CN117789444A

CN117789444A - Parking lot data matching method, device, equipment, medium and vehicle

Info

Publication number: CN117789444A
Application number: CN202211135734.1A
Authority: CN
Inventors: 李江龙; 罗金辉; 单乐
Original assignee: Beijing Chusudu Technology Co ltd
Current assignee: Beijing Chusudu Technology Co ltd
Priority date: 2022-09-19
Filing date: 2022-09-19
Publication date: 2024-03-29
Also published as: WO2024060420A1

Abstract

The embodiment of the invention discloses a method, a device, equipment, a medium and a vehicle for matching parking lot data, wherein the method comprises the following steps: randomly selecting vehicle running tracks acquired for different times from parking lot running tracks acquired for multiple times, taking one of the vehicle running tracks as a current reference track, and taking the other vehicle running track as a current track to be matched; based on alignment frames of the current reference track matched with the vehicle pose in the current track to be matched, the relative pose among the track frames to be matched and the absolute position of the vehicle outside the parking lot, sequentially carrying out pose adjustment on the track frames to be matched, and carrying out pose alignment on the track frames subjected to pose adjustment and the corresponding reference track frames to obtain a current alignment result; and returning to execute the matching operation of the current reference track and the current track to be matched until the matching operation of the running tracks of the preset number of parking lots is completed. By adopting the technical scheme, the parking lot track acquired for multiple times is quickly and accurately associated.

Description

Parking lot data matching method, device, equipment, medium and vehicle

Technical Field

The embodiment of the invention relates to the technical field of automatic driving, in particular to a method, a device, equipment, a medium and a vehicle for matching parking lot data.

Background

At present, the functions of autonomous parking and automatic navigation of an unmanned vehicle in a parking lot are more and more emphasized, and a high-precision map is required to be relied on when the autonomous parking and automatic navigation of the parking lot are realized, so that the accurate map construction of the parking lot is particularly important.

A multi-story large parking lot often requires multiple data acquisitions to cover completely. The parking lot is often built underground without GPS (Global Positioning System ) signals, and the parking lot is multi-layer, and scenes among the layers are very similar, so that great difficulty is brought to the construction of the map of the parking lot.

In the related art, for parking lot data acquired multiple times, track data acquired multiple times is usually manually associated. Because the data volume of the vehicle track frame in the data acquired for many times is larger, the efficiency and the accuracy of the mode of manually carrying out data association are lower.

Disclosure of Invention

The embodiment of the invention provides a method, a device, equipment, a medium and a vehicle for matching parking lot data, so as to realize rapid and accurate association of parking lot tracks acquired for multiple times.

The specific technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides a method for matching parking lot data, including:

selecting vehicle running tracks acquired for different times at will from parking lot running tracks acquired for multiple times, taking one of the vehicle running tracks as a current reference track, and taking the other vehicle running track as a current track to be matched, wherein each running track comprises a front track entering a parking lot, a track in the parking lot and a rear track exiting the parking lot;

matching the current reference track with the current track to be matched, and in the matching process, carrying out pose adjustment on each track frame to be matched in sequence based on an alignment frame of matching the current reference track with the pose of the vehicle in the current track to be matched, the relative pose among all track frames to be matched in the current track to be matched and absolute position information of the vehicle before entering a parking lot and after exiting the parking lot, and carrying out pose alignment on the track frames after pose adjustment and the corresponding reference track frames in the current reference track to obtain a current alignment result corresponding to the current matching process;

and taking the running track corresponding to the current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is completed, so as to obtain the target matching result of the vehicle pose.

According to the scheme, in the process of matching the reference tracks acquired at different times with the tracks to be matched, the track frames to be matched can be subjected to pose adjustment in sequence based on the alignment frames of the reference tracks matched with the vehicle poses in the tracks to be matched, the relative poses among the track frames to be matched in the current tracks to be matched and the absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, and the track frames subjected to pose adjustment are subjected to pose alignment with the corresponding reference track frames in position, so that the running tracks acquired at multiple times can be accurately and rapidly associated together. In the association process, the vehicle pose alignment can be more accurate by adopting a progressive pose adjustment mode for each track frame to be matched.

Optionally, based on an alignment frame in which the current reference track is matched with the pose of the vehicle in the current track to be matched, the relative pose between frames of each track to be matched in the current track to be matched, and absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, pose adjustment is sequentially performed on each track frame to be matched, and pose alignment is performed on the track frame after pose adjustment and a reference track frame corresponding to the current reference track, so as to obtain a current alignment result corresponding to the current matching process, where the method includes:

Traversing each track frame to be matched in the current track to be matched in sequence, and determining a current reference track frame corresponding to the position of the track frame to be matched in the current reference track for each current track frame to be matched;

matching the current reference track frame with the current track frame to be matched;

if the matching is successful, determining a first relative pose between the current track frame to be matched and the corresponding current reference track frame;

based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track, and the second relative pose among the track frames to be matched in the current running track to be matched, pose optimization is carried out on all track frames to be matched successfully, and the track frames to be matched after pose optimization and the reference track frames corresponding to the positions where the matching is successful are used as alignment frames;

based on the current alignment frame, adjusting the pose of the track frame to be matched adjacent to the current alignment frame, taking the adjacent track frame with the adjusted pose as a new current track frame to be matched, and returning to execute the matching operation of the new current track frame to be matched and the current reference track frame corresponding to the position until all track frames to be matched in the current track to be matched are traversed, so as to obtain the current alignment result of the vehicle pose in the current matching process.

Optionally, after determining the first relative pose between the current track frame to be matched and the corresponding current reference track frame, the method provided by the embodiment of the present invention further includes:

and carrying out pose optimization on all the reference track frames in the current reference track based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track and the second relative pose among the track frames to be matched in the current running track to be matched, and taking the track frames to be matched after pose optimization and the reference track frames after pose optimization which are successfully matched as alignment frames.

According to the technical scheme, when the topological graph corresponding to the current track frame to be matched is optimized, pose optimization can be performed on all the reference track frames in the current reference track at the same time, so that more accurate pose alignment between the reference track and the track to be matched is realized.

Optionally, matching the current reference track frame with the current track frame to be matched includes:

local feature information of a parking lot is acquired, wherein the local feature information comprises: traffic sign board, wall body mark, road mark and lane mark;

if the local characteristic information of the parking lot corresponding to the current reference track frame is consistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched, the current reference track frame and the current track frame to be matched are successfully matched; or,

If the local characteristic information of the parking lot corresponding to the current reference track frame is inconsistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched, determining that the current reference track frame and the current track frame to be matched fail to be matched.

Optionally, the method provided by the embodiment of the invention further includes:

if the matching of the current reference track frame and the current track frame to be matched fails, temporarily storing the current track frame to be matched which fails to be matched;

after the current alignment result is obtained, sequentially traversing temporary-stored target track frames with failed matching, and repeatedly executing the matching operation of the target track frames with optimized positions and the reference track frames corresponding to the positions on the basis of the positions of the target track frames with optimized positions for any target track frame with failed matching until all temporary-stored target track frames with failed matching are traversed, so as to obtain a new current alignment result;

correspondingly, taking the running track corresponding to the current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched, wherein the method comprises the following steps:

And taking the running track corresponding to the new current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched.

