ADAS point location information thinning method and device, electronic equipment and storage medium

Abstract

The invention relates to an ADAS point location information thinning method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining initial ADAS point location information obtained by carrying out ADAS point location collection on a road; performing preliminary thinning on the initial ADAS point location information to obtain candidate ADAS point location information; the candidate ADAS point location information is thinned to obtain target ADAS point location information, and the target ADAS point location information is used for generating road topological data corresponding to roads in the vehicle-mounted high-precision map. According to the scheme, after the collected initial ADAS point location information is subjected to preliminary thinning, further thinning is carried out, the target ADAS point location information subjected to two-time thinning is obtained to generate the road topology data of the high-precision map, the data size of the ADAS point location information is further compressed through two-time thinning, the storage size of the high-precision map is reduced, and the data storage efficiency of the high-precision map is improved. And the occupation of the high-precision map on the storage space of the vehicle domain controller is reduced.

Description

Technical Field

[0001] The present disclosure relates to the field of intelligent driving technology, and in particular, to a method, apparatus, electronic device, and storage medium for thinning ADAS point information. Background Art

[0002] As an important part of high-level intelligent driving technology, high-precision maps (hereinafter collectively referred to as high-precision maps) have become a hot research topic. In the production process of high-precision maps, a high-precision map collection vehicle first needs to drive on the road to collect Advanced Driver Assistance System (ADAS) point information of the road. After the collection is completed, the collected ADAS point information is uploaded to the cloud, and the cloud will merge and cut the ADAS point information to generate road topology data in the high-precision map.

[0003] Generally, the collection points of the high-precision map collection vehicle on the road are too dense. In order to save space, when map manufacturers produce in-vehicle high-precision maps in the cloud, they will thin the collected ADAS point information and use the thinned data to construct the high-precision map.

[0004] Currently, in related technologies, the ADAS point information of a road is thinned by thinning at a certain interval. However, this thinning method is not thorough enough, and the data volume of the thinned ADAS point information is still large, resulting in the in-vehicle high-precision map still occupying a large storage space on the vehicle domain controller, seriously encroaching on the available space of other in-vehicle application programs. Summary of the Invention

[0005] In order to solve the above technical problems or at least partially solve the above technical problems, at least one embodiment of the present disclosure provides a method, apparatus, electronic device, and storage medium for thinning ADAS point information.

[0006] In a first aspect, the present disclosure provides a method for thinning ADAS point information, including:

[0007] Obtaining initial ADAS point information obtained by collecting ADAS points on a road;

[0008] Performing preliminary thinning on the initial ADAS point information to obtain candidate ADAS point information;

[0009] Performing thinning on the candidate ADAS point information to obtain target ADAS point information, where the target ADAS point information is used to generate road topology data corresponding to the road in the in-vehicle high-precision map.

[0010] In a second aspect, the present disclosure provides a device for thinning ADAS point information, including:

[0011] An acquisition module, configured to acquire initial ADAS point information obtained by collecting ADAS points on a road;

[0012] A first thinning module, configured to perform preliminary thinning on the initial ADAS point information to obtain candidate ADAS point information;

[0013] A second thinning module, configured to thin the candidate ADAS point information to obtain target ADAS point information, where the target ADAS point information is used to generate road topology data corresponding to the road in a vehicle-mounted high-precision map.

[0014] In a third aspect, the present disclosure provides an electronic device, including: a processor and a memory;

[0015] The processor is configured to execute the thinning method of the ADAS point information as described in the first aspect by calling a program or instruction stored in the memory.

[0016] In a fourth aspect, the present disclosure provides a computer-readable storage medium, where the computer-readable storage medium stores a program or instruction, and the program or instruction causes a computer to execute the thinning method of the ADAS point information as described in the first aspect.

[0017] In a fifth aspect, the present disclosure provides a computer program product, where the computer program product is configured to execute the thinning method of the ADAS point information as described in the first aspect.

[0018] The technical solution provided by the embodiments of the present disclosure has at least the following advantages compared with the prior art:

[0019] In the embodiments of the present disclosure, initial ADAS point information obtained by collecting ADAS points on a road is acquired. First, the initial ADAS point information is preliminarily thinned to obtain candidate ADAS point information, and then the candidate ADAS point information is thinned to obtain target ADAS point information, where the target ADAS point information is used to generate road topology data corresponding to the road in a vehicle-mounted high-precision map. By adopting the above technical solution, after preliminarily thinning the collected initial ADAS point information, further thinning is performed to obtain the target ADAS point information after two times of thinning to generate the road topology data of the high-precision map. Through two times of thinning, the data volume of the ADAS point information is further compressed, thereby reducing the storage size of the high-precision map, reducing the occupation of the storage space of the vehicle domain controller by the high-precision map, and also reducing the data volume of the ADAS point information that needs to be loaded when the vehicle-end intelligent driving-related functional program loads the high-precision map, thereby helping to improve the performance and efficiency of the vehicle-end intelligent driving-related functional program using the high-precision map. Description of the Drawings

[0020] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure.

[0021] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or in the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0022] Figure 1 Schematic flow diagram of the thinning method for ADAS point position information provided by an exemplary embodiment of the present disclosure;

[0023] Figure 2 Schematic flow diagram of the thinning method for ADAS point position information provided by another exemplary embodiment of the present disclosure;

[0024] Figure 3 Schematic flow diagram of the thinning method for ADAS point position information provided by yet another exemplary embodiment of the present disclosure;

[0025] Figure 4 Schematic structural diagram of the thinning device for ADAS point position information provided by an embodiment of the present disclosure. Detailed implementation manners

[0026] In order to more clearly understand the above objects, features, and advantages of the present disclosure, the following will further elaborate on the present disclosure in conjunction with the accompanying drawings and embodiments. It can be understood that the described embodiments are part of the embodiments of the present disclosure, rather than all of the embodiments. The specific embodiments described herein are only used to explain the present disclosure, rather than limiting the present disclosure. Without conflict, the embodiments of the present disclosure and the features in the embodiments can be combined with each other. Based on the described embodiments of the present disclosure, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present disclosure.

