Memory parking initialization method and device, electronic equipment and storage medium

Abstract

The invention discloses a memory parking initialization method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining an initialization reference key frame through similarity calculation and a preset threshold value, and carrying out the initialization of the current frame according to the pose of the current frame relative to the initialization reference key frame and the pose of the initialization reference key frame in a map; the pose of the current frame relative to the map is obtained, pose initialization of the current frame is completed, and quick pose initialization of the current frame in the map can be achieved even under the conditions that no positioning signal exists and the vehicle does not drive to the starting point of the map.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and more specifically, to a memory parking initialization method, device, electronic device and storage medium. Background Art

[0002] At present, all major car companies are competing fiercely in the field of autonomous driving, which shows that intelligent driving has become increasingly important in people's lives today, and it is also a hot research topic for researchers. One of the more important functions in intelligent driving is memory automatic parking (Home-zone Parking Assist, HPA). HPA refers to the automatic parking system that automatically completes all instructions for the vehicle to automatically park in and out of a parking space from the parking lot entrance to the parking space and from the parking space to the parking lot exit without human intervention in a specific parking space in a specific area.

[0003] In the current HPA algorithm, for a specific local map, the user needs to drive to the vicinity of the starting point of the target map, or start at a location with GPS signal. If you need to locate any point in the target map trajectory without GPS signal, you need to build a map index or traverse the entire map to initialize the positioning, but this requires a lot of computing resources.

[0004] In the prior art, there is a method of prompting the user to drive to the vicinity of the starting point of the target map through navigation information, so as to quickly initialize the positioning. However, this method requires the user to drive the vehicle to the starting area of ​​the target map. It is not possible to enter the HPA tracking stage from any point in the HPA memory trajectory, nor is it possible to initialize the positioning during movement.

[0005] There is also a method of initializing positioning based on a global coordinate system in the prior art, but this method requires the user to drive to an area with GPS signals in advance to obtain global information, and it is impossible to perform positioning initialization directly in an underground parking lot.

[0006] In view of this, the present invention provides a memory parking initialization method, device, electronic device and storage medium to achieve rapid positioning initialization from any point near the HPA target map construction trajectory. Summary of the invention

[0007] In view of this, the present invention provides a memory parking initialization method, device, electronic device and storage medium to achieve rapid positioning initialization from any point near the HPA target map construction trajectory.

[0008] In one aspect, the present invention provides a memory parking initialization method, comprising:

[0009] S101: Acquire a map, where the map includes at least two key frames;

[0010] S102: extracting feature points after dedistortion processing of the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame;

[0011] S103: collecting a laser point cloud corresponding to the current frame, extracting feature points after dedistortion processing of the laser point cloud corresponding to the current frame, and obtaining a feature description set of the current frame;

[0012] S104: traversing the feature description sets of all the key frames, matching them with the feature description sets of the current frame respectively to obtain corresponding matching point pair sets and recording them, and performing similarity calculation on all the matching point pair sets respectively to obtain corresponding similarity values;

[0013] S105: Determine whether the maximum similarity value is greater than or equal to a preset threshold;

[0014] If the maximum similarity value is greater than or equal to the preset threshold, the key frame corresponding to the maximum similarity value is used as the initialization reference key frame; the posture of the current frame relative to the map is obtained based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map, and the posture initialization of the current frame is completed.

[0015] Optionally, determining whether the maximum similarity value is greater than or equal to the preset threshold further includes:

[0016] If the maximum similarity value is less than the preset threshold, the next frame is updated as the current frame, and steps S103 to S105 are repeated.

[0017] Optionally, the dedistortion processing of the laser point cloud corresponding to the key frame includes: selecting a coordinate system where the laser point cloud corresponding to any key frame is located as a reference coordinate system; converting the laser point clouds corresponding to all the key frames to the reference coordinate system;

[0018] The laser point cloud corresponding to the current frame is dedistorted, including: converting the laser point cloud corresponding to the current frame to the reference coordinate system.

