A vehicle positioning method, device, electronic equipment and storage medium
By acquiring and filtering the observation vectors of feature points in vehicle positioning, matching feature points are determined and their positions are adjusted, thus solving the problem of high feature point matching error rate and improving the accuracy of vehicle positioning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA AUTOMOTIVE INNOVATION CORP
- Filing Date
- 2024-03-29
- Publication Date
- 2026-08-04
AI Technical Summary
In existing technologies, the high error rate of feature point matching leads to low accuracy in vehicle positioning.
By acquiring the observation vector and position information of each feature point in the parking area map corresponding to the target vehicle, and combining it with the observation direction of the current environmental image, feature point projection and filtering are performed to determine the matching third feature point. The position information of these feature points in the parking area map is then used to adjust the position and achieve vehicle positioning.
This improved the accuracy and efficiency of feature point matching, thereby enhancing the accuracy of vehicle positioning.
Smart Images
Figure CN118196453B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of autonomous driving, and more specifically, to a vehicle positioning method, apparatus, electronic device, and storage medium. Background Technology
[0002] In the field of autonomous driving parking localization, feature point localization is a commonly used method. The basic process of feature point localization is to extract some repeatable corner points from the original image, perform 3D reconstruction on these corner points, save their features, and correspondingly, the saved map is used as a feature point map.
[0003] The combination of corner points extracted from a single frame of an image and descriptors obtained by processing images extracted near those corner points using a descriptor calculation function is called a feature point.
[0004] Currently, the common storage methods for feature point maps are either to directly access the optimized 3D coordinates and the feature descriptor observed in the first frame, or to select and store the descriptor with the smallest distance among all remaining descriptors.
[0005] However, in reality, some feature points may be located on the edges or vertices of objects. These points appear to have different background elements when viewed from different angles, leading to instability in the height of the descriptors corresponding to the feature points. If the angles used to calculate the descriptors for the feature point map differ from those used during observation, these point pairs become difficult to match, resulting in wasted feature points and affecting the accuracy of feature point matching. Consequently, the accuracy of vehicle localization is low. Therefore, a new vehicle localization method is needed to achieve accurate vehicle positioning. Summary of the Invention
[0006] In view of this, embodiments of the present invention provide a vehicle positioning method, device, electronic device, and storage medium to solve the problem of low vehicle positioning accuracy caused by high feature point matching error rate in automatic parking technology.
[0007] One aspect of the present invention provides a vehicle positioning method, the method comprising the following steps:
[0008] In response to a positioning request for a target vehicle, the system acquires a first observation vector, a first location information, and a current image observation direction corresponding to the current environment image for each second feature point in the map of the parking area corresponding to the target vehicle; the first location information is the location information of the first camera when the current environment image is acquired.
[0009] Project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image to obtain the feature point projection position corresponding to each second feature point within the current image observation range;
[0010] Based on the projection position of the feature point and the first position information, determine the current observation vector corresponding to each second feature point within the current image observation range;
[0011] Based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, the second feature points within the current image observation range are filtered to obtain a third feature point that matches the first feature point in the current environment image;
[0012] Based on the fourth location information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the position of the target vehicle is adjusted to obtain the target position of the target vehicle.
[0013] Another aspect of the present invention provides a bus description file error correction processing apparatus, the apparatus comprising:
[0014] The first information acquisition module is used to respond to a positioning request for a target vehicle by acquiring a first observation vector, a first location information, and a current image observation direction corresponding to the current environment image for each second feature point in the map of the parking area corresponding to the target vehicle; the first location information is the location information of the first camera when the current environment image is acquired.
[0015] The second information acquisition module is used to project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image, and to acquire the feature point projection position corresponding to each second feature point within the current image observation range;
[0016] The second determining module is used to determine the current observation vector corresponding to each second feature point within the current image observation range based on the feature point projection position and the first position information;
[0017] The matching module is used to filter the second feature points within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, so as to obtain a third feature point that matches the first feature point in the current environment image.
[0018] The position adjustment module is used to adjust the position of the target vehicle based on the fourth position information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, so as to obtain the target position of the target vehicle.
[0019] Another aspect of the present invention provides an electronic device, comprising:
[0020] processor;
[0021] Memory used to store the processor's executable instructions;
[0022] The processor is configured to execute the instructions to implement the vehicle positioning method described in any one of the above descriptions.
[0023] Another aspect of the present invention provides a computer-readable storage medium that, when the instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the vehicle positioning method described in any one of the preceding descriptions.
