Parking lot road network generation method, device, system and equipment, computer storage medium and automobile

Abstract

The invention relates to a road network generation method, device, system and equipment for a parking lot, a computer storage medium and an automobile, and the method comprises the steps: determining lane lines in the parking lot according to semantic point cloud map data and driving track points under the condition that the semantic point cloud map data and the driving track points of the parking lot are obtained, wherein the lane line has a direction; determining a lane in the parking lot according to the lane line; and determining an adjacent relationship and a successive relationship between different lanes to obtain a road network of the parking lot, the successive relationship being used for representing whether the lanes are continuous. The road network generation efficiency is improved, and the problem of low road network generation efficiency in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the technical field of automobile accessories, and in particular to a method, device, system, equipment, computer storage medium and automobile for generating a road network of a parking lot. Background Art

[0002] Different from the parking lot map, the parking lot road network is a mesh path structure generated based on the parking lot map and driving trajectory data. It determines the driving route and is a key link in the intelligent transportation system and autonomous driving technology to facilitate vehicle navigation, positioning and autonomous parking processes.

[0003] There are already methods for generating road networks in the prior art. However, if the methods for generating road networks in the prior art are to generate road networks with adjacent relationships and predecessor and successor relationships, it involves collecting multiple data and requires a large amount of calculation to generate the road network. Therefore, the efficiency of generating the road network is low. Summary of the invention

[0004] The present application provides a method, device, system, equipment, computer storage medium and automobile for generating a road network of a parking lot, so as to solve the problem of low efficiency of road network generation.

[0005] In a first aspect, the present application provides a method for generating a road network of a parking lot, comprising: upon obtaining semantic point cloud map data and driving trajectory points of the parking lot, determining lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determining lanes in the parking lot according to the lane lines; determining adjacent relationships and successor relationships between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0006] In a second aspect, the present application provides a road network generation device for a parking lot, comprising: a first determination module, for determining the lane lines in the above-mentioned parking lot according to the semantic point cloud map data and the above-mentioned driving trajectory points after obtaining the semantic point cloud map data and the above-mentioned driving trajectory points, wherein the above-mentioned lane lines have directions; a second determination module, for determining the lanes in the above-mentioned parking lot according to the above-mentioned lane lines; a third determination module, for determining the adjacent relationship and the predecessor and successor relationship between different lanes to obtain the road network of the above-mentioned parking lot, wherein the above-mentioned predecessor and successor relationship is used to indicate whether the lanes are continuous.

[0007] In a third aspect, the present application provides a road network generation system for a parking lot, comprising: a memory for storing computer-executable instructions; a controller for performing the following steps when executing the computer-executable instructions: upon obtaining the semantic point cloud map data and driving trajectory points of the parking lot, determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determining the lanes in the parking lot according to the lane lines; determining the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous; and a cloud server for providing cloud services.

[0008] In a fourth aspect, the present application provides a car, comprising: a controller, for determining, upon obtaining semantic point cloud map data and driving trajectory points of the parking lot, lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determining lanes in the parking lot according to the lane lines; determining adjacent relationships and successor relationships between different lanes to obtain a road network of the parking lot, wherein the successor relationships are used to indicate whether the lanes are continuous.

[0009] In a fifth aspect, the present application provides a road network generation device for a parking lot, comprising: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; at least one memory connected to the at least one bus, wherein the processor is configured to: upon obtaining the semantic point cloud map data and driving trajectory points of the parking lot, determine the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determine the lanes in the parking lot according to the lane lines; determine the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0010] In a sixth aspect, the present application further provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute any of the above-mentioned methods for generating a parking lot road network in the present application.

[0011] Beneficial effects of this application:

[0012] In the process of determining the road network of the parking lot, the present application can determine the lane lines according to the semantic point cloud map data and the driving trajectory points after obtaining the semantic point cloud map data and the driving trajectory points, so as to determine the lanes according to the lane lines, and then determine the adjacent relationship and the successor relationship between the lanes, so that a road network including the successor relationship and the adjacent relationship of the lanes can be established according to the lane lines and the driving trajectory points, thereby improving the efficiency of generating the road network, and thus solving the problem of low efficiency of road network generation in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 A schematic diagram of a road network generation method of the present application;

[0014] Figure 2 A lane line schematic diagram of this application;

[0015] Figure 3 A schematic diagram for determining lane lines for this application;

[0016] Figure 4 A schematic diagram of lane line point cloud data for this application;

[0017] Figure 5 A schematic diagram for determining a lane for this application;

[0018] Figure 6 A lane centerline schematic diagram of the present application;

[0019] Figure 7 A flowchart of a road network generation method of the present application;

[0020] Figure 8 A schematic diagram of a road network generating device of the present application;

[0021] Fig. 9 A schematic diagram of a road network generation device of the present application. DETAILED DESCRIPTION

[0022] The following will describe the implementation methods of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, not for limiting the scope of protection of the present application.

[0023] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0024] For ease of description, spatial relative terms may be used in the text to describe the relative position or movement of an element or feature as shown in the figure relative to another element or feature, such as "inside", "outside", "inner side", "outer side", "below", "below", "above", "above", "front", "back", etc. Such spatial relative terms are intended to include different orientations of the device in use or operation in addition to the orientation depicted in the figure. For example, if the device in the figure undergoes a position flip or a change in posture or a change in motion state, then these directional indications will also change accordingly, for example: an element described as "below other elements or features" or "below other elements or features" will subsequently be oriented as "above other elements or features" or "above other elements or features". Therefore, the example term "below..." can include both upper and lower orientations. The device can be oriented otherwise (rotated 90 degrees or in other directions) and the spatial relative descriptors used in the text are interpreted accordingly.

[0025] In order to solve the technical problem of low road network generation efficiency in the prior art, the present application provides a road network generation method, device, system, equipment, computer storage medium and automobile for a parking lot, which can achieve the effect of improving the road network generation efficiency.

[0026] Figure 1 A schematic diagram of a road network generation method according to an embodiment of the present application is shown in FIG. Figure 1 As shown, including:

[0027] S101, when the semantic point cloud map data and the driving trajectory points of the parking lot are obtained, determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions;

[0028] S102, determining a lane in the parking lot according to the lane line;

[0029] S103, determining the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0030] In order to solve the problem of low efficiency in generating road networks in the prior art, the present application proposes this method, which aims to determine a road network including the adjacent relationship and the successor relationship of lanes through lane lines and driving trajectory points, thereby improving the efficiency of road network generation.

[0031] In this application, a mobile mapping device can be used to move in the parking lot to obtain semantic point cloud map data in the parking lot. For example, various sensors on the mobile mapping device, such as cameras, laser radars, millimeter wave radars, global positioning systems (GPS), etc., are used to collect semantic point cloud map data and driving trajectories in the parking lot. The driving trajectory includes the movement trajectory of the mobile mapping device and the direction of the movement trajectory.

