A method, device and equipment for recognizing a curve based on an experience map
By filtering and grouping waypoints in the experience map, false curves are identified and eliminated, improving the accuracy and efficiency of curve recognition for autonomous vehicles and ensuring safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SINO TRUK JINAN POWER CO LTD
- Filing Date
- 2023-10-07
- Publication Date
- 2026-06-02
Smart Images
Figure CN117346799B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of autonomous driving technology, and in particular to a curve recognition method, apparatus and device based on experience maps. Background Technology
[0002] Experience maps are created by collecting and processing vast amounts of driving data, providing beyond-line-of-sight perception capabilities that allow vehicles to anticipate information beyond sensor limitations. Therefore, in the field of autonomous driving, experience maps are commonly used for navigation and route planning. The detailed environmental information provided by experience maps assists vehicles in navigation and route planning.
[0003] However, "false curves" may exist on experience maps, which are curves displayed on the map but are not actually curves. False curves may occur due to data acquisition errors or other reasons, such as driver error causing the steering wheel to turn, or inherent noise in the sensors themselves during the experience map data acquisition process.
[0004] For autonomous vehicles that navigate based on experience maps, whether it's a real curve or a "false curve," the vehicle will adopt corresponding driving strategies, such as slowing down or adjusting the heading angle, to smoothly pass through the curve. However, these "false curves" mislead the autonomous driving system, leading to unnecessary vehicle control behaviors, thereby affecting the stability and safety of the vehicle's driving, as well as the working efficiency of autonomous vehicles in operation. Summary of the Invention
[0005] This application provides a method, apparatus, and device for identifying curves based on empirical maps, in order to solve the problem of how to more accurately identify real curves in empirical maps.
[0006] In a first aspect, this application provides a curve recognition method based on an empirical map, comprising:
[0007] Based on the path to be identified in the experience map, the initial instantaneous curvature of each path point is obtained in a preset order. According to the preset filtering rule, the initial instantaneous curvature of the path points that match the preset filtering rule is filtered to obtain the filtered instantaneous curvature of the path points that match the preset filtering rule and the initial instantaneous curvature of the path points that do not match the preset filtering rule.
[0008] Based on the preset instantaneous curvature threshold, the path points corresponding to the filtered instantaneous curvature and the path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules are filtered to obtain multiple suspected curved path points.
[0009] The multiple suspected curve path points are grouped, and the number of suspected curve path points contained in the suspected curve obtained by the grouping process is used to determine whether the suspected curve is a real curve, so as to realize the identification of all curves in the path to be identified in the experience map.
[0010] In one possible design, the step of filtering the initial instantaneous curvature of path points matching the preset filtering rule, to obtain the filtered instantaneous curvature of path points matching the preset filtering rule and the initial instantaneous curvature of path points not matching the preset filtering rule, includes:
[0011] Based on a preset filter length threshold, filter start point and filter end point are obtained; wherein, the filter start point is the distance from the start point of the current path to be identified to the filter length threshold, and the filter end point is the distance from the end point of the current path to be identified to the filter length threshold.
[0012] Based on the filter start point and the filter end point, the following formula is used:
[0013]
[0014] The instantaneous curvature of the filter at all path points from the starting point to the ending point of the filter is obtained sequentially.
[0015] Where, γ i1 γ represents the filtered instantaneous curvature of path point i. i0 γ represents the initial instantaneous curvature of path point i. j Let represent the initial instantaneous curvature of path point j, and L represent the filter length threshold.
[0016] In one possible design, obtaining the initial instantaneous curvature of each path point sorted in a preset order based on the path to be identified in the empirical map includes:
[0017] If the path point is the starting point of the path to be identified, then the initial instantaneous curvature of the starting point is the initial instantaneous curvature of the path points adjacent to the starting point.
[0018] or,
[0019] If the path point is the endpoint of the path to be identified, then the initial instantaneous curvature of the endpoint path point is the initial instantaneous curvature of the path points adjacent to the endpoint path point.
[0020] or,
[0021] If the path point is neither the starting path point nor the ending path point, then the following formula is used:
[0022]
[0023] The initial instantaneous curvature γ of the path point is obtained. i0 ;
[0024] Where, φ (i-1) φ represents the heading angle of path point i-1. (i+1) Let x represent the heading angle of path point i+1. (i-1) The x-coordinate of path point i-1 is represented by x. (i+1) The x-coordinate of path point i+1 is represented by y. (i-1) The y-coordinate of path point i-1 is represented by... (i+1) This represents the ordinate of the path point i+1.
[0025] In one possible design, the path points corresponding to the filtered instantaneous curvature and the path points corresponding to the initial instantaneous curvature that do not match the preset filtering rule are filtered according to a preset instantaneous curvature threshold to obtain multiple suspected curve path points, including:
[0026] If the instantaneous curvature of the filter is greater than the instantaneous curvature threshold, then the path point corresponding to the instantaneous curvature of the filter is a suspected curved path point;
[0027] or,
[0028] If the initial instantaneous curvature is greater than the instantaneous curvature threshold, then the path point corresponding to the initial instantaneous curvature is a suspected curve path point.
[0029] In one possible design, before obtaining the initial instantaneous curvature of each path point sorted in a preset order in the path to be identified based on the experience map, the method includes:
[0030] Obtain the path information to be identified from the experience map. The path information to be identified includes multiple discrete path points and the coordinate information and heading information corresponding to the path points.
[0031] The discrete path points are sorted according to a preset order to obtain multiple path points sorted in the preset order.
[0032] In one possible design, the grouping of the multiple suspected curve path points includes:
[0033] Configure a corresponding index number for each path point sorted according to the preset order;
[0034] Based on the preset sorting order, according to the index number configured for the path point, the index number difference between the suspected curve path point and its adjacent suspected curve path point is calculated sequentially, and it is determined whether the index number difference is less than the preset difference threshold.
[0035] If the difference in index number is less than a preset difference threshold, then the suspected curve path point and its adjacent suspected curve path points belong to the same suspected curve.