[0024] This invention provides a vehicle positioning method, device, electronic device, and storage medium. By acquiring the first observation vector, first position information, current image observation direction corresponding to the current environment image, and the feature point projection position corresponding to each second feature point within the current image observation range, and the current observation vector corresponding to each second feature point within the current image observation range determined by the first position information, the second feature points within the current image observation range are filtered to obtain a third feature point matching the first feature point in the current environment image. This reduces the error in determining the third feature point and improves the accuracy and efficiency of feature point matching. Furthermore, by utilizing the fourth position information of the third feature point in the parking area map and the feature point projection position corresponding to the third feature point, the position of the target vehicle is adjusted to obtain the target position of the target vehicle, further improving the accuracy of vehicle positioning. Attached Figure Description
[0025] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0026] Figure 1 This is a flowchart illustrating a vehicle positioning method according to an exemplary embodiment;
[0027] Figure 2 This is a schematic diagram of a process for constructing a parking area map according to an exemplary embodiment;
[0028] Figure 3 This is a flowchart illustrating a process for obtaining a first observation vector corresponding to each second feature point in a map of a parking area corresponding to a target vehicle, according to an exemplary embodiment.
[0029] Figure 4 This is a schematic diagram of a process for determining a third feature point that matches a first feature point, according to an exemplary embodiment.
[0030] Figure 5 This is a flowchart illustrating a process for determining the angle between multiple observation vectors corresponding to each second feature point within the current image observation range, according to an exemplary embodiment.
[0031] Figure 6 This is a schematic diagram of a process for filtering each second feature point within the current image observation range based on the included angle of multiple observation vectors corresponding to each second feature point within the current salient observation range, according to an exemplary embodiment.
[0032] Figure 7 This is a schematic diagram of a process for determining the target position of a target vehicle based on the fourth position information corresponding to the third feature point and the projection point position information corresponding to the third feature point, according to an exemplary embodiment.
[0033] Figure 8 This is a schematic diagram of a process for adjusting the position of a target vehicle based on error term data, according to an exemplary embodiment.
[0034] Figure 9 This is a schematic diagram of the structure of a vehicle positioning device according to an exemplary embodiment. Detailed Implementation
[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0036] It should be noted that the terms "first," "second," etc., in the specification and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.
[0037] Figure 1This is a flowchart illustrating a vehicle positioning method according to an exemplary embodiment. This specification provides method operation steps as shown in the embodiments or flowcharts, but based on conventional or non-inventive labor, more or fewer operation steps may be included. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only possible execution order. In actual system or server product execution, the method can be executed in the order shown in the embodiments or drawings, or in parallel (e.g., in a parallel processor or multi-threaded processing environment). Specifically, as shown... Figure 1 The illustration shows an embodiment of a vehicle positioning method according to this application. The method may include:
[0038] S101: In response to a positioning request for the target vehicle, obtain the first observation vector, the first location information, and the current image observation direction corresponding to the current environment image for each second feature point in the map of the parking area corresponding to the target vehicle.
[0039] In a specific embodiment, each second feature point can be a feature description point extracted from the parking area map; the current environmental image can be an environmental image around the target vehicle captured by the first camera on the target vehicle when the positioning request is triggered; the first observation vector can be the direction vector between the position information of each second feature point in multiple environmental images corresponding to the parking area map and the position information of the first camera when multiple environmental images are captured; the first position information can be the position information of the first camera when the current environmental image is captured.
[0040] Figure 2 This is a schematic diagram of a process for constructing a parking area map according to an exemplary embodiment. In an optional embodiment, such as... Figure 2 As shown, the above method may further include:
[0041] S201: Acquire multiple environmental images;
[0042] S203: Construct a parking area map based on multiple environmental images.
[0043] In one specific embodiment, each time the target vehicle receives a positioning request, it sends a shooting command to the first camera, obtains the third location information of the first camera and stores the third location information. After receiving the shooting command, the first camera will take pictures of the environment around the target vehicle. Specifically, it obtains environmental images of the target vehicle at the time of each positioning request, and further, it obtains multiple environmental images.
