Method, device, equipment and storage medium for collecting turning driving data

Abstract

The present invention relates to the field of autonomous driving technology, and discloses a method, device, equipment and storage medium for collecting turning driving data, which are used to accurately and efficiently collect turning driving data. The method for collecting turning driving data comprises: obtaining multiple frames of driving data of a main vehicle, and filtering the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle; performing main vehicle turn recognition on the multiple frames of intersection driving data corresponding to each intersection to obtain multiple frames of first turning driving data corresponding to at least one turn; based on the obstacle information around the main vehicle in each frame of the first turning driving data, performing second turn discrimination on the multiple frames of first turning driving data corresponding to each turn to obtain target turning driving data.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a method, device, equipment and storage medium for collecting turning driving data. Background Art

[0002] With the development of autonomous driving technology, autonomous vehicles can drive autonomously in different scenarios, such as red light scenarios and intersection scenarios classified according to road conditions, overtaking scenarios and lane merging scenarios classified according to driving behavior. Among them, turning and changing lanes between the main vehicle and the obstacle vehicle is a relatively complex and error-prone algorithm scenario in autonomous driving.

[0003] The existing technology usually collects turning driving data by manually annotating data to train classification models. However, this method is too inefficient to collect target scenes by manually checking all scenes due to the huge amount of unmanned driving data. The results obtained by the existing judgment logic have too much noise, and most scenes are concentrated in more crowded roads, which are not typical turning scenes. As a result, it is impossible to accurately and efficiently collect turning scene data. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for collecting turning driving data, which are used for accurately and efficiently collecting turning driving data.

[0005] A first aspect of the present invention provides a method for collecting turning driving data, comprising:

[0006] Acquire multiple frames of driving data of the main vehicle, and perform intersection range data screening on the multiple frames of driving data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle;

[0007] Performing host vehicle turn recognition on multiple frames of intersection driving data corresponding to each intersection, and obtaining multiple frames of first turning driving data corresponding to at least one turn;

[0008] Based on the obstacle information around the host vehicle in each frame of the first turning driving data, a second turning discrimination is performed on a plurality of frames of the first turning driving data corresponding to each turn to obtain target turning driving data.

[0009] Optionally, the acquiring of multiple frames of driving data of the main vehicle and filtering the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection respectively include:

[0010] Acquire multiple frames of driving data of the main vehicle, each frame of driving data includes intersection range identification information, and the intersection range identification information is used to indicate whether the main vehicle is in the intersection range in the corresponding frame of driving data;

[0011] The intersection range data of the host vehicle is filtered for each frame of driving data using the intersection range identification information to obtain multiple frames of intersection driving data corresponding to at least one intersection.

[0012] Optionally, the filtering of the main vehicle intersection range data for each frame of driving data by using the intersection range identification information to obtain multiple frames of intersection driving data corresponding to at least one intersection respectively includes:

[0013] Setting any one frame of the multiple frames of driving data as the current frame;

[0014] The intersection range identification information is used to sequentially identify the intersection range of the main vehicle for the current frame and the next frame of the current frame until all frames complete the intersection range identification of the main vehicle, thereby obtaining a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range;

[0015] A plurality of frames of intersection driving data corresponding to at least one intersection in the plurality of frames of driving data are determined through the first target frame and the second target frame.

[0016] Optionally, the performing of main vehicle turn recognition on multiple frames of intersection driving data corresponding to each intersection to obtain multiple frames of first turn driving data corresponding to at least one turn, includes:

[0017] Calculate the number of frames of intersection driving data corresponding to each intersection to obtain the number of intersection frames corresponding to each intersection, and determine whether the number of intersection frames corresponding to each intersection is less than a preset intersection frame number threshold;

[0018] If the intersection frame number is less than the preset intersection frame number threshold, the difference in the initial and final directions of the multi-frame intersection driving data of the corresponding intersection is calculated to obtain the main vehicle direction difference corresponding to each intersection, and the multi-frame intersection driving data of the corresponding intersection is determined as the main vehicle turning driving data according to the main vehicle direction difference corresponding to each intersection, and multiple frames of first turning driving data corresponding to at least one turn are obtained.

[0019] Optionally, the calculation of the difference in the start and end orientations of multiple frames of intersection driving data corresponding to the intersection to obtain the orientation difference of the main vehicle corresponding to each intersection, and determining the multiple frames of intersection driving data corresponding to the intersection as the main vehicle turning driving data according to the orientation difference of the main vehicle corresponding to each intersection, and obtaining multiple frames of first turning driving data corresponding to at least one turn, respectively, includes:

[0020] Obtain the heading angles of the main vehicle at the starting position and the ending position of the intersection in multiple frames of intersection driving data, respectively, to obtain the starting heading angle and the ending heading angle corresponding to each intersection;

[0021] Calculate the orientation difference of the main vehicle for the starting orientation angle and the ending orientation angle corresponding to each intersection, obtain the orientation difference of the main vehicle corresponding to each intersection, and determine whether the orientation difference of the main vehicle corresponding to each intersection is within the preset orientation difference range;

[0022] If the main vehicle's orientation difference is within a preset orientation difference range, the multi-frame intersection driving data of the corresponding intersection is determined to be a multi-frame main vehicle turning driving data, and a multi-frame first turning driving data corresponding to at least one turn is obtained.

[0023] Optionally, the second turn discrimination is performed on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the main vehicle in each frame of the first turn driving data to obtain the target turn driving data, including:

[0024] Obtaining the vehicle obstacle distance and the vehicle obstacle direction in the obstacle information around the main vehicle in the first turning driving data of each frame;

[0025] By using the vehicle obstacle distance and the vehicle obstacle direction, determining whether the obstacle information around the main vehicle in each frame of the first turning driving data corresponding to each turn satisfies the preset second turning judgment condition;

[0026] If the obstacle information around the main vehicle in the turning driving data meets the preset second turn judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, and the second turning driving data corresponding to at least one second turning process is obtained, and the target turning driving data is determined according to the total number of frames of the second turning driving data corresponding to each second turning process.

[0027] Optionally, judging whether the obstacle information around the main vehicle in each frame of the first turning driving data corresponding to each turn satisfies a preset second turning judgment condition by using the vehicle obstacle distance and the vehicle obstacle direction includes:

[0028] Determine whether the vehicle obstacle distance corresponding to the first turning driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold;

[0029] If the vehicle obstacle distance corresponding to the turning driving data is less than the preset following vehicle distance threshold, the difference between the vehicle obstacle orientation in the turning driving data of the corresponding frame and the main vehicle orientation in the turning driving data of the corresponding frame is calculated to obtain the turning orientation difference;

[0030] Determining whether the turning direction difference is less than a preset direction difference threshold;

[0031] If the turning direction difference is less than a preset direction difference threshold, it is determined that the obstacle information around the host vehicle in the turning driving data of the corresponding frame meets the preset second turning judgment condition.