[0032] It should be noted that, in this embodiment, the mobile surveying and mapping device can collect data, and the data can be handed over to the processor or the cloud for processing to generate a road network. The mobile surveying and mapping device can also collect data and process the data by itself to generate a road network. The mobile surveying and mapping device can be a mobile intelligent robot with a sensor, or a car with a sensor. The car can be a car dedicated to collecting data, or a car driven by a user. If it is a car driven by a user, the user's consent is required to obtain the data collected by the car. In this case, the data can be obtained and handed over to the processor or uploaded to the cloud for processing to generate a road network. If it is a car dedicated to collecting data, the data can be collected by the car and handed over to the processor or uploaded to the cloud for processing, or the car can process the data by itself to generate a road network.

[0033] In the above example, if the road network is generated by acquiring data collected by the user's car, since the routes of different users' cars in the parking lot are different, the data collected by one user's car can be used to generate part of the road network, and the data collected by other users' cars can be used to complete the road network. If a mobile surveying and mapping device is used to collect data, the mobile surveying and mapping device is controlled to traverse the parking lot to collect data and generate a road network.

[0034] In this embodiment, the lane lines in the parking lot can be determined based on the semantic point cloud map data and the driving trajectory points. The lane lines in the parking lot can be lines drawn on the road surface of the parking lot, which are generally white or yellow, but can also be other colors, and can be straight lines or curves, and can be solid lines or intermittent dashed lines. In some cases, a lane line is composed of two lines. For the case of two lines, in one example, the road network generation process can be performed as one line. Or, in another example, the road network generation process is performed as two lines. If the road network is generated as one line, the center lines of the two lines can be used instead of the two lines. If the road network is generated as two lines, one of the two lines forms a lane with the other adjacent lines, and the other line forms another lane with the other adjacent lines. The gap between the two lines is too small, so if it is identified as a lane, it is actively corrected. For example, Figure 2 As shown in the figure, a lane line includes two lines, where 201 and 202 are two lane lines respectively, and 203 is two lines of a lane line. There is a gap between the two lines. If the gap between the two lines is identified as a lane, error correction is performed. Lane line 201 and the left line of lane line 203 form a lane, and lane line 202 and the right line of lane line 203 form a lane. The gap between the two lines of lane 203 is not used as a lane after correction.

[0035] Generally speaking, there is a lane line on each side of a lane (except for the case where a lane line consists of two lines as mentioned above), and two adjacent lanes share a lane line (if the lane line consists of two lines as mentioned above, the two adjacent lanes each occupy one line of a lane line).

[0036] After the lane lines are determined, the lanes can be determined based on the lane lines. Since a lane line and adjacent lane lines can form a lane, and the lane lines are obtained through point cloud data, the point cloud data of the lane lines can be marked. The lane lines can be used to determine how many lanes there are, whether the lanes are adjacent to each other, and whether the successor relationship is continuous.

[0037] In the process of determining the road network of the parking lot, the present application can determine the lane lines according to the semantic point cloud map data and the driving trajectory points after obtaining the semantic point cloud map data and the driving trajectory points, so as to determine the lanes according to the lane lines, and then determine the adjacent relationship and the successor relationship between the lanes, so that a road network including the successor relationship and the adjacent relationship of the lanes can be established according to the lane lines and the driving trajectory points, thereby improving the efficiency of generating the road network, and thus solving the problem of low efficiency of road network generation in the prior art.

[0038] In one example, if Figure 3As shown, the step S101 of determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points may include:

[0039] S301, obtaining a lane line point cloud of the lane line of the parking lot from the semantic point cloud map;

[0040] S302, connecting the lane line point clouds into lane lines without directions;

[0041] S303, determining the moving direction of the movable surveying and mapping device that generates the driving trajectory points according to the driving trajectory points;

[0042] S304, determining the direction of the lane line according to the above moving direction, and obtaining a lane line with a direction.

[0043] In this embodiment, the lane lines in the parking lot can be determined based on the acquired semantic point cloud map data and driving trajectory points. When acquiring the semantic point cloud map data and driving trajectory points, the semantic point cloud data in the parking lot can be collected based on the car's sensors such as cameras and lidars. For driving trajectory points, the positions of the vehicles can be collected as trajectory points at predetermined intervals during the driving of the vehicle, and the overlapping or too close trajectory points can be deleted, and the non-overlapping trajectory points can be retained. Since the car has a direction when driving, the driving trajectory points have a pointing direction.

[0044] After obtaining the semantic point cloud map data and the driving trajectory points, the lane line point cloud of the lane line can be obtained from the semantic point cloud map data. That is, the redundant point clouds in the semantic point cloud map data are filtered out, and only the lane line point cloud of the lane line is retained. In another example, when the car collects semantic point cloud map data, only the lane point cloud can be collected. For example, the camera collects the image of the lane line on the ground, and the point cloud data generation algorithm is used to generate the point cloud data based on the image. For example, when the car is driving or stopped, the position of the car is determined, and then the lane line on the ground is photographed by the camera, the shooting angle of the camera is determined, and the position of the lane line in the scene is determined according to the position of the car and the shooting angle of the camera, as well as the image of the lane line taken by the camera, thereby obtaining the point cloud data of the lane line.

[0045] After obtaining the point cloud data of the lane, the point cloud data of the lane is represented by multiple points, such as Figure 4 As an example, if it is a solid line, it is shown as a1, and if it is a dotted line, it is shown as b1. If the lane line is two lines, the solid line is shown as c1, and the dotted line is shown as d1. In this embodiment, the point cloud of the lane line can be abstracted into a line, as shown from a2 to d2.

[0046] Since lane lines do not have directions, the direction of the lane lines can be determined in combination with driving trajectory points. Driving trajectory points are generated by the mobile surveying and mapping device in chronological order during movement, so the driving direction can be determined. The direction of the lane lines is determined based on the driving direction, thereby determining the direction of the lane lines.

[0047] In this embodiment, the lane line is determined by the above method. Not only can the position of the lane line be determined by point cloud data, but the direction of the lane line can also be determined according to the driving trajectory points, thereby determining the lane line with direction and improving the accuracy of the lane line.

[0048] In one example, if Figure 5 As shown, for step S102, determining the lane in the parking lot according to the lane line may include:

[0049] S501, starting from a first lane line among the lane lines, traversing an adjacent lane line of the first lane line to obtain a second lane line;

[0050] S502, determining a lane centerline according to the first lane line and the second lane line;

[0051] S503: Determine the lane center line, the first lane line, and the second lane line as lines on a lane to determine a lane, wherein two adjacent lanes share a lane line.

[0052] In this embodiment, since the lane lines in the parking lot have been determined in the above process, the lanes can be determined based on the lane lines. For lane determination, it is possible to start from one lane line and traverse adjacent lane lines. For example, Figure 6 As shown ( Figure 6 , lane lines are all represented by solid lines), a lane line 601 is randomly selected, and lane lines are traversed from the lane line to the left and right. The adjacent lane lines traversed form a lane with the current lane line. If the adjacent lane lines are traversed continuously, a lane is traversed continuously. For each traversed lane, the lane center line 602 is determined, such as Figure 6 As shown in the dashed line 602, each lane corresponds to a lane centerline, and vice versa, a lane centerline also corresponds to a lane.