[0044] In a specific embodiment, when multiple environmental images are acquired, each second feature point in the multiple environmental images, the second feature description information corresponding to each second feature point in the environmental images, and the second location information corresponding to each second feature point in the environmental images can be acquired accordingly. Further, the multiple second location information corresponding to each second feature point in the multiple environmental images is spatially transformed and stored. That is, the multiple second location information corresponding to each second feature point in the multiple environmental images is converted into corresponding spatial location information. Further, the spatial location information corresponding to each second feature point is obtained. Further, based on each second feature point in the multiple environmental images, a parking area feature map is constructed using a map building algorithm and combined with the obtained spatial location information of each second feature point.
[0045] In the above embodiments, by using multiple environmental images to construct a parking area map, the application scenario is combined with the actual situation, thereby improving the practicality of the present invention.
[0046] In one specific embodiment, when the target vehicle receives a location request again, the target vehicle will obtain the first observation vector, the first location information, and the current image observation direction corresponding to each second feature point in the map of the parking area corresponding to the target vehicle.
[0047] Figure 3 This is a flowchart illustrating a process for obtaining a first observation vector corresponding to each second feature point in a map of a parking area corresponding to a target vehicle, according to an exemplary embodiment. In an optional embodiment, such as... Figure 3 As shown, before obtaining the first observation vector corresponding to each second feature point in the map of the parking area corresponding to the target vehicle, the above method may further include:
[0048] S301: Obtain multiple second location information and multiple third location information corresponding to each second feature point in the map of the parking area corresponding to the target vehicle;
[0049] S303: Determine the first observation vector corresponding to each second feature point based on multiple second position information and multiple third position information.
[0050] In a specific embodiment, the multiple second location information can be the location information of each second feature point in multiple environmental images corresponding to the parking area map; the multiple third location information can be the location information of the first camera when multiple environmental images are collected respectively; the multiple environmental images are multiple environmental images used to construct the parking area map.
[0051] In a specific embodiment, after the target vehicle responds to the positioning request, it acquires multiple second location information corresponding to each second feature point in multiple environmental images stored in the target vehicle, multiple second feature description information and third location information corresponding to each second feature point in the regional feature map, and acquires the first location information of the first camera.
[0052] In a specific embodiment, each second location information corresponding to each second feature point corresponds one-to-one with the third location information of the first camera when acquiring the environmental image corresponding to the second location information. Optionally, multiple environmental images are acquired in this invention. Further, multiple second location information of each second feature point can be obtained. Optionally, multiple third location information of the first camera corresponding to multiple environmental images can be obtained. Further, by connecting the second feature points with the first camera, a first observation vector corresponding to each second feature point is obtained. Optionally, the first camera captures an environmental image A, which contains multiple second feature points. Further, multiple second location information of each second feature point in environmental image A is obtained. Further, the third location information of the first camera when capturing environmental image A is obtained. Further, the first camera and each second feature point in environmental image A are connected by position. A line is drawn to obtain the first observation vector corresponding to each second feature point in environmental image A. Optionally, an environmental image B is captured by the first camera, and environmental image B contains multiple second feature points (the multiple second feature points in environmental image B are the same as those in environmental image A). Further, multiple second position information of each second feature point in environmental image B is obtained. Further, the third position information of the first camera when capturing environmental image B is obtained. Further, the position of the first camera and each second feature point in environmental image B is connected to obtain the first observation vector corresponding to each second feature point in environmental image B. Further, the first observation vector corresponding to each second feature point in the parking area map is obtained. Optionally, the first observation vector corresponding to each second feature point in environmental image B and the first observation vector corresponding to each second feature point in environmental image A can be used as the first observation vector corresponding to each second feature point in the parking area map.
[0053] In the above embodiments, by determining the first direction vector between the position information of each second feature point in multiple environmental images corresponding to the parking area map and the position information of the first camera when acquiring multiple environmental images, a foundation is laid for subsequent feature point matching, thereby improving the stability and accuracy of matching.
[0054] S103: Project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image, and obtain the feature point projection position corresponding to each second feature point within the current image observation range.
[0055] In a specific embodiment, when capturing the current environment image, the current image observation direction corresponding to the current environment image is obtained. Accordingly, based on the current image observation direction corresponding to the current environment image, the current image observation range in the parking area map can be determined. Further, multiple second feature points in the current image observation range of the parking area map are projected onto the current environment image, and the feature point projection positions of multiple second feature points in the current image observation range of the parking area map in the current environment image are further obtained. Optionally, multiple second feature points in the parking area map within the current image observation range are the same feature points as each second feature point in the current environment image. However, due to spatial transformation and other reasons, the projection position information of each second feature point within the current image observation range is not the same as the position information of each second feature point in the current environment image.