[0032] Optionally, if the obstacle information around the main vehicle in the turning driving data satisfies a preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, and the second turning driving data corresponding to at least one second turning process is obtained, and the target turning driving data is determined according to the total number of frames of the second turning driving data corresponding to each second turning process, including:

[0033] If the obstacle information around the host vehicle in the turning driving data satisfies the preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data corresponding to the turn, and the second turning driving data corresponding to at least one second turning process is obtained;

[0034] Calculate the total number of frames of the second turning driving data corresponding to each second turning process to obtain the number of second turning frames corresponding to each second turning process;

[0035] Determining whether the second turning frame number corresponding to each second turning process meets a preset second turning threshold condition;

[0036] If the second turning frame number meets the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data.

[0037] Optionally, after performing second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the main vehicle in each frame of first turn driving data to obtain target turn driving data, the method for collecting turn driving data further includes:

[0038] According to the second turning frame number, the target turning driving data corresponding to each second turning process are sorted by scene relevance to obtain a second turning driving data sequence.

[0039] A second aspect of the present invention provides a device for collecting turning driving data, comprising:

[0040] A screening module is used to obtain multiple frames of driving data of the main vehicle, and to screen the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle;

[0041] An identification module is used to identify the main vehicle turning on the multiple frames of intersection driving data corresponding to each intersection, and obtain multiple frames of first turning driving data corresponding to at least one turn;

[0042] The discrimination module is used to perform second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on obstacle information around the main vehicle in each frame of first turn driving data to obtain target turn driving data.

[0043] Optionally, the screening module includes:

[0044] An acquisition unit, used to acquire multiple frames of driving data of the main vehicle, each frame of driving data includes intersection range identification information, and the intersection range identification information is used to indicate whether the main vehicle is in the intersection range in the corresponding frame of driving data;

[0045] The screening unit is used to screen the intersection range data of the main vehicle for each frame of driving data according to the intersection range identification information, and obtain multiple frames of intersection driving data corresponding to at least one intersection.

[0046] Optionally, the screening unit is specifically used for:

[0047] Setting any one frame of the multiple frames of driving data as the current frame;

[0048] The intersection range identification information is used to sequentially identify the intersection range of the main vehicle for the current frame and the next frame of the current frame until all frames complete the intersection range identification of the main vehicle, thereby obtaining a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range;

[0049] A plurality of frames of intersection driving data corresponding to at least one intersection in the plurality of frames of driving data are determined through the first target frame and the second target frame.

[0050] Optionally, the identification module includes:

[0051] A calculation unit, used to calculate the number of frames of intersection driving data corresponding to each intersection, obtain the number of intersection frames corresponding to each intersection, and determine whether the number of intersection frames corresponding to each intersection is less than a preset intersection frame number threshold;

[0052] A determination unit is used to calculate the difference in the initial and final directions of multiple frames of intersection driving data of the corresponding intersection if the number of intersection frames is less than a preset intersection frame number threshold, obtain the difference in the direction of the main vehicle corresponding to each intersection, and determine that the multiple frames of intersection driving data of the corresponding intersection are the turning driving data of the main vehicle according to the difference in the direction of the main vehicle corresponding to each intersection, and obtain multiple frames of first turning driving data corresponding to at least one turn.

[0053] Optionally, the determining unit is specifically configured to:

[0054] Obtain the heading angles of the main vehicle at the starting position and the ending position of the intersection in multiple frames of intersection driving data, respectively, to obtain the starting heading angle and the ending heading angle corresponding to each intersection;

[0055] Calculate the orientation difference of the main vehicle for the starting orientation angle and the ending orientation angle corresponding to each intersection, obtain the orientation difference of the main vehicle corresponding to each intersection, and determine whether the orientation difference of the main vehicle corresponding to each intersection is within the preset orientation difference range;

[0056] If the main vehicle's orientation difference is within a preset orientation difference range, the multi-frame intersection driving data of the corresponding intersection is determined to be a multi-frame main vehicle turning driving data, and a multi-frame first turning driving data corresponding to at least one turn is obtained.

[0057] Optionally, the identification module includes:

[0058] An extraction unit, used to obtain the vehicle obstacle distance and the vehicle obstacle direction in the obstacle information around the main vehicle in the first turning driving data of each frame;

[0059] A judgment unit, configured to judge whether the obstacle information around the host vehicle in each frame of the first turning driving data corresponding to each turn satisfies a preset second turning judgment condition based on the vehicle obstacle distance and the vehicle obstacle direction;

[0060] A determination unit is used to determine the corresponding frame of turning driving data as the second turning driving data of the corresponding turn if the obstacle information around the main vehicle in the turning driving data meets the preset second turning determination condition, obtain the second turning driving data corresponding to at least one second turning process, and determine the target turning driving data according to the total number of frames of the second turning driving data corresponding to each second turning process.

[0061] Optionally, the judging unit is specifically configured to:

[0062] Determine whether the vehicle obstacle distance corresponding to the first turning driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold;

[0063] If the vehicle obstacle distance corresponding to the turning driving data is less than the preset following vehicle distance threshold, the difference between the vehicle obstacle orientation in the turning driving data of the corresponding frame and the main vehicle orientation in the turning driving data of the corresponding frame is calculated to obtain the turning orientation difference;

[0064] Determining whether the turning direction difference is less than a preset direction difference threshold;

[0065] If the turning direction difference is less than a preset direction difference threshold, it is determined that the obstacle information around the host vehicle in the turning driving data of the corresponding frame meets the preset second turning judgment condition.

[0066] Optionally, the determination unit is specifically used for:

[0067] If the obstacle information around the host vehicle in the turning driving data satisfies the preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data corresponding to the turn, and the second turning driving data corresponding to at least one second turning process is obtained;

[0068] Calculate the total number of frames of the second turning driving data corresponding to each second turning process to obtain the number of second turning frames corresponding to each second turning process;

[0069] Determining whether the second turning frame number corresponding to each second turning process meets a preset second turning threshold condition;

[0070] If the second turning frame number meets the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data.

[0071] Optionally, the device for collecting turning driving data further includes:

[0072] The sorting module is used to sort the target turning driving data corresponding to each second turning process according to the second turning frame number, so as to obtain a second turning driving data sequence.

[0073] The third aspect of the present invention provides a device for collecting turning driving data, comprising: a memory and at least one processor, wherein a computer program is stored in the memory; the at least one processor calls the computer program in the memory so that the device for collecting turning driving data executes the above-mentioned method for collecting turning driving data.

[0074] A fourth aspect of the present invention provides a computer-readable storage medium, in which a computer program is stored. When the computer-readable storage medium is run on a computer, the computer executes the above-mentioned method for collecting turning driving data.