According to the technical scheme, the target track frame which fails to be matched is queried and matched with the reference track frame at the corresponding position again, so that the alignment result of the vehicle pose is more accurate.

Optionally, determining the current reference track frame corresponding to the current track frame position to be matched in the current reference track includes:

under the condition that the current track frame to be matched belongs to a track before entering a parking lot or a track after exiting the parking lot:

acquiring an absolute position corresponding to a current track frame to be matched, and searching in a first set distance range from the absolute position in a current reference track to obtain a current reference track frame corresponding to the current track frame to be matched;

or, in the case that the current track frame to be matched belongs to the track in the parking lot:

searching in a first set distance range from the current position information in the current reference track according to the current position information corresponding to the current track frame to be matched to obtain a candidate reference track frame;

Judging whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot or not;

and if the candidate reference track frames belong to the same layer, taking the candidate reference track frames as current reference track frames corresponding to the positions of the current track frames to be matched.

Optionally, determining whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot includes:

determining the shortest topological path from the current track frame to be matched to the candidate reference track frame;

detecting whether a ramp section exists in the shortest topological path;

if the ramp section does not exist, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot; or,

if a ramp section exists, and if the difference value between the height information of the ascending road section and the height information of the descending road section in the ramp section is in a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot; or,

and if the difference value between the height information of the ascending road section and the height information of the descending road section in the ramp road section exceeds a second set distance range, determining that the candidate reference track frame and the current track frame to be matched do not belong to the same layer of the parking lot.

According to the technical scheme, the shortest topological path from the current track frame to be matched to the candidate reference track frame is determined, so that the complex situation between various paths from the current track frame to be matched to the candidate reference track frame is avoided. By detecting the ramp section in the shortest topological path and according to the height information of the ascending ramp section and the descending ramp section in the ramp section, whether the candidate reference track frame in the current reference track and the current track frame to be matched belong to the same layer of parking lot or not can be effectively judged, so that the situation that track frames which do not belong to the same layer of parking lot are mismatched is avoided, and the problem of cross-layer data error association is solved.

Optionally, detecting whether a ramp section exists in the shortest topology path includes:

judging whether the pitch angle of the vehicle relative to the horizontal plane is larger than a preset angle or not;

and taking a road section corresponding to the track frame with the pitch angle being larger than or equal to the preset angle as a ramp road section, and taking a road section corresponding to the track frame with the pitch angle being smaller than the preset angle as a non-ramp road section.

Optionally, performing pose optimization on all successfully matched track frames based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track, and the second relative pose between each track frame to be matched in the current track to be matched, including:

Constructing a first topological graph corresponding to a current track to be matched and a second topological graph corresponding to a current reference track, wherein nodes of the first topological graph are frames of the tracks to be matched, adjacent frames of the tracks to be matched are connected by edges, and nodes of the second topological graph are frames of the reference tracks, and adjacent frames of the reference tracks are connected by edges;

and optimizing the first topological graph by a pose graph optimizing method based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each driving track and the second relative pose among the track frames to be matched in the current driving track to be matched, and taking the track frames to be matched after pose optimization and the successfully matched reference track frames as alignment frames, wherein the alignment frame nodes in the first topological graph and the second topological graph are connected by edges.

According to the technical scheme, the topological graph corresponding to the current track frame to be matched and the topological graph corresponding to the reference track are optimized by adopting the pose graph (posagraph) optimization method, so that the vehicle pose can be accurately registered, the track to be referenced and the track to be matched are accurately associated together, and the problem of insufficient data association degree is solved.

Optionally, before matching the current reference track with the current track to be matched, the method provided by the embodiment of the invention further includes:

and respectively optimizing the first topological graph and the second topological graph based on absolute position information before entering the parking lot and after exiting the parking lot so as to reduce the deviation between the current track to be matched and the current reference track in absolute position.

Optionally, the vehicle pose corresponding to each reference track frame or the vehicle pose corresponding to each track frame to be matched is obtained through any one of the following sensor data, or is obtained through a mode of fusing the following various sensor data:

inertial measurement unit IMU data, image data, radar point cloud data, and odometry data.

In a second aspect, an embodiment of the present invention further provides a device for matching parking lot data, including:

the system comprises a driving track selection module, a driving track selection module and a control module, wherein the driving track selection module is configured to randomly select vehicle driving tracks collected for different times from parking lot driving tracks collected for multiple times, take one of the driving tracks as a current reference track, and take the other driving track as a current track to be matched, and each driving track comprises a front track entering a parking lot, a track in the parking lot and a rear track exiting the parking lot;

The matching module is configured to match the current reference track with the current track to be matched, and in the matching process, based on an alignment frame of the current reference track matched with the vehicle pose in the current track to be matched, the relative pose among all track frames to be matched in the current track to be matched and absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, pose adjustment is sequentially carried out on all track frames to be matched, and pose alignment is carried out on the track frames after pose adjustment and the corresponding reference track frames in the current reference track, so that a current alignment result corresponding to the current matching process is obtained;

the data association module is configured to take a running track corresponding to the current alignment result as a new current reference track, select another unmatched running track from the parking lot running tracks acquired for multiple times as a new current track to be matched, and return to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is completed, so as to obtain a target matching result of the vehicle pose.

Optionally, the matching module includes:

the current reference track frame determining unit is configured to sequentially traverse each track frame to be matched in the current track to be matched, and for each current track frame to be matched, the current reference track frame corresponding to the current track frame position to be matched is determined in the current reference track;

The matching unit is configured to match the current reference track frame with the current track frame to be matched;

the relative pose determining unit is configured to determine a first relative pose between the current track frame to be matched and the corresponding current reference track frame if the current reference track frame is successfully matched with the current track frame to be matched;

the first pose optimization unit is configured to perform pose optimization on all successfully matched track frames based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each driving track, and the second relative pose among all track frames to be matched in the current track to be matched, and takes the track frames to be matched after pose optimization and the reference track frames corresponding to the successfully matched positions as alignment frames;

the pose alignment unit is configured to adjust the pose of the track frame to be matched adjacent to the current alignment frame based on the current alignment frame, take the adjacent track frame with the adjusted pose as a new current track frame to be matched, and return to execute the matching operation of the new current track frame to be matched and the current reference track frame corresponding to the position until all track frames to be matched in the current track to be matched are traversed, so as to obtain the current alignment result of the vehicle pose in the current matching process.

Optionally, the device provided by the embodiment of the present invention further includes:

the second pose optimization unit is configured to, after determining a first relative pose between the current track frame to be matched and the corresponding current reference track frame, optimize the pose of all the reference track frames in the current reference track based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each driving track, and the second relative pose between the track frames to be matched in the current track to be matched, and take the track frames to be matched after pose optimization and the reference track frames after pose optimization after successful matching as alignment frames.