[0053] By identifying lanes using the method of this embodiment, the center line of each lane can be determined. After the road network is constructed, the driving of cars in the road network can be controlled based on the center lines of the lanes, so that the cars stay at the center lines of the lanes while driving and do not deviate from the lanes.

[0054] In one example, after the lanes are identified, step S103 may be executed to determine the adjacent relationship and the preceding and succeeding relationship of the lanes.

[0055] For determining the adjacent relationship, a unique number can be assigned to the identified lanes or lane centerlines. For example, starting from a lane centerline, traverse the adjacent lane centerlines; determine the lane where the traversed adjacent lane centerline is located as the lane adjacent to the current lane; assign a unique number to each lane to record the adjacent relationship between lanes.

[0056] The lane centerline can be assigned a unique number to record the lane centerline and the corresponding lane. If the lane line includes two lines, the gap between the two lines may be determined as a lane, such as Figure 2 The gap between lane lines 203 in the image is shown in FIG. At this time, error correction is performed, and it can be determined whether a lane is formed according to the distance between the lane lines. If the distance is too small, it is impossible to form a lane.

[0057] In this embodiment, the unique number can be a letter, a number, a symbol, or a combination of any two or all of the three. The relationship between lanes can be reflected by the unique number of the lane centerline. If the unique number is a number, the unique number of a lane can be 0, the adjacent lane on the left can be -1, and the adjacent lane on the right can be 1. According to the numbering, it can be known that lanes numbered 2 and 3 are adjacent, and lanes numbered 1 and 3 are separated by a lane.

[0058] The adjacent relationship of lanes is determined by this embodiment, so that whether lanes are adjacent and how many lanes are separated from each other can be determined based on the unique numbers, and data support can be provided for lane changes of vehicles in the road network.

[0059] To determine the predecessor and successor relationship, the above-mentioned driving trajectory points can be traversed to determine the lane where each of the above-mentioned driving trajectory points is located; each trajectory point is marked with the above-mentioned unique number of the lane where it is located; when the unique numbers of two adjacent trajectory points indicate that the lanes are not adjacent, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship, or when the unique numbers of two adjacent trajectory points are the same, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship.

[0060] In this embodiment, the predecessor and successor relationship is used to determine whether the lanes are still continuous after passing through an intersection or turning. For example, if the three lanes before the intersection are merged into two lanes after passing through an intersection, the predecessor and successor relationship of some lanes has changed and no longer exists, while the predecessor and successor relationship of some lanes exists.

[0061] In order to determine whether the successor relationship of the lanes has changed, each track point in the driving trajectory points can be marked with the unique number of the lane in which it is located. Then, each track point in the driving trajectory points corresponds to a unique number, and as the first track point in the driving trajectory points is traversed backward, if the unique number of the next track point changes, it means that the lane is different from the lane where the previous track point is located. It means that the car has changed lanes. If the lane is changed, it means that the lane where the next track point is located does not have a successor relationship with the lane where the previous track point is located. In other words, in this embodiment, whether the lanes have a successor relationship can be determined based on whether the unique number has changed.

[0062] Figure 7 This is a flow chart of the present embodiment. First, the semantic point cloud map data and driving trajectory points of the target parking lot acquired by the vehicle are loaded on the vehicle side, and then the semantic point cloud map data is vectorized to obtain vector lane lines. The vectorization process is to determine the direction of the point cloud data of the lane lines in the semantic point cloud map data. The direction of the driving trajectory points is the direction of the point cloud data of the lane lines. The obtained vector lane lines are a series of discrete vectorized points stored in sequence. All lane line vector points are constructed into a KDTree tree A. KDTree tree A is used to store lane line vector points and can quickly search for lane line vector points. Repeat the following steps for each lane line:

[0063] 1. Determine the direction of one side, traverse the vector points sequentially, find the adjacent lane line vector point on this side in tree A, and generate the lane center point (the point on the lane center line, also called the lane center line point) from the two vector points;

[0064] 2 After traversing this side, construct the corresponding lane based on the generated lane center point;

[0065] 3 Repeat steps 1 and 2 on the other side to build another lane.

[0066] After the lanes are constructed, each lane corresponds to a lane centerline, which is composed of multiple lane center points. Construct a KDTree tree B with the lane center points of all lane centerlines. KDTree tree B is used to store lane center points and can quickly search for lane center points. Repeat the following steps for each lane:

[0067] a For the current lane, traverse the points of the center line of its lane in sequence, and find the points of the center line of the adjacent lane in tree B;

[0068] b. Construct the adjacent relationship between the current lane and the adjacent lane based on the found centerline points of the adjacent lane.

[0069] After determining the relationship between lanes, traverse the driving trajectory points in sequence, find the nearest lane center point in tree B, and record the corresponding lane of the driving trajectory point; according to the order of driving trajectory points, determine whether to build the predecessor and successor relationship of the lanes: if the lane numbers corresponding to two consecutive trajectory points already have an adjacent relationship, then the predecessor and successor relationship will not be built; if the lanes corresponding to two consecutive trajectory points do not have an adjacent relationship, it means that the lane numbers corresponding to the two consecutive trajectory points are the same, then the predecessor and successor relationship will be built. If the car adds another driving trajectory point, then according to the newly added driving trajectory point, repeatedly search for the nearest lane center point in tree B, and record the corresponding lane of the driving trajectory point; according to the order of driving trajectory points, determine whether to build the predecessor and successor relationship of the lane.

[0070] Figure 8 : is a schematic diagram of a parking lot road network generation device of this embodiment, including:

[0071] The first determination module 801 is used to determine the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points when the semantic point cloud map data and the driving trajectory points of the parking lot are obtained, wherein the lane lines have directions;

[0072] The second determination module 802 is used to determine the lane in the parking lot according to the lane line;

[0073] The third determination module 803 is used to determine the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0074] In order to solve the problem of low efficiency in generating road networks in the prior art, the present application proposes the present device, which is intended to determine a road network including adjacent relationships and successor relationships of lanes through lane lines and driving trajectory points, thereby improving the efficiency of road network generation.

[0075] In this application, a mobile mapping device can be used to move in the parking lot to obtain semantic point cloud map data in the parking lot. For example, various sensors on the mobile mapping device, such as cameras, laser radars, millimeter wave radars, global positioning systems (GPS), etc., are used to collect semantic point cloud map data and driving trajectories in the parking lot. The driving trajectory includes the movement trajectory of the mobile mapping device and the direction of the movement trajectory.