[0036] If the difference in index number is not less than a preset difference threshold, then the suspected curve path point and its adjacent suspected curve path points do not belong to the same suspected curve.
[0037] In one possible design, determining whether a suspected curve is a true curve based on the number of suspected curve path points included in the suspected curve obtained from the grouping process includes:
[0038] Determine whether the number of suspected curve path points contained in the suspected curve road is greater than a preset number threshold;
[0039] If the value is greater than the value, then the suspected curve is a true curve;
[0040] If the value is not greater than the value, then the suspected curve is a false curve.
[0041] In one possible design, the method further includes:
[0042] Based on a preset sorting order, the path point information corresponding to the starting point and the ending point of the true curve is obtained, and the path point information corresponding to the starting point and the ending point of the curve is sent to the driving planning module of the autonomous vehicle to realize the driving control of the autonomous driving.
[0043] Secondly, the application itself provides a curve recognition device based on an experience map, including:
[0044] The curvature calculation module is used to obtain the initial instantaneous curvature of each path point sorted in a preset order based on the path to be identified in the experience map, and to filter the initial instantaneous curvature of the path points that match the preset filtering rules according to the preset filtering rules, so as to obtain the filtered instantaneous curvature of the path points that match the preset filtering rules and the initial instantaneous curvature of the path points that do not match the preset filtering rules.
[0045] The path point filtering module is used to filter path points corresponding to the filtered instantaneous curvature and path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules according to a preset instantaneous curvature threshold, thereby obtaining multiple suspected curve path points.
[0046] The curve identification module is used to group the multiple suspected curve path points and, based on the number of suspected curve path points contained in the suspected curves obtained by the grouping process, to identify all curves in the path to be identified in the experience map.
[0047] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;
[0048] The memory stores computer-executed instructions;
[0049] The processor executes computer execution instructions stored in the memory to implement the above-described method.
[0050] Fourthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the above-described method.
[0051] This application obtains the initial instantaneous curvature of path points, providing fundamental information for subsequent curve detection. Based on preset filtering rules, the initial instantaneous curvature of the corresponding path points is filtered to remove noise or outliers, resulting in a smoother and more stable final instantaneous curvature that more accurately reflects the characteristics of the road curve. Furthermore, based on a preset instantaneous curvature threshold, the instantaneous curvature of all path points is filtered to identify potential curve path points. These potential curve path points are grouped, and the number of potential curve path points within a group determines whether it is a real curve, eliminating some suspected cases and ensuring that only genuine curved roads are identified as curves. Using real-world conditions as the basis for judgment closely reflects actual conditions, leading to more accurate curve identification results. The data processing method is simple and efficient, avoiding the system's operational pressure caused by overly complex algorithms, enabling faster and more efficient results and improving recognition efficiency. Furthermore, accurate curve identification, while ensuring the safe and stable driving of autonomous vehicles, can reduce unnecessary deceleration and steering operations during autonomous driving, improving vehicle efficiency. Attached Figure Description
[0052] To more clearly illustrate the technical solutions 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0053] Figure 1A schematic diagram of a curve recognition method based on an experience map provided in one embodiment of this application;
[0054] Figure 2 A schematic diagram illustrating a method for obtaining multiple suspected curve path points according to an embodiment of this application;
[0055] Figure 3 A schematic diagram of a specific path to be identified is provided for a particular embodiment of this application;
[0056] Figure 4 This is a schematic diagram illustrating the results of filtering and selecting remaining path points according to a preset threshold, provided in a specific embodiment of this application.
[0057] Figure 5 This is a schematic diagram illustrating the result of obtaining the remaining path points without filtering and based solely on a preset threshold, according to a specific embodiment of this application.
[0058] Figure 6 This is a schematic diagram illustrating a method for obtaining multiple path points ordered in a preset order according to one embodiment of this application;
[0059] Figure 7 This is a schematic diagram illustrating a method for grouping multiple suspected curve path points according to an embodiment of this application.
[0060] Figure 8 A schematic diagram of grouping results provided for a specific embodiment of this application;
[0061] Figure 9 A schematic diagram illustrating a method for determining whether a suspected curve is a true curve, provided in one embodiment of this application.
[0062] Figure 10 A schematic diagram of the remaining suspected curves obtained after filtering by a quantity threshold, provided as a specific embodiment of this application;
[0063] Figure 11 A schematic diagram showing the details of intra-segment blanking provided for a specific embodiment of this application;
[0064] Figure 12 A schematic diagram of the curve recognition device based on an experience map provided in an embodiment of this application;
[0065] Figure 13 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0066] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.
[0067] First, the relevant concepts or terms involved in this application will be explained:
[0068] Instantaneous curvature: In mathematics, curvature is a quantity that describes the degree of bending of a geometric object; for a curve, curvature describes the degree to which the curve deviates from a straight line; instantaneous curvature can be understood as the curvature at a specific point in time. The concept of instantaneous curvature is similar to that of instantaneous velocity. Instantaneous velocity is the velocity of an object at a specific point in time, while instantaneous curvature is the curvature of a curve at a specific point in time.
[0069] Experience maps are maps based on historical data and experience, containing a wealth of information about the environment and routes. These maps are generated by collecting extensive historical driving data and can help autonomous driving systems better understand and adapt to their environment.
[0070] Figure 1 This diagram illustrates a curve recognition method based on an experience map, as provided in one embodiment of this application. Specifically, it includes steps S11-S13:
[0071] S11. Based on the path to be identified in the experience map, obtain the initial instantaneous curvature of each path point sorted in a preset order, and filter the initial instantaneous curvature of the path points that match the preset filtering rules according to the preset filtering rules, so as to obtain the filtered instantaneous curvature of the path points that match the preset filtering rules and the initial instantaneous curvature of the path points that do not match the preset filtering rules.