[0019] Optionally, the pose of the current frame relative to the initialization reference key frame is obtained in the following manner:

[0020] Obtaining the matching point pair set corresponding to the initialization reference key frame;

[0021] The pose of the current frame relative to the initialization reference key frame is calculated according to the set of matching point pairs corresponding to the initialization reference key frame.

[0022] Optionally, traversing the feature description sets of all the key frames and matching them with the feature description sets of the current frame to obtain the corresponding matching point pair sets includes:

[0023] The key frame feature description set includes key frame feature points, and the key frame feature points include corresponding key frame high-dimensional feature description vector sets;

[0024] The feature description set of the current frame includes feature points of the current frame, and the feature points of the current frame include corresponding high-dimensional feature description vector sets of the current frame;

[0025] Calculate the maximum bipartite graph matching of all the key frame high-dimensional feature description vector sets and all the current frame high-dimensional feature description vector sets in the feature description set of the key frame;

[0026] The key frame feature points corresponding to the key frame high-dimensional feature description vector set and the current frame feature points corresponding to the current frame high-dimensional feature description vector set that matches the key frame high-dimensional feature description vector set constitute matching point pairs, and all the matching point pairs constitute the matching point pair set.

[0027] Optionally, calculating the pose of the current frame relative to the initialization reference key frame according to the set of matching point pairs corresponding to the initialization reference key frame includes:

[0028] Constructing an overdetermined equation according to the set of matching point pairs corresponding to the initialization reference key frame;

[0029] The overdetermined equation is solved by the least square method to obtain the pose of the current frame relative to the initialization reference key frame.

[0030] Optionally, the overdetermined equation is calculated in the following manner:

[0031] Q1=T×P1

[0032] …

[0033] Q i =T×P i

[0034] …

[0035] Q n =T×P n

[0036] Where T is the transformation matrix, (Q i , P i ) is the i-th matching point pair in the matching point pair set corresponding to the initialization reference key frame, P i is the feature point of the i-th current frame, Q iIt is the i-th key frame feature point in the feature description set of the initialization reference key frame, where 1≤i≤n, n≥4 and n is an integer.

[0037] On the other hand, the present invention also provides a memory parking initialization device, comprising:

[0038] A map providing module, used for obtaining a map, wherein the map includes at least two key frames;

[0039] A data processing module, used to process the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame and to process the laser point cloud corresponding to the current frame to obtain a feature description set of the current frame;

[0040] A similarity calculation module is used to traverse the feature description sets of all the key frames, match them with the feature description set of the current frame respectively to obtain corresponding matching point pair sets and record them, and perform similarity calculation on all the matching point pair sets to obtain corresponding similarity values;

[0041] An initialization module is used to use the key frame corresponding to the maximum similarity value greater than or equal to a preset threshold as an initialization reference key frame, obtain the posture of the current frame relative to the map based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map, and complete the posture initialization of the current frame.

[0042] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the electronic device implements any one of the above-mentioned memory parking initialization methods.

[0043] In yet another aspect, the present invention further provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any of the above-mentioned memory parking initialization methods.

[0044] Compared with the prior art, the memory parking initialization method, device, electronic device and storage medium provided by the present invention achieve at least the following beneficial effects:

[0045] The memory parking initialization method, device, electronic device and storage medium provided by the present invention obtain an initialization reference key frame through similarity calculation and a preset threshold, obtain the posture of the current frame relative to the map according to the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map, and complete the posture initialization of the current frame. Even in the absence of a positioning signal and the vehicle has not reached the starting point of the map, the posture of the current frame on the map can be quickly initialized.

[0046] Of course, any product implementing the present invention does not necessarily need to achieve all of the technical effects described above at the same time.

[0047] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0049] Figure 1 It is a flow chart of the memory parking initialization method provided by the present invention;

[0050] Figure 2 It is another flow chart of the memory parking initialization method provided by the present invention.

[0051] Figure 3 It is a structural schematic diagram of the memory parking initialization device provided by the present invention.