Optionally, the matching unit includes:

a local feature information acquisition subunit configured to acquire local feature information of a parking lot, the local feature information including: traffic sign board, wall body mark, road mark and lane mark;

the matching subunit is configured to determine that the current reference track frame and the current track frame to be matched are successfully matched if the local characteristic information of the parking lot corresponding to the current reference track frame is consistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched; or,

the temporary storage module of the matching failure frame is configured to temporarily store the current track frame to be matched, which is failed to be matched, if the current reference track frame is failed to be matched with the current track frame to be matched;

the re-matching module is configured to sequentially traverse the temporarily-stored target track frames with failed matching, repeatedly execute the matching operation of the target track frames with the optimized positions and the reference track frames corresponding to the positions on the basis of the positions of the target track frames with optimized positions for any target track frame with failed matching until all temporarily-stored target track frames with failed matching are traversed, and obtain a new current alignment result;

accordingly, the data association module is specifically configured to:

and taking the running track corresponding to the new current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is completed, so as to obtain the target matching result of the vehicle pose.

Optionally, the current reference track frame determining unit includes:

the first searching subunit is configured to acquire an absolute position corresponding to the current track frame to be matched under the condition that the current track frame to be matched belongs to a track before entering a parking lot or a track after exiting the parking lot, and search the current reference track within a first set distance range from the absolute position to obtain the current reference track frame corresponding to the current track frame to be matched;

or,

the second searching subunit is configured to search in a first set distance range from the current position information in the current reference track according to the current position information corresponding to the current track frame to be matched under the condition that the current track frame to be matched belongs to the track in the parking lot, so as to obtain a candidate reference track frame;

the same-layer judging subunit is configured to judge whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot or not;

and the current reference track frame determining subunit is configured to take the candidate reference track frame as the current reference track frame corresponding to the current track frame to be matched if the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot.

Optionally, the same-layer judging subunit includes:

a shortest topology path determining component configured to determine a shortest topology path from a current track frame to be matched to the candidate reference track frame;

a ramp segment detection component configured to detect whether a ramp segment exists in the shortest topological path;

the same-layer determining component is configured to determine that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot if no ramp section exists in the shortest topological path;

or if the ramp section exists in the shortest topological path and if the difference value between the height information of the ascending ramp section and the height information of the descending ramp section in the ramp section is within a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot;

or if the difference value between the height information of the ascending road section and the height information of the descending road section in the ramp road section exceeds the second set distance range, determining that the candidate reference track frame and the track frame to be matched do not belong to the same layer of the parking lot.

Optionally, the peer-layer determining component is specifically configured to:

And taking a road section corresponding to the track frame with the pitch angle being greater than or equal to the preset angle as a ramp road section, and taking a road section corresponding to the track frame with the pitch angle being smaller than the preset angle as a non-ramp road section.

Optionally, the first pose optimization unit is specifically configured to:

Optionally, the apparatus provided in this embodiment further includes:

and the initial optimization module is configured to optimize the first topological graph and the second topological graph based on absolute position information before entering the parking lot and after exiting the parking lot before matching the current reference track with the current track to be matched so as to reduce the deviation between the current track to be matched and the current reference track in absolute position.

In a third aspect, an embodiment of the present invention provides an electronic device, including:

one or more processors;

storage means for storing one or more programs,

the one or more programs, when executed by the one or more processors, cause the one or more processors to implement a method of matching parking lot data as provided by any embodiment of the present invention.

In a fourth aspect, an embodiment of the present invention provides a storage medium having a computer program stored thereon, wherein the program when executed by a processor implements the method for matching parking lot data as provided in any embodiment of the present invention.

In a fifth aspect, an embodiment of the present invention provides a vehicle, where the vehicle includes the matching device for parking lot data provided in any embodiment of the present invention, or includes the electronic device provided in any embodiment of the present invention.

In a sixth aspect, embodiments of the present invention provide a computer program comprising program instructions which, when executed by a computer, implement a method for matching parking lot data as provided in any of the embodiments of the present invention.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1a is a flowchart of a method for matching parking lot data according to an embodiment of the present invention;

FIG. 1b is a schematic diagram of a vehicle track including a parking lot of a next floor according to an embodiment of the present invention;

FIG. 1c is a schematic diagram of a vehicle track including an underground two-layer parking garage according to an embodiment of the present invention;

fig. 1d is a topological diagram after initial pose optimization of a reference driving track and a track to be matched according to a first embodiment of the present invention;

Fig. 1e is a topology diagram of matching a track to be matched with a reference track according to a first embodiment of the present invention;

FIG. 1f is a schematic diagram of a matching method for matching a trace to be matched with a reference trace according to a first embodiment of the present invention;

fig. 2 is a flowchart of a method for matching parking lot data according to a second embodiment of the present invention;

fig. 3 is a block diagram of a device for matching parking lot data according to a third embodiment of the present invention;

fig. 4 is a block diagram of an electronic device according to a fourth embodiment of the present invention;

fig. 5 is a schematic diagram of a vehicle according to a fifth embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without any inventive effort, are intended to be within the scope of the invention.

It should be noted that the terms "comprising" and "having" and any variations thereof in the embodiments of the present invention and the accompanying drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those listed steps or elements but may include other steps or elements not listed or inherent to such process, method, article, or apparatus.

The embodiment of the invention discloses a method, a device, equipment, a medium and a vehicle for matching parking lot data. The following will describe in detail.

Example 1

Fig. 1a is a flowchart of a method for matching parking lot data according to an embodiment of the present invention, which can be applied to a vehicle-mounted terminal such as a vehicle-mounted computer, a vehicle-mounted industrial control computer (Industrial personal Computer, IPC), etc., and can also be applied to a server, which is not limited in the embodiment of the present invention. The method provided by the embodiment can be applied to an application scene of correlating the multi-layer parking lot driving tracks acquired for multiple times. The method provided by the embodiment can be performed by a matching device of parking lot data, and the device can be realized by software and/or hardware. As shown in fig. 1a, the method provided in this embodiment specifically includes:

s110, randomly selecting vehicle running tracks collected for different times from parking lot running tracks collected for multiple times, taking one of the vehicle running tracks as a current reference track, and taking the other vehicle running track as a current track to be matched.

Among the vehicle running tracks collected at different times, the vehicle pose corresponding to each frame track can be obtained through the vehicle-mounted sensor, or can be obtained through a mode of fusing various sensor data. The in-vehicle sensors may include inertial measurement units (Inertial Measurement Unit, IMU), image sensors, radar and odometer, among others.

In this embodiment, the vehicle travel tracks collected multiple times include a travel track before entering the parking lot, a travel track in the parking lot, and a travel track after exiting the parking lot. Wherein a GPS (Global Positioning System ) signal is present both before the vehicle enters the parking lot and after it exits the parking lot. The three tracks, namely the track before entering the parking lot, the track during running in the parking lot and the track after exiting the parking lot, are continuous tracks. By applying GPS constraint to the two tracks entering and exiting the parking lot and simultaneously utilizing the continuity of the three tracks, the whole track can obtain the positioning accuracy of the absolute position at the meter level.

In this embodiment, the matching of the current reference track with the current track to be matched actually matches the current reference track with the vehicle pose corresponding to each track frame in the current track frame to be matched, and aligns the track to be matched of each frame with the reference track of the corresponding position by optimizing the vehicle pose. In order to reduce the calculation amount, a plurality of key frames can be selected from the current reference track as reference track frames, and a plurality of key frames can be selected from the current track to be matched as track frames to be matched. Wherein, the selection of each key frame can be: and taking the track frame with the driving distance being the preset distance as a key frame. Wherein the preset distance can be any distance in the range of 1-3 meters.