[0072] In this embodiment, the initial instantaneous curvature of multiple discrete path points on the path to be identified is obtained based on the path to be identified in the experience map. The initial instantaneous curvature of the path points provides information on the curvature of the path to be identified from historical driving data records. In autonomous driving and navigation systems, instantaneous curvature is a key indicator of the degree of road curvature, which helps to more accurately plan the vehicle's driving path and driving strategy, ensuring that the vehicle safely and smoothly passes through curves. For example, if the vehicle is about to enter a curve with a large curvature, the control strategy can be adjusted in advance (such as deceleration, adjusting the steering angle, etc.). It should be further noted that the path points on the path to be identified are ordered according to a preset order. The preset order is usually set according to the vehicle's driving direction or task requirements. For example, if the vehicle is traveling from the starting point to the end point, then the preset order is from the starting point to the end point; if the vehicle is patrolling a certain area, then the preset order may be a closed loop.
[0073] Because sensor measurements and map data may contain noise and errors, these can affect the initial instantaneous curvature of waypoints. To obtain a more accurate curvature estimate, the initial instantaneous curvature is filtered. Filtering reduces noise, errors, and uncertainties, improving data quality and reliability. Filtering rules can be based on the physical characteristics of the path, navigation requirements, or other factors. If a waypoint meets the filtering rules, a filtering algorithm can be applied to adjust its curvature value. These filtering algorithms include, but are not limited to, mean filtering, median filtering, Kalman filtering, and Bayesian filtering. Through filtering, filtered instantaneous curvatures of waypoints matching the preset filtering rules can be obtained. Meanwhile, for waypoints that do not match the preset filtering rules, their initial instantaneous curvatures are retained. At this point, each waypoint on the path to be identified has one and only one corresponding instantaneous curvature: either the filtered instantaneous curvature or the initial instantaneous curvature.
[0074] S12, based on the preset instantaneous curvature threshold, filter the path points corresponding to the filtered instantaneous curvature and the path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules to obtain multiple suspected curve path points.
[0075] In this embodiment, after filtering the two sets of data obtained in step S11 according to a preset instantaneous curvature threshold, all resulting path points are suspected curve path points. Setting the instantaneous curvature threshold can exclude path segments with small curvature, thereby reducing the complexity of subsequent calculations and analysis. Furthermore, filtering out path points with significant curvature can effectively detect potential curve locations, facilitating more accurate analysis of the path's curve characteristics. The instantaneous curvature threshold can be set according to application scenario requirements and road characteristics. This threshold can be a constant or dynamically adjusted based on actual conditions. All suspected curve path points that meet the criteria are recorded for subsequent curve identification, path planning, or navigation decisions.
[0076] S13, group multiple suspected curve path points, and determine whether a suspected curve is a real curve based on the number of suspected curve path points contained in the suspected curve obtained by grouping, so as to realize the identification of all curves in the path to be identified in the experience map.
[0077] In this embodiment, multiple suspected curve path points are grouped, meaning the suspected curve path points selected in step S12 are grouped according to certain rules. Grouping can be based on the distance between path points or other relevant characteristics. This will group suspected curve path points with the same characteristics into the same group. The number of suspected curve path points included in the grouped suspected curves determines whether a suspected curve is a true curve. True curves are usually formed due to road ruggedness, and the path length corresponding to such true curves is relatively long, meaning that the number of suspected curve path points corresponding to abnormal instantaneous curvature within the curve will be relatively large. Conversely, for false curves formed due to errors or noise, the number of suspected curves in the group will be relatively small. Therefore, the number of suspected curve path points included in a suspected curve can be used to determine whether a suspected curve is a true curve. Identifying suspected curves based on the number of path points with abnormal instantaneous curvature within a curve in real-world conditions can reduce the complexity of calculation and analysis and improve the efficiency of curve identification. The number can be flexibly set to better reflect the real situation and environment, which is more in line with the actual environment of the current path, reduces false alarms, and improves the accuracy of curve recognition.
[0078] By acquiring the initial instantaneous curvature of path points, foundational information is provided for subsequent curve detection. Based on preset filtering rules, the initial instantaneous curvature of the corresponding path points is filtered to remove noise or outliers, resulting in a smoother and more stable final instantaneous curvature that more accurately reflects the characteristics of the road curve. Furthermore, based on a preset instantaneous curvature threshold, the instantaneous curvature of all path points is filtered to identify potential curve path points. These potential curve path points are grouped, and the number of potential curve path points within a group determines whether it is a genuine curve, eliminating some suspected cases and ensuring that only genuine curved roads are identified as curves. Using real-world conditions as the basis for judgment, the results are closer to actual conditions, leading to more accurate curve identification. The data processing method is simple and efficient, avoiding the system's operational pressure caused by overly complex algorithms, enabling faster and more efficient results and improving identification efficiency. Furthermore, accurate curve identification, while ensuring the safe and stable driving of autonomous vehicles, can reduce unnecessary deceleration and steering operations during autonomous driving, improving vehicle efficiency.
[0079] In one embodiment, a method for obtaining the initial instantaneous curvature of each path point ordered in a preset order is provided, which is a specific limitation on obtaining the initial instantaneous curvature in step S11, specifically including steps S111-S113:
[0080] S111, If the path point is the starting path point of the path to be identified, then the initial instantaneous curvature of the starting path point is the initial instantaneous curvature of the path points adjacent to the starting path point.
[0081] S112, If the path point is the end point of the path to be identified, then the initial instantaneous curvature of the end point is the initial instantaneous curvature of the path point adjacent to the end point.
[0082] S113, If a path point is neither the starting path point nor the ending path point, then the following formula is used:
[0083]
[0084] Obtain the initial instantaneous curvature γ of the path point i0 ; where φ (i-1) φ represents the heading angle of path point i-1. (i+1) Let x represent the heading angle of path point i+1. (i-1) The x-coordinate of path point i-1 is represented by x. (i+1) The x-coordinate of path point i+1 is represented by y. (i-1) Let y represent the ordinate of path point i-1. (i+1) This represents the ordinate of the path point i+1.