[0052] In the figure: 100, memory parking initialization device; 1, image supply module; 2, data processing module; 3, similarity calculation module; 4, initialization module. DETAILED DESCRIPTION

[0053] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0054] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0055] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0056] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0057] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0058] Example 1

[0059] Combination Figure 1 , Figure 1 : is a flow chart of the memory parking initialization method provided by the present invention, to illustrate a specific embodiment of the memory parking initialization method provided by the present invention, including:

[0060] S101: Acquire a map, where the map includes at least two key frames, and each key frame includes a corresponding laser point cloud;

[0061] S102: extracting feature points after dedistortion processing of the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame;

[0062] S103: collecting a laser point cloud corresponding to the current frame, extracting feature points after dedistortion processing of the laser point cloud corresponding to the current frame, and obtaining a feature description set of the current frame;

[0063] S104: traversing the feature description sets of all key frames, matching them with the feature description set of the current frame to obtain corresponding matching point pair sets and recording them, and performing similarity calculation on all matching point pair sets to obtain corresponding similarity values;

[0064] S105: Determine whether the maximum similarity value is greater than or equal to a preset threshold;

[0065] S1051: If the maximum similarity value is greater than or equal to a preset threshold, the key frame corresponding to the maximum similarity value is used as the initialization reference key frame; the posture of the current frame relative to the map is obtained based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map, and the posture initialization of the current frame is completed.

[0066] It should be noted that in step S101, when establishing a map, the movement frequency can be preset, and key frames can be selected according to the movement frequency interval. Of course, it is not limited to this, and the selection of key frames can be selected according to actual needs. In step S103, the laser point cloud of the current frame is a set of point cloud data collected by the radar within a preset time period from the current moment. In step S105, a preset threshold is set to ensure that the feature description set of the filtered key frame has a high similarity with the feature description set of the current frame, thereby ensuring the reliability of the positioning result and avoiding the feature description set of the key frame corresponding to the maximum similarity value having a low similarity with the feature description set of the current frame, resulting in errors in subsequent conversions. The value of the preset threshold can be set according to actual needs, and this embodiment does not make specific limitations on this.

[0067] It is understandable that for the laser instant positioning and map construction technical route, it is necessary to extract the features of the laser point cloud. Therefore, no additional computing power is required to obtain the feature description set of the key frame. Then, the feature description set corresponding to the current frame is obtained and quickly matched with the feature description set of the key frame for similarity matching to provide the initial pose for positioning initialization. The overall computational requirements are relatively small.

[0068] Compared with the prior art, the memory parking initialization method provided by the present invention obtains an initialization reference key frame through similarity calculation and a preset threshold, and obtains the posture of the current frame relative to the map according to the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map to complete the posture initialization of the current frame. Even in the absence of a positioning signal (GPS signal) and the vehicle has not reached the starting point of the map, the posture of the current frame on the map can be quickly initialized.

[0069] Example 2

[0070] Reference Figure 2 , Figure 2 FIG. 1 is another flow chart of the memory parking initialization method provided by the present invention, to illustrate another specific embodiment of the memory parking initialization method provided by the present invention, including:

[0071] S101: Acquire a map, where the map includes at least two key frames, and each key frame includes a corresponding laser point cloud.

[0072] S102: extracting feature points after dedistortion processing of the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame;

[0073] In step S102, the laser point cloud corresponding to the key frame is subjected to dedistortion processing, including: selecting the coordinate system where the laser point cloud corresponding to any key frame is located as the reference coordinate system; converting the laser point clouds corresponding to all key frames to the reference coordinate system. Feature points are extracted from the laser point cloud corresponding to the key frame, and the extraction method includes but is not limited to the intrinsic shape signatures (ISS) method or the scale invariant feature transform 3D (SIFT3D) method.