[0075] In the technical solution provided by the present invention, multiple frames of driving data of the main vehicle are obtained, and the multiple frames of driving data are screened for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data includes obstacle information around the main vehicle; the main vehicle turn recognition is performed on the multiple frames of intersection driving data corresponding to each intersection to obtain multiple frames of first turn driving data corresponding to at least one turn; based on the obstacle information around the main vehicle in each frame of the first turn driving data, the multiple frames of first turn driving data corresponding to each turn are judged for second turn to obtain target turn driving data. In the embodiment of the present invention, since the main vehicle is at the intersection is a prerequisite for turning, therefore, the multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data are extracted through the intersection range data screening, and whether the main vehicle turns when passing the intersection, the multiple frames of first turn driving data corresponding to at least one turn in the multiple frames of intersection driving data corresponding to each intersection are obtained through the main vehicle turn recognition, and finally, the obstacle information around the main vehicle is used to judge whether the main vehicle and the vehicle obstacle make a second turn, thereby obtaining the target turn driving data, and the present invention can accurately and efficiently collect the second turn driving data. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 A schematic diagram of an embodiment of a method for collecting turning driving data in an embodiment of the present invention;

[0077] Figure 2 A schematic diagram of another embodiment of a method for collecting turning driving data in an embodiment of the present invention;

[0078] Figure 3 A schematic diagram of an embodiment of a device for collecting turning driving data in an embodiment of the present invention;

[0079] Figure 4 A schematic diagram of another embodiment of a device for collecting turning driving data according to an embodiment of the present invention;

[0080] Figure 5 The figure is a schematic diagram of an embodiment of a device for collecting turning driving data in an embodiment of the present invention. DETAILED DESCRIPTION

[0081] The embodiments of the present invention provide a method, device, equipment and storage medium for collecting turning driving data, which are used to accurately and efficiently collect second turning driving data.

[0082] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0083] It is understandable that the execution subject of the present invention may be a device for collecting turning driving data, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking a server as the execution subject as an example.

[0084] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the method for collecting turning driving data in an embodiment of the present invention includes:

[0085] 101. Acquire multiple frames of driving data of the main vehicle, and filter the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, wherein each frame of driving data includes obstacle information around the main vehicle;

[0086] It should be noted that the main vehicle refers to the automatic driving control vehicle, and the multi-frame driving data of the main vehicle can be the actual driving data of the main vehicle or the road test data of the main vehicle. The multi-frame driving data includes the main vehicle driving control data of a preset duration, and each frame of driving data includes but is not limited to the basic information of the main vehicle, the position information of the main vehicle, the orientation information of the main vehicle, the intersection range identification information of the main vehicle, and the obstacle information around the main vehicle, among which the obstacle information around the main vehicle is used to indicate the vehicle information that causes an obstacle effect on the main vehicle, and the obstacle information around the main vehicle includes but is not limited to the position information and orientation information of the vehicle obstacle.

[0087] In one embodiment, in order to control the amount of data for the collection operation, that is, to control the amount of data of multiple frames of driving data, multiple frames of driving data of the main vehicle are obtained through a preset collection time range and collection vehicle range, thereby improving the collection efficiency of turning driving data.

[0088] It should be noted that since the main vehicle in the multi-frame driving data may pass through one or more intersections, in order to improve the accuracy of collecting turning driving data and reduce the noise of the collected data, in one embodiment, multi-frame intersection driving data corresponding to at least one intersection is extracted from the multi-frame driving data. Specifically, based on the position information of the main vehicle in each frame of driving data, it is determined whether the main vehicle is at the intersection. If the main vehicle is at the intersection, the corresponding frame driving data is determined to be the intersection driving data of the corresponding intersection, and multi-frame intersection driving data corresponding to at least one intersection is obtained.

[0089] 102. Performing host vehicle turn recognition on multiple frames of intersection driving data corresponding to each intersection, and obtaining multiple frames of first turning driving data corresponding to at least one turn;

[0090] It is understandable that, since there is a situation where the main vehicle passes through the intersection but does not necessarily turn, in order to improve the accuracy of turning driving data collection and reduce data noise, the main vehicle turn identification is performed on the multiple frames of intersection driving data corresponding to each intersection to obtain multiple frames of first turning driving data corresponding to at least one turn. In one embodiment, the main vehicle orientation difference of the multiple frames of intersection driving data corresponding to each intersection is calculated through the orientation information of the main vehicle in each frame of intersection driving data to obtain the main vehicle orientation difference corresponding to each intersection, and then the main vehicle orientation difference corresponding to each intersection is determined to determine whether the main vehicle turns at the corresponding intersection. If the main vehicle turns at the corresponding intersection, the multiple frames of intersection driving data corresponding to the corresponding intersection are determined as multiple frames of first turning driving data to obtain multiple frames of first turning driving data corresponding to at least one turn. Further, by determining whether the main vehicle orientation difference corresponding to each intersection is greater than a preset orientation difference threshold, it is determined whether the main vehicle turns at the corresponding intersection, wherein the main vehicle orientation difference is used to indicate the orientation angle difference between the main vehicle at the starting position and the end position of the intersection. For example, assuming that there are multiple frames of intersection driving data corresponding to intersections A, intersection B and intersection C respectively, through the orientation information of the main vehicle in each frame of intersection driving data, it is calculated that the orientation difference of the main vehicle corresponding to intersection A is 20 degrees, the orientation difference of the main vehicle corresponding to intersection B is 90 degrees, and the orientation difference of the main vehicle corresponding to intersection C is 190 degrees. Assuming that the preset orientation difference threshold is 60 degrees, it is determined that the main vehicle turns at intersections B and intersection C. Then, the multiple frames of intersection driving data corresponding to intersection B are determined as the multi-frame first turn driving data of one turn, and the multi-frame intersection driving data corresponding to intersection C are determined as the multi-frame first turn driving data of another turn, and the multi-frame first turn driving data corresponding to the two turns are obtained.

[0091] In one embodiment, since the turning scene under congestion is not a typical scene data, it is difficult to be applied to other typical turning scenes, such as test data for the development or iteration of autonomous driving algorithms. Therefore, in order to reduce the noise of turning driving data collection and improve the accuracy of turning driving data collection, after obtaining multiple frames of intersection driving data corresponding to at least one intersection, before performing main vehicle turn recognition on the multiple frames of intersection driving data corresponding to each intersection, the intersection driving time is calculated for the multiple frames of intersection driving data corresponding to each intersection to obtain the intersection driving time corresponding to each intersection, and it is determined whether the intersection driving time corresponding to each intersection is less than a preset intersection driving time threshold. If the intersection driving time is less than the preset intersection driving time threshold, the multiple frames of intersection driving data corresponding to the corresponding intersection are determined as multiple frames of smooth driving data, and multiple frames of smooth driving data corresponding to at least one intersection are obtained. The main vehicle turn recognition is performed on the multiple frames of smooth driving data corresponding to each intersection to obtain multiple frames of first turning driving data corresponding to at least one turn.