[0027] Many specific details are set forth in the following description in order to fully understand the present disclosure, but the present disclosure can also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0028] With the continuous development of technology, intelligent driving of automobiles has increasingly become the key R & D direction of major automobile manufacturers. How to pursue more advanced (L3+) intelligent driving functions is the direction that major automobile manufacturers unanimously pursue. According to industry consensus and relevant information, high-precision maps have become an important part of high-level intelligent driving technology.

[0029] In the production process of high-precision maps, high-precision map collection vehicles first hit the road to collect road information. After the collection is completed, the collected data is uploaded to the cloud, and the cloud will merge and cut the collected data to generate road topology data. The most important data in the road topology is the ADAS point information. Among them, the ADAS point information is the collection point information of the high-precision map collection vehicle on the road. The ADAS point information mainly includes: collection point coordinates, collection point offset, collection point heading, collection point cross slope, collection point longitudinal slope, and collection point curvature. After the produced high-precision map is sent to the vehicle terminal for use, in-vehicle electronic horizon programs such as the Electronic Horizon Provider (EHP) and the Electronic Horizon Reconstructor (EHR) will load the ADAS point data of the road topology from the in-vehicle high-precision map, and then send it to in-vehicle intelligent driving related function programs such as Navigate on Autopilot (NOA). Finally, the in-vehicle intelligent driving related function programs will accurately control the vehicle steering, speed increase and decrease, and vehicle body attitude adjustment according to the data contained in the ADAS point information of the high-precision map.

[0030] Generally, the collection points of high-precision map collection vehicles on the road are too dense. To save space, almost all map manufacturers will thin out the collection points when making in-vehicle high-precision maps in the cloud. Generally, the ADAS point information is thinned out at a certain interval. Usually, two sets of thinning interval standards are adopted for straight roads and curves. The thinning interval for straight roads is larger, and the thinning interval for curves is smaller. The thinning interval standards of each map manufacturer are usually different.

[0031] However, by using the method of thinning out at intervals, the data volume of the thinned ADAS point information is still large, resulting in a large space occupied by the high-precision map data file sent to the vehicle on the vehicle domain controller, seriously occupying the available space of other in-vehicle application programs.

[0032] In the process of researching high-precision maps, the inventors of the present application found that, considering the usage effect of the ADAS point information in the high-precision map by the in-vehicle intelligent driving related functional programs, on a straight and flat road section, the intelligent driving functional programs can also complete intelligent driving with fewer ADAS point information in the high-precision map, and its performance is exactly the same as that when using all the ADAS point information on the road. Therefore, there is still room for further thinning of the ADAS point information thinned according to the thinning interval. In addition, the in-vehicle electronic horizon programs (EHP, EHR) and the intelligent driving related functional programs will also have lower loads and power consumptions when loading and using fewer ADAS point information in the high-precision map. Based on this, the present disclosure provides a method for thinning ADAS point information. After thinning the collected ADAS point information according to the interval, further thinning is performed, which can further compress and reduce the storage size of the high-precision map, and solve the problem that the high-precision map data occupies a large storage space on the vehicle domain controller; and, by reducing the data volume of the ADAS point information of the roads in the high-precision map, the vehicle-end intelligent driving related functional programs only need to load less high-precision map data, thereby reducing the loads and power consumptions of the vehicle-end intelligent driving related functional programs when using the high-precision map, and improving the performance and efficiency of the vehicle-end intelligent driving related functional programs in using the high-precision map.

[0033] The following will explain in detail the specific implementation manners of the method, device, electronic device, and storage medium for thinning ADAS point information according to the present disclosure with reference to the accompanying drawings.

[0034] Figure 1 FIG. is a schematic flowchart of a method for thinning ADAS point information provided by an exemplary embodiment of the present disclosure. This method can be executed by the ADAS point information thinning device provided by the embodiments of the present disclosure. The ADAS point information thinning device can be implemented by software and / or hardware, and can be integrated in an electronic device.

[0035] As Figure 1 shown, the method for thinning ADAS point information may include the following steps:

[0036] Step 101, obtain the initial ADAS point information obtained by collecting ADAS points on a road.

[0037] Among them, the initial ADAS point information is obtained by the high-precision map collection vehicle driving on the road. In the high-precision map data, the most important data is the ADAS point information. The ADAS point is the collection point of the high-precision map collection vehicle on the road. One collection point corresponds to one ADAS point information. One ADAS point information mainly includes: collection point coordinates, collection point offset, collection point heading, collection point cross slope, collection point longitudinal slope, and collection point curvature. These 6 data are a group.

[0038] In this embodiment, for any road, multiple ADAS point information is usually collected. All the ADAS point information of a road collected is the initial ADAS point information of that road. The initial ADAS point information of a road is arranged in the order of the collection sequence of each point. That is, the first ADAS point information is the information of the point earliest collected on that road, corresponding to the starting point of that road, and the last ADAS point information is the information of the last point collected on that road, corresponding to the ending point of that road. Since making a high-precision map requires data of many roads, collecting ADAS point information for many roads will generate a large amount of data. Therefore, it is necessary to thin out the initial ADAS point information of the road to reduce the data volume.

[0039] Step 102: Thin out the initial ADAS point information preliminarily to obtain candidate ADAS point information.

[0040] Exemplarily, the initial ADAS point information of a road can be thinned out preliminarily according to a preset thinning interval. The ADAS point information obtained after thinning is called candidate ADAS point information. Among them, the thinning interval can be set in advance according to actual needs. For example, the thinning interval is set to 1 ADAS point, that is, one ADAS point information is extracted every 1 ADAS point information. Thus, compared with the initial ADAS point information, the data volume of the obtained candidate ADAS point information is reduced by half.