[0076] It should be noted that, in this embodiment, the mobile surveying and mapping device can collect data, and the data can be processed by this device to generate a road network. The mobile surveying and mapping device can also collect data and process the data by itself through the installed device to generate a road network. The mobile surveying and mapping device can be a mobile intelligent robot with a sensor, or a car with a sensor. The car can be a car dedicated to collecting data, or a car driven by a user. If it is a car driven by a user, the user's consent is required to obtain the data collected by the car. In this case, the data can be obtained and handed over to the processor or uploaded to the cloud, and processed by this device to generate a road network. If it is a car dedicated to collecting data, the data can be collected by the car, handed over to the processor or uploaded to the cloud, and processed by this device, or the car can process the data by itself to generate a road network.

[0077] In the above example, if the road network is generated by acquiring data collected by the user's car, since the routes of different users' cars in the parking lot are different, the data collected by one user's car can be used to generate part of the road network, and the data collected by other users' cars can be used to complete the road network. If a mobile surveying and mapping device is used to collect data, the mobile surveying and mapping device is controlled to traverse the parking lot to collect data and generate a road network.

[0078] In this embodiment, the lane lines in the parking lot can be determined based on the semantic point cloud map data and the driving trajectory points. The lane lines in the parking lot can be lines drawn on the road surface of the parking lot, which are generally white or yellow, but can also be other colors, and can be straight lines or curves, and can be solid lines or intermittent dashed lines. In some cases, a lane line is composed of two lines. For the case of two lines, in one example, the road network generation process can be performed as one line. Or, in another example, the road network generation process is performed as two lines. If the road network is generated as one line, the center lines of the two lines can be used instead of the two lines. If the road network is generated as two lines, one of the two lines forms a lane with the other adjacent lines, and the other line forms another lane with the other adjacent lines. The gap between the two lines is too small, so if it is identified as a lane, it is actively corrected. For example, Figure 2 As shown in the figure, a lane line includes two lines, where 201 and 202 are two lane lines respectively, and 203 is two lines of a lane line. There is a gap between the two lines. If the gap between the two lines is identified as a lane, error correction is performed. Lane line 201 and the left line of lane line 203 form a lane, and lane line 202 and the right line of lane line 203 form a lane. The gap between the two lines of lane 203 is not used as a lane after correction.

[0079] Generally speaking, there is a lane line on each side of a lane (except for the case where a lane line consists of two lines as mentioned above), and two adjacent lanes share a lane line (if the lane line consists of two lines as mentioned above, the two adjacent lanes each occupy one line of a lane line).

[0080] After the lane lines are determined, the lanes can be determined based on the lane lines. Since a lane line and adjacent lane lines can form a lane, and the lane lines are obtained through point cloud data, the point cloud data of the lane lines can be marked. The lane lines can be used to determine how many lanes there are, whether the lanes are adjacent to each other, and whether the successor relationship is continuous.

[0081] In the process of determining the road network of the parking lot, the present application can determine the lane lines according to the semantic point cloud map data and the driving trajectory points after obtaining the semantic point cloud map data and the driving trajectory points, so as to determine the lanes according to the lane lines, and then determine the adjacent relationship and the successor relationship between the lanes, so that a road network including the successor relationship and the adjacent relationship of the lanes can be established according to the lane lines and the driving trajectory points, thereby improving the efficiency of generating the road network, and thus solving the problem of low efficiency of road network generation in the prior art.

[0082] In one example, if Figure 3 As shown, the step S101 of determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points may include:

[0083] S301, obtaining a lane line point cloud of the lane line of the parking lot from the semantic point cloud map;

[0084] S302, connecting the lane line point clouds into lane lines without directions;

[0085] S303, determining the moving direction of the movable surveying and mapping device that generates the driving trajectory points according to the driving trajectory points;

[0086] S304, determining the direction of the lane line according to the above moving direction, and obtaining a lane line with a direction.

[0087] In this embodiment, the lane lines in the parking lot can be determined based on the acquired semantic point cloud map data and driving trajectory points. When acquiring the semantic point cloud map data and driving trajectory points, the semantic point cloud data in the parking lot can be collected based on the car's sensors such as cameras and lidars. For driving trajectory points, the positions of the vehicles can be collected as trajectory points at predetermined intervals during the driving of the vehicle, and the overlapping or too close trajectory points can be deleted, and the non-overlapping trajectory points can be retained. Since the car has a direction when driving, the driving trajectory points have a pointing direction.

[0088] After obtaining the semantic point cloud map data and the driving trajectory points, the lane line point cloud of the lane line can be obtained from the semantic point cloud map data. That is, the redundant point clouds in the semantic point cloud map data are filtered out, and only the lane line point cloud of the lane line is retained. In another example, when the car collects semantic point cloud map data, only the lane point cloud can be collected. For example, the camera collects the image of the lane line on the ground, and the point cloud data generation algorithm is used to generate the point cloud data based on the image. For example, when the car is driving or stopped, the position of the car is determined, and then the lane line on the ground is photographed by the camera, the shooting angle of the camera is determined, and the position of the lane line in the scene is determined according to the position of the car and the shooting angle of the camera, as well as the image of the lane line taken by the camera, thereby obtaining the point cloud data of the lane line.

[0089] After obtaining the point cloud data of the lane, the point cloud data of the lane is represented by multiple points, such as Figure 4 As an example, if it is a solid line, it is shown as a1, and if it is a dotted line, it is shown as b1. If the lane line is two lines, the solid line is shown as c1, and the dotted line is shown as d1. In this embodiment, the point cloud of the lane line can be abstracted into a line, as shown from a2 to d2.

[0090] Since lane lines do not have directions, the direction of the lane lines can be determined in combination with driving trajectory points. Driving trajectory points are generated by the mobile surveying and mapping device in chronological order during movement, so the driving direction can be determined. The direction of the lane lines is determined based on the driving direction, thereby determining the direction of the lane lines.

[0091] This embodiment determines the lane line through the above method. Not only can the position of the lane line be determined through the point cloud, but the direction of the lane line can also be determined based on the driving trajectory points, thereby determining the lane line with direction and improving the accuracy of the lane line.

[0092] In one example, if Figure 5 As shown, for step S102, determining the lane in the parking lot according to the lane line may include:

[0093] S501, starting from a first lane line among the lane lines, traversing an adjacent lane line of the first lane line to obtain a second lane line;

[0094] S502, determining a lane centerline according to the first lane line and the second lane line;

[0095] S503: Determine the lane center line, the first lane line, and the second lane line as lines on a lane to determine a lane, wherein two adjacent lanes share a lane line.

[0096] In this embodiment, since the lane lines in the parking lot have been determined in the above process, the lanes can be determined based on the lane lines. For lane determination, it is possible to start from one lane line and traverse adjacent lane lines. For example, Figure 6 As shown ( Figure 6 , lane lines are all represented by solid lines), a lane line 601 is randomly selected, and lane lines are traversed from the lane line to the left and right. The adjacent lane lines traversed form a lane with the current lane line. If the adjacent lane lines are traversed continuously, a lane is traversed continuously. For each traversed lane, the lane center line 602 is determined, such as Figure 6 As shown in the dashed line 602, each lane corresponds to a lane centerline, and vice versa, a lane centerline also corresponds to a lane.