[0074] S103: collecting a laser point cloud corresponding to the current frame, extracting feature points after dedistortion processing of the laser point cloud corresponding to the current frame, and obtaining a feature description set of the current frame;

[0075] In step S103, the laser point cloud corresponding to the current frame is subjected to dedistortion processing, including: converting the laser point cloud corresponding to the current frame to a reference coordinate system. Similarly, feature points are extracted from the laser point cloud corresponding to the current frame, and the extraction method includes but is not limited to the intrinsic shape signatures (ISS) method or the scale invariant feature transform 3D (SIFT3D) method.

[0076] S104: Traverse the feature description sets of all key frames, match them with the feature description set of the current frame respectively to obtain corresponding matching point pair sets and record them, and perform similarity calculation on all matching point pair sets to obtain corresponding similarity values.

[0077] In step S104, the feature description sets of all key frames are traversed and matched with the feature description set of the current frame to obtain corresponding matching point pair sets, including:

[0078] The feature description set of the key frame includes key frame feature points, and the key frame feature points include corresponding key frame high-dimensional feature description vector sets;

[0079] The feature description set of the current frame includes feature points of the current frame, and the feature points of the current frame include a corresponding high-dimensional feature description vector set of the current frame;

[0080] Calculate the maximum bipartite graph matching of all key frame high-dimensional feature description vector sets and all current frame high-dimensional feature description vector sets in the feature description set of the key frame;

[0081] The key frame feature points corresponding to the key frame high-dimensional feature description vector set and the current frame feature points corresponding to the current frame high-dimensional feature description vector set that matches the key frame high-dimensional feature description vector set constitute matching point pairs, and all matching point pairs constitute a matching point pair set.

[0082] Specifically, the Hungarian algorithm may be used to calculate the maximum bipartite graph matching of all key frame high-dimensional feature description vector sets and all current frame high-dimensional feature description vector sets in the key frame feature description set, but of course, it is not limited thereto.

[0083] S105: Determine whether the maximum similarity value is greater than or equal to a preset threshold.

[0084] S1051: If the maximum similarity value is greater than or equal to a preset threshold, the key frame corresponding to the maximum similarity value is used as the initialization reference key frame; the pose of the current frame relative to the map is obtained based on the pose of the current frame relative to the initialization reference key frame and the pose of the initialization reference key frame on the map, and the pose initialization of the current frame is completed;

[0085] In step S1051, the pose of the current frame relative to the initialization reference key frame is obtained in the following manner: a matching point pair set corresponding to the initialization reference key frame is obtained. Specifically, the matching point pair set recorded when traversing the feature description sets of all key frames can be directly queried.

[0086] The pose of the current frame relative to the initialization reference key frame is calculated based on the set of matching point pairs corresponding to the initialization reference key frame.

[0087] The pose of the current frame relative to the initialization reference key frame is calculated based on the matching point pair set corresponding to the initialization reference key frame, including:

[0088] An overdetermined equation is constructed according to a matching point pair set corresponding to an initialization reference key frame, wherein the number of matching point pairs in the matching point pair set corresponding to the initialization reference key frame is greater than or equal to 4 pairs;

[0089] The least squares method is used to solve the overdetermined equations to obtain the pose of the current frame relative to the initialization reference key frame.

[0090] Specifically, the overdetermined equation is calculated as follows:

[0091] Q1=T×P1

[0092] …

[0093] Q i =T×P i

[0094] …

[0095] Q n =T×P n

[0096] Where T is the transformation matrix, (Q i , P i ) is the i-th matching point pair in the matching point pair set corresponding to the initial reference key frame, P i is the feature point of the i-th current frame, Q i To initialize the i-th key frame feature point in the feature description set of the reference key frame, 1≤i≤n, n≥4 and n is an integer.

[0097] S1052: If the maximum similarity value is less than the preset threshold, the next frame is updated as the current frame, and steps S103 to S105 are repeated.

[0098] It should be noted that when the radar is scanning, the vehicle where the radar is located has a certain degree of rotation and translation. Therefore, the laser spot beams emitted at different time points are not in a strictly unified coordinate system, and de-distortion processing is required to transform the laser point cloud into a coordinate system corresponding to a unified timestamp. The coordinate system can be the coordinate system corresponding to the laser point cloud corresponding to any key frame. For example, the coordinate system corresponding to the laser point cloud corresponding to the first determined key frame is selected. Of course, it is not limited to this. The laser point cloud de-distortion can be carried out using existing technology, and this embodiment will not go into details.