[0056] S105: Based on the feature point projection position and the first position information, determine the current observation vector corresponding to each second feature point within the current image observation range.
[0057] In a specific embodiment, the feature point projection position corresponding to each second feature point within the current image observation range is connected with the first position information corresponding to each second feature point in the current environment image, and the current observation vector corresponding to each second feature point within the current image observation range can be determined accordingly.
[0058] In the above embodiments, by determining the first observation vector corresponding to each second feature point and the current observation vector corresponding to each second feature point, the foundation for subsequent vector calculation is laid, and the calculation efficiency is further improved. While reducing the amount of calculation, the matching efficiency of feature points is correspondingly improved.
[0059] S107: Based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, filter the second feature points within the current image observation range to obtain the third feature points that match the first feature points in the current environment image.
[0060] In a specific embodiment, each first feature point can be a feature description point extracted from the current environmental image; the first feature description information can be the feature description information describing each first feature point in the current environmental image.
[0061] Figure 3 This is a schematic flowchart illustrating a process for determining a third feature point that matches a first feature point, according to an exemplary embodiment. In an optional embodiment, such as... Figure 4As shown, the above-mentioned filtering of the second feature points within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range to obtain a third feature point that matches the first feature point in the current environment image may include:
[0062] S401: Based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, determine the angle between multiple observation vectors corresponding to each second feature point within the current image observation range;
[0063] S403: Based on the angle between multiple observation vectors corresponding to each second feature point within the current image observation range, filter the second feature points within the current image observation range to obtain the fourth feature point;
[0064] S405: Based on the first feature description information corresponding to each first feature point and the second feature description information corresponding to the fourth feature point, determine the fourth feature point that has the same first feature description information as the first feature point from the fourth feature points, and use it as the third feature point.
[0065] In a specific embodiment, after receiving the current shooting command, the first camera will take pictures of the environment around the target vehicle. Specifically, it will obtain the current environment image and further obtain each first feature point in the current environment image and the first feature description information corresponding to each first feature point.
[0066] In a specific embodiment, the angle between the current observation vector corresponding to each second feature point in the parking area map within the current image observation range and the first observation vector corresponding to each second feature point in the parking area map within the current image observation range is calculated. Further, multiple vector angles between the current observation vectors corresponding to each second feature point in the current environment image are obtained.
[0067] Figure 5 This is a schematic flowchart illustrating a process for determining the angle between multiple observation vectors corresponding to each second feature point within the current image observation range, according to an exemplary embodiment. In an optional embodiment, such as... Figure 5 As shown, determining the angle between multiple observation vectors corresponding to each second feature point within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range can include:
[0068] S501: Perform vector normalization on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range to obtain the current observation unit vector corresponding to each second feature point within the current image observation range and multiple second observation unit vectors corresponding to each second feature point within the current image observation range.
[0069] S503: Based on the current observation unit vector corresponding to each second feature point within the current image observation range and the second observation unit vector corresponding to each second feature point within the current image observation range, calculate the cosine value of the included angle to determine the included angle of multiple observation vectors corresponding to each second feature point within the current image observation range.
[0070] In a specific embodiment, in the process of calculating the included angle of multiple vectors of the current observation vector corresponding to each second feature point in the parking area map within the current image observation range, it is necessary to first perform vector normalization processing on the current observation vector corresponding to each second feature point in the current image and the first observation vector corresponding to each second feature point. Further, the current observation unit vector and multiple second observation unit vectors corresponding to each second feature point in the parking area map within the current image observation range can be obtained. Further, based on the current observation unit vector and multiple second observation unit vectors corresponding to each second feature point in the parking area map within the current image observation range, the inner product length is calculated. Further, the cosine value of the included angle corresponding to each second feature point within the current image observation range is obtained. Correspondingly, the cosine value of the included angle corresponding to each second feature point within the current image observation range is used as the included angle of multiple observation vectors corresponding to each second feature point within the current image observation range.
[0071] Figure 6 This is a schematic flowchart illustrating a process for filtering each second feature point within the current image observation range based on the included angle of multiple observation vectors corresponding to each second feature point within the current salient observation range, according to an exemplary embodiment. In an optional embodiment, such as Figure 6 As shown, the above-mentioned filtering of the second feature points within the current image observation range based on the angle between multiple observation vectors corresponding to each second feature point within the current image observation range to obtain the fourth feature point may include:
[0072] S601: Determine the minimum angle between the observation vectors corresponding to each second feature point within the current image observation range from the angles between the multiple observation vectors corresponding to each second feature point within the current image observation range;
[0073] S603: Among the second feature points within the current image observation range, the second feature point whose minimum observation vector angle is less than a preset threshold is taken as the fourth feature point.