[0085] In this embodiment, the initial instantaneous curvature is the curvature value of a point on the path. In path planning, the initial instantaneous curvature of the start and end path points is usually calculated using a simplified method: if a path point is the start point of the path to be identified, its initial instantaneous curvature can be equal to the initial instantaneous curvature of its adjacent next path point. This is because the start path point usually does not require additional information to calculate its curvature, and its curvature is highly correlated with the curvature of the next path point. Similarly, if a path point is the end point of the path to be identified, its initial instantaneous curvature can be equal to the initial instantaneous curvature of its adjacent previous path point. This is also because the curvature of the end path point is usually closely related to the curvature of the previous path point.
[0086] If a path point is neither the starting nor ending path point, its initial instantaneous curvature is obtained by using the curvature of the corresponding curvature circles of its immediate and adjacent points. The length of the straight line segment between the preceding and following path points is the chord length, and the difference in heading angle between them is the central angle corresponding to the chord length. The radius of the curvature circle can be calculated from the relationship between the chord length, central angle, and radius. The reciprocal of the radius is the curvature of the curvature circle, which is the initial instantaneous curvature of the intermediate point (the current path point). For this purpose, information about path point i-1 (the preceding path point) and path point i+1 (the following path point) needs to be obtained, including their heading angles and coordinates. The initial instantaneous curvature provides crucial information about the path's geometric characteristics, providing concrete data to support the control decisions of autonomous vehicles.
[0087] Furthermore, a method is provided for filtering the initial instantaneous curvature of path points that match a preset filtering rule. This is a specific limitation on step S11 above, specifically including steps S114-S115:
[0088] S114. Based on the preset filter length threshold, filter start point and filter end point are obtained; wherein, the length of the filter start point from the start point to the start point of the current path to be identified is the filter length threshold, and the length of the filter end point from the end point of the current path to be identified is the filter length threshold.
[0089] In this embodiment, the filtering length threshold is a pre-set parameter, which can be set according to the specific application requirements and path characteristics. Based on the filtering length threshold, measurements are taken from the starting point of the path towards the ending point. When the measured length reaches the threshold, the current point is determined as the filtering starting point. The distance of this path point from the starting point of the current path to be identified is equal to the filtering length threshold. Similarly, measurements are taken from the ending point of the path towards the starting point. When the measured length reaches the filtering length threshold, the current point is determined as the filtering ending point. The distance of this path point from the ending point of the current path to be identified is equal to the filtering length threshold.
[0090] S115, based on the filter start point and filter end point, uses the following formula:
[0091]
[0092] The instantaneous curvature of the filter at all path points from the starting path point to the ending path point is obtained sequentially; where γ i1 γ represents the filtered instantaneous curvature of path point i. i0 γ represents the initial instantaneous curvature of path point i. j Let represent the initial instantaneous curvature of path point j, and L represent the filter length threshold.
[0093] In this embodiment, the selection of the filter start point path point is used to determine the starting position of the path segment to be filtered; similarly, the selection of the filter end point path point is used to determine the ending position of the path segment to be filtered. It is necessary to calculate the instantaneous curvature of the filtered path points for all path points from the filter start point path point to the filter end point path point. Wherein, γ i1 γ represents the instantaneous curvature of the path point i, which is the final result to be calculated; i0 γ represents the initial instantaneous curvature of path point i, i.e., the curvature value before any filtering or processing (the initial instantaneous curvature obtained in step S11); L represents the filter length threshold, which represents the radius or window size of the filter, and determines the number of adjacent path points considered when calculating the instantaneous curvature of the filter, i.e., the number of path points selected before and after. Specifically, for path point i, path points from iL to i-1 and from i+1 to i+L will be considered; j Let represent the initial instantaneous curvature of path point j, which is the initial instantaneous curvature of each path point from iL to i-1 and from i+1 to i+L.
[0094] As can be seen from the formula, when calculating the instantaneous curvature of the filter at each path point, the curvature γ of the current path point i is considered. i0 It also takes into account the curvature γ of the path points in its vicinity. j(Adjacent path points within the sliding window). The purpose of this formula is to smooth and filter the curvature of path points to reduce noise and abrupt changes, resulting in instantaneous curvature of path points that more closely resembles the actual path.
[0095] In one embodiment, Figure 2 This is a schematic diagram illustrating a method for obtaining multiple suspected curve path points according to an embodiment of this application. It provides a specific limitation on step S12, such as... Figure 2 As shown, the specific steps include S21-S22:
[0096] S21, If the instantaneous curvature of the filter is greater than the instantaneous curvature threshold, then the path point corresponding to the instantaneous curvature of the filter is a suspected curved path point;
[0097] S22, if the initial instantaneous curvature is greater than the instantaneous curvature threshold, then the path point corresponding to the initial instantaneous curvature is a suspected curve path point.
[0098] In this embodiment, the instantaneous curvature corresponding to the path points on the current path to be identified includes two types: one is the filtered instantaneous curvature obtained after filtering, and the other is the initial instantaneous curvature without filtering. Furthermore, each path point on the current path to be identified has one and only one corresponding instantaneous curvature that needs to be filtered according to this embodiment; that is, the instantaneous curvature corresponding to the path point is either the filtered instantaneous curvature or the initial instantaneous curvature. When the filtered instantaneous curvature is greater than the instantaneous curvature threshold, the corresponding path point is considered a suspected curved path point. Similarly, when the initial instantaneous curvature is greater than the instantaneous curvature threshold, the corresponding path point is considered a suspected curved path point. The instantaneous curvature threshold setting under different scenarios can be determined according to specific application requirements to improve system performance and reliability. When the filtered instantaneous curvature is greater than the instantaneous curvature threshold, i.e., the filtered instantaneous curvature is high, it means that the path has a significant curvature change near the path point, and there may be a curved section of the road. These path points may require additional control strategies, such as deceleration or adjustment of vehicle direction.