[0097] By identifying lanes using the method of this embodiment, the center line of each lane can be determined. After the road network is constructed, the driving of cars in the road network can be controlled based on the center lines of the lanes, so that the cars stay at the center lines of the lanes while driving and do not deviate from the lanes.

[0098] In one example, after the lanes are identified, step S103 may be executed to determine the adjacent relationship and the preceding and succeeding relationship of the lanes.

[0099] For determining the adjacent relationship, a unique number can be assigned to the identified lanes or lane centerlines. For example, starting from a lane centerline, traverse the adjacent lane centerlines; determine the lane where the traversed adjacent lane centerline is located as the lane adjacent to the current lane; assign a unique number to each lane to record the adjacent relationship between lanes.

[0100] The lane centerline can be assigned a unique number to record the lane centerline and the corresponding lane. If the lane line includes two lines, the gap between the two lines may be determined as a lane, such as Figure 2 The gap between lane lines 203 in the image is shown in FIG. At this time, error correction is performed, and it can be determined whether a lane is formed according to the distance between the lane lines. If the distance is too small, it is impossible to form a lane.

[0101] In this embodiment, the unique number can be a letter, a number, a symbol, or a combination of any two or all of the three. The relationship between lanes can be reflected by the unique number of the lane centerline. If the unique number is a number, the unique number of a lane can be 0, the adjacent lane on the left can be -1, and the adjacent lane on the right can be 1. According to the numbering, it can be known that lanes numbered 2 and 3 are adjacent, and lanes numbered 1 and 3 are separated by a lane.

[0102] The adjacent relationship of lanes is determined by this embodiment, so that whether lanes are adjacent and how many lanes are separated from each other can be determined based on the unique numbers, and data support can be provided for lane changes of vehicles in the road network.

[0103] To determine the predecessor and successor relationship, the above-mentioned driving trajectory points can be traversed to determine the lane where each of the above-mentioned driving trajectory points is located; each trajectory point is marked with the above-mentioned unique number of the lane where it is located; when the unique numbers of two adjacent trajectory points indicate that the lanes are not adjacent, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship, or when the unique numbers of two adjacent trajectory points are the same, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship.

[0104] In this embodiment, the predecessor and successor relationship is used to determine whether the lanes are still continuous after passing through an intersection or turning. For example, if the three lanes before the intersection are merged into two lanes after passing through an intersection, the predecessor and successor relationship of some lanes has changed and no longer exists, while the predecessor and successor relationship of some lanes exists.

[0105] In order to determine whether the successor relationship of the lanes has changed, each track point in the driving trajectory points can be marked with the unique number of the lane in which it is located. Then, each track point in the driving trajectory points corresponds to a unique number, and as the first track point in the driving trajectory points is traversed backward, if the unique number of the next track point changes, it means that the lane is different from the lane where the previous track point is located. It means that the car has changed lanes. If the lane is changed, it means that the lane where the next track point is located does not have a successor relationship with the lane where the previous track point is located. In other words, in this embodiment, whether the lanes have a successor relationship can be determined based on whether the unique number has changed.

[0106] For other examples of this embodiment, please refer to the above method examples, which will not be repeated here.

[0107] The present embodiment also provides a road network generation system for a parking lot, including: a memory for storing computer-executable instructions; a controller for performing the following steps when executing the computer-executable instructions: upon obtaining semantic point cloud map data and driving trajectory points of the parking lot, determining lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determining lanes in the parking lot according to the lane lines; determining the adjacent relationship and the predecessor and successor relationship between different lanes to obtain the road network of the parking lot, wherein the predecessor and successor relationship is used to indicate whether the lanes are continuous; and a cloud server for providing cloud services.

[0108] In this embodiment, the controller can be located in the car, in a device near the parking lot, or in a cloud server, and can obtain semantic point cloud map data and driving trajectory points to generate a road network.

[0109] In this embodiment, the lane lines in the parking lot can be determined based on the semantic point cloud map data and the driving trajectory points. The lane lines in the parking lot can be lines drawn on the road surface of the parking lot, which are generally white or yellow, but can also be other colors, and can be straight lines or curves, and can be solid lines or intermittent dashed lines. In some cases, a lane line is composed of two lines. For the case of two lines, in one example, the road network generation process can be performed as one line. Or, in another example, the road network generation process is performed as two lines. If the road network is generated as one line, the center lines of the two lines can be used instead of the two lines. If the road network is generated as two lines, one of the two lines forms a lane with the other adjacent lines, and the other line forms another lane with the other adjacent lines. The gap between the two lines is too small, so if it is identified as a lane, it is actively corrected. For example, Figure 2 As shown in the figure, a lane line includes two lines, where 201 and 202 are two lane lines respectively, and 203 is two lines of a lane line. There is a gap between the two lines. If the gap between the two lines is identified as a lane, error correction is performed. Lane line 201 and the left line of lane line 203 form a lane, and lane line 202 and the right line of lane line 203 form a lane. The gap between the two lines of lane 203 is not used as a lane after correction.

[0110] Generally speaking, there is a lane line on each side of a lane (except for the case where a lane line consists of two lines as mentioned above), and two adjacent lanes share a lane line (if the lane line consists of two lines as mentioned above, the two adjacent lanes each occupy one line of a lane line).

[0111] After the lane lines are determined, the lanes can be determined based on the lane lines. Since a lane line and adjacent lane lines can form a lane, and the lane lines are obtained through point cloud data, the point cloud data of the lane lines can be marked. The lane lines can be used to determine how many lanes there are, whether the lanes are adjacent to each other, and whether the successor relationship is continuous.

[0112] In the process of determining the road network of the parking lot, the present application can determine the lane lines according to the semantic point cloud map data and the driving trajectory points after obtaining the semantic point cloud map data and the driving trajectory points, so as to determine the lanes according to the lane lines, and then determine the adjacent relationship and the successor relationship between the lanes, so that a road network including the successor relationship and the adjacent relationship of the lanes can be established according to the lane lines and the driving trajectory points, thereby improving the efficiency of generating the road network, and thus solving the problem of low efficiency of road network generation in the prior art.

[0113] In one example, if Figure 3 As shown, the step S101 of determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points may include:

[0114] S301, obtaining a lane line point cloud of the lane line of the parking lot from the semantic point cloud map;

[0115] S302, connecting the lane line point clouds into lane lines without directions;

[0116] S303, determining the moving direction of the movable surveying and mapping device that generates the driving trajectory points according to the driving trajectory points;

[0117] S304, determining the direction of the lane line according to the above moving direction, and obtaining a lane line with a direction.