[0099] The feature description set of a key frame includes key frame feature points, and the key frame feature points include corresponding key frame high-dimensional feature description vector sets. The key frame high-dimensional feature description vector set includes N key frame high-dimensional feature description vectors, where N is a positive number. The key frame high-dimensional feature description vector can be a high-dimensional feature description of features such as normal, curvature, and neighborhood characteristics. The number of types of key frame high-dimensional feature description vectors corresponding to each key frame feature point is the same. Similarly, the feature description set of the current frame includes the current frame feature point, and the current frame feature point has a corresponding current frame high-dimensional feature description vector set. The current frame high-dimensional feature description vector set includes N key frame high-dimensional feature description vectors, and the number of types of key frame high-dimensional feature description vectors corresponding to each current frame feature point is the same, and the number of types of key frame high-dimensional feature description vectors corresponding to the key frame feature point is also the same, which is convenient for similarity calculation and bipartite graph matching. Specifically, the feature description set of each key frame includes at least 4 key frame feature points, and the feature description set of the current frame includes at least 4 current frame feature points, ensuring that the matching point pair set calculated using the Hungarian algorithm includes at least 4 matching point pairs, thereby realizing the construction of overdetermined equations.

[0100] It can be understood that the coordinate system conversion calculation of the current frame and the initialization reference key frame is performed according to the set of matching point pairs to obtain the pose of the current frame relative to the initialization reference key frame. According to the pose of the current frame relative to the initialization reference key frame and the pose of the initialization reference key frame on the map, the pose of the current frame relative to the map is obtained to complete the pose initialization of the current frame. Even in the absence of a positioning signal and the vehicle has not reached the starting point of the map, the pose of the current frame on the map can be quickly initialized.

[0101] Example 3

[0102] This embodiment describes another specific embodiment of the memory parking initialization method provided by the present invention, including:

[0103] Determine a key frame at a certain distance and record the pose information A corresponding to each key frame i Constitute the laser point cloud set P.

[0104] The laser point cloud set P is subjected to dedistortion processing, 3D feature extraction and feature description. Specifically, for the laser point cloud of the i-th key frame, several 3D feature points FP are obtained after processing. ij The corresponding high-dimensional feature description vector FD ij , forming the feature description set S corresponding to the i-th key frame i The feature description sets of all key frames constitute a set S = {S1, S2, ..., S N} and save it in the system.

[0105] The laser point cloud collected in the current frame is subjected to dedistortion, feature point extraction and feature description according to the same process, and the feature description set S of the current frame is generated. now .

[0106] Traverse the set S of key frame feature descriptions, for the i-th feature description set S i The high-dimensional feature description vector set FD i , and the feature description set S of the current frame now The high-dimensional feature description vector set FD now Perform bipartite graph matching and calculate the similarity value based on the matching results. The key frame with the highest similarity value that exceeds a certain threshold is used as the initialization reference key frame. The n pairs of matching point pairs obtained by matching the initialization reference key frame with the current frame bipartite graph are saved in the matching point pair set Pair, where n is a positive number and n is greater than or equal to 4. The two points of the matching point pair are located in the local coordinate system of the current frame and the local coordinate system of the initialization reference key frame respectively. The two points need to be transformed to the same coordinate system, so the equation is constructed:

[0107] Q1=T×P1

[0108] …

[0109] Q i =T×P i

[0110] …

[0111] Q n =T×P n

[0112] Where T is the transformation matrix, (Q i , P i ) is the i-th matching point pair in the matching point pair set Pair, P i is the feature point of the i-th current frame, Q i is the i-th key frame feature point in the feature description set of the initial reference key frame, 1≤i≤n, n≥4 and n is an integer. When n≥4, the equation is an overdetermined equation, and Q i =T×P iIt is an equation established from three dimensions: x-direction, y-direction and z-direction.