[0092] 103. Based on the obstacle information around the host vehicle in each frame of the first turning driving data, a second turning discrimination is performed on a plurality of frames of the first turning driving data corresponding to each turn to obtain target turning driving data.

[0093] It can be understood that the second turn can be a parallel turn between the main vehicle and the surrounding obstacle vehicles, or a follow-up turn between the main vehicle and the surrounding obstacle vehicles, or a turn in which the main vehicle maintains a specific turning relationship with the surrounding vehicle obstacles, that is, the target turning driving data can be parallel turning driving data, follow-up turning driving data, and other turning driving data in which the main vehicle maintains a specific turning relationship with the surrounding vehicle obstacles, among which parallel turning means that the main vehicle and the surrounding obstacle vehicles maintain a certain degree of relative parallelism during the turning process, and follow-up turning means that the main vehicle and the surrounding obstacle vehicles maintain a following relationship during the turning process. In order to accurately collect turning driving data, based on the obstacle information around the main vehicle in each frame of the first turning driving data, the second turn discrimination is performed on the multiple frames of first turning driving data corresponding to each turn to obtain the target turning driving data. In one embodiment, based on the orientation information and position information of the main vehicle in each frame of the first turning driving data, and the orientation information and position information in the obstacle information around the main vehicle, second turning vehicle obstacle identification is performed on multiple frames of first turning driving data corresponding to each turn to obtain an identification result, wherein the second turning vehicle obstacle is used to indicate a vehicle obstacle that maintains a specific turning relationship with the main vehicle, such as a parallel turning vehicle obstacle, a following turning vehicle obstacle, etc., and the identification result is used to indicate the number of second turning vehicle obstacles in the corresponding frame of turning driving data. If the identification result indicates that the number of second turning vehicle obstacles in the corresponding frame of turning driving data is greater than a preset number threshold, the corresponding frame is determined to be second turning driving data, and at least one frame of second turning driving data corresponding to each turn is obtained, and the target turning driving data is determined according to the at least one frame of second turning driving data corresponding to each turn.

[0094] Furthermore, based on at least one frame of second turn driving data corresponding to each turn, determining the target turn driving data includes: calculating the number of frames of the second turn driving data corresponding to each turn to obtain the second turn frame number corresponding to each turn, and judging whether the second turn frame number satisfies a preset second turn threshold condition; if the second turn frame number satisfies the preset second turn threshold condition, determining the second turn driving data of the corresponding turn as the second turn driving data corresponding to a second turn to obtain the target turn driving data.

[0095] In the embodiment of the present invention, since being at an intersection is a prerequisite for turning, multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data are extracted through intersection range data screening, and whether the main vehicle turns when passing through the intersection is determined by obtaining multiple frames of first turning driving data corresponding to at least one turn in the multiple frames of intersection driving data corresponding to each intersection through main vehicle turn recognition, and finally, through the obstacle information around the main vehicle, it is determined whether the main vehicle and the vehicle obstacle make a second turn, thereby obtaining the target turning driving data. The present invention can accurately and efficiently collect turning driving data.

[0096] See also Figure 2 Another embodiment of the method for collecting turning driving data in the embodiment of the present invention includes:

[0097] 201. Acquire multiple frames of driving data of the main vehicle, and filter the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, wherein each frame of driving data includes obstacle information around the main vehicle;

[0098] Specifically, step 201 includes: obtaining multiple frames of driving data of the main vehicle, each frame of driving data includes intersection range identification information, and the intersection range identification information is used to indicate whether the main vehicle is in the intersection range in the corresponding frame of driving data; filtering the main vehicle intersection range data of each frame of driving data through the intersection range identification information, and obtaining multiple frames of intersection driving data corresponding to at least one intersection. In this embodiment, in order to quickly filter the intersection range data of multiple frames of driving data and efficiently filter out the driving data of the main vehicle in the intersection range, the intersection range identification information in each frame of driving data is obtained, wherein the intersection range identification information is used to indicate whether the main vehicle is in the intersection range in the corresponding frame of driving data, for example, if the intersection range identification information is 1, it means that the main vehicle is in the intersection range in the frame of driving data, and if the intersection range identification information is 0, it means that the main vehicle is not in the intersection range in the frame of driving data, which is not specifically limited here. Through the intersection range identification information, the intersection driving data in the multiple frames of driving data is extracted, and the multiple frames of intersection driving data corresponding to at least one intersection are obtained.

[0099] Further, each frame of driving data is filtered for the main vehicle intersection range data through the intersection range identification information to obtain multiple frames of intersection driving data corresponding to at least one intersection, including: setting any frame in the multiple frames of driving data as the current frame; using the intersection range identification information to sequentially determine the main vehicle intersection range for the current frame and the next frame of the current frame until all frames complete the main vehicle intersection range determination, to obtain a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range; through the first target frame and the second target frame, determine the multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data. In this embodiment, by setting any frame in the multiple frames of driving data as the current frame, the main vehicle intersection range is determined for each frame in the multiple frames of driving data, to obtain a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range, and then according to the first target frame and the second target frame, the multiple frames of driving data are divided into intersection data to obtain multiple frames of intersection driving data corresponding to at least one intersection. This implementation can quickly determine the intersection driving data of different intersections, thereby improving the collection efficiency of turning driving data.

[0100] 202. Perform host vehicle turn recognition on multiple frames of intersection driving data corresponding to each intersection, and obtain multiple frames of first turning driving data corresponding to at least one turn;

[0101] Specifically, step 202 includes: calculating the number of frames of the multi-frame intersection driving data corresponding to each intersection, obtaining the number of intersection frames corresponding to each intersection, and determining whether the number of intersection frames corresponding to each intersection is less than the preset intersection frame number threshold; if the number of intersection frames is less than the preset intersection frame number threshold, calculating the difference between the beginning and end orientations of the multi-frame intersection driving data corresponding to the intersection, and obtaining the difference in orientation of the main vehicle corresponding to each intersection. For example, in the multi-frame intersection driving data corresponding to a certain intersection, the orientation information when the main vehicle is located at the starting position of the intersection is 10 degrees, and the orientation information when the main vehicle is located at the end position of the intersection is 180 degrees. Then, the difference between the beginning and end orientations corresponding to the intersection is 180-10=170 degrees, that is, the difference in orientation of the main vehicle corresponding to the intersection is 170 degrees. Then, according to the difference in orientation of the main vehicle corresponding to each intersection, it is determined that the multi-frame intersection driving data corresponding to the intersection is the turning driving data of the main vehicle, and a multi-frame first turning driving data corresponding to at least one turn is obtained.