[0074] In a specific embodiment, the preset threshold can be the maximum value corresponding to the minimum observation vector angle. Optionally, the preset threshold can be set by the user. Specifically, the multiple observation vector angles corresponding to each second feature point in the parking area map within the current image observation range are sorted according to their numerical values. Further, the minimum observation vector angle among the multiple observation vector angles corresponding to each second feature point in the parking area map within the current image observation range can be obtained. Further, the minimum observation vector angle is compared with the preset threshold. Further, the second feature point corresponding to the minimum observation vector angle that is less than the preset threshold is taken as the fourth feature point.
[0075] In a specific embodiment, after determining the fourth feature point, the second feature description information corresponding to the fourth feature point is obtained. Specifically, the second feature description information corresponding to the fourth feature point in the current observation direction is obtained. Further, the first feature description information corresponding to the first feature point is matched with the second feature description information corresponding to the fourth feature point. The fourth feature point whose first feature description information and the second feature description information are the same is designated as the third feature point. Accordingly, multiple third feature points can be determined.
[0076] In the above embodiments, by filtering the second feature point within the current image observation range multiple times to determine the third feature point for position adjustment, the error in determining the third feature point is reduced, the amount of computation is reduced, and thus the accuracy of feature point matching is improved.
[0077] S109: Based on the fourth location information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the position of the target vehicle is adjusted to obtain the target position of the target vehicle.
[0078] Figure 7 This is a schematic flowchart illustrating a process for determining the target position of a target vehicle based on fourth position information corresponding to a third feature point and the projection point position information corresponding to the third feature point, according to an exemplary embodiment. In an optional embodiment, such as... Figure 7 As shown, the above-mentioned adjustment of the target vehicle's position based on the fourth location information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, to obtain the target vehicle's target position, may include:
[0079] S701: Based on the fourth position information corresponding to the third feature point and the projected position of the feature point corresponding to the third feature point, calculate the position error term and obtain the error term data;
[0080] S703: Based on the error term data, adjust the position of the target vehicle to obtain the target position of the target vehicle.
[0081] In a specific embodiment, by obtaining the third feature point, the fourth location information corresponding to the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point are obtained. The PNP bundle adjustment algorithm is used to adjust the position of the target vehicle. Specifically, based on the fourth location information corresponding to the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the reprojection error is calculated, that is, the position error term is calculated. Accordingly, the error term data is obtained. Further, based on the error term data, the position of the target vehicle is adjusted.
[0082] Figure 8 This is a schematic diagram illustrating a process for adjusting the position of a target vehicle based on error term data, according to an exemplary embodiment. In an optional embodiment, such as... Figure 7 As shown, the above-mentioned position adjustment of the target vehicle based on the error term data to obtain the target position of the target vehicle may include:
[0083] S801: Based on the error term data, optimize the first position information to obtain the first camera target position corresponding to the first camera;
[0084] S803: Construct the preset relationship information between the target vehicle and the first camera;
[0085] S805: Based on the current location information of the target vehicle, the target position of the first camera, and the preset relationship information, the position of the target vehicle is adjusted to obtain the target position of the target vehicle.
[0086] In a specific embodiment, the preset relationship information may include preset angle relationship information and preset position relationship information between the target vehicle and the first camera. Based on error term data, the position of the first camera is optimized, that is, the first position information corresponding to the first camera is optimized, and the target position of the first camera object is obtained accordingly. Furthermore, when the target vehicle obtains a positioning request, the preset relationship information between the first camera and the target vehicle will be constructed accordingly, and the current position information of the target vehicle will be obtained through existing positioning technology. Furthermore, based on the target position of the first camera, the current position information of the target vehicle, and the preset relationship information, the position of the target vehicle will be adjusted accordingly, and the target position of the target vehicle will be obtained accordingly. Furthermore, the target position of the target vehicle will be obtained through existing positioning technology.
[0087] In the above embodiments, the position of the target vehicle is adjusted by combining the fourth position information corresponding to the third feature point with the projected position of the feature point corresponding to the third feature point, thereby improving the accuracy of position adjustment and further improving the accuracy of positioning.