[0099] Specifically, Figure 3 This is a schematic diagram of a specific path to be identified according to a specific embodiment of this application. After filtering the initial instantaneous curvature of path points that match the preset filtering rules as described in the above embodiment, obtaining the filtered instantaneous curvature of path points that match the preset filtering rules and the initial instantaneous curvature of path points that do not match the preset filtering rules, further filtering is performed according to a preset instantaneous curvature threshold, resulting in the following... Figure 4 As shown, Figure 4 This is a schematic diagram illustrating the result of filtering and selecting the remaining path points according to a preset threshold, provided as a specific embodiment of this application. Figure 4 Medium-thick or dark lines represent the remaining path points, while thin lines or light-colored areas represent path points that have been filtered out.
[0100] Another specific embodiment is provided herein. Figure 5 This diagram illustrates the result of obtaining remaining path points by filtering based on a preset threshold without filtering, according to a specific embodiment of this application. To further highlight the handling of data noise, errors, and the instantaneous curvature of abnormal path points caused by erroneous operations by filtering, in the specific path to be identified described above, filtering is performed directly based on the preset threshold without filtering, resulting in the following... Figure 5 The results are shown. Thick or dark lines represent the remaining path points, while thin or light lines represent the path points that have been filtered out. Figure 4 and Figure 5 The comparison shows that filtering can eliminate most of the path points with instantaneous curvature anomalies caused by noise, errors, or misoperations.
[0101] In one embodiment, Figure 6 This is a schematic diagram illustrating a method for obtaining multiple path points ordered in a preset order according to one embodiment of this application. It further supplements this method and represents the operation prior to step S11. Figure 6 As shown, the specific steps include S61-S62:
[0102] S61, Obtain the path information to be identified from the experience map. The path information to be identified includes multiple discrete path points and the coordinate and heading information of the path points.
[0103] S62, sort multiple discrete path points according to a preset order to obtain multiple path points sorted according to the preset order.
[0104] In this embodiment, the path information to be identified in the experience map is one of the fundamental data for the path planning and navigation system. The experience map can include various types of roads and paths, such as urban roads, highways, rural roads, and trails. The path information to be identified on the experience map includes multiple discrete path points and their corresponding coordinate and heading information. The coordinate information includes the x-coordinate and y-coordinate of the path point. This coordinate information describes the precise location of the path point on the map, enabling the autonomous vehicle to accurately locate itself. The heading information refers to the orientation or direction of the path point. It is usually expressed as a heading angle, and the heading information indicates the direction of travel to the next path point.
[0105] Sort the waypoints to be identified to ensure that autonomous vehicles can follow the path in a specific order. The preset order is usually determined based on task requirements or path planning strategies. The sorting strategy can be rationally planned based on factors such as physical distance, time priority, road grade, and traffic rules. Each waypoint is sorted according to the selected sorting strategy. For example, if sorted by distance, the Euclidean distance formula can be used to calculate the distance from each waypoint to the starting point, thus sorting the waypoints according to their distance from the starting point. Preferably, the order in which the information of multiple waypoints on the path to be identified is used for sorting. Since the reading order is the driving order of vehicles in historical data, it is more valuable for vehicles traveling on the same road in the same direction. Furthermore, sorting can be achieved by comparing time, which is faster, reduces computational complexity, and alleviates the pressure on system operation. The waypoints are arranged in a preset order so that the subsequent navigation system can follow these waypoints sequentially. There should be multiple waypoints on curved road sections, and adjacent waypoints should have some correlation. After sorting, grouping related waypoints together is more conducive to curve identification. Furthermore, the sorted waypoints provide a clear navigation route for vehicle travel. Additionally, the sorted waypoints can be used to create visual navigation routes on a map, allowing drivers or operators to clearly see the path.
[0106] In one embodiment, Figure 7 This is a schematic diagram illustrating a method for grouping multiple suspected curve path points according to an embodiment of this application, further defining the grouping process in step S13. Figure 7 As shown, the specific steps include S71-S74:
[0107] S71, Configure the corresponding index number for each path point sorted in the preset order;
[0108] S72, based on the preset order sorting, calculate the index number difference between the suspected curve path point and its adjacent suspected curve path point according to the index number configured for the path point, and determine whether the index number difference is less than the preset difference threshold.
[0109] S73, if the index sequence difference is less than the preset difference threshold, then the suspected curve path point and its adjacent suspected curve path point belong to the same suspected curve.
[0110] S74, if the index difference is not less than the preset difference threshold, then the suspected curve path point and its adjacent suspected curve path points do not belong to the same suspected curve.
[0111] In this embodiment, it is necessary to determine whether two adjacent suspected curved road path points belong to the same suspected curve. Generally, when the driving distance between two adjacent suspected curved road path points is relatively short, it is considered that these two suspected curved road path points belong to the same suspected curve. Therefore, it is possible to calculate the driving distance between two adjacent suspected path points and determine whether they belong to the same suspected curve through a preset distance threshold. In order to reduce the amount of data calculation, it is proposed to determine whether they belong to the same suspected curve by calculating the difference in the index numbers between two adjacent suspected path points. That is, corresponding index numbers are configured for each path point sorted in the preset order in sequence, and then the relationship between the suspected curved road path points is determined according to the difference in these index numbers. This calculation has a relatively small amount of calculation and is simple and clear. It should be further noted here that this is only a preferred specific implementation manner and does not limit this method.
[0112] Specifically, corresponding index numbers are assigned to each path point sorted in the preset order. These index numbers are usually consecutive integers, starting from the starting point of the path to be recognized and gradually increasing in sequence. The assignment of index numbers can adopt a simple iterative method to ensure that each path point has a unique index number. For the path points sorted in the preset order, calculate the difference between the index number of the suspected curved road path point and the index number of the adjacent suspected curved road path point. The index number difference reflects the relative interval between two adjacent suspected curved road path points. If the index number difference is less than the preset difference threshold, then the corresponding two suspected curved road path points are considered to belong to the same suspected curve. This indicates that two suspected curved road path points on the path are very close in space and may correspond to a curved road or a curve.