[0118] In this embodiment, the lane lines in the parking lot can be determined based on the acquired semantic point cloud map data and driving trajectory points. When acquiring the semantic point cloud map data and driving trajectory points, the semantic point cloud data in the parking lot can be collected based on the car's sensors such as cameras and lidars. For driving trajectory points, the positions of the vehicles can be collected as trajectory points at predetermined intervals during the driving of the vehicle, and the overlapping or too close trajectory points can be deleted, and the non-overlapping trajectory points can be retained. Since the car has a direction when driving, the driving trajectory points have a pointing direction.

[0119] After obtaining the semantic point cloud map data and the driving trajectory points, the lane line point cloud of the lane line can be obtained from the semantic point cloud map data. That is, the redundant point clouds in the semantic point cloud map data are filtered out, and only the lane line point cloud of the lane line is retained. In another example, when the car collects semantic point cloud map data, only the lane point cloud can be collected. For example, the camera collects the image of the lane line on the ground, and the point cloud data generation algorithm is used to generate the point cloud data based on the image. For example, when the car is driving or stopped, the position of the car is determined, and then the lane line on the ground is photographed by the camera, the shooting angle of the camera is determined, and the position of the lane line in the scene is determined according to the position of the car and the shooting angle of the camera, as well as the image of the lane line taken by the camera, thereby obtaining the point cloud data of the lane line.

[0120] After obtaining the point cloud data of the lane, the point cloud data of the lane is represented by multiple points, such as Figure 4 As an example, if it is a solid line, it is shown as a1, and if it is a dotted line, it is shown as b1. If the lane line is two lines, the solid line is shown as c1, and the dotted line is shown as d1. In this embodiment, the point cloud of the lane line can be abstracted into a line, as shown from a2 to d2.

[0121] Since lane lines do not have directions, the direction of the lane lines can be determined in combination with driving trajectory points. Driving trajectory points are generated by the mobile surveying and mapping device in chronological order during movement, so the driving direction can be determined. The direction of the lane lines is determined based on the driving direction, thereby determining the direction of the lane lines.

[0122] In this embodiment, the lane line is determined by the above method. Not only can the position of the lane line be determined through point cloud data, but the direction of the lane line can also be determined based on the driving trajectory points, thereby determining the lane line with direction and improving the accuracy of the lane line.

[0123] In one example, if Figure 5 As shown, for step S102, determining the lane in the parking lot according to the lane line may include:

[0124] S501, starting from a first lane line among the lane lines, traversing an adjacent lane line of the first lane line to obtain a second lane line;

[0125] S502, determining a lane centerline according to the first lane line and the second lane line;

[0126] S503: Determine the lane center line, the first lane line, and the second lane line as lines on a lane to determine a lane, wherein two adjacent lanes share a lane line.

[0127] In this embodiment, since the lane lines in the parking lot have been determined in the above process, the lanes can be determined based on the lane lines. For lane determination, it is possible to start from one lane line and traverse adjacent lane lines. For example, Figure 6 As shown ( Figure 6 , lane lines are all represented by solid lines), a lane line 601 is randomly selected, and lane lines are traversed from the lane line to the left and right. The adjacent lane lines traversed form a lane with the current lane line. If the adjacent lane lines are traversed continuously, a lane is traversed continuously. For each traversed lane, the lane center line 602 is determined, such as Figure 6 As shown in the dashed line 602, each lane corresponds to a lane centerline, and vice versa, a lane centerline also corresponds to a lane.

[0128] By identifying lanes using the method of this embodiment, the center line of each lane can be determined. After the road network is constructed, the driving of cars in the road network can be controlled based on the center lines of the lanes, so that the cars stay at the center lines of the lanes while driving and do not deviate from the lanes.

[0129] In one example, after the lanes are identified, step S103 may be executed to determine the adjacent relationship and the preceding and succeeding relationship of the lanes.

[0130] For determining the adjacent relationship, a unique number can be assigned to the identified lanes or lane centerlines. For example, starting from a lane centerline, traverse the adjacent lane centerlines; determine the lane where the traversed adjacent lane centerline is located as the lane adjacent to the current lane; assign a unique number to each lane to record the adjacent relationship between lanes.

[0131] The lane centerline can be assigned a unique number to record the lane centerline and the corresponding lane. If the lane line includes two lines, the gap between the two lines may be determined as a lane, such as Figure 2 The gap between lane lines 203 in the image is shown in FIG. At this time, error correction is performed, and it can be determined whether a lane is formed according to the distance between the lane lines. If the distance is too small, it is impossible to form a lane.

[0132] In this embodiment, the unique number can be a letter, a number, a symbol, or a combination of any two or all of the three. The relationship between lanes can be reflected by the unique number of the lane centerline. If the unique number is a number, the unique number of a lane can be 0, the adjacent lane on the left can be -1, and the adjacent lane on the right can be 1. According to the numbering, it can be known that lanes numbered 2 and 3 are adjacent, and lanes numbered 1 and 3 are separated by a lane.

[0133] The adjacent relationship of lanes is determined by this embodiment, so that whether lanes are adjacent and how many lanes are separated from each other can be determined based on the unique numbers, and data support can be provided for lane changes of vehicles in the road network.

[0134] To determine the predecessor and successor relationship, the above-mentioned driving trajectory points can be traversed to determine the lane where each of the above-mentioned driving trajectory points is located; each trajectory point is marked with the above-mentioned unique number of the lane where it is located; when the unique numbers of two adjacent trajectory points indicate that the lanes are not adjacent, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship, or when the unique numbers of two adjacent trajectory points are the same, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship.

[0135] In this embodiment, the predecessor and successor relationship is used to determine whether the lanes are still continuous after passing through an intersection or turning. For example, if the three lanes before the intersection are merged into two lanes after passing through an intersection, the predecessor and successor relationship of some lanes has changed and no longer exists, while the predecessor and successor relationship of some lanes exists.

[0136] In order to determine whether the successor relationship of the lanes has changed, each track point in the driving trajectory points can be marked with the unique number of the lane in which it is located. Then, each track point in the driving trajectory points corresponds to a unique number, and as the first track point in the driving trajectory points is traversed backward, if the unique number of the next track point changes, it means that the lane is different from the lane where the previous track point is located. It means that the car has changed lanes. If the lane is changed, it means that the lane where the next track point is located does not have a successor relationship with the lane where the previous track point is located. In other words, in this embodiment, whether the lanes have a successor relationship can be determined based on whether the unique number has changed.

[0137] For other examples of this embodiment, please refer to the above method examples, which will not be repeated here.

[0138] A car is also provided in this embodiment, including: a controller, for determining lane lines in the parking lot according to the semantic point cloud map data and driving trajectory points of the parking lot, wherein the lane lines have directions; determining lanes in the parking lot according to the lane lines; determining the adjacent relationship and the successor relationship between different lanes to obtain a road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0139] The car in this embodiment can be used to obtain semantic point cloud map data and driving trajectory points of a parking lot, and then process the semantic point cloud map data and driving trajectory points of the parking lot to generate a road network. In another example, the above-mentioned car can obtain the semantic point cloud map data and driving trajectory points of the parking lot obtained by other cars and devices, and then generate a road network.