[0113] The transformation matrix is ​​a homogeneous matrix and is calculated as follows:

[0114]

[0115] Among them, R 11 , R 12 and R 13 P i Rotation parameter around the x direction, R 21 , R 22 and R 23 P i Rotation parameter around the y direction, R 31 , R 32 and R 33 P i Rotation parameter around the z direction, T1 is P i The movement variable along the x direction, T2 is P i The movement variable along the y direction, T3 is P i The movement variable in the z direction.

[0116] Substitute the transformation matrix into the overdetermined equation for solution. At this time, only R 11 , R 12 , R 13 , R 21 , R 22 , R 23 , R 31 , R 32 , R 33 , T1, T2 and T3 are unknown numbers, and the least square method is used to solve them to obtain the pose of the current frame relative to the initialization reference keyframe. Since the pose of the initialization reference keyframe on the map is known, combined with the pose of the current frame relative to the initialization reference keyframe, the pose of the current frame relative to the map can be obtained, thereby completing the pose initialization of the current frame, without the need for the vehicle to be in a position with a positioning signal or to drive to the starting point of the map.

[0117] Example 4

[0118] Based on the same invention idea, refer to Figure 3 , Figure 3 1 is a structural schematic diagram of a memory parking initialization device provided by the present invention, to illustrate a specific embodiment of the memory parking initialization device 100 provided by the present invention, including:

[0119] A map providing module 1, coupled to the data processing module 2, is used to obtain a map, wherein the map includes at least two key frames;

[0120] The data processing module 2 is coupled to the similarity calculation module 3 and is used to process the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame and to process the laser point cloud corresponding to the current frame to obtain a feature description set of the current frame;

[0121] The similarity calculation module 3 is coupled to the initialization module 4 and is used to traverse the feature description sets of all key frames and perform similarity calculations with the feature description set of the current frame to obtain similarity values;

[0122] The initialization module 4 is coupled to the key frame determination module 1, and is used to use the key frame corresponding to the maximum similarity value greater than or equal to a preset threshold as the initialization reference key frame, and complete the initialization by obtaining the posture of the current frame relative to the map based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map.

[0123] Example 5

[0124] Based on the same inventive concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the electronic device implements a memory parking initialization method as described in any one of the above embodiments.

[0125] Example 6

[0126] Based on the same inventive concept, the present invention further provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the memory parking initialization method as described in any one of the above embodiments is implemented.

[0127] It can be seen from the above embodiments that the memory parking initialization method, device, electronic device and storage medium provided by the present invention achieve at least the following beneficial effects:

[0128] The memory parking initialization method, device, electronic device and storage medium provided by the present invention obtain an initialization reference key frame through similarity calculation and a preset threshold, and complete the initialization by obtaining the posture of the current frame relative to the map based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map. Even in the absence of a positioning signal and the vehicle has not reached the starting point of the map, the posture of the current frame on the map can be quickly initialized.

[0129] Although some specific embodiments of the present invention have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are only for illustration and not for limiting the scope of the present invention. In the above embodiments of the present invention, the description of each embodiment has its own emphasis, and for the part that is not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

Claims

1. A memory parking initialization method, characterized in that: include: S101: Acquire a map, where the map includes at least two key frames; S102: extracting feature points after dedistortion processing of the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame; S103: collecting a laser point cloud corresponding to the current frame, extracting feature points after dedistortion processing of the laser point cloud corresponding to the current frame, and obtaining a feature description set of the current frame; S104: traversing the feature description sets of all the key frames, matching them with the feature description sets of the current frame respectively to obtain corresponding matching point pair sets and recording them, and performing similarity calculation on all the matching point pair sets respectively to obtain corresponding similarity values; S105: Determine whether the maximum similarity value is greater than or equal to a preset threshold; If the largest similarity value is greater than or equal to the preset threshold, taking the key frame corresponding to the largest similarity value as the initialization reference key frame; The pose of the current frame relative to the map is obtained based on the pose of the current frame relative to the initialization reference key frame and the pose of the initialization reference key frame on the map, thereby completing the pose initialization of the current frame.