[0041] It should be noted that in this embodiment, when thinning out the initial ADAS point information preliminarily, the candidate ADAS point information obtained should include the ADAS point information corresponding to the starting point and the ending point of the road.

[0042] Step 103: Thin out the candidate ADAS point information to obtain target ADAS point information, and the target ADAS point information is used to generate the road topology data corresponding to the road in the in-vehicle high-precision map.

[0043] In this embodiment, after obtaining the candidate ADAS point information by preliminary thinning, the candidate ADAS point information can be further thinned out to obtain the target ADAS point information. For example, thinning can be performed according to the specific data differences of each ADAS point information in the candidate ADAS point information. Another example is that the candidate ADAS point information can be thinned out again according to a preset thinning interval. Compared with the candidate ADAS point information, the data volume of the thinned-out target ADAS point information is further reduced.

[0044] The method for thinning ADAS point information in the embodiments of the present disclosure obtains the initial ADAS point information collected by performing ADAS point collection on a road. First, the initial ADAS point information is preliminarily thinned to obtain candidate ADAS point information, and then the candidate ADAS point information is thinned to obtain target ADAS point information. The target ADAS point information is used to generate the road topology data corresponding to the road in the in-vehicle high-precision map. By adopting the above technical solution, after preliminarily thinning the collected initial ADAS point information and then performing further thinning, the target ADAS point information after two thinnings is obtained to generate the road topology data of the high-precision map. Through the two thinnings, the data volume of the ADAS point information is further compressed, thereby reducing the storage size of the high-precision map, reducing the occupation of the storage space of the vehicle domain controller by the high-precision map, and also reducing the data volume of the ADAS point information that needs to be loaded when the in-vehicle intelligent driving related functional programs load the high-precision map, thereby helping to improve the performance and efficiency of the in-vehicle intelligent driving related functional programs using the high-precision map.

[0045] Generally, in a high-precision map, an ADAS point information includes the position information and road condition information corresponding to the collection point. The position information includes the position coordinates and the offset. The road condition information includes the heading, cross slope, longitudinal slope, and curvature. Among them, the position coordinates of the collection point represent the reference point for the vehicle during intelligent driving on the road, and the offset of the collection point represents the position offset distance of the collection point coordinates relative to the coordinates of the road / lane starting point. Specifically, the offset can be the sum of the distance between the current collection point and the adjacent previous collection point and the offset of the previous collection point. The main functions of these two data, the position coordinates and the offset, are to provide a reference for the driving trajectory to the intelligent driving function. Whether thinned or not, these two data of different ADAS point information are inconsistent. The four data, the heading, cross slope, longitudinal slope, and curvature of the collection point, mainly function to assist in controlling the vehicle steering, speed increase and decrease, and vehicle body attitude adjustment in the intelligent driving function. On a straight and flat section of the road, the vehicle intelligent driving function only needs to keep the current vehicle body attitude and drive straight. The number of ADAS point information does not affect the performance of the intelligent driving function on a straight and flat section of the road. Therefore, for a road including a straight and flat section, it can be further thinned according to the road condition information of its ADAS point information. Thus, in an optional embodiment of the present disclosure, the candidate ADAS point information can be thinned according to the road condition information to obtain the target ADAS point information, where the target ADAS point information includes the first ADAS point information and the last ADAS point information in the candidate ADAS point information, that is, the first and last points in the candidate ADAS point information will not be deleted during thinning to ensure the topological continuity between roads in the high-precision map.

[0046] In the embodiments of the present disclosure, when determining whether a road section is straight and has no undulations, it can be determined based on the road condition information in the ADAS point information. Specifically, it can be compared whether the road condition information (heading, cross slope, longitudinal slope, curvature) in adjacent ADAS point information is exactly the same. If it is exactly the same, then it can be determined that the adjacent ADAS points are connected to form a straight and non-undulating road section, and further thinning can be performed on these ADAS point information. Thus, in an alternative embodiment of the present disclosure, as Figure 2 shown, based on the above embodiments, the step of thinning the candidate ADAS point information according to the road condition information to obtain the target ADAS point information may include the following sub-steps:

[0047] Step 201, determine a first set and a second set from the candidate ADAS point information, where the ADAS point information in the first set needs to be thinned, and the ADAS point information in the second set does not need to be thinned.

[0048] In this embodiment, for the obtained candidate ADAS point information, it can be divided into two sets. The ADAS point information in one set (for the convenience of description and distinction, called the first set) needs to be further thinned, and the ADAS point information in the other set (for the convenience of description and distinction, called the second set) is the data reserved for generating the road topology and does not need to be thinned again. To avoid the problem that the continuity between road topologies is difficult to guarantee due to the deletion of the first and last points, the first and last points in the candidate ADAS point information can be divided into the second set.

[0049] As an example, a small part of the ADAS point information located at both ends can be determined from the candidate ADAS point information to form the second set, and the remaining large part of the ADAS point information located in the middle part can be formed into the first set for further thinning.

[0050] As another example, the first ADAS point information can be selected from the candidate ADAS point information as the starting point information, and the last ADAS point information can be selected as the end point information. Then, based on the starting point information and the end point information, a second set is formed, and based on the remaining ADAS point information in the candidate ADAS point information except the starting point information and the end point information, a first set is formed. That is to say, the first ADAS point information and the last ADAS point information in the candidate ADAS point information form the second set without abstraction, and the remaining all ADAS point information forms the first set for further abstraction. Thus, it is ensured that the first and last points will not be deleted due to thinning, and the point information of these two points is retained for generating the road topology to ensure the topological continuity between roads in the high-precision map.

[0051] Step 202: Traverse the ADAS point information in the first set in the order of acquisition of each ADAS point information in the first set.