[0140] In this embodiment, the lane lines in the parking lot can be determined based on the semantic point cloud map data and the driving trajectory points. The lane lines in the parking lot can be lines drawn on the road surface of the parking lot, which are generally white or yellow, but can also be other colors, and can be straight lines or curves, and can be solid lines or intermittent dashed lines. In some cases, a lane line is composed of two lines. For the case of two lines, in one example, the road network generation process can be performed as one line. Or, in another example, the road network generation process is performed as two lines. If the road network is generated as one line, the center lines of the two lines can be used instead of the two lines. If the road network is generated as two lines, one of the two lines forms a lane with the other adjacent lines, and the other line forms another lane with the other adjacent lines. The gap between the two lines is too small, so if it is identified as a lane, it is actively corrected. For example, Figure 2 As shown in the figure, a lane line includes two lines, where 201 and 202 are two lane lines respectively, and 203 is two lines of a lane line. There is a gap between the two lines. If the gap between the two lines is identified as a lane, error correction is performed. Lane line 201 and the left line of lane line 203 form a lane, and lane line 202 and the right line of lane line 203 form a lane. The gap between the two lines of lane 203 is not used as a lane after correction.

[0141] Generally speaking, there is a lane line on each side of a lane (except for the case where a lane line consists of two lines as mentioned above), and two adjacent lanes share a lane line (if the lane line consists of two lines as mentioned above, the two adjacent lanes each occupy one line of a lane line).

[0142] After the lane lines are determined, the lanes can be determined based on the lane lines. Since a lane line and adjacent lane lines can form a lane, and the lane lines are obtained through point cloud data, the point cloud data of the lane lines can be marked. The lane lines can be used to determine how many lanes there are, whether the lanes are adjacent to each other, and whether the successor relationship is continuous.

[0143] In the process of determining the road network of the parking lot, the present application can determine the lane lines according to the semantic point cloud map data and the driving trajectory points after obtaining the semantic point cloud map data and the driving trajectory points, so as to determine the lanes according to the lane lines, and then determine the adjacent relationship and the successor relationship between the lanes, so that a road network including the successor relationship and the adjacent relationship of the lanes can be established according to the lane lines and the driving trajectory points, thereby improving the efficiency of generating the road network, and thus solving the problem of low efficiency of road network generation in the prior art.

[0144] In one example, if Figure 3As shown, the step S101 of determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points may include:

[0145] S301, obtaining a lane line point cloud of the lane line of the parking lot from the semantic point cloud map;

[0146] S302, connecting the lane line point clouds into lane lines without directions;

[0147] S303, determining the moving direction of the movable surveying and mapping device that generates the driving trajectory points according to the driving trajectory points;

[0148] S304, determining the direction of the lane line according to the above moving direction, and obtaining a lane line with a direction.

[0149] In this embodiment, the lane lines in the parking lot can be determined based on the acquired semantic point cloud map data and driving trajectory points. When acquiring the semantic point cloud map data and driving trajectory points, the semantic point cloud data in the parking lot can be collected based on the car's sensors such as cameras and lidars. For driving trajectory points, the positions of the vehicles can be collected as trajectory points at predetermined intervals during the driving of the vehicle, and the overlapping or too close trajectory points can be deleted, and the non-overlapping trajectory points can be retained. Since the car has a direction when driving, the driving trajectory points have a pointing direction.

[0150] After obtaining the semantic point cloud map data and the driving trajectory points, the lane line point cloud of the lane line can be obtained from the semantic point cloud map data. That is, the redundant point clouds in the semantic point cloud map data are filtered out, and only the lane line point cloud of the lane line is retained. In another example, when the car collects semantic point cloud map data, only the lane point cloud can be collected. For example, the camera collects the image of the lane line on the ground, and the point cloud data generation algorithm is used to generate the point cloud data based on the image. For example, when the car is driving or stopped, the position of the car is determined, and then the lane line on the ground is photographed by the camera, the shooting angle of the camera is determined, and the position of the lane line in the scene is determined according to the position of the car and the shooting angle of the camera, as well as the image of the lane line taken by the camera, thereby obtaining the point cloud data of the lane line.

[0151] After obtaining the point cloud data of the lane, the point cloud data of the lane is represented by multiple points, such as Figure 4 As an example, if it is a solid line, it is shown as a1, and if it is a dotted line, it is shown as b1. If the lane line is two lines, the solid line is shown as c1, and the dotted line is shown as d1. In this embodiment, the point cloud of the lane line can be abstracted into a line, as shown from a2 to d2.

[0152] Since lane lines do not have directions, the direction of the lane lines can be determined in combination with driving trajectory points. Driving trajectory points are generated by the mobile surveying and mapping device in chronological order during movement, so the driving direction can be determined. The direction of the lane lines is determined based on the driving direction, thereby determining the direction of the lane lines.

[0153] In this embodiment, the lane line is determined by the above method. Not only can the position of the lane line be determined through point cloud data, but the direction of the lane line can also be determined based on the driving trajectory points, thereby determining the lane line with direction and improving the accuracy of the lane line.

[0154] In one example, if Figure 5 As shown, for step S102, determining the lane in the parking lot according to the lane line may include:

[0155] S501, starting from a first lane line among the lane lines, traversing an adjacent lane line of the first lane line to obtain a second lane line;

[0156] S502, determining a lane centerline according to the first lane line and the second lane line;

[0157] S503: Determine the lane center line, the first lane line, and the second lane line as lines on a lane to determine a lane, wherein two adjacent lanes share a lane line.

[0158] In this embodiment, since the lane lines in the parking lot have been determined in the above process, the lanes can be determined based on the lane lines. For lane determination, it is possible to start from one lane line and traverse adjacent lane lines. For example, Figure 6 As shown ( Figure 6 , lane lines are all represented by solid lines), a lane line 601 is randomly selected, and lane lines are traversed from the lane line to the left and right. The adjacent lane lines traversed form a lane with the current lane line. If the adjacent lane lines are traversed continuously, a lane is traversed continuously. For each traversed lane, the lane center line 602 is determined, such as Figure 6 As shown in the dashed line 602, each lane corresponds to a lane centerline, and vice versa, a lane centerline also corresponds to a lane.

[0159] By identifying lanes using the method of this embodiment, the center line of each lane can be determined. After the road network is constructed, the driving of cars in the road network can be controlled based on the center lines of the lanes, so that the cars stay at the center lines of the lanes while driving and do not deviate from the lanes.

[0160] In one example, after the lanes are identified, step S103 may be executed to determine the adjacent relationship and the preceding and succeeding relationship of the lanes.

