point that is closest to the second average position among the multiple first lane line points as the center point.

15. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the lane line extraction method according to any one of claims 1 to 7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable the computer to execute the lane line extraction method according to any one of claims 1-7.

17. A computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the lane line extraction method according to any one of claims 1 to 7.

Images