[0102] It should be noted that in order to exclude turning driving data in excessively congested conditions and improve the accuracy of turning driving data collection, the frame number of multiple frames of intersection driving data corresponding to each intersection is calculated to obtain the intersection frame number corresponding to each intersection and determine whether the intersection frame number corresponding to each intersection is less than the preset intersection frame number threshold. If the intersection frame number is less than the preset intersection frame number threshold, it means that when the main vehicle passes through the intersection, the intersection is not in a congested state and the intersection turn is completed within the preset intersection frame number threshold. Then, the difference in the start and end direction of the multiple frames of intersection driving data of the corresponding intersection is calculated to obtain the main vehicle direction difference corresponding to each intersection, and the multiple frames of intersection driving data of the corresponding intersection are determined to be the main vehicle turning driving data according to the main vehicle direction difference corresponding to each intersection, and multiple frames of first turning driving data corresponding to at least one turn are obtained. If the intersection frame number is greater than or equal to the preset intersection frame number threshold, it means that when the main vehicle passes through the intersection, the intersection is in a congested state, and the intersection turn is not completed within the preset intersection frame number threshold frame number, then the multi-frame intersection driving data corresponding to the intersection will not be included in the subsequent processing, thereby excluding the second turning driving data under congestion, so that the accuracy of turning driving data collection is improved.

[0103] In one embodiment, if the number of intersection frames is less than a preset intersection frame number threshold, the orientation information of the start frame and the end frame in the multi-frame intersection driving data corresponding to each intersection is used to calculate the difference in the beginning and end orientations of the multi-frame intersection driving data corresponding to each intersection, and the main vehicle orientation difference corresponding to each intersection is obtained, and it is determined whether the main vehicle orientation difference corresponding to each intersection is greater than the preset orientation difference threshold. If the main vehicle orientation difference is greater than the preset orientation difference threshold, the multi-frame intersection driving data of the corresponding intersection is used as the main vehicle turning driving data, and the multi-frame first turning driving data corresponding to at least one turn is obtained. This embodiment can quickly identify whether the main vehicle is turning at the intersection, thereby improving the efficiency of turning driving data collection.

[0104] Furthermore, the difference in the starting and ending orientations of multiple frames of intersection driving data at the corresponding intersection is calculated to obtain the orientation difference of the main vehicle corresponding to each intersection, and the multiple frames of intersection driving data at the corresponding intersection are determined to be the turning driving data of the main vehicle according to the orientation difference of the main vehicle corresponding to each intersection, and multiple frames of first turning driving data corresponding to at least one turn are obtained, including: respectively obtaining the orientation angles of the main vehicle at the starting position and the end position of the intersection in the multiple frames of intersection driving data to obtain the starting orientation angle and the ending orientation angle corresponding to each intersection; calculating the orientation difference of the main vehicle for the starting orientation angle and the ending orientation angle corresponding to each intersection to obtain the orientation difference of the main vehicle corresponding to each intersection, and judging whether the orientation difference of the main vehicle corresponding to each intersection is within a preset orientation difference range; if the orientation difference of the main vehicle is within the preset orientation difference range, determining that the multiple frames of intersection driving data at the corresponding intersection are multiple frames of turning driving data of the main vehicle, and obtaining multiple frames of first turning driving data corresponding to at least one turn. In this embodiment, the starting heading angle corresponding to each intersection is determined by the heading angle of the main vehicle at the starting position of the intersection in the multi-frame intersection driving data, and the ending heading angle corresponding to each intersection is determined by the heading angle of the main vehicle at the end position of the intersection in the multi-frame intersection driving data. Then, the main vehicle heading difference corresponding to each intersection is calculated through the starting heading angle and the ending heading angle corresponding to each intersection, and it is determined whether the main vehicle heading difference corresponding to each intersection is within a preset heading difference range; if the main vehicle heading difference is within the preset heading difference range, it means that the main vehicle is turning at the intersection, then the multi-frame intersection driving data of the corresponding intersection is determined to be a multi-frame main vehicle turning driving data, and a multi-frame first turning driving data corresponding to at least one turn is obtained. As an example but not limitation, assuming that the preset heading difference range is 60-190 degrees, if the starting heading angle corresponding to a certain intersection is 30 degrees and the ending heading angle is 200 degrees, then the main vehicle heading difference corresponding to the intersection is 170 degrees, which falls within the range of 60-190 degrees, then the multi-frame intersection driving data corresponding to the intersection is determined to be multi-frame first turn driving data.

[0105] 203. Obtaining the vehicle obstacle distance and the vehicle obstacle direction in the obstacle information around the main vehicle in the first turning driving data of each frame;

[0106] It can be understood that the obstacle information around the main vehicle in each frame of the first turn driving data includes the vehicle obstacle distance and the vehicle obstacle direction, wherein the vehicle obstacle distance is used to indicate the distance between the vehicle obstacle and the main vehicle, and the vehicle obstacle direction is used to indicate the direction information of the vehicle obstacle.

[0107] 204. Determine whether the obstacle information around the host vehicle in each frame of the first turning driving data corresponding to each turn satisfies a preset second turning determination condition based on the vehicle obstacle distance and the vehicle obstacle direction;

[0108] Specifically, step 204 includes: determining whether the vehicle obstacle distance corresponding to the first turning driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold; if the vehicle obstacle distance corresponding to the turning driving data is less than a preset following vehicle distance threshold, calculating the difference between the vehicle obstacle orientation in the turning driving data of the corresponding frame and the main vehicle orientation in the turning driving data of the corresponding frame to obtain a turning orientation difference; determining whether the turning orientation difference is less than a preset orientation difference threshold; if the turning orientation difference is less than the preset orientation difference threshold, determining that the obstacle information around the main vehicle in the turning driving data of the corresponding frame meets the preset second turning judgment condition. In one embodiment, the preset second turn determination condition is used to indicate a parallel turn determination condition. In order to accurately identify a vehicle obstacle that makes a parallel turn with the main vehicle, it is first determined whether the vehicle obstacle distance corresponding to the first turn driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold. If the vehicle obstacle distance corresponding to the turn driving data is less than a preset follower vehicle distance threshold, it means that the vehicle obstacle is relatively close to the main vehicle and there is a probability of parallel turn driving. Then, the difference between the vehicle obstacle orientation in the turn driving data of the corresponding frame and the main vehicle orientation in the turn driving data of the corresponding frame is calculated to obtain a turn orientation difference, and it is determined whether the turn orientation difference is less than a preset orientation difference threshold to determine whether the vehicle obstacle that is relatively close to the main vehicle also has a similar orientation to the main vehicle. If the turn orientation difference is less than the preset orientation difference threshold, it means that the vehicle obstacle also has a similar orientation to the main vehicle. Then, it is determined that the obstacle information around the main vehicle in the turn driving data of the corresponding frame meets the preset second turn determination condition, and it is determined that the vehicle obstacle in the turn driving data of the corresponding frame makes a parallel turn with the main vehicle.