[0161] For determining the adjacent relationship, a unique number can be assigned to the identified lanes or lane centerlines. For example, starting from a lane centerline, traverse the adjacent lane centerlines; determine the lane where the traversed adjacent lane centerline is located as the lane adjacent to the current lane; assign a unique number to each lane to record the adjacent relationship between lanes.

[0162] The lane centerline can be assigned a unique number to record the lane centerline and the corresponding lane. If the lane line includes two lines, the gap between the two lines may be determined as a lane, such as Figure 2 The gap between lane lines 203 in the image is shown in FIG. At this time, error correction is performed, and it can be determined whether a lane is formed according to the distance between the lane lines. If the distance is too small, it is impossible to form a lane.

[0163] In this embodiment, the unique number can be a letter, a number, a symbol, or a combination of any two or all of the three. The relationship between lanes can be reflected by the unique number of the lane centerline. If the unique number is a number, the unique number of a lane can be 0, the adjacent lane on the left can be -1, and the adjacent lane on the right can be 1. According to the numbering, it can be known that lanes numbered 2 and 3 are adjacent, and lanes numbered 1 and 3 are separated by a lane.

[0164] The adjacent relationship of lanes is determined by this embodiment, so that whether lanes are adjacent and how many lanes are separated from each other can be determined based on the unique numbers, and data support can be provided for lane changes of vehicles in the road network.

[0165] To determine the predecessor and successor relationship, the above-mentioned driving trajectory points can be traversed to determine the lane where each of the above-mentioned driving trajectory points is located; each trajectory point is marked with the above-mentioned unique number of the lane where it is located; when the unique numbers of two adjacent trajectory points indicate that the lanes are not adjacent, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship, or when the unique numbers of two adjacent trajectory points are the same, it is determined that the lanes where the two adjacent trajectory points are located have a predecessor and successor relationship.

[0166] In this embodiment, the predecessor and successor relationship is used to determine whether the lanes are still continuous after passing through an intersection or turning. For example, if the three lanes before the intersection are merged into two lanes after passing through an intersection, the predecessor and successor relationship of some lanes has changed and no longer exists, while the predecessor and successor relationship of some lanes exists.

[0167] In order to determine whether the successor relationship of the lanes has changed, each track point in the driving trajectory points can be marked with the unique number of the lane in which it is located. Then, each track point in the driving trajectory points corresponds to a unique number, and as the first track point in the driving trajectory points is traversed backward, if the unique number of the next track point changes, it means that the lane is different from the lane where the previous track point is located. It means that the car has changed lanes. If the lane is changed, it means that the lane where the next track point is located does not have a successor relationship with the lane where the previous track point is located. In other words, in this embodiment, whether the lanes have a successor relationship can be determined based on whether the unique number has changed.

[0168] For other examples of this embodiment, please refer to the above method examples, which will not be repeated here.

[0169] like Fig. 9 As shown, in an embodiment of the present application, a parking lot road network generation device is provided, including a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114.

[0170] Memory 113, used for storing computer programs;

[0171] In one embodiment of the present application, the processor 111, when executing the program stored in the memory 113, implements the method for generating a road network of a parking lot provided in any of the aforementioned method embodiments, including: when the semantic point cloud map data and driving trajectory points of the parking lot are obtained, determining the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determining the lanes in the parking lot according to the lane lines; determining the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

[0172] A computer-readable storage medium is also provided in an embodiment of the present application, on which a computer program is stored. When the above computer program is executed by a processor, the steps of the method for generating a road network of a parking lot provided in any of the above method embodiments are implemented.

[0173] The device embodiments described above are illustrative, wherein the units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0174] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.

[0175] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0176] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A method for generating a road network of a parking lot, characterized in that: include: When the semantic point cloud map data and the driving trajectory points of the parking lot are obtained, the lane lines in the parking lot are determined according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; Determining a lane in the parking lot according to the lane line; The adjacent relationship and the successor relationship between different lanes are determined to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

2. The method according to claim 1, characterized in that The determining of the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points comprises: Acquire a lane line point cloud of the lane line of the parking lot from the semantic point cloud map; Connecting the lane line point clouds into lane lines without directions; Determine the moving direction of the movable surveying and mapping device that generates the driving trajectory points according to the driving trajectory points; The direction of the lane line is determined according to the moving direction to obtain a lane line with a direction.

3. The method according to claim 1, characterized in that Determining the lane in the parking lot according to the lane line includes: Starting from a first lane line among the lane lines, traversing an adjacent lane line of the first lane line to obtain a second lane line; Determine a lane centerline according to the first lane line and the second lane line; The lane center line, the first lane line, and the second lane line are determined as lines on a lane to determine a lane, wherein two adjacent lanes share a lane line.

4. The method according to claim 3, characterized in that Determining the adjacent relationship between different lanes includes: Starting from the center line of one lane, traverse the center lines of the adjacent lanes; The lane where the center line of the traversed adjacent lane is located is determined as the lane adjacent to the current lane; A unique number is assigned to each lane to record the adjacent relationship between lanes.

5. The method according to claim 4, characterized in that Determining the successor relationship between different lanes includes: traversing the driving trajectory points to determine the lane where each of the driving trajectory points is located; Mark each track point with the unique number of the lane where it is located; When the unique numbers of two adjacent trajectory points indicate that the lanes are not adjacent, it is determined that the lanes where the two adjacent trajectory points are located have a successor relationship, or when the unique numbers of the two adjacent trajectory points are the same, it is determined that the lanes where the two adjacent trajectory points are located have a successor relationship.

6. A parking lot road network generation device, characterized in that: include: A first determination module is used to determine the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points when the semantic point cloud map data and the driving trajectory points of the parking lot are acquired, wherein the lane lines have directions; A second determination module, configured to determine a lane in the parking lot according to the lane line; The third determination module is used to determine the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

7. A parking lot road network generation system, characterized in that: include: A memory for storing computer executable instructions; The controller is configured to perform the following steps when executing the computer executable instructions: when the semantic point cloud map data and the driving trajectory points of the parking lot are acquired, determine the lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determine the lanes in the parking lot according to the lane lines; Determine the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous; Cloud server, used to provide cloud services.

8. A car, characterized in that: include: A controller is used to, upon obtaining semantic point cloud map data and driving trajectory points of a parking lot, determine lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determine lanes in the parking lot according to the lane lines; and determine adjacent relationships and successor relationships between different lanes to obtain a road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

9. A parking lot road network generation device, characterized in that: It includes at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; and at least one memory connected to the at least one bus, wherein the processor is configured to: upon obtaining semantic point cloud map data and driving trajectory points of the parking lot, determine lane lines in the parking lot according to the semantic point cloud map data and the driving trajectory points, wherein the lane lines have directions; determine lanes in the parking lot according to the lane lines; determine the adjacent relationship and the successor relationship between different lanes to obtain the road network of the parking lot, wherein the successor relationship is used to indicate whether the lanes are continuous.

10. A computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the method for generating a parking lot road network as described in any one of claims 1 to 5 of the present application.

Images