[0113] In a specific embodiment, Figure 8 This is a schematic diagram of the grouping result provided for a specific embodiment of this application. Here, the difference threshold is set as GT, and the index difference G is calculated for each suspected path point to its previous suspected path point. If the index number of the current suspected curved road path point is 32 and the index number of its previous suspected curved road path point is 10, then at this time G = 22. If G < GT, the distance between the two suspected path points is too close and they belong to the same suspected curve. If G ≥ GT, the distance between the two points is too far and they belong to different suspected curves. According to this comparison rule, traverse Figure 4 the suspected curved road path points shown, and the result is as Figure 8 shown. It can be seen that it is divided into segments 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12.
[0114] In an embodiment, Figure 9 This is a schematic diagram of the method for determining whether a suspected curve is a true curve provided for an embodiment of this application, and it is a description of the specific implementation of the suspected curve recognition in step S13. As Figure 9 As shown, the specific steps include S91-S93:
[0115] S91, determine whether the number of suspected curved path points contained in the suspected curved road is greater than a preset number threshold.
[0116] If S92 is greater than 2, then the suspected curve is actually a curve.
[0117] If S93 is not greater than 1, then the suspected curve is a false curve.
[0118] In this embodiment, the number of pathpoints within a suspected curve is obtained and compared with a preset threshold to determine whether the suspected curve is a true or false curve. Curves are typically the curved sections of a road. After filtering and instantaneous curvature thresholding, true curves are characterized by a high density of pathpoints with abnormal instantaneous curvature; that is, true curves usually contain more pathpoints with instantaneous curvature exceeding the instantaneous curvature threshold. Therefore, the number of pathpoints within a suspected curve is an important metric. The preset threshold is predefined. A low threshold may lead to too many suspected curves being identified as true curves, reducing accuracy. Conversely, a high threshold may cause true curves to be mistakenly identified as false curves, reducing system performance. Therefore, the threshold can be adjusted according to different scenarios and road conditions to adapt to the characteristics of curves under different circumstances. Different thresholds may be needed for different types of roads, such as urban roads and highways. Flexible threshold adjustment can balance the accuracy and performance of identification.
[0119] For a set of pathpoints identified as belonging to the same suspected curve, the number of suspected curve pathpoints within it is calculated. This calculated number is then compared to a preset threshold. If the number of suspected curve pathpoints exceeds the threshold, the suspected curve is identified as a genuine curve; otherwise, it is identified as a false curve. While threshold comparison is a relatively simple process, it is a method based on real-world conditions, more closely resembling real-world scenarios, and possessing high accuracy. It can accurately identify curves and take appropriate measures to address complex road conditions, ensuring safe and efficient driving and navigation in various situations.
[0120] In one specific embodiment, Figure 10 This is a schematic diagram illustrating the remaining suspected curves obtained after quantity threshold filtering, provided as a specific embodiment of this application. Based on Figure 8 The segmentation results are used to determine whether the number of suspected pathpoints in each segment is greater than a preset threshold, and segments with a number less than the preset threshold are removed. The result is as follows. Figure 10 As shown, the remaining segments are 1, 2, 4, 5, 6, 8, 10, and 12.
[0121] In one embodiment, the method of this application further includes: obtaining path point information corresponding to the starting point of a true curve and path point information corresponding to the ending point of a curve based on a preset order, and sending the path point information corresponding to the starting point of the curve and the path point information corresponding to the ending point of the curve to the driving planning module of the autonomous vehicle to realize driving control of autonomous driving.
[0122] In this embodiment, the pathpoint information at the start and end of the curve is transmitted to the driving planning module of the autonomous vehicle. This module is responsible for formulating the vehicle's driving strategy and trajectory planning. This information enables the autonomous vehicle to know the curve's coordinates and heading. The autonomous vehicle obtains the path information for the entire curve based on the pathpoint information corresponding to the start and end points. Specifically, each pathpoint is assigned an index number according to a preset order. Once the index number of the curve's start point is determined, the coordinate information of that pathpoint can be obtained. This coordinate information includes the horizontal (x) and vertical (y) coordinates, used to accurately locate the vehicle's position on the road. Furthermore, the pathpoint information at the curve's start point also includes heading information. The heading angle is the direction the vehicle is facing, crucial for navigation and control. Heading information helps the vehicle determine when to begin turning to follow the curve. Similar to the curve's start point, after determining the index number of the curve's end point, the system can extract the coordinate information of that pathpoint. This coordinate information marks the end of the curve, helping the vehicle smoothly transition to the next road segment. Simultaneously, the heading information at this waypoint helps the vehicle know when to end the turn and gradually adjust to the next direction of travel. Furthermore, considering the characteristics of the road curve, the driving planning module adjusts the vehicle's speed to adapt to the curve radius and the vehicle's dynamic performance. The driving planning module is also responsible for heading control to ensure that the vehicle can turn correctly on curves and follow the road curve.
[0123] Based on this, Figure 11 This is a detailed schematic diagram illustrating the intra-segment blanking process provided in a specific embodiment of this application. Since the pathpoint information corresponding to the start and end points of the curve is already known, the corresponding locations can be found in the original empirical map, and all pathpoints from the start to the end of the curve can be marked for blanking the curve path segment. Specifically, the pathpoints corresponding to the blank portions within the segment are filled with the pathpoint information from the corresponding locations on the empirical map. Figure 11 (a) and Figure 11 (b) A comparison of paragraph 5 before and after the addition; Figure 11 (c) and Figure 11(d) Comparison of paragraph 8 before and after the supplementation; Figure 11 (e) and Figure 11 (f) Comparison of paragraph 10 before and after the supplement.
[0124] Furthermore, the completed segment can be filled in, and the information of all waypoints within that segment can be transmitted to the autonomous vehicle. This directly provides the vehicle with complete curve waypoint information, reducing the time consumed in information acquisition and further enhancing the vehicle's safety, stability, and navigation performance, thereby achieving safer, more efficient, and reliable autonomous driving.