2. The memory parking initialization method according to claim 1, characterized in that: Determining whether the maximum similarity value is greater than or equal to the preset threshold value further includes: If the maximum similarity value is less than the preset threshold, the next frame is updated as the current frame, and steps S103 to S105 are repeated.

3. The memory parking initialization method according to claim 1, characterized in that: The laser point cloud corresponding to the key frame is subjected to dedistortion processing, including: selecting a coordinate system where the laser point cloud corresponding to any key frame is located as a reference coordinate system; converting the laser point clouds corresponding to all the key frames to the reference coordinate system; The laser point cloud corresponding to the current frame is dedistorted, including: converting the laser point cloud corresponding to the current frame to the reference coordinate system.

4. The memory parking initialization method according to claim 1, characterized in that: The pose of the current frame relative to the initialization reference key frame is obtained in the following manner: Obtaining the matching point pair set corresponding to the initialization reference key frame; The pose of the current frame relative to the initialization reference key frame is calculated according to the set of matching point pairs corresponding to the initialization reference key frame.

5. The memory parking initialization method according to claim 1, characterized in that: Traversing the feature description sets of all the key frames and matching them with the feature description sets of the current frame to obtain the corresponding matching point pair sets, including: The key frame feature description set includes key frame feature points, and the key frame feature points include corresponding key frame high-dimensional feature description vector sets; The feature description set of the current frame includes feature points of the current frame, and the feature points of the current frame include corresponding high-dimensional feature description vector sets of the current frame; Calculate the maximum bipartite graph matching of all the key frame high-dimensional feature description vector sets and all the current frame high-dimensional feature description vector sets in the feature description set of the key frame; The key frame feature points corresponding to the key frame high-dimensional feature description vector set and the current frame feature points corresponding to the current frame high-dimensional feature description vector set that matches the key frame high-dimensional feature description vector set constitute matching point pairs, and all the matching point pairs constitute the matching point pair set.

6. The memory parking initialization method according to claim 4, characterized in that: Calculating the pose of the current frame relative to the initialization reference key frame according to the set of matching point pairs corresponding to the initialization reference key frame includes: Constructing an overdetermined equation according to the set of matching point pairs corresponding to the initialization reference key frame; The overdetermined equation is solved by the least square method to obtain the pose of the current frame relative to the initialization reference key frame.

7. The memory parking initialization method according to claim 6, characterized in that: The overdetermined equation is calculated as follows: Q1=T×P1 … Q i =T×P i … Q n =T×Pn Where T is the transformation matrix, (Q i , P i ) is the i-th matching point pair in the matching point pair set corresponding to the initialization reference key frame, P i is the feature point of the i-th current frame, Q i It is the i-th key frame feature point in the feature description set of the initialization reference key frame, where 1≤i≤n, n≥4 and n is an integer.

8. A memory parking initialization device, characterized in that: include: A map providing module, used for obtaining a map, wherein the map includes at least two key frames; A data processing module, used to process the laser point cloud corresponding to the key frame to obtain a feature description set of the key frame and to process the laser point cloud corresponding to the current frame to obtain a feature description set of the current frame; A similarity calculation module is used to traverse the feature description sets of all the key frames, match them with the feature description set of the current frame respectively to obtain corresponding matching point pair sets and record them, and perform similarity calculation on all the matching point pair sets to obtain corresponding similarity values; An initialization module is used to use the key frame corresponding to the maximum similarity value greater than or equal to a preset threshold as an initialization reference key frame, obtain the posture of the current frame relative to the map based on the posture of the current frame relative to the initialization reference key frame and the posture of the initialization reference key frame on the map, and complete the posture initialization of the current frame.

9. An electronic device, characterized in that: The electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the electronic device implements the memory parking initialization method as claimed in any one of claims 1 to 7.

10. A storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the memory parking initialization method according to any one of claims 1 to 7 is implemented.

Images