In this embodiment, in order to achieve accurate alignment of the pose of the current reference track and the current track to be matched, initial optimization may be performed on the pose of each reference track frame and each track frame to be matched. Specifically, a first topological graph corresponding to a current track to be matched and a second topological graph corresponding to a current reference track can be constructed, wherein nodes of the first topological graph are frames of the tracks to be matched, adjacent frames of the tracks to be matched are connected by edges, nodes of the second topological graph are frames of the reference tracks, and adjacent frames of the reference tracks are connected by edges. Before the current reference track is matched with the current track to be matched, the first topological graph and the second topological graph are respectively optimized based on absolute position information before entering the parking lot and after exiting the parking lot, so that deviation between the current track to be matched and the current reference track in absolute position can be reduced.

Specifically, fig. 1B is a topology diagram of a vehicle running track including a parking lot (B1) of the first floor according to the embodiment of the present invention, and fig. 1c is a topology diagram of a vehicle running track including a parking lot (B2) of the second floor according to the embodiment of the present invention. As shown in fig. 1b and 1c, the nodes in each travel track are vehicle travel track frames, i.e., optimization variables. The adjacent nodes are connected by edges, and the adjacent nodes are restrained by relative pose. The GPS signals generated by GPS measurements before entering the parking lot and after exiting the parking lot are constrained. One of the running tracks in fig. 1b and fig. 1c is taken as a track to be matched, the other running track is taken as a reference track, and fig. 1d is a topological diagram after initial pose optimization of the reference running track and the track to be matched according to the first embodiment of the present invention. After pose map optimization (posmagraph) is performed on fig. 1b and 1c, initial optimization of vehicle pose can be achieved, so as to reduce deviation between the track to be matched and the reference track in absolute position. As shown in fig. 1d, after the initial optimization is completed, the nodes of each track frame in the reference driving track and the track to be matched are not completely aligned, and the positions of the nodes are still in meter-level errors, so that pose adjustment is required to be performed on each track frame through subsequent steps to realize pose alignment between each track frame in the track to be matched and each track frame in the reference track frame.

S120, matching the current reference track with the current track to be matched, and in the matching process, carrying out pose adjustment on each track frame to be matched in sequence based on an alignment frame of matching the current reference track with the pose of the vehicle in the current track to be matched, the relative pose between each track frame to be matched in the current track to be matched and absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, and carrying out pose alignment on the track frames after pose adjustment and the corresponding reference track frames in the current reference track to obtain a current alignment result corresponding to the current matching process.

In this embodiment, for each track frame to be matched in the tracks to be matched, pose optimization can be performed on each track frame to be matched in a progressive optimization manner. Specifically, the method can be realized by the following steps (one) - (three).

Traversing each track frame to be matched in the current track to be matched in turn, and determining a current reference track frame corresponding to the position of the track frame to be matched in the current reference track for each current track frame to be matched.

The current reference track frame corresponding to the current track frame to be matched is a reference track frame which belongs to the same layer of parking lot as the current track frame to be matched and is close to the current track frame to be matched in position, namely, the distance is within a first set distance range. The number of reference trajectory frames satisfying the above condition may be one or a plurality of.

In this embodiment, the track frame to be matched currently may belong to a track before entering the parking lot or a track after exiting the parking lot, or may also belong to a track in the parking lot.

Under the condition that the current track frame to be matched belongs to a track before entering a parking lot or a track after exiting the parking lot, the absolute position information of the vehicle can be determined based on the received GPS signal, so that the search of the reference track frame corresponding to the position can be directly performed according to the absolute position information, and the method can be realized specifically as follows:

and acquiring an absolute position corresponding to the current track frame to be matched, and searching in a first set distance range from the absolute position in the current reference track to obtain a current reference track frame corresponding to the current track frame to be matched. Wherein the first set distance range may be 8-12 meters.

Or, in the case that the current track frame to be matched belongs to the track in the parking lot, the vehicle cannot receive the GPS signal or the received GPS signal is weak, at this time, the current reference track frame corresponding to the current track frame position may be determined through the following steps a to C.

A. And searching in the current reference track within a first set distance range from the current position information according to the current position information corresponding to the current track frame to be matched, so as to obtain a candidate reference track frame.

B. And judging whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot, and if so, taking the candidate reference track frame as the current reference track frame corresponding to the current track frame position.

The method for judging whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot is various, in this embodiment, the ramp section is detected from the shortest topology path, and whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot is judged by judging the height difference between the uphill section and the downhill section in the ramp section, specifically, the method can be realized as follows:

determining the shortest topological path from the current track frame to be matched to the candidate reference track frame; detecting whether a ramp section exists in the shortest topological path; if the ramp section does not exist, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot; or if the ramp section exists and the difference value of the height information of the ascending road section and the height information of the descending road section in the ramp section is in a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot; or if the difference value between the height information of the ascending road section and the height information of the descending road section in the ramp road section exceeds the second set distance range, determining that the candidate reference track frame and the track frame to be matched do not belong to the same layer of the parking lot. Wherein the second set distance range may be 3-5 meters.

In this embodiment, whether the road section is a ramp road section may be determined by a pitch angle of the vehicle relative to a horizontal plane, and specifically, a road section corresponding to a track frame with a pitch angle greater than or equal to a preset angle may be used as the ramp road section, and a road section corresponding to a track frame with a pitch angle less than the preset angle may be used as the non-ramp road section. Wherein the range of the preset angle can be 5-10 degrees. The pitch angle of the vehicle relative to the horizontal plane may be determined by the IMU or the attitude of the vehicle may be determined based on the data from the multisensor fusion to determine the pitch angle of the vehicle relative to the horizontal plane. The multisensor may include an IMU, an odometer, an image sensor, and the like, among others.

In this embodiment, the height information of the uphill road section or the height information of the downhill road section may be calculated from a height difference between a vehicle pose of the vehicle at the top of the uphill road section and a vehicle pose of the vehicle at the bottom of the uphill road section.

In this embodiment, for the candidate reference track frame belonging to the same layer of parking lot as the current track to be matched, the candidate reference track frame is used as the current reference track frame corresponding to the current track frame position to be matched. And for the candidate reference track frames which do not belong to the same layer of parking lot as the current track to be matched, not matching the current track to be matched. In this case, the next new track frame to be matched is continuously traversed backward, and the operation of determining the current reference track frame corresponding to the new current track frame to be matched is performed back, i.e., the operation of performing step (one) is performed back.

For example, fig. 1e is a topology diagram of matching a track to be matched with a reference track according to a first embodiment of the present invention. As shown in fig. 1e, K2, K4, K8, K7, and K6 are reference track frames in the current reference track. K1, K3, K9, K10, K5, and K11 are track frames to be matched in the current track to be matched, where K1 and K2, K3, and K4 are alignment frames in which the pose is aligned. In the progressive positioning matching process, the GPS signal cannot be received after entering the parking lot, so that the vehicle track has obvious drift in absolute position and even can generate staggered layers. For example, in fig. 1e, since the track where K5 is located gradually drifts, its position is very close to K6, which may lead to erroneous association between the two, and this situation may lead to failure of the entire mapping. To avoid this problem, the shortest topology path between two frames needs to be queried from the topology map each time two frames are initiated, e.g. shortest topology path_0 between K5 and K6 is: K5-K10-K9-K3-K4-K8-K7-K6.