[0125] Building upon this foundation, complete curve information, including curve lines and markings, can be presented as explicit images on the human-machine interface. This visually indicates the location and shape of the curve. These lines and markings can use dashed lines, color variations, or other visual effects to highlight the curve. This allows operators or drivers to gain a more intuitive and objective understanding of the curve information along the path. Displaying complete curve information helps users understand when the vehicle needs to slow down, turn, or maintain speed, thereby increasing user trust in the autonomous driving system. In emergency situations, such as encountering sudden events or changes in road conditions, the displayed curve information can help users react more quickly, such as taking evasive action or interacting effectively with the autonomous driving system.
[0126] As can be seen from the above specific embodiments, instantaneous curvature is a method for measuring path curvature. By obtaining instantaneous curvature, the curvature of the road can be understood. This provides basic information for subsequent curve detection, enabling the system to identify curved sections of the road. Further, based on preset filtering rules, the initial instantaneous curvature of path points matching these rules is filtered. This removes some noise or outliers, making the final instantaneous curvature smoother and more stable. The filtered instantaneous curvature reflects more accurate road curve characteristics. Simultaneously, based on a preset instantaneous curvature threshold, the filtered instantaneous curvature and the unfiltered initial instantaneous curvature are screened. Path points that may be located on curves are identified. If the instantaneous curvature is greater than the threshold, it indicates that the path point may be on a curve; or if the initial instantaneous curvature is greater than the threshold, it will also be considered a suspected curve path point. Suspected curve path points are grouped. These groups are based on the difference in index numbers between path points to determine whether they belong to the same curve. This helps organize suspected curve path points into curve segments, providing structured data for subsequent analysis. Based on the number of suspected curve path points in each group, it determines whether a suspected curve is a genuine curve. By eliminating some suspected cases, it ensures that only genuine curved roads are identified as curves, effectively identifying curves from the paths to be identified in the experience map. This provides crucial information for vehicle path planning and control, thereby improving driving safety and efficiency. It reliably detects and identifies curves under different road conditions to support the decision-making and operation of autonomous vehicles.
[0127] In this embodiment of the invention, electronic devices or main control devices can be divided into functional modules according to the above method examples. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional module. It should be noted that the module division in this embodiment of the invention is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.
[0128] Figure 12 This is a schematic diagram of the curve recognition device based on an experience map provided in an embodiment of this application. Figure 12 As shown, the curve recognition device 12 includes:
[0129] The curvature calculation module 121 is used to obtain the initial instantaneous curvature of each path point sorted in a preset order based on the path to be identified in the experience map, and to filter the initial instantaneous curvature of the path points that match the preset filtering rules according to the preset filtering rules, so as to obtain the filtered instantaneous curvature of the path points that match the preset filtering rules and the initial instantaneous curvature of the path points that do not match the preset filtering rules.
[0130] The path point filtering module 122 is used to filter path points corresponding to the filtered instantaneous curvature and path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules according to the preset instantaneous curvature threshold, so as to obtain multiple suspected curve path points.
[0131] The curve recognition module 123 is used to group multiple suspected curve path points and, based on the number of suspected curve path points contained in the suspected curves obtained through grouping, to recognize all curves within the path to be recognized in the experience map.
[0132] The electronic device provided in this embodiment can execute the curve recognition method based on experience map described in the above embodiment. Its implementation principle and technical effect are similar, and will not be described again in this embodiment.
[0133] In the aforementioned specific implementation, each module can be implemented as a processor, which can execute computer execution instructions stored in memory, thereby enabling the processor to perform the methods described above.
[0134] Figure 13 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 13 As shown, the electronic device 13 includes at least one processor 131 and a memory 132. The electronic device 13 also includes a communication component 133. The processor 131, the memory 132, and the communication component 133 are connected via a bus 134.
[0135] In the specific implementation process, at least one processor 131 executes computer execution instructions stored in memory 132, causing at least one processor 131 to execute the curve recognition method based on experience map as executed on the electronic device side as described above.
[0136] The specific implementation process of processor 131 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.
[0137] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.
[0138] The memory may include high-speed RAM, and may also include non-volatile storage (NVM), such as at least one disk storage.
[0139] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.
[0140] The above description of the functions implemented by electronic devices and main control devices has introduced the solutions provided by the embodiments of the present invention. It is understood that, in order to implement the above functions, the electronic device or main control device includes hardware structures and / or software modules corresponding to the execution of each function. By combining the units and algorithm steps of the various examples described in the embodiments of the present invention, the embodiments of the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the technical solutions of the embodiments of the present invention.
[0141] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described curve recognition method based on an experience map.
[0142] The aforementioned computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.
[0143] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in an electronic device or a host device.
[0144] This application also provides a computer program product, comprising: a computer program stored in a readable storage medium, wherein at least one processor of an electronic device can read the computer program from the readable storage medium, and the at least one processor executes the computer program to cause the electronic device to perform the scheme provided in any of the above embodiments.
[0145] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.