[0088] This invention provides a vehicle positioning method. First, the target vehicle uses a first camera to capture multiple environmental images of the current application scenario. Then, the obtained environmental images are combined with an algorithm to construct a parking area map of the current application scenario. Further, when a positioning request is responded to again, the method obtains first feature description information, first location information, and the current image observation direction corresponding to each first feature point in the current environmental image. It also obtains multiple second location information, multiple second feature description information, and multiple third location information corresponding to each second feature point in the parking area map corresponding to the target vehicle. Further, based on the second location information and the multiple third location information, a position connection is made, correspondingly obtaining a first observation vector corresponding to each second feature point. Further, each second feature point in the parking area map within the current image observation range corresponding to the current observation direction is projected onto the current environmental image. Further, the feature point projection position corresponding to each second feature point in the parking area map within the current image observation range is obtained. Further, based on the aforementioned feature point projection positions and the first location information corresponding to the first camera, a position connection is made to obtain the current image observation range. The current observation vector corresponding to each second feature point within the current image observation range is further combined with the current observation vector corresponding to each second feature point and the first observation vector corresponding to each second feature point to filter the second feature points within the current image observation range. Correspondingly, a first filtering is performed by the angle between the observation vectors of the current observation vector corresponding to each second feature point and the first observation vector. Accordingly, the second feature point corresponding to the smallest observation vector angle less than a preset threshold is obtained, i.e., the fourth feature point. Further, a second filtering is performed by matching the first feature description information corresponding to the first feature point with the second feature description information corresponding to the terrain feature point. Accordingly, the fourth feature point among the fourth feature points that is the same as the first feature description information corresponding to the first feature point is taken as the third feature point. Further, based on the fourth position information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the target vehicle is adjusted using the bundle algorithm to obtain the target position of the target vehicle. This invention can achieve accurate positioning of the target vehicle, improve the high stability of feature description information and the accuracy of feature point matching, and thus improve the positioning accuracy.
[0089] Figure 9 This is a schematic diagram of a vehicle positioning device according to an exemplary embodiment. The following describes an embodiment of a vehicle positioning device according to this application, specifically, as follows: Figure 9 As shown, the device includes:
[0090] The first information acquisition module 901 is used to respond to a positioning request for a target vehicle by acquiring a first observation vector, a first position information, and a current image observation direction corresponding to the current environment image for each second feature point in the map of the parking area corresponding to the target vehicle; the first position information is the position information of the first camera when the current environment image is acquired.
[0091] The second information acquisition module 903 is used to project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image, and to acquire the feature point projection position corresponding to each second feature point within the current image observation range;
[0092] The second determining module 905 is used to determine the current observation vector corresponding to each second feature point within the current image observation range based on the feature point projection position and the first position information;
[0093] The matching module 907 is used to filter the second feature points in the current image observation range based on the current observation vector corresponding to each second feature point in the current image observation range and the first observation vector corresponding to each second feature point in the current image observation range, so as to obtain a third feature point that matches the first feature point in the current environment image.
[0094] The position adjustment module 909 is used to adjust the position of the target vehicle based on the fourth position information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, so as to obtain the target position of the target vehicle.
[0095] In an optional embodiment, the first information acquisition module 901 includes:
[0096] The unit for acquiring second feature point information is used to acquire multiple second location information and multiple third location information corresponding to each second feature point in the parking area map corresponding to the target vehicle; the multiple second location information are the location information of each second feature point in multiple environmental images corresponding to the parking area map, and the multiple third location information are the location information of the first camera when multiple environmental images are acquired respectively.
[0097] A first observation vector unit is determined, which is used to determine the first observation vector corresponding to each second feature point based on the plurality of second position information and the plurality of third position information.
[0098] In an optional embodiment, the matching module 907 includes:
[0099] Multiple angle determination units are used to determine the angle between multiple observation vectors corresponding to each second feature point in the current image observation range based on the current observation vector corresponding to each second feature point in the current image observation range and the first observation vector corresponding to each second feature point in the current image observation range.
[0100] The fourth feature point determination unit is used to filter the second feature points within the current image observation range based on the angle between multiple observation vectors corresponding to each second feature point within the current image observation range, and obtain the fourth feature point;
[0101] The third feature point determination unit is used to determine, based on the first feature description information corresponding to each first feature point and the second feature description information corresponding to the fourth feature point, a fourth feature point that has the same first feature description information as the first feature point, and to use it as the third feature point.