And inquiring the ascending and descending slope condition on the shortest topological path in the shortest topological path path_0 according to the gesture of the track frame to be matched. If in fig. 1e, K3-K9 is directly lowered from outside the parking lot to the B2 layer, for example, 6m is lowered, K4-K8 is lowered from outside the parking lot to the B1 layer, for example, 3m is lowered, it may be determined that the current track frame K5 to be matched and the candidate reference track frame K6 are not in the same layer, and matching between K5 and K6 is not required, so that an erroneous matching condition is eliminated. At this time, the candidate reference track frame is searched again for the next track frame K11 to be matched of the current track frame to be matched, if the candidate reference track frame K6 is searched, whether K11 and K6 belong to the same layer is continuously judged, that is, the up-down slope condition on the shortest topology path in K11 to K6 is calculated. For example, K4-K8 falls by 3m, K3-K9 falls by 6m, K5-K11 rises by 3m again, so that K11 and K6 can be determined to be on the same floor, at this time, the reference track frame K6 can be used as the current reference track frame corresponding to the position of the key frame K11 to be currently matched, and step (two) of matching K6 and K11 using the parking lot local feature information and connecting K6 and K11 in the topology after K6 and K11 become aligned frames with matching poses can be continued.

In this embodiment, by determining the shortest topology path from the current track frame to be matched to the candidate reference track frame, the complex situation between various paths from the current track frame to be matched to the candidate reference track frame is avoided. By detecting the ramp section in the shortest topological path and according to the height information of the ascending ramp section and the descending ramp section in the ramp section, whether the candidate reference track frame in the current reference track and the current track frame to be matched belong to the same layer of parking lot or not can be effectively judged, so that the situation that track frames which do not belong to the same layer of parking lot are mismatched is avoided, and the problem of cross-layer data error association is solved.

And secondly, matching the current reference track frame with the current track frame to be matched, and if the matching is successful, determining a first relative pose between the current track frame to be matched and the corresponding current reference track frame.

It can be understood by those skilled in the art that the vehicle tracks obtained by different times of running of the vehicle in the same parking lot have deviations in pose, but the local characteristic information of the parking lot corresponding to the track frames corresponding to the positions in different running tracks is fixed. The local characteristic information of the parking lot can comprise traffic signs, wall marks, pavement marks, lane line marks and the like. The local characteristic information of the parking lot can be acquired by an on-board sensor (such as an image sensor, a radar and the like).

Based on the above reasons, in this embodiment, the current reference track frame may be matched with the current track frame to be matched based on the obtained local feature information of the parking lot, and if the local feature information of the parking lot corresponding to the current reference track frame is consistent with the local feature information of the parking lot corresponding to the current track frame to be matched, it is determined that the current reference track frame is successfully matched with the current track frame to be matched; if the local characteristic information of the parking lot corresponding to the current reference track frame is inconsistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched, determining that the current reference track frame and the current track frame to be matched fail to be matched.

In this embodiment, for a current track frame to be matched successfully and a corresponding current reference track frame, the two frames are calculated in a relative pose manner, so as to correlate the two frames. And temporarily storing the current track frames to be matched, which are failed to be matched, and re-matching the track frames to be matched, which are failed to be matched, after the traversal of all the track frames to be matched is completed.

And thirdly, performing pose optimization on all successfully matched track frames based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track, and the second relative pose among all track frames to be matched in the current track to be matched, and taking the track frames to be matched after pose optimization and the reference track frames corresponding to the successfully matched positions as alignment frames.

In the embodiment, the pose of the vehicle is optimized by constructing a pose topological graph and optimizing the pose topological graph, so that the pose alignment of the track frame to be matched and the reference track frame corresponding to the successfully matched position is realized. The method is concretely realized by the following steps 1-2:

1. and constructing a first topological graph corresponding to the current track to be matched and a second topological graph corresponding to the current reference track.

For example, the topological graph shown in fig. 1b is used as the topological graph corresponding to the current reference track, and the topological graph shown in fig. 1c is used as the topological graph corresponding to the current track to be matched. As shown in fig. 1b, the nodes of the second topological graph corresponding to the current reference track are reference track frames, and adjacent reference track frames are connected by edges. As shown in fig. 1c, the nodes of the first topological graph corresponding to the current track to be matched are track frames to be matched, adjacent track frames to be matched are connected by edges,

2. and optimizing the first topological graph by a pose graph optimizing method based on a first relative pose between the current track frame to be matched and the corresponding current reference track frame, absolute position information of the vehicle in each driving track before entering the parking lot and after exiting the parking lot, and a second relative pose between the track frames to be matched in the current track to be matched, and taking the track frames to be matched after pose optimization and the successfully matched reference track frames as alignment frames, wherein the alignment frame nodes in the first topological graph and the second topological graph are connected by edges.

In this embodiment, by optimizing the topological graph corresponding to the current track frame to be matched by using a pose graph (posagraph) optimization method, the variables to be optimized, i.e., the vehicle pose, as represented by the nodes in fig. 1c can be optimized. Specifically, in the optimization process, if the nth track frame to be matched is traversed currently, the track frame subjected to pose optimization is from the 1 st track frame to be matched to the current nth track frame to be matched, the optimized positions corresponding to the 1 st to the N nodes are obtained, and the N track frames subjected to pose optimization and the successfully matched reference track frame can be used as alignment frames. In the embodiment, by adopting a pose chart optimization mode, the vehicle pose in the track to be matched and the reference track can be accurately registered, so that the track to be matched and the reference track are accurately associated together, and the problem of insufficient data association degree is solved.

Further, in order to enable the pose alignment of the current track frame to be matched with the current reference track frame to be more accurate, when the topological graph corresponding to the current track frame to be matched is optimized, pose optimization can be conducted on all reference track frames in the current reference track based on absolute position information of vehicles before entering a parking lot and after exiting the parking lot in first relative pose running tracks between the current track frame to be matched and the corresponding current reference track frame, and second relative pose between the track frames to be matched in the current track to be matched, so that more accurate pose alignment between the reference track and the track to be matched is achieved.

For example, fig. 1f is a topology diagram of matching a track to be matched with a reference track according to a first embodiment of the present invention. On the basis of fig. 1d, fig. 1f shows a further optimization procedure for the vehicle pose. For the first track frame K1 to be matched in the current track data2 to be matched, the track frame K1 is the track frame before entering the parking lot. According to the absolute position information of the track frame K1 before entering the parking lot, a reference track frame K2 corresponding to the pose of the track frame K1 to be matched in a first set distance range can be queried in the reference running track data 1. And matching the K1 with the K2 by utilizing the parking lot local characteristic information corresponding to the K1 and the K2, if the matching is successful, calculating the relative pose of the vehicle corresponding to the K1 track frame and the K2 track frame, and after the calculation of the relative pose is completed, optimizing the first topological graph and the second topological graph, so that more accurate pose configuration between the K1 and the K2 can be realized. The current track frame K1 to be matched after pose optimization and the corresponding reference track frame K2 are alignment frames. In the topological graph, the nodes corresponding to K1 and K2 are connected, so that the current track data2 to be matched and the current reference track data1 acquire initial registration.