[0052] In this embodiment, for the first set, the ADAS point information in the first set can be traversed in the order of acquisition of each ADAS point information in the first set. That is to say, start traversing from the first ADAS point information in the first set.

[0053] Step 203: Compare the currently traversed ADAS point information with the reference ADAS point information. Among them, at the first comparison, the reference ADAS point information is the first ADAS point information in the first set, and the currently traversed ADAS point information is the second ADAS point information in the first set.

[0054] In this embodiment, when starting to traverse the ADAS point information in the first set, the first ADAS point information in the first set is determined as the reference ADAS point information, and the second ADAS point information in the first set is determined as the currently traversed ADAS point information. Compare the currently traversed ADAS point information with the reference ADAS point information, and select to delete or retain the currently traversed ADAS point information according to the comparison result.

[0055] Step 204: In the case where the road condition information of the currently traversed ADAS point information is the same as the road condition information of the reference ADAS point information, delete the currently traversed ADAS point information, and use the next ADAS point information adjacent to the currently traversed ADAS point information as the new currently traversed ADAS point information to compare with the reference ADAS point information.

[0056] In this embodiment, the currently traversed ADAS point information is compared with the reference ADAS point information. If the road condition information of the currently traversed ADAS point information is the same as that of the reference ADAS point information, that is, the heading, cross slope, longitudinal slope, and curvature of both are the same, then the currently traversed ADAS point information is deleted, and the traversal continues to the next ADAS point information adjacent to the currently traversed ADAS point information. The next ADAS point information is determined as the new currently traversed ADAS point information, and the new currently traversed ADAS point information is compared with the reference ADAD point information to determine whether their road condition information is the same.

[0057] Step 205, in the case where at least one of the road condition information in the currently traversed ADAS point information is inconsistent with the reference ADAS point information, retain the currently traversed ADAS point information, and use the currently traversed ADAS point information as the new reference ADAS point information. Use the next ADAS point information adjacent to the new reference ADAS point information as the new currently traversed ADAS point information and compare it with the new reference ADAS point information.

[0058] In this embodiment, the currently traversed ADAS point information is compared with the reference ADAS point information. If the road condition information of the currently traversed ADAS point information is inconsistent with the reference ADAS point information, that is, at least one of the heading, cross slope, longitudinal slope, and curvature of both is different, then retain the currently traversed ADAS point information, and use the currently traversed ADAS point information as the new reference ADAS point information. Use the next ADAS point information adjacent to the new reference ADAS point information as the new currently traversed ADAS point information, and then compare the new currently traversed ADAS point information with the new reference ADAS point information.

[0059] Step 206, when the traversal of the ADAS point information in the first set is completed, determine the target ADAS point information based on the ADAS point information retained in the first set and the ADAS point information in the second set.

[0060] In this embodiment, each ADAS point information is traversed in sequence according to the order of collection of each ADAS point information in the first set. The first ADAS point information in the first set is first used as the reference ADAS point information, and the second ADAS point information is used as the currently traversed ADAS point information. The road condition information of the two ADAS point information is compared to see whether they are completely consistent. If they are completely consistent, the currently traversed ADAS point information is deleted, and the next ADAS point information is obtained as the new currently traversed ADAS point information, while the reference ADAS point information is kept unchanged. If the road condition information of the two ADAS point information involved in the comparison is not completely consistent, the currently traversed ADAS point information is retained and determined as the new reference ADAS point information, and the next ADAS point information is obtained as the new currently traversed ADAS point information, and the road condition information of the currently traversed ADAS point information is compared again to see whether they are completely consistent. Repeat the above process until the currently traversed ADAS point information is the last ADAS point information in the first set. After comparing it with the current reference ADAS point information, delete or retain the currently traversed ADAS point information. At this point, all ADAS point information in the first set has been traversed, and the ADAS point information in the first set has been further thinned out to obtain the ADAS point information retained after the first set is thinned out. Finally, the ADAS point information retained after thinning out in the first set and the ADAS point information in the second set are determined as the target ADAS point information.

[0061] For a straight road without undulations, the road condition information of all ADAS point information corresponding to this road may be the same. In this case, only retaining the ADAS point information corresponding to the starting point and the ending point of this road can usually complete intelligent driving well. Therefore, in order to further reduce the data volume of ADAS point information, in an alternative implementation manner of the present disclosure, for the case where the first ADAS point information is selected from the candidate ADAS point information as the starting point information, and the last ADAS point information is selected as the ending point information, and the starting point information and the ending point information are used to form a second set, when determining the target ADAS point information based on the ADAS point information retained in the first set and the ADAS point information in the second set, the road condition information of the first ADAS point information in the ADAS point information retained in the first set (for the convenience of description and distinction, it is called the target road condition information) can be obtained first, and whether the obtained target road condition information is consistent with the road condition information of the starting point information in the second set. When the target road condition information is consistent with the road condition information of the starting point information in the second set (that is, their headings, cross slopes, longitudinal slopes, and curvatures are all the same), the first ADAS point information in the ADAS point information retained in the first set is deleted to obtain a third set; furthermore, the ADAS point information in the third set and the ADAS point information in the second set are determined as the target ADAS point information. Thus, further thinning is achieved, and the data volume of the ADAS point information used to generate the high-precision map is further reduced, thereby further reducing the storage size of the high-precision map. Additionally, optionally, the road condition information of the last ADAS point information in the ADAS point information retained in the first set can also be obtained and compared with the road condition information of the ending point information in the second set. If they are completely consistent, the last ADAS point information in the ADAS point information retained in the first set is deleted to further reduce the data volume of the ADAS point information used to generate the high-precision map.