[0102] In an optional embodiment, the fourth feature point determination unit includes:
[0103] The minimum included angle determination subunit is used to determine the minimum included angle of the observation vector corresponding to each second feature point in the current image observation range from the included angles of multiple observation vectors corresponding to each second feature point in the current image observation range;
[0104] The minimum angle feature point determination sub-unit is used to identify the second feature point within the current image observation range whose minimum observation vector angle is less than a preset threshold as the fourth feature point.
[0105] In an optional embodiment, the third feature point determination unit includes:
[0106] The vector normalization subunit is used to perform vector normalization processing on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, to obtain the current observation unit vector corresponding to each second feature point within the current image observation range and multiple second observation unit vectors corresponding to each second feature point within the current image observation range;
[0107] Angle calculation subunit is used to calculate the cosine value of the angle based on the current observation unit vector corresponding to each second feature point within the current image observation range and the second observation unit vector corresponding to each second feature point within the current image observation range, and to determine the angle between multiple observation vectors corresponding to each second feature point within the current image observation range.
[0108] In an optional embodiment, the device position adjustment module 811 includes:
[0109] The error term calculation unit is used to calculate the position error term based on the fourth position information corresponding to the third feature point and the projection position of the feature point corresponding to the third feature point, and obtain the error term data.
[0110] The position adjustment unit is used to adjust the position of the target vehicle based on the error term data to obtain the target position of the target vehicle.
[0111] In an optional embodiment, the above-mentioned position adjustment unit includes:
[0112] The first camera target position determination subunit is used to optimize the first position information based on the error term data to obtain the first camera target position corresponding to the first camera.
[0113] A preset relationship subunit is constructed to construct preset relationship information between the target vehicle and the first camera. The preset relationship information includes preset angle relationship information and preset position relationship information between the target vehicle and the first camera.
[0114] The vehicle position adjustment subunit is used to adjust the position of the target vehicle based on the current position information of the target vehicle, the target position of the first camera, and preset relationship information to obtain the target position of the target vehicle.
[0115] In an optional embodiment, the above-described apparatus further includes:
[0116] A multi-environment image acquisition unit is used to acquire the multiple environment images;
[0117] A parking map subunit is constructed to build the parking area map based on the multiple environmental images.
[0118] In an exemplary embodiment, an electronic device is also provided, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the vehicle positioning method as described in the embodiments of this disclosure.
[0119] In an exemplary embodiment, a computer-readable storage medium is also provided, wherein instructions in the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the vehicle positioning method of the present disclosure embodiments.
[0120] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the vehicle positioning method provided in the various optional implementations described above.
[0121] It is understood that in the specific embodiments of the present invention, user-related data is involved. When the above embodiments of the present invention are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0122] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.
[0123] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.
[0124] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.
Claims
1. A vehicle positioning method, characterized in that, The method includes: In response to a positioning request for a target vehicle, the system acquires a first observation vector, a first location information, and a current image observation direction corresponding to the current environment image for each second feature point in the parking area map corresponding to the target vehicle; the first location information is the location information of the first camera when the current environment image is acquired; the second feature point is a feature description point extracted from the parking area map. Project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image to obtain the feature point projection position corresponding to each second feature point within the current image observation range; Based on the projection position of the feature point and the first position information, determine the current observation vector corresponding to each second feature point within the current image observation range; Based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, the second feature points within the current image observation range are filtered to obtain a third feature point that matches the first feature point in the current environment image; the first feature point is a feature description point extracted from the current environment image; the third feature point is a fourth feature point whose first feature description information corresponding to the first feature point is the same as the second feature description information corresponding to the fourth feature point; the fourth feature point is a second feature point within the current image observation range whose minimum observation vector angle is less than a preset threshold. Based on the fourth location information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the position of the target vehicle is adjusted to obtain the target position of the target vehicle; the fourth location information is the location information of the third feature point in the parking area map.
2. The method according to claim 1, characterized in that, The step of obtaining the first observation vector corresponding to each second feature point in the parking area map of the target vehicle includes: Acquire multiple second location information and multiple third location information corresponding to each second feature point in the parking area map corresponding to the target vehicle; the multiple second location information are the location information of each second feature point in multiple environmental images corresponding to the parking area map, and the multiple third location information are the location information of the first camera when multiple environmental images are collected respectively. Based on the plurality of second location information and the plurality of third location information, a first observation vector corresponding to each second feature point is determined.