And fourthly, based on the current alignment frame, adjusting the pose of the track frame to be matched adjacent to the current alignment frame, taking the adjacent track frame with the adjusted pose as a new current track frame to be matched, and returning to execute the matching operation of the new current track frame to be matched and the current reference track frame corresponding to the position until all track frames to be matched in the current track to be matched are traversed, so as to obtain the current alignment result of the vehicle pose in the current matching process.

As can be appreciated by those skilled in the art, since the relative pose between two adjacent track frames in the same driving track frame is fixed, after the optimized position of the nth track frame to be matched in the current traversal is obtained through topological graph optimization, the pose of the next track frame to be matched, that is, the n+1th track frame, adjacent to the nth track frame to be matched in the current traversal is also required to be correspondingly adjusted, the n+1th track frame to be matched after the pose adjustment is taken as the current new track frame to be matched, the operation of executing the step (one) is returned to obtain the current reference track frame corresponding to the new current track frame to be matched, the matching operation of the current reference track frame corresponding to the new current track frame to be matched in the step (two) is continuously performed, and then the steps (three) and (four) are continuously performed until all track frames to be matched in the current track to obtain the current alignment result of the vehicle pose in the current matching process is traversed.

Specifically, as shown in fig. 1f, the track frame to be matched K1 and the track frame to be matched K2 are already aligned, and at this time, the pose of the track frame to be matched K1 is already optimized. Because the relative pose between the track frame to be matched K1 and the adjacent track frame K3 is fixed, after the pose of the track frame to be matched K1 is adjusted, the adjacent track frame K3 (solid black circle) can update the pose to the black dotted line circle according to the relative pose between the track frame to be matched K1, namely, the position of the adjacent track frame K3 is adjusted. And taking the track frame K3 with the adjusted pose as a new track frame to be matched currently, searching again according to the position to obtain a corresponding reference track frame K4 in the reference track, matching the track frame K4 with the track frame K3 by using local features of a parking lot, and calculating relative poses after the matching is successful to realize the correlation of the track frame K4 and the track frame K3. And repeating the process for all subsequent frames of the track to be matched until the processing of all frames to be matched is completed.

S130, taking the running track corresponding to the current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is completed, so as to obtain the target matching result of the vehicle pose.

The predetermined number may be set according to the number of actually collected parking lot travel tracks, for example, may be set to be the total number of collected parking lot travel tracks, which is not particularly limited in this embodiment.

According to the technical scheme provided by the embodiment, in the process of matching the reference tracks acquired for different times with the tracks to be matched, based on the alignment frames of the reference tracks, which are matched with the vehicle pose in the tracks to be matched, the relative pose among the frames of each track to be matched in the current tracks to be matched and the absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, the pose adjustment can be sequentially carried out on each track frame to be matched, and the track frames after the pose adjustment are aligned with the corresponding reference track frames in position, so that the running tracks acquired for multiple times can be accurately associated together. In the association process, the vehicle pose alignment can be more accurate by adopting a progressive pose adjustment mode for each track frame to be matched. Compared with the mode of manually correlating the collected parking lot data in the related art, the method provided by the embodiment improves the efficiency and accuracy of correlation of the parking lot data.

Example two

Fig. 2 is a flowchart of a matching method of parking lot data provided by a second embodiment of the present invention, where on the basis of the foregoing embodiment, the situation that the matching between a current reference track frame and a current track frame to be matched fails is refined, as shown in fig. 2, the method provided by the present embodiment includes:

s200, randomly selecting vehicle running tracks collected for different times from parking lot running tracks collected for multiple times, taking one of the vehicle running tracks as a current reference track, and taking the other vehicle running track as a current track to be matched.

S210, matching the current reference track with the current track to be matched, traversing each track frame to be matched in the current track to be matched in sequence in the matching process, and determining a current reference track frame corresponding to the current track frame position to be matched in the current reference track for each current track frame to be matched.

S220, matching the current reference track frame with the current track frame to be matched, judging whether the matching is successful, and if so, executing a step S230; otherwise, step S240 is performed.

S230, determining a first relative pose between the current track frame to be matched and the corresponding current reference track frame, and continuing to execute the step S250.

S240, temporarily storing the current track frame to be matched, which is failed to be matched, and continuously executing the step S270.

S250, performing pose optimization on all successfully matched track frames and all reference track frames in the current reference track based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each driving track, and the second relative pose among all track frames to be matched in the current track to be matched, and continuously executing the step S260 by taking the pose-optimized track frames to be matched and the pose-optimized reference track frames which are successfully matched as alignment frames.

In this embodiment, when the pose optimization method is used to perform pose optimization, pose optimization is performed on all the track frames to be matched that have been successfully matched and all the reference track frames in the current reference track, so that the pose alignment of the current track frame to be matched and the current reference track frame can be more accurate.

And S260, based on the current alignment frame, adjusting the pose of the track frame to be matched adjacent to the current alignment frame, taking the adjacent track frame with the adjusted pose as a new current track frame to be matched, returning to the step S210, until all track frames to be matched in the current track to be matched are traversed, obtaining a current alignment result corresponding to the current matching process, and continuing to execute the step S270.

And S270, sequentially traversing the temporarily-stored target track frames with failed matching, acquiring the optimized pose of any one target track frame with failed matching, returning to the step S210 until all temporarily-stored target track frames with failed matching are traversed, obtaining a new current alignment result, and continuing to execute the step S280.

S280, taking the running track corresponding to the new current alignment result as a new current reference track, selecting another unmatched running track from the parking lot running tracks acquired for multiple times as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched, namely returning to execute the step S210 until all the parking lot running tracks are traversed, so as to obtain the target matching result of the vehicle pose.

In this embodiment, after the matching of the current reference track and the current track to be matched is completed, there are already many frames to be matched to achieve correct matching, and the relative pose between the track to be matched and the reference track is more accurate than before matching. In addition, by taking the matching result obtained each time as a new reference track and repeatedly executing the mode of matching the new reference track with a new track to be matched, the accurate association of the multiple acquired running tracks is realized. Compared with the mode of manually correlating the collected parking lot data in the related art, the method provided by the embodiment further improves the efficiency and accuracy of correlation of the parking lot data.

Example III

Fig. 3 is a block diagram of a device for matching parking lot data according to a third embodiment of the present invention, as shown in fig. 3, the device includes: a travel track selection module 310, a matching module 320, and a data association module 330, wherein,

the driving track selection module 310 is configured to arbitrarily select vehicle driving tracks collected for different times from parking lot driving tracks collected for multiple times, and take one of the driving tracks as a current reference track and the other driving track as a current track to be matched, wherein each driving track comprises a front track entering a parking lot, a track in the parking lot and a rear track exiting the parking lot;

the matching module 320 is configured to match the current reference track with the current track to be matched, and in the matching process, based on an alignment frame of the current reference track matching with the vehicle pose in the current track to be matched, the relative pose between each track frame to be matched in the current track to be matched, and absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, pose adjustment is sequentially performed on each track frame to be matched, and pose alignment is performed on the track frame after pose adjustment and the corresponding reference track frame in the current reference track, so as to obtain a current alignment result corresponding to the current matching process;

The data association module 330 is configured to take a running track corresponding to the current alignment result as a new current reference track, select another unmatched running track from the parking lot running tracks acquired for multiple times as a new current track to be matched, and return to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is completed, thereby obtaining the target matching result of the vehicle pose.