[0062] For example, assume that ADAS points are collected for a road, and after initially thinning the collected initial ADAS point information, there are still 10 pieces of ADAS point information remaining. The specific data of these 10 ADAS points are shown in Table 1 below. Each ADAS point includes 6 attribute data, namely: coordinates, offset, heading, cross slope, longitudinal slope, and curvature. Among them, {x, y, z} is used to replace the real coordinate values. When thinning the ADAS point information of the road, the first and last two ADAS points need to be excluded because at least the first and last two points are required on a road to ensure the topological continuity between roads in the high-precision map. In Table 1, after excluding the first and last two ADAS points, the remaining 8 ADAS points can be thinned. The specific thinning logic is as follows: Determine whether the heading, cross slope, longitudinal slope, and curvature values between adjacent ADAS points are exactly the same. If they are the same, retain the first ADAS point of the adjacent ADAS points and delete the remaining ADAS points of the adjacent ADAS points; if they are different, retain the compared ADAS points. Traverse and process all ADAS points of a road according to the above thinning logic, and finally complete the further thinning of the ADAS point information on a road. After thinning the ADAS point information shown in Table 1, except for the first and last two ADAS point information (numbers 1 and 10), only the ADAS points numbered 2, 4, and 7 are retained, and the ADAS points numbered 3, 5, 6, 8, and 9 are deleted. The thinning effect is shown in Table 2. Finally, the ADAS point information retained after further thinning the original data (Table 1) in the high-precision map is shown in Table 3. Compared with the ADAS point information in Table 1, the number of ADAS points after further thinning is reduced by half, thereby reducing the occupied space size of the high-precision map data file.

[0063] Table 1

[0064] ADAS Point Number 1 2 3 4 5 Coordinate <![CDATA[{x1,y1,z1}]]> <![CDATA[{x2,y2,z2}]]> <![CDATA[{x3,y3,z3}]]> <![CDATA[{x4,y4,z4}]]> <![CDATA[{x5,y5,z5}]]> Offset 0 4502 8003 13008 19009 Course 184045 184045 184045 184045 184045 Cross Slope 16 12 12 16 16 Longitudinal Slope -2 -3 -3 -2 -2 Curvature 0 0 0 0 0 ADAS Point Number 6 7 8 9 10 Coordinate <![CDATA[{x6,y6,z6}]]> <![CDATA[{x7,y7,z7}]]> <![CDATA[{x8,y8,z8}]]> <![CDATA[{x9,y9,z9}]]> <![CDATA[{x 10 ,y 10 ,z 10}]]> Offset 24011 29013 33082 43582 63095 Course 184045 184045 184045 184045 184044 Cross Slope 16 18 18 18 18 Longitudinal Slope -2 -2 -2 -2 -2 Curvature 0 0 0 0 0

[0065] Table 2

[0066] ADAS Point Number 1 2 3 4 5 Coordinate <![CDATA[{x1,y1,z1}]]> <![CDATA[{x2,y2,z2}]]> <![CDATA[{x3,y3,z3}]]> <![CDATA[{x4,y4,z4}]]> <![CDATA[{x5,y5,z5}]]> Offset 0 4502 8003 13008 19009 Course 184045 184045 184045 184045 184045 Cross Slope 16 12 12 16 16 Longitudinal Slope -2 -3 -3 -2 -2 Curvature 0 0 0 0 0 Thinning Strategy Keep the First Point Keep Delete Keep Delete ADAS Point Number 6 7 8 9 10 Coordinate <![CDATA[{x6,y6,z6}]]> <![CDATA[{x7,y7,z7}]]> <![CDATA[{x8,y8,z8}]]> <![CDATA[{x9,y9,z9}]]> <![CDATA[{x 10 ,y 10 ,z 10}]]> Offset 24011 29013 33082 43582 63095 Course 184045 184045 184045 184045 184044 Cross Slope 16 18 18 18 18 Longitudinal Slope -2 -2 -2 -2 -2 Curvature 0 0 0 0 0 Thinning Strategy Delete Keep Delete Delete Keep the Last Point

[0067] Table 3

[0068] ADAS Point Number 1 2 4 7 10 Coordinate <![CDATA[{x1,y1,z1}]]> <![CDATA[{x2,y2,z2}]]> <![CDATA[{x4,y4,z4}]]> <![CDATA[{x7,y7,z7}]]> <![CDATA[{x 10 ,y 10 ,z 10}]]> Offset 0 4502 13008 29013 63095 Course 184045 184045 184045 184045 184044 Cross Slope 16 12 16 18 18 Longitudinal Slope -2 -3 -2 -2 -2 Curvature 0 0 0 0 0

[0069] The method for thinning the ADAS point information of this embodiment determines a first set that needs to be further thinned and a second set that does not need to be thinned from the candidate ADAS point information. According to the acquisition sequence of each ADAS point information in the first set, the ADAS point information in the first set is traversed in turn, and the currently traversed ADAS point information is compared with the reference ADAS point information. Among them, when comparing for the first time, the reference ADAS point information is the first ADAS point information in the first set, and the currently traversed ADAS point information is the second ADAS point information in the first set; when the road condition information of the currently traversed ADAS point information is the same as the road condition information of the reference ADAS point information, the currently traversed ADAS point information is deleted, and the next ADAS point information adjacent to the currently traversed ADAS point information is used as the new currently traversed ADAS point information to be compared with the reference ADAS point information; when at least one of the road condition information of the currently traversed ADAS point information is different from the reference ADAS point information, the currently traversed ADAS point information is retained, and the currently traversed ADAS point information is used as the new reference ADAS point information, and the next ADAS point information adjacent to the new reference ADAS point information is used as the new currently traversed ADAS point information to be compared with the new reference ADAS point information; when the ADAS point information in the first set is traversed, the target ADAS point information is determined based on the ADAS point information retained in the first set and the ADAS point information in the second set. Thus, it realizes further thinning of the candidate ADAS point information after preliminary thinning, and by traversing the ADAS point information one by one for thinning, it can completely delete the subsequent ADAS point information with the same road condition information as the previous ADAS point information, minimizing the number of retained ADAS point information to the greatest extent and reducing the storage size of the high-precision map.