[0109] 205. If the obstacle information around the main vehicle in the turning driving data satisfies the preset second turn judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, and the second turning driving data corresponding to at least one second turning process is obtained, and the target turning driving data is determined according to the total number of frames of the second turning driving data corresponding to each second turning process.

[0110] It should be noted that, since there may be turning driving data satisfying the preset second turning determination condition in the non-second turning driving data, in order to avoid the one-sidedness of determining as the second turning driving data only if there is turning driving data satisfying the preset second turning determination condition, after obtaining the second turning driving data corresponding to at least one second turning process, it is determined whether the second turning driving data corresponding to each second turning process satisfies the preset second turning threshold condition according to the total number of frames of the second turning driving data corresponding to each second turning process, and the target turning driving data is determined according to the determination result. For example, the total number of frames of the second turning driving data needs to be greater than 50 frames to satisfy the preset second turning threshold condition.

[0111] Specifically, step 205 includes: if the obstacle information around the main vehicle in the turning driving data meets the preset second turn judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, and the second turning driving data corresponding to at least one second turning process is obtained; the total number of frames of the second turning driving data corresponding to each second turning process is calculated to obtain the second turning frame number corresponding to each second turning process; whether the second turning frame number corresponding to each second turning process meets the preset second turning threshold condition; if the second turning frame number meets the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data. In one embodiment, after obtaining the second turning driving data corresponding to at least one second turning process, the total number of frames of the second turning driving data corresponding to each second turning process is calculated to obtain the second turning frame number corresponding to each second turning process, and the ratio of the second turning frame number corresponding to each second turning process to the total number of frames is calculated to obtain the second turning ratio corresponding to each second turning process, and then it is determined whether the second turning frame number and the second turning ratio corresponding to each second turning process meet the preset second turning threshold condition. If the second turning frame number and the second turning ratio meet the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data. For example, when the second turning frame number is greater than 50 and the second turning ratio is greater than 70%, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data.

[0112] Furthermore, based on the obstacle information around the main vehicle in each frame of the first turning driving data, the second turning discrimination is performed on the multiple frames of the first turning driving data corresponding to each turn, and after obtaining the target turning driving data, the method further includes: sorting the target turning driving data corresponding to each second turning process according to the second turning frame number, and obtaining a second turning driving data sequence. Since the second turning frame number and the second turning ratio can indicate the scene relevance of the corresponding target turning driving data, that is, the more the second turning driving frames are, the more consistent it is with the second turning driving scene. Therefore, in order to obtain the second turning driving data with different scene relevance, the scene relevance of the target turning driving data corresponding to each second turning process is sorted in descending order according to the second turning frame number or the second turning ratio, and the second turning driving data sequence is obtained, thereby satisfying the acquisition of data with different scene relevance.

[0113] In the embodiment of the present invention, since being at an intersection is a prerequisite for turning, multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data are extracted through intersection range data screening, and whether the main vehicle turns when passing through the intersection is determined by obtaining multiple frames of first turning driving data corresponding to at least one turn in the multiple frames of intersection driving data corresponding to each intersection through main vehicle turn recognition, and finally, through the vehicle obstacle distance and vehicle obstacle direction in the obstacle information around the main vehicle, it is determined whether the main vehicle and the vehicle obstacle make a second turn, thereby obtaining target turning driving data. The present invention can accurately and efficiently collect turning driving data.

[0114] The above describes the method for collecting turning driving data in the embodiment of the present invention. The following describes the device for collecting turning driving data in the embodiment of the present invention. Figure 3 , an embodiment of a device for collecting turning driving data in an embodiment of the present invention includes:

[0115] The screening module 301 is used to obtain multiple frames of driving data of the main vehicle, and screen the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle;

[0116] The identification module 302 is used to identify the main vehicle turn for the multiple frames of intersection driving data corresponding to each intersection, and obtain multiple frames of first turning driving data corresponding to at least one turn;

[0117] The discrimination module 303 is used to perform second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the host vehicle in each frame of the first turn driving data to obtain target turn driving data.

[0118] In the embodiment of the present invention, since being at an intersection is a prerequisite for turning, multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data are extracted through intersection range data screening, and whether the main vehicle turns when passing through the intersection is determined by obtaining multiple frames of first turning driving data corresponding to at least one turn in the multiple frames of intersection driving data corresponding to each intersection through main vehicle turn recognition, and finally, through the obstacle information around the main vehicle, it is determined whether the main vehicle and the vehicle obstacle make a second turn, thereby obtaining the target turning driving data. The present invention can accurately and efficiently collect turning driving data.

[0119] See also Figure 4 Another embodiment of the device for collecting turning driving data in the embodiment of the present invention includes:

[0120] The screening module 301 is used to obtain multiple frames of driving data of the main vehicle, and screen the multiple frames of driving data for intersection range data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle;

[0121] The identification module 302 is used to identify the main vehicle turn for the multiple frames of intersection driving data corresponding to each intersection, and obtain multiple frames of first turning driving data corresponding to at least one turn;

[0122] The discrimination module 303 is used to perform second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the host vehicle in each frame of the first turn driving data to obtain target turn driving data.

[0123] Optionally, the screening module 301 includes:

[0124] An acquisition unit 3011 is used to acquire multiple frames of driving data of the main vehicle, each frame of driving data includes intersection range identification information, and the intersection range identification information is used to indicate whether the main vehicle is in the intersection range in the corresponding frame of driving data;

[0125] The screening unit 3012 is used to screen the intersection range data of the host vehicle for each frame of driving data according to the intersection range identification information, and obtain multiple frames of intersection driving data corresponding to at least one intersection.

[0126] Optionally, the screening unit 3012 is specifically used for:

[0127] Setting any one frame of the multiple frames of driving data as the current frame;

[0128] The intersection range identification information is used to sequentially identify the intersection range of the main vehicle for the current frame and the next frame of the current frame until all frames complete the intersection range identification of the main vehicle, thereby obtaining a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range;

[0129] A plurality of frames of intersection driving data corresponding to at least one intersection in the plurality of frames of driving data are determined through the first target frame and the second target frame.

[0130] Optionally, the identification module 302 includes:

[0131] The calculation unit 3021 is used to calculate the frame number of multiple frames of intersection driving data corresponding to each intersection, obtain the intersection frame number corresponding to each intersection, and determine whether the intersection frame number corresponding to each intersection is less than a preset intersection frame number threshold;

[0132] The determination unit 3022 is used to calculate the difference in the initial and final directions of the multi-frame intersection driving data of the corresponding intersection if the intersection frame number is less than a preset intersection frame number threshold, obtain the main vehicle direction difference corresponding to each intersection, and determine the multi-frame intersection driving data of the corresponding intersection as the main vehicle turning driving data according to the main vehicle direction difference corresponding to each intersection, and obtain multi-frame first turning driving data corresponding to at least one turn.