Optionally, the matching module 320 includes:

Optionally, the matching unit includes:

Accordingly, the data association module is specifically configured to:

Optionally, the current reference track frame determining unit includes:

or,

Optionally, the same-layer judging subunit includes:

the same-layer determining component is configured to determine that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot if no ramp section exists in the shortest topological path; or,

if a ramp section exists in the shortest topological path, and if the difference value between the height information of the ascending ramp section and the height information of the descending ramp section in the ramp section is within a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot; or,

Optionally, the peer-layer determining component is specifically configured to:

Optionally, the first pose optimization unit is specifically configured to:

Optionally, the apparatus provided in this embodiment further includes:

The device for matching the parking lot data provided by the embodiment of the invention can execute the method for matching the parking lot data provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method. Technical details not described in detail in the above embodiments may be referred to the method for matching parking lot data provided in any embodiment of the present invention.

Example IV

Fig. 4 is a block diagram of an electronic device according to a fourth embodiment of the present invention, as shown in fig. 4, where the electronic device includes:

A memory 510 storing executable program code;

a processor 520 coupled to the memory 510;

the processor 520 invokes the executable program code stored in the memory 510 to perform the method for matching the parking lot data provided by any embodiment of the present invention.

Example five

Based on the above embodiments, a further embodiment of the present invention provides a vehicle comprising an apparatus according to any of the embodiments above, or comprising an electronic device as described above.

Fig. 5 is a schematic diagram of a vehicle according to a fifth embodiment of the present invention. As shown in fig. 5, the vehicle includes a speed sensor 61, an ECU (Electronic Control Unit ) 62, a GPS (Global Positioning System, global positioning system) positioning device 63, a T-Box (Telematics Box) 64. Wherein the speed sensor 61 is used for measuring the speed of the vehicle and taking the speed of the vehicle as an empirical speed for model training; the GPS positioning device 63 is used for acquiring the current geographic position of the vehicle; T-Box64 may communicate with a server as a gateway; the ECU62 may perform the above-described matching method of the parking lot data.

In addition, the vehicle may further include: V2X (Vehicle-to-Evering) module 65, radar 66 and camera 67. The V2X module 65 is used for communication with other vehicles, road side equipment, etc.; the radar 66 or the camera 67 is used for sensing road environment information in front and/or other directions to obtain original point cloud data; the radar 66 and/or camera 67 may be disposed at the front and/or rear of the vehicle body.

Based on the above method embodiments, another embodiment of the present invention provides a storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to implement a method for matching parking lot data as described in any of the above embodiments.

Those of ordinary skill in the art will appreciate that: the drawing is a schematic diagram of one embodiment and the modules or flows in the drawing are not necessarily required to practice the invention.

Those of ordinary skill in the art will appreciate that: the modules in the apparatus of the embodiments may be distributed in the apparatus of the embodiments according to the description of the embodiments, or may be located in one or more apparatuses different from the present embodiments with corresponding changes. The modules of the above embodiments may be combined into one module, or may be further split into a plurality of sub-modules.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for matching parking lot data, comprising:

matching the current reference track with the current track to be matched, and in the matching process, carrying out pose adjustment on each track frame to be matched in sequence based on an alignment frame of the current reference track matched with the vehicle pose in the current track to be matched, the relative pose among all track frames to be matched in the current track to be matched and absolute position information of the vehicle before entering a parking lot and after exiting the parking lot, and carrying out pose alignment on the track frames after pose adjustment and the corresponding reference track frames in the current reference track to obtain a current alignment result corresponding to the current matching process;

2. The method according to claim 1, wherein the step of sequentially performing pose adjustment on each track frame to be matched based on an alignment frame in which the current reference track is matched with the pose of the vehicle in the current track to be matched, the relative pose between each track frame to be matched in the current track to be matched, and absolute position information of the vehicle before entering the parking lot and after exiting the parking lot, and performing pose alignment on the track frame after pose adjustment and a corresponding reference track frame in the current reference track to obtain a current alignment result corresponding to the current matching process includes:

traversing each track frame to be matched in the current track to be matched in sequence, and determining a current reference track frame corresponding to the current track frame position to be matched in the current reference track for each current track frame to be matched;

based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track, and second relative pose among track frames to be matched in the current track to be matched, carrying out pose optimization on all track frames to be matched successfully, and taking the track frames to be matched after pose optimization and a reference track frame corresponding to the position where matching is successful as an alignment frame;

Based on the current alignment frame, adjusting the pose of a track frame to be matched adjacent to the current alignment frame, taking the adjacent track frame with the adjusted pose as a new current track frame to be matched, and returning to execute the matching operation of the new current track frame to be matched and the current reference track frame corresponding to the position until all track frames to be matched in the current track to be matched are traversed, so as to obtain the current alignment result of the vehicle pose in the current matching process.

3. The method of claim 2, wherein after determining the first relative pose between the current to-be-matched trajectory frame and the corresponding current reference trajectory frame, the method further comprises:

and carrying out pose optimization on all the reference track frames in the current reference track based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track and the second relative pose among the track frames to be matched in the current running track to be matched, and taking the track frames to be matched after pose optimization and the reference track frames after pose optimization successfully matched as alignment frames.

4. The method of claim 2, wherein the matching the current reference trajectory frame with the current trajectory frame to be matched comprises:

if the local characteristic information of the parking lot corresponding to the current reference track frame is consistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched, determining that the current reference track frame is successfully matched with the current track frame to be matched; or,

if the local characteristic information of the parking lot corresponding to the current reference track frame is inconsistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched, determining that the current reference track frame fails to be matched with the current track frame to be matched.

5. The method according to any one of claims 2-4, further comprising:

after the current alignment result is obtained, sequentially traversing temporary-stored target track frames with failed matching, and repeatedly performing matching operation of the target track frames with optimized positions and reference track frames corresponding to the positions on the basis of the positions of the target track frames with optimized positions for any target track frame with failed matching until all temporary-stored target track frames with failed matching are traversed, so as to obtain a new current alignment result;

Correspondingly, the step of taking the running track corresponding to the current alignment result as a new current reference track, selecting another unmatched running track from the parking lot running tracks acquired for multiple times as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched comprises the following steps:

6. The method of claim 2, wherein the determining a current reference track frame in the current reference track that corresponds to the current track frame position to be matched comprises:

acquiring an absolute position corresponding to a current track frame to be matched, and searching in the current reference track within a first set distance range from the absolute position to obtain a current reference track frame corresponding to the current track frame to be matched;

searching in the current reference track within a first set distance range from the current position information according to the current position information corresponding to the current track frame to be matched to obtain a candidate reference track frame;

judging whether the candidate reference track frame and the current track frame to be matched belong to the same layer of a parking lot or not;

and if the candidate reference track frames belong to the same layer, taking the candidate reference track frames as current reference track frames corresponding to the current track frame position to be matched.