[0070] In an alternative embodiment of the present disclosure, as Figure 3 shown, on the basis of the foregoing embodiment, the step of thinning the candidate ADAS point information according to the road condition information to obtain the target ADAS point information may include the following sub-steps:

[0071] Step 301, perform curve fitting based on the position coordinates of the candidate ADAS point information to obtain a road simulation curve.

[0072] Among them, a currently commonly used curve fitting algorithm or curve fitting tool can be used to perform curve fitting based on the position coordinates of the candidate ADAS point information to obtain a road simulation curve. For example, according to the position coordinates of each candidate ADAS point information, a road simulation curve can be fitted using the least squares fitting method.

[0073] Step 302: Determine the ADAS point information located outside the road simulation curve from the candidate ADAS point information as the first reserved ADAS point information, and determine the ADAS point information located on the road simulation curve as the ADAS point information to be thinned.

[0074] In this embodiment, after obtaining the road simulation curve, the ADAS point information located on the road simulation curve and the ADAS point information located outside the road simulation curve in the candidate ADAS point information can be screened out by using the road simulation curve. The ADAS point information outside the road simulation curve is determined as the first reserved ADAS point information, and these ADAS point information do not need to be further thinned; and the ADAS point information located on the road simulation curve is determined as the ADAS point information to be thinned, and these ADAS point information need to be further thinned.

[0075] Step 303: Thin the ADAS point information to be thinned based on the road condition information of the ADAS point information to be thinned, and obtain the second reserved ADAS point information.

[0076] In this embodiment, for the screened ADAS point information to be thinned, the ADAS point information to be thinned can be further thinned according to the road condition information of the ADAS point information to be thinned, and the reserved ADAS point information after thinning is obtained (for the convenience of description and distinction, it is called the second reserved ADAS point information).

[0077] As an example, it can be compared whether the road condition information of adjacent ADAS point information in the ADAS point information to be thinned is exactly the same. If it is exactly the same, the first of the adjacent ADAS point information is reserved, and the remaining ADAS point information in the adjacent ADAS point information is deleted to obtain the second reserved ADAS point information.

[0078] As another example, two adjacent ADAS point location information in the ADAS point location information to be decimated can be traversed, and the road condition information of these two ADAS point location information can be compared to determine whether the difference in at least one of the course, cross slope, longitudinal slope, and curvature between the two ADAS point location information is greater than a preset difference. Herein, the preset difference can be preset according to actual requirements. The course, cross slope, longitudinal slope, and curvature can correspond to the same preset difference, or can be respectively set with their own corresponding preset differences; in the case where the difference in at least one of the road condition information of the two ADAS point location information is greater than the preset difference, these two ADAS point location information are retained; in the case where the difference in each of the road condition information of the two ADAS point location information is not greater than the preset difference, the previous ADAS point location information of these two ADAS point location information is retained, and the latter ADAS point location information is deleted; the retained ADAS point location information is determined as the second retained ADAS point location information. That is to say, in this example, for any two adjacent ADAS point location information in the ADAS point location information to be decimated, if the difference in at least one of the course, cross slope, longitudinal slope, and curvature between these two ADAS point location information is greater than the preset difference, these two ADAS point location information are retained, otherwise the latter one of these two ADAS point location information is deleted, and the previous one is retained. Then, the next set of two adjacent ADAS point location information is traversed. In these two ADAS point location information, if the previous one was retained in the previous set of two adjacent ADAS point location information, the previous one in the currently traversed two ADAS point location information is still the previous one retained last time; if both were retained in the previous set of two adjacent ADAS point location information, the previous one in the currently traversed two ADAS point location information is the latter one retained last time.

[0079] For example, assume that there are 3 ADAS point location information in the ADAS point location information to be decimated, which are A, B, and C in sequence. First, compare the road condition information of A and B. If the difference in at least one of the course, cross slope, longitudinal slope, and curvature between these two point location information is greater than the preset difference, for example, the difference in the course between A and B is greater than the preset difference, then retain A and B, and next compare the road condition information of B and C. If the differences in the course, cross slope, longitudinal slope, and curvature between A and B are not greater than the preset difference, then retain A and delete B, and next compare the road condition information of A and C.

[0080] In this embodiment, by traversing two adjacent ADAS point information in the ADAS point information to be thinned, it is determined whether the difference in at least one of the course, cross slope, longitudinal slope, and curvature between the two ADAS point information is greater than a preset difference. When the difference is greater than the preset difference, the two ADAS point information are retained; otherwise, only the first ADAS point information of the two ADAS point information is retained, and the second ADAS point information is deleted. Thus, it is possible to filter out the point information with relatively small differences in road conditions between adjacent ADAS point information, and only retain the point information with relatively large differences in road conditions, achieving a greater degree of thinning and greatly reducing the amount of data of the ADAS point information used to generate the high-precision map.

[0081] Step 304: Determine the target ADAS point information based on the first retained ADAS point information and the second retained ADAS point information.

[0082] In this embodiment, after determining the first retained ADAS point information and the second retained ADAS point information, all the ADAS point information in the first retained ADAS point information and the second ADAS point information can be merged in the order of acquisition to obtain the target ADAS point information.

[0083] The ADAS point information thinning method of this embodiment obtains a road simulation curve by performing curve fitting based on the position coordinates of the candidate ADAS point information, determines the ADAS point information located outside the road simulation curve from the candidate ADAS point information as the first retained ADAS point information, and determines the ADAS point information located on the road simulation curve as the ADAS point information to be thinned. Furthermore, the ADAS point information to be thinned is thinned based on the road condition information of the ADAS point information to be thinned to obtain the second retained ADAS point information. Finally, the target ADAS point information is determined based on the first retained ADAS point information and the second retained ADAS point information. Thus, further thinning of the candidate ADAS point information is achieved, the amount of data of the ADAS point information used to generate the high-precision map is reduced, and thus the storage size of the high-precision map is reduced.