3. The method according to claim 1, characterized in that, The step of filtering the second feature points within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range to obtain third feature points that match the first feature points in the current environment image includes: Based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, determine the angle between multiple observation vectors corresponding to each second feature point within the current image observation range; Based on the angle between multiple observation vectors corresponding to each second feature point within the current image observation range, the second feature points within the current image observation range are filtered to obtain the fourth feature point; Based on the first feature description information corresponding to each first feature point and the second feature description information corresponding to the fourth feature point, a fourth feature point that has the same first feature description information as the first feature point is determined from the fourth feature points and is used as the third feature point.
4. The method according to claim 3, characterized in that, The step of filtering the second feature points within the current image observation range based on the angle between multiple observation vectors corresponding to each second feature point within the current image observation range to obtain the fourth feature point includes: From the angles between multiple observation vectors corresponding to each second feature point within the current image observation range, determine the smallest angle between the observation vectors corresponding to each second feature point within the current image observation range; The second feature point whose minimum observation vector angle is less than a preset threshold among the second feature points within the current image observation range is taken as the fourth feature point.
5. The method according to claim 3, characterized in that, The step of determining the angle between multiple observation vectors corresponding to each second feature point within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range includes: The current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range are subjected to vector normalization processing to obtain the current observation unit vector corresponding to each second feature point within the current image observation range and multiple second observation unit vectors corresponding to each second feature point within the current image observation range. Based on the current observation unit vector corresponding to each second feature point within the current image observation range and the second observation unit vector corresponding to each second feature point within the current image observation range, the cosine value of the included angle is calculated to determine the included angle of multiple observation vectors corresponding to each second feature point within the current image observation range.
6. The method according to claim 1, characterized in that, Based on the fourth location information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, the target vehicle's position is adjusted to obtain the target position of the target vehicle, including: Based on the fourth position information corresponding to the third feature point and the projected position of the feature point corresponding to the third feature point, the position error term is calculated to obtain the error term data. Based on the error term data, the position of the target vehicle is adjusted to obtain the target position of the target vehicle.
7. The method according to claim 6, characterized in that, The step of adjusting the current position of the target vehicle based on the error term data to obtain the target position of the target vehicle includes: Based on the error term data, the first position information is optimized to obtain the first camera target position corresponding to the first camera. Construct preset relationship information between the target vehicle and the first camera, the preset relationship information including preset angle relationship information and preset position relationship information between the target vehicle and the first camera; Based on the current location information of the target vehicle, the target position of the first camera, and the preset relationship information, the position of the target vehicle is adjusted to obtain the target position of the target vehicle.
8. The method according to any one of claims 1 to 7, characterized in that, The method further includes: Acquire the multiple environmental images; Based on the multiple environmental images, a parking area map is constructed.
9. A vehicle positioning device, characterized in that, The device includes: The first information acquisition module is used to respond to a positioning request for a target vehicle by acquiring a first observation vector, a first location information, and a current image observation direction corresponding to the current environment image for each second feature point in the parking area map corresponding to the target vehicle; the first location information is the location information of the first camera when the current environment image is acquired; the second feature point is a feature description point extracted from the parking area map. The second information acquisition module is used to project multiple second feature points within the current image observation range corresponding to the current image observation direction onto the current environment image, and to acquire the feature point projection position corresponding to each second feature point within the current image observation range; The second determining module is used to determine the current observation vector corresponding to each second feature point within the current image observation range based on the feature point projection position and the first position information; The matching module is used to filter the second feature points within the current image observation range based on the current observation vector corresponding to each second feature point within the current image observation range and the first observation vector corresponding to each second feature point within the current image observation range, to obtain a third feature point that matches the first feature point in the current environment image; the first feature point is a feature description point extracted from the current environment image; the third feature point is a fourth feature point whose first feature description information corresponding to the first feature point is the same as the second feature description information corresponding to the fourth feature point; the fourth feature point is a second feature point within the current image observation range whose minimum observation vector angle is less than a preset threshold. The position adjustment module is used to adjust the position of the target vehicle based on the fourth position information of the third feature point in the parking area map and the projection position of the feature point corresponding to the third feature point, so as to obtain the target position of the target vehicle; the fourth position information is the position information of the third feature point in the parking area map.
10. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the vehicle positioning method as described in any one of claims 1 to 8.
11. A computer-readable storage medium, characterized in that, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the vehicle positioning method as described in any one of claims 1 to 8.