7. The method of claim 6, wherein the determining whether the candidate reference trajectory frame and the current trajectory frame to be matched belong to the same layer of a parking lot comprises:

detecting whether a ramp section exists in the shortest topological path;

if a ramp section exists, and if the difference value between the height information of the ascending ramp section and the height information of the descending ramp section in the ramp section is in a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of a parking lot;

And if the difference value between the height information of the ascending road section and the height information of the descending road section of the ramp road section exceeds a second set distance range, determining that the candidate reference track frame and the current track frame to be matched do not belong to the same layer of the parking lot.

8. The method of claim 7, wherein the detecting whether a ramp segment exists in the shortest topology path comprises:

9. The method according to claim 2, wherein the performing pose optimization on all the track frames to be matched that have been successfully matched based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each track, and the second relative pose between each track frame to be matched in the current track to be matched comprises:

constructing a first topological graph corresponding to the current track to be matched and a second topological graph corresponding to the current reference track, wherein nodes of the first topological graph are frames of the track to be matched, adjacent frames of the track to be matched are connected by edges, and nodes of the second topological graph are frames of the reference track, and adjacent frames of the reference track are connected by edges;

And optimizing the first topological graph through a pose graph optimizing method based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each driving track and the second relative pose among the track frames to be matched in the current driving track to be matched, and taking the track frames to be matched after pose optimization and the successfully matched reference track frames as alignment frames, wherein the alignment frame nodes in the first topological graph and the second topological graph are connected by edges.

10. The method of claim 9, wherein prior to matching the current reference trajectory with the current trajectory to be matched, the method further comprises:

11. The method according to claim 1, wherein the vehicle pose corresponding to each reference track frame or the vehicle pose corresponding to each track frame to be matched is obtained by any one of the following sensor data, or is obtained by fusing the following sensor data:

12. A matching device for parking lot data, comprising:

the matching module is configured to match the current reference track with the current track to be matched, and in the matching process, based on an alignment frame for matching the current reference track with the vehicle pose in the current track to be matched, the relative pose among all track frames to be matched in the current track to be matched, and absolute position information of the vehicle before entering a parking lot and after exiting the parking lot, pose adjustment is sequentially performed on all track frames to be matched, and pose alignment is performed on the track frames after pose adjustment and the corresponding reference track frames in the current reference track, so that a current alignment result corresponding to the current matching process is obtained;

The data association module is configured to take the running track corresponding to the current alignment result as a new current reference track, select another unmatched running track from the parking lot running tracks acquired for multiple times as a new current track to be matched, and return to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks of a preset number is completed, so as to obtain a target matching result of the vehicle pose.

13. The apparatus of claim 12, wherein the matching module comprises:

a current reference track frame determining unit, configured to match the current reference track with the current track to be matched, and in the matching process, sequentially traverse each track frame to be matched in the current track to be matched, and for each current track frame to be matched, determine a current reference track frame corresponding to the current track frame to be matched in the current reference track;

a relative pose determining unit configured to determine a first relative pose between the current track frame to be matched and the corresponding current reference track frame if the current reference track frame is successfully matched with the current track frame to be matched;

The first pose optimization unit is configured to perform pose optimization on all successfully matched track frames based on the first relative pose, absolute position information of the vehicle before entering the parking lot and after exiting the parking lot in each running track, and second relative pose among all track frames to be matched in the current track to be matched, and takes the track frames to be matched after pose optimization and the reference track frames corresponding to the successfully matched positions as alignment frames;

14. The apparatus of claim 13, wherein the apparatus further comprises:

the second pose optimization unit is configured to, after determining a first relative pose between the current track frame to be matched and the corresponding current reference track frame, perform pose optimization on all reference track frames in the current reference track based on the first relative pose, absolute position information of vehicles before entering a parking lot and after exiting the parking lot in each driving track, and a second relative pose between the track frames to be matched in the current track to be matched, and take the track frames to be matched after pose optimization and the reference track frames after pose optimization successfully match as alignment frames.

15. The apparatus of claim 13, wherein the matching unit comprises:

the matching subunit is configured to determine that the current reference track frame is successfully matched with the current track frame to be matched if the local characteristic information of the parking lot corresponding to the current reference track frame is consistent with the local characteristic information of the parking lot corresponding to the current track frame to be matched; or,

16. The apparatus according to any one of claims 13-15, wherein the apparatus further comprises:

correspondingly, the data association module is specifically configured to:

and taking the running track corresponding to the new current alignment result as a new current reference track, selecting another unmatched running track from the multiple-time collected parking lot running tracks as a new current track to be matched, and returning to execute the matching operation of the new current reference track and the new current track to be matched until the matching operation of the parking lot running tracks is finished, so as to obtain the target matching result of the vehicle pose.

17. The apparatus of claim 13, wherein the current reference trajectory frame determination unit comprises:

the first searching subunit is configured to acquire an absolute position corresponding to the current track frame to be matched under the condition that the current track frame to be matched belongs to a track before entering a parking lot or a track after exiting the parking lot, and search the current reference track within a first set distance range from the absolute position to acquire the current reference track frame corresponding to the current track frame to be matched;

Or,

the second searching subunit is configured to search in the current reference track within a first set distance range from the current position information according to the current position information corresponding to the current track frame to be matched under the condition that the current track frame to be matched belongs to the track in the parking lot, so as to obtain a candidate reference track frame;

the same-layer judging subunit is configured to judge whether the candidate reference track frame and the current track frame to be matched belong to the same layer of the parking lot;

and the current reference track frame determining subunit is configured to take the candidate reference track frame as a current reference track frame corresponding to the current track frame to be matched if the candidate reference track frame and the current track frame to be matched belong to the same layer of a parking lot.

18. The apparatus of claim 17, wherein the peer determination subunit comprises:

a shortest topology path determining component configured to determine a shortest topology path from the current track frame to be matched to the candidate reference track frame;

The same-layer determining component is configured to determine that the candidate reference track frame and the current track frame to be matched belong to the same layer of a parking lot if no ramp section exists in the shortest topological path;

or if a ramp section exists in the shortest topological path, and if the difference value between the height information of the ascending ramp section and the height information of the descending ramp section in the ramp section is within a second set distance range, determining that the candidate reference track frame and the current track frame to be matched belong to the same layer of a parking lot;

or if the difference value between the height information of the ascending road section and the height information of the descending road section in the ramp road section exceeds a second set distance range, determining that the candidate reference track frame and the current track frame to be matched do not belong to the same layer of the parking lot.

19. The apparatus of claim 18, wherein the peer determination component is specifically configured to:

20. The apparatus according to claim 13, wherein the first pose optimization unit is specifically configured to:

21. An electronic device, comprising:

one or more processors;

storage means for storing one or more programs,

when executed by the one or more processors, causes the one or more processors to implement the method of any of claims 1-11.

22. A storage medium having stored thereon a computer program, which when executed by a processor, implements the method of any of claims 1-11.

23. A vehicle comprising the system of any one of claims 1-11 or the electronic device of claim 21.