[0084] To implement the above embodiment, the present disclosure also provides an ADAS point information thinning device, which can be implemented by software and / or hardware and can be integrated in an electronic device.

[0085] Figure 4 As shown in the structural schematic diagram of the ADAS point information thinning device provided by an embodiment of the present disclosure, Figure 4 as shown, the ADAS point information thinning device 30 may include: an acquisition module 310, a first thinning module 320, and a second thinning module 330.

[0086] Among them, the acquisition module 310 is used to acquire the initial ADAS point information obtained by collecting ADAS points on the road.

[0087] The first thinning module 320 is used to initially thin the initial ADAS point information to obtain candidate ADAS point information.

[0088] The second thinning module 330 is used to thin the candidate ADAS point information to obtain target ADAS point information, and the target ADAS point information is used to generate the road topology data corresponding to the road in the vehicle-mounted high-precision map.

[0089] Optionally, the ADAS point information includes the position information and road condition information corresponding to the collection point. The position information includes position coordinates and offsets, and the road condition information includes heading, cross slope, longitudinal slope, and curvature. The second thinning module 330 includes:

[0090] The thinning sub-module is used to thin the candidate ADAS point information according to the road condition information to obtain target ADAS point information, where the target ADAS point information includes the first ADAS point information and the last ADAS point information in the candidate ADAS point information.

[0091] Further optionally, the thinning sub-module includes:

[0092] The first division unit is used to determine a first set and a second set from the candidate ADAS point information, where the ADAS point information in the first set needs to be thinned, and the ADAS point information in the second set does not need to be thinned.

[0093] The traversal unit is used to sequentially traverse the ADAS point information in the first set according to the acquisition sequence of each ADAS point information in the first set.

[0094] The comparison unit is used to compare the currently traversed ADAS point information with the reference ADAS point information. Among them, when comparing for the first time, the reference ADAS point information is the first ADAS point information in the first set, and the currently traversed ADAS point information is the second ADAS point information in the first set.

[0095] The first update unit is used to delete the currently traversed ADAS point information when the road condition information of the currently traversed ADAS point information is consistent with the road condition information of the reference ADAS point information, and use the next ADAS point information adjacent to the currently traversed ADAS point information as the new currently traversed ADAS point information to compare with the reference ADAS point information.

[0096] A second update unit, configured to, when at least one piece of road condition information in the currently traversed ADAS point information is inconsistent with the reference ADAS point information, retain the currently traversed ADAS point information, use the currently traversed ADAS point information as the new reference ADAS point information, and use the next ADAS point information adjacent to the new reference ADAS point information as the new currently traversed ADAS point information to compare with the new reference ADAS point information;

[0097] A first determination unit, configured to, when the traversal of the ADAS point information in the first set is completed, determine target ADAS point information based on the ADAS point information retained in the first set and the ADAS point information in the second set.

[0098] Optionally, the first partitioning unit is further configured to:

[0099] Screen out the first ADAS point information from the candidate ADAS point information as the starting point information, and screen out the last ADAS point information as the ending point information;

[0100] Based on the starting point information and the ending point information, form a second set;

[0101] Based on the remaining ADAS point information in the candidate ADAS point information except the starting point information and the ending point information, form a first set.

[0102] Optionally, the first determination unit is further configured to:

[0103] Obtain the target road condition information of the first ADAS point information in the ADAS point information retained in the first set;

[0104] When the target road condition information is consistent with the road condition information of the starting point information in the second set, delete the first ADAS point information in the ADAS point information retained in the first set to obtain a third set;

[0105] Determine the ADAS point information in the third set and the ADAS point information in the second set as the target ADAS point information.

[0106] Optionally, the thinning sub-module includes:

[0107] A curve fitting unit, configured to perform curve fitting based on the position coordinates of the candidate ADAS point information to obtain a road simulation curve;

[0108] A second partitioning unit, configured to determine, from the candidate ADAS point information, the ADAS point information located outside the road simulation curve as the first reserved ADAS point information, and determine the ADAS point information located on the road simulation curve as the ADAS point information to be thinned;

[0109] A thinning unit, configured to thin the ADAS point information to be thinned based on the road condition information of the ADAS point information to be thinned, so as to obtain the second reserved ADAS point information;

[0110] A second determination unit, configured to determine the target ADAS point information based on the first reserved ADAS point information and the second reserved ADAS point information.

[0111] Further optionally, the thinning unit is further configured to:

[0112] Traverse two adjacent ADAS point information in the ADAS point information to be thinned;

[0113] Compare the road condition information of the two ADAS point information, and determine whether the difference of at least one of the heading, cross slope, longitudinal slope and curvature between the two ADAS point information is greater than a preset difference;

[0114] When the difference of at least one of the road condition information of the two ADAS point information is greater than the preset difference, retain the two ADAS point information;

[0115] When the difference of each of the road condition information of the two ADAS point information is not greater than the preset difference, retain the previous ADAS point information of the two ADAS point information and delete the latter ADAS point information;

[0116] Determine the retained ADAS point information as the second reserved ADAS point information.

[0117] The ADAS point information thinning device applicable to an electronic device provided by an embodiment of the present disclosure can execute the ADAS point information thinning method provided by an embodiment of the present disclosure, and has corresponding functional modules and beneficial effects for executing the method. The content not described in detail in the device embodiment of the present disclosure can be referred to the description in any method embodiment of the present disclosure.

[0118] An embodiment of the present disclosure further provides an electronic device, including a processor and a memory; the processor is configured to execute the steps of each embodiment of the foregoing ADAS point information thinning method by calling a program or instruction stored in the memory. To avoid repeated description, it will not be elaborated here.