[0133] Optionally, the determining unit 3022 is specifically configured to:

[0134] Obtain the heading angles of the main vehicle at the starting position and the ending position of the intersection in multiple frames of intersection driving data, respectively, to obtain the starting heading angle and the ending heading angle corresponding to each intersection;

[0135] Calculate the orientation difference of the main vehicle for the starting orientation angle and the ending orientation angle corresponding to each intersection, obtain the orientation difference of the main vehicle corresponding to each intersection, and determine whether the orientation difference of the main vehicle corresponding to each intersection is within the preset orientation difference range;

[0136] If the main vehicle's orientation difference is within a preset orientation difference range, the multi-frame intersection driving data of the corresponding intersection is determined to be a multi-frame main vehicle turning driving data, and a multi-frame first turning driving data corresponding to at least one turn is obtained.

[0137] Optionally, the identification module 303 includes:

[0138] The extraction unit 3031 is used to obtain the vehicle obstacle distance and the vehicle obstacle direction in the obstacle information around the main vehicle in the first turning driving data of each frame;

[0139] The judging unit 3032 is used to judge whether the obstacle information around the host vehicle in each frame of the first turning driving data corresponding to each turn meets the preset second turning judgment condition according to the vehicle obstacle distance and the vehicle obstacle direction;

[0140] The determination unit 3033 is used to determine the corresponding frame turning driving data as the second turning driving data of the corresponding turn if the obstacle information around the main vehicle in the turning driving data meets the preset second turning determination condition, obtain the second turning driving data corresponding to at least one second turning process, and determine the target turning driving data according to the total number of frames of the second turning driving data corresponding to each second turning process.

[0141] Optionally, the judging unit 3032 is specifically configured to:

[0142] Determine whether the vehicle obstacle distance corresponding to the first turning driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold;

[0143] If the vehicle obstacle distance corresponding to the turning driving data is less than the preset following vehicle distance threshold, the difference between the vehicle obstacle orientation in the turning driving data of the corresponding frame and the main vehicle orientation in the turning driving data of the corresponding frame is calculated to obtain the turning orientation difference;

[0144] Determining whether the turning direction difference is less than a preset direction difference threshold;

[0145] If the turning direction difference is less than a preset direction difference threshold, it is determined that the obstacle information around the host vehicle in the turning driving data of the corresponding frame meets the preset second turning judgment condition.

[0146] Optionally, the determining unit 3033 is specifically used for:

[0147] If the obstacle information around the host vehicle in the turning driving data satisfies the preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data corresponding to the turn, and the second turning driving data corresponding to at least one second turning process is obtained;

[0148] Calculate the total number of frames of the second turning driving data corresponding to each second turning process to obtain the number of second turning frames corresponding to each second turning process;

[0149] Determining whether the second turning frame number corresponding to each second turning process meets a preset second turning threshold condition;

[0150] If the second turning frame number meets the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data.

[0151] Optionally, the device for collecting turning driving data further includes:

[0152] The sorting module 304 is used to sort the target turning driving data corresponding to each second turning process according to the second turning frame number, so as to obtain a second turning driving data sequence.

[0153] In the embodiment of the present invention, since being at an intersection is a prerequisite for turning, multiple frames of intersection driving data corresponding to at least one intersection in the multiple frames of driving data are extracted through intersection range data screening, and whether the main vehicle turns when passing through the intersection is determined by obtaining multiple frames of first turning driving data corresponding to at least one turn in the multiple frames of intersection driving data corresponding to each intersection through main vehicle turn recognition, and finally, through the vehicle obstacle distance and vehicle obstacle direction in the obstacle information around the main vehicle, it is determined whether the main vehicle and the vehicle obstacle make a second turn, thereby obtaining target turning driving data. The present invention can accurately and efficiently collect turning driving data.

[0154] above Figure 3 and Figure 4 The device for collecting turning driving data in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The device for collecting turning driving data in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0155] Figure 5 1 is a schematic diagram of the structure of a device for collecting turning driving data provided by an embodiment of the present invention. The device 500 for collecting turning driving data may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of computer program operations in the device 500 for collecting turning driving data. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of computer program operations in the storage medium 530 on the device 500 for collecting turning driving data.

[0156] The turning driving data collection device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 5The structure of the device for collecting turning driving data shown does not constitute a limitation on the device for collecting turning driving data, and may include more or less components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0157] The present invention also provides a computer device, which includes a memory and a processor. The memory stores a computer-readable computer program. When the computer-readable computer program is executed by the processor, the processor executes the steps of the method for collecting turning driving data in the above-mentioned embodiments.

[0158] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. A computer program is stored in the computer-readable storage medium. When the computer program runs on a computer, the computer executes the steps of the method for collecting turning driving data.

[0159] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0160] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several computer programs to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0161] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for collecting turning driving data, characterized in that: The method for collecting turning driving data comprises: Acquire multiple frames of driving data of the host vehicle of a specified data volume, and perform intersection range data screening on the multiple frames of driving data to obtain multiple frames of intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the host vehicle; Performing host vehicle turn identification on multiple frames of smooth driving data corresponding to each intersection, and obtaining multiple frames of first turn driving data corresponding to at least one turn; wherein the multiple frames of smooth driving data are multiple frames of intersection driving data corresponding to intersections whose driving time is less than a preset intersection driving time threshold, and the host vehicle turn is identified by the difference in the orientation angle between the host vehicle at the starting position and the end position of the intersection; Based on the obstacle information around the main vehicle in each frame of the first turn driving data, a second turn discrimination is performed on multiple frames of the first turn driving data corresponding to each turn to obtain target turn driving data, wherein the second turn is a turn in which the main vehicle maintains a specific turning relationship with the surrounding vehicle obstacles, and the specific turning relationship includes a parallel turning relationship or a following vehicle turning relationship maintained between the main vehicle and the surrounding obstacle vehicles.

2. The method for collecting turning driving data according to claim 1, characterized in that: The method of acquiring a plurality of frames of driving data of the host vehicle of a specified data volume and filtering the plurality of frames of driving data for intersection range data to obtain a plurality of frames of intersection driving data corresponding to at least one intersection respectively includes: Acquire multiple frames of driving data of a host vehicle of a specified data volume, each frame of driving data including intersection range identification information, wherein the intersection range identification information is used to indicate whether the host vehicle in the corresponding frame of driving data is within the intersection range; The intersection range data of the host vehicle is filtered for each frame of driving data using the intersection range identification information to obtain multiple frames of intersection driving data corresponding to at least one intersection.

3. The method for collecting turning driving data according to claim 2, characterized in that: The method of filtering the main vehicle intersection range data for each frame of driving data by using the intersection range identification information to obtain multiple frames of intersection driving data corresponding to at least one intersection, includes: Setting any one frame of the multiple frames of driving data as the current frame; The intersection range identification information is used to sequentially identify the intersection range of the main vehicle for the current frame and the next frame of the current frame until all frames complete the intersection range identification of the main vehicle, thereby obtaining a first target frame in which the main vehicle is within the intersection range and a second target frame in which the main vehicle is not within the intersection range; A plurality of frames of intersection driving data corresponding to at least one intersection in the plurality of frames of driving data are determined through the first target frame and the second target frame.