[0146] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A curve recognition method based on empirical maps, characterized in that, include: Based on the path to be identified in the experience map, the initial instantaneous curvature of each path point is obtained in a preset order. According to the preset filtering rule, the initial instantaneous curvature of the path points that match the preset filtering rule is filtered to obtain the filtered instantaneous curvature of the path points that match the preset filtering rule and the initial instantaneous curvature of the path points that do not match the preset filtering rule. Based on the preset instantaneous curvature threshold, the path points corresponding to the filtered instantaneous curvature and the path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules are filtered to obtain multiple suspected curved path points. The multiple suspected curve path points are grouped and processed, and the number of suspected curve path points contained in the suspected curve obtained by the grouping process is used to determine whether the suspected curve is a real curve, so as to realize the identification of all curves in the path to be identified in the experience map. The step of filtering the initial instantaneous curvature of path points matching the preset filtering rule according to the preset filtering rule, to obtain the filtered instantaneous curvature of path points matching the preset filtering rule and the initial instantaneous curvature of path points not matching the preset filtering rule, includes: Based on a preset filter length threshold, filter start point and filter end point are obtained; wherein, the filter start point is the distance from the start point of the current path to be identified to the filter length threshold, and the filter end point is the distance from the end point of the current path to be identified to the filter length threshold. Based on the filter start point and the filter end point, the following formula is used: The instantaneous curvature of the filtered path points from the starting point to the ending point is obtained sequentially; where γ i1 γ represents the filtered instantaneous curvature of path point i. i0 γ represents the initial instantaneous curvature of path point i. j Let represent the initial instantaneous curvature of path point j, and L represent the filter length threshold. The process of obtaining the initial instantaneous curvature of each path point sorted in a preset order based on the path to be identified in the empirical map includes: If the path point is the starting point of the path to be identified, then the initial instantaneous curvature of the starting point is the initial instantaneous curvature of the path points adjacent to the starting point. or, If the path point is the endpoint of the path to be identified, then the initial instantaneous curvature of the endpoint path point is the initial instantaneous curvature of the path points adjacent to the endpoint path point. or, If the path point is neither the starting path point nor the ending path point, then the following formula is used: The initial instantaneous curvature γ of the path point is obtained. i0 ; where φ (i-1) φ represents the heading angle of path point i-1. (i+1) Let x represent the heading angle of path point i+1. (i-1) The x-coordinate of path point i-1 is represented by x. (i+1) The x-coordinate of path point i+1 is represented by y. (i-1) The y-coordinate of path point i-1 is represented by... (i+1) This represents the ordinate of the path point i+1.
2. The method according to claim 1, characterized in that, The method involves filtering path points corresponding to the filtered instantaneous curvature and path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules, based on a preset instantaneous curvature threshold, to obtain multiple suspected curve path points, including: If the instantaneous curvature of the filter is greater than the instantaneous curvature threshold, then the path point corresponding to the instantaneous curvature of the filter is a suspected curved path point; or, If the initial instantaneous curvature is greater than the instantaneous curvature threshold, then the path point corresponding to the initial instantaneous curvature is a suspected curve path point.
3. The method according to claim 1, characterized in that, Before obtaining the initial instantaneous curvature of each path point sorted in a preset order in the path to be identified based on the empirical map, the method includes: Obtain the path information to be identified from the experience map. The path information to be identified includes multiple discrete path points and the coordinate information and heading information corresponding to the path points. The multiple discrete path points are sorted according to a preset order to obtain multiple path points sorted in the preset order.
4. The method according to claim 3, characterized in that, The grouping process for the multiple suspected curve path points includes: Configure a corresponding index number for each path point sorted according to the preset order; Based on the preset sorting order, according to the index number configured for the path point, the index number difference between the suspected curve path point and its adjacent suspected curve path point is calculated sequentially, and it is determined whether the index number difference is less than the preset difference threshold. If the difference in index number is less than a preset difference threshold, then the suspected curve path point and its adjacent suspected curve path points belong to the same suspected curve. If the difference in index number is not less than a preset difference threshold, then the suspected curve path point and its adjacent suspected curve path points do not belong to the same suspected curve.
5. The method according to claim 1 or 3, characterized in that, The step of determining whether a suspected curve is a true curve based on the number of suspected curve path points included in the suspected curve obtained from the grouping process includes: Determine whether the number of suspected curve path points contained in the suspected curve road is greater than a preset number threshold; If the value is greater than the value, then the suspected curve is a true curve; If the value is not greater than the value, then the suspected curve is a false curve.
6. The method according to claim 5, characterized in that, The method further includes: Based on a preset sorting order, the path point information corresponding to the starting point and the ending point of the true curve is obtained, and the path point information corresponding to the starting point and the ending point of the curve is sent to the driving planning module of the autonomous vehicle to realize the driving control of the autonomous driving.
7. A curve recognition device based on an experience map, characterized in that, include: The curvature calculation module is used to obtain the initial instantaneous curvature of each path point sorted in a preset order based on the path to be identified in the experience map, and to filter the initial instantaneous curvature of the path points that match the preset filtering rules according to the preset filtering rules, so as to obtain the filtered instantaneous curvature of the path points that match the preset filtering rules and the initial instantaneous curvature of the path points that do not match the preset filtering rules. The path point filtering module is used to filter path points corresponding to the filtered instantaneous curvature and path points corresponding to the initial instantaneous curvature that do not match the preset filtering rules according to a preset instantaneous curvature threshold, thereby obtaining multiple suspected curve path points. The curve identification module is used to group the multiple suspected curve path points and, based on the number of suspected curve path points contained in the suspected curve obtained by the grouping process, to identify all curves in the path to be identified in the experience map. The curvature calculation module is specifically used to filter and obtain the starting point and ending point of the filter path according to a preset filter length threshold; wherein, the length of the starting point of the filter path from the starting point of the current path to be identified is the filter length threshold, and the length of the ending point of the filter path from the ending point of the current path to be identified is the filter length threshold. Based on the filter start point and the filter end point, the following formula is used: The instantaneous curvature of the filtered path points from the starting point to the ending point is obtained sequentially; where γ i1 γ represents the filtered instantaneous curvature of path point i. i0 γ represents the initial instantaneous curvature of path point i. j Let represent the initial instantaneous curvature of path point j, and L represent the filter length threshold. The curvature calculation module is further specifically used to determine the initial instantaneous curvature of the starting path point as the initial instantaneous curvature of the path points adjacent to the starting path point if the path point is the starting path point of the path to be identified. or, If the path point is the endpoint of the path to be identified, then the initial instantaneous curvature of the endpoint path point is the initial instantaneous curvature of the path points adjacent to the endpoint path point. or, If the path point is neither the starting path point nor the ending path point, then the following formula is used: The initial instantaneous curvature γ of the path point is obtained. i0 ; where φ (i-1) φ represents the heading angle of path point i-1. (i+1) Let x represent the heading angle of path point i+1. (i-1) The x-coordinate of path point i-1 is represented by x. (i+1) The x-coordinate of path point i+1 is represented by y. (i-1) The y-coordinate of path point i-1 is represented by... (i+1) This represents the ordinate of the path point i+1.
8. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 6.