[0119] Embodiments of the present disclosure also provide a computer-readable storage medium. The computer-readable storage medium is non-transitory. The computer-readable storage medium stores a program or instructions, and the program or instructions cause a computer to execute the steps of the embodiments of the foregoing thinning method for ADAS point information. To avoid repetitive description, details are not elaborated herein.

[0120] Embodiments of the present disclosure also provide a computer program product, which is used to execute the steps of the embodiments of the foregoing thinning method for ADAS point information.

[0121] It should be noted that, in this document, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0122] The above are only specific embodiments of the present disclosure, enabling those skilled in the art to understand or implement the present disclosure. Various modifications to these embodiments will be obvious 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 disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for thinning ADAS point position information, characterized in that, The method includes: Obtaining initial ADAS point information obtained by collecting ADAS points on a road; Performing preliminary thinning on the initial ADAS point information to obtain candidate ADAS point information; Performing thinning on the candidate ADAS point information to obtain target ADAS point information, where the target ADAS point information is used to generate road topology data corresponding to the road in a vehicle-mounted high-precision map.

2. The method according to claim 1, characterized in that, The ADAS point information includes position information and road condition information corresponding to the collection point. The position information includes position coordinates and offsets, and the road condition information includes heading, cross slope, longitudinal slope, and curvature; The performing thinning on the candidate ADAS point information to obtain target ADAS point information includes: Performing thinning on the candidate ADAS point information according to the road condition information to obtain the target ADAS point information, where the target ADAS point information includes the first ADAS point information and the last ADAS point information in the candidate ADAS point information.

3. The method according to claim 2, wherein The performing thinning on the candidate ADAS point information according to the road condition information to obtain the target ADAS point information includes: Determining a first set and a second set from the candidate ADAS point information, where the ADAS point information in the first set needs to be thinned, and the ADAS point information in the second set does not need to be thinned; Traversing the ADAS point information in the first set in sequence according to the collection sequence of each ADAS point information in the first set; Comparing the currently traversed ADAS point information with reference ADAS point information. When comparing for the first time, the reference ADAS point information is the first ADAS point information in the first set, and the currently traversed ADAS point information is the second ADAS point information in the first set; When the road condition information of the currently traversed ADAS point information is the same as the road condition information of the reference ADAS point information, deleting the currently traversed ADAS point information, and using the next ADAS point information adjacent to the currently traversed ADAS point information as the new currently traversed ADAS point information to compare with the reference ADAS point information; When at least one of the road condition information of the currently traversed ADAS point information is different from the road condition information of the reference ADAS point information, retaining the currently traversed ADAS point information, using the currently traversed ADAS point information as the new reference ADAS point information, and using the next ADAS point information adjacent to the new reference ADAS point information as the new currently traversed ADAS point information to compare with the new reference ADAS point information; When the traversal of the ADAS point information in the first set is completed, determining the target ADAS point information based on the ADAS point information retained in the first set and the ADAS point information in the second set.

4. The method according to claim 3, wherein Determining a first set and a second set from the candidate ADAS point information includes: Screening out the first ADAS point information from the candidate ADAS point information as the starting point information, and screening out the last ADAS point information as the ending point information; Based on the starting point information and the ending point information, constituting the second set; Based on the remaining ADAS point information in the candidate ADAS point information except the starting point information and the ending point information, constituting the first set.

5. The method according to claim 4, wherein Determining the target ADAS point information based on the ADAS point information retained in the first set and the ADAS point information in the second set includes: Obtaining the target road condition information of the first ADAS point information in the ADAS point information retained in the first set; When the target road condition information is consistent with the road condition information of the starting point information in the second set, deleting the first ADAS point information in the ADAS point information retained in the first set to obtain a third set; Determining the ADAS point information in the third set and the ADAS point information in the second set as the target ADAS point information.

6. The method according to claim 2, wherein Thinning the candidate ADAS point information according to the road condition information to obtain the target ADAS point information, including: Performing curve fitting based on the position coordinates of the candidate ADAS point information to obtain a road simulation curve; Determining the ADAS point information located outside the road simulation curve from the candidate ADAS point information as the first retained ADAS point information, and determining the ADAS point information located on the road simulation curve as the ADAS point information to be thinned; Thinning the ADAS point information to be thinned based on the road condition information of the ADAS point information to be thinned to obtain the second retained ADAS point information; Based on the first retained ADAS point information and the second retained ADAS point information, determining the target ADAS point information.

7. The method according to claim 6, characterized in that, Thinning the ADAS point information to be thinned based on the road condition information of the ADAS point information to be thinned to obtain the second retained ADAS point information, including: Traversing two adjacent ADAS point information in the ADAS point information to be thinned; Comparing the road condition information of the two ADAS point information, and determining whether the difference in at least one of the heading, cross slope, longitudinal slope, and curvature between the two ADAS point information is greater than a preset difference; When the difference in at least one of the road condition information of the two ADAS point information is greater than the preset difference, retaining the two ADAS point information; When the difference in each of the road condition information of the two ADAS point information is not greater than the preset difference, retaining the previous ADAS point information of the two ADAS point information and deleting the latter ADAS point information; Determine the reserved ADAS point information as the second reserved ADAS point information.

8. A thinning device for ADAS point position information, characterized in that, It includes: An acquisition module, configured to acquire initial ADAS point information obtained by collecting ADAS points on a road; A first thinning module, configured to perform preliminary thinning on the initial ADAS point information to obtain candidate ADAS point information; A second thinning module, configured to thin the candidate ADAS point information to obtain target ADAS point information, where the target ADAS point information is used to generate road topology data corresponding to the road in a vehicle-mounted high-precision map.

9. An electronic device, characterized in that, It includes: A processor and a memory; The processor is configured to execute the ADAS point information thinning method according to any one of claims 1 to 7 by calling a program or instruction stored in the memory.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instruction, and the program or instruction causes a computer to execute the ADAS point information thinning method according to any one of claims 1 to 7.

Images