4. The method for collecting turning driving data according to claim 1, characterized in that: The method of performing host vehicle turn recognition on multiple frames of intersection driving data corresponding to each intersection to obtain multiple frames of first turning driving data corresponding to at least one turn includes: Calculate the number of frames of intersection driving data corresponding to each intersection to obtain the number of intersection frames corresponding to each intersection, and determine whether the number of intersection frames corresponding to each intersection is less than a preset intersection frame number threshold; If the intersection frame number is less than the preset intersection frame number threshold, the difference in the initial and final directions of the multi-frame intersection driving data of the corresponding intersection is calculated to obtain the main vehicle direction difference corresponding to each intersection, and the multi-frame intersection driving data of the corresponding intersection is determined as the main vehicle turning driving data according to the main vehicle direction difference corresponding to each intersection, and multiple frames of first turning driving data corresponding to at least one turn are obtained.

5. The method for collecting turning driving data according to claim 4, characterized in that: The method of calculating the difference between the start and end directions of multiple frames of intersection driving data corresponding to the intersection, obtaining the difference in the direction of the main vehicle corresponding to each intersection, and determining that the multiple frames of intersection driving data corresponding to the intersection are the turning driving data of the main vehicle according to the difference in the direction of the main vehicle corresponding to each intersection, and obtaining multiple frames of first turning driving data corresponding to at least one turn, respectively, includes: Obtain the heading angles of the main vehicle at the starting position and the ending position of the intersection in multiple frames of intersection driving data, respectively, to obtain the starting heading angle and the ending heading angle corresponding to each intersection; Calculate the orientation difference of the main vehicle for the starting orientation angle and the ending orientation angle corresponding to each intersection, obtain the orientation difference of the main vehicle corresponding to each intersection, and determine whether the orientation difference of the main vehicle corresponding to each intersection is within the preset orientation difference range; If the main vehicle's orientation difference is within a preset orientation difference range, the multi-frame intersection driving data of the corresponding intersection is determined to be a multi-frame main vehicle turning driving data, and a multi-frame first turning driving data corresponding to at least one turn is obtained.

6. The method for collecting turning driving data according to claim 1, characterized in that: The method of performing second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the main vehicle in each frame of first turn driving data to obtain target turn driving data includes: Obtaining the vehicle obstacle distance and the vehicle obstacle direction in the obstacle information around the main vehicle in the first turning driving data of each frame; By using the vehicle obstacle distance and the vehicle obstacle direction, determining whether the obstacle information around the main vehicle in each frame of the first turning driving data corresponding to each turn satisfies the preset second turning judgment condition; If the obstacle information around the main vehicle in the turning driving data meets the preset second turn judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, and the second turning driving data corresponding to at least one second turning process is obtained, and the target turning driving data is determined according to the total number of frames of the second turning driving data corresponding to each second turning process.

7. The method for collecting turning driving data according to claim 6, characterized in that: The determining, by using the vehicle obstacle distance and the vehicle obstacle orientation, whether the obstacle information around the host vehicle in each frame of the first turning driving data corresponding to each turn satisfies a preset second turning determination condition includes: Determine whether the vehicle obstacle distance corresponding to the first turning driving data of each frame corresponding to each turn is less than a preset vehicle obstacle distance threshold; If the vehicle obstacle distance corresponding to the turning driving data is less than the preset following vehicle distance threshold, the difference between the vehicle obstacle orientation in the turning driving data of the corresponding frame and the main vehicle orientation in the turning driving data of the corresponding frame is calculated to obtain the turning orientation difference; Determining whether the turning direction difference is less than a preset direction difference threshold; If the turning direction difference is less than a preset direction difference threshold, it is determined that the obstacle information around the host vehicle in the turning driving data of the corresponding frame meets the preset second turning judgment condition.

8. The method for collecting turning driving data according to claim 6, characterized in that: If the obstacle information around the host vehicle in the turning driving data satisfies the preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data of the corresponding turn, the second turning driving data corresponding to at least one second turning process is obtained, and the target turning driving data is determined according to the total number of frames of the second turning driving data corresponding to each second turning process, including: If the obstacle information around the host vehicle in the turning driving data satisfies the preset second turning judgment condition, the corresponding frame turning driving data is determined as the second turning driving data corresponding to the turn, and the second turning driving data corresponding to at least one second turning process is obtained; Calculate the total number of frames of the second turning driving data corresponding to each second turning process to obtain the number of second turning frames corresponding to each second turning process; Determining whether the second turning frame number corresponding to each second turning process meets a preset second turning threshold condition; If the second turning frame number meets the preset second turning threshold condition, the multiple frames of first turning driving data corresponding to the second turning process are determined as the target turning driving data.

9. The method for collecting turning driving data according to claim 8, characterized in that: After performing second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the main vehicle in each frame of first turn driving data to obtain target turn driving data, the method for collecting turn driving data further includes: According to the second turning frame number, the target turning driving data corresponding to each second turning process are sorted by scene relevance to obtain a second turning driving data sequence.

10. A device for collecting turning driving data, characterized in that: The device for collecting turning driving data comprises: A screening module is used to obtain a specified amount of multi-frame driving data of the main vehicle, and to screen the multi-frame driving data for intersection range data to obtain a multi-frame intersection driving data corresponding to at least one intersection, each frame of driving data including obstacle information around the main vehicle; An identification module is used to identify a turn of a main vehicle for a plurality of frames of smooth driving data corresponding to each intersection, and obtain a plurality of frames of first turning driving data corresponding to at least one turn; wherein the plurality of frames of smooth driving data are a plurality of frames of intersection driving data corresponding to an intersection whose driving time is less than a preset intersection driving time threshold, and the turning of the main vehicle is identified by the difference in the orientation angle between the main vehicle at the starting position and the end position of the intersection; A discrimination module is used to perform a second turn discrimination on multiple frames of first turn driving data corresponding to each turn based on the obstacle information around the main vehicle in each frame of the first turn driving data to obtain target turn driving data, wherein the second turn is a turn in which the main vehicle maintains a specific turning relationship with the surrounding vehicle obstacles, and the specific turning relationship includes a parallel turning relationship or a following turning relationship maintained between the main vehicle and the surrounding obstacle vehicles.

11. A device for collecting turning driving data, characterized in that: The device for collecting turning driving data includes: a memory and at least one processor, wherein a computer program is stored in the memory; The at least one processor calls the computer program in the memory to enable the turning driving data collection device to execute the turning driving data collection method as described in any one of claims 1-9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for collecting turning driving data as described in any one of claims 1 to 9 is implemented.

Images