CN114067563A

CN114067563A - Intersection identification method and corresponding storage medium, product, model and reminding method and equipment

Info

Publication number: CN114067563A
Application number: CN202111313716.3A
Authority: CN
Inventors: 崔俊涛
Original assignee: Shanghai Wanwei Technology Co ltd
Current assignee: Shanghai Wanwei Technology Co ltd
Priority date: 2021-11-08
Filing date: 2021-11-08
Publication date: 2022-02-18
Anticipated expiration: 2041-11-08
Also published as: CN114067563B

Abstract

The invention discloses an intersection identification method, a computer readable storage medium, a computer program product, an intersection data model, an intersection reminding method and an intersection reminding device, which mainly comprise the following steps: obtaining trajectory data of a vehicle within a target area; analyzing the track data to obtain position data sets { Li } and { Ri } of left-turn and right-turn actions; calculating intersection center position sets { LOi }, { ROi } of the position data sets { Li }, { Ri }, and constructing a data set of the determined/to-be-determined intersection, wherein one intersection center position LOi or ROi corresponds to data of the determined intersection or the to-be-determined intersection. Compared with the prior art, the method has the advantages that the intersection position is obtained by analyzing the track data of the vehicle, so that the cost is reduced; the left-turn analysis result and the right-turn analysis result are distinguished, and the left-turn analysis result is used as the determined intersection, so that the accuracy is improved.

Description

Intersection identification method and corresponding storage medium, product, model and reminding method and equipment

Technical Field

The invention belongs to the field of intelligent traffic, and particularly relates to an intersection identification method, a corresponding storage medium, a product, a model, a reminding method and corresponding equipment.

Background

Intersection identification is one of the important tasks in constructing map data. For common public transport roads, particularly municipal roads in cities, a map construction method adopted by common commercial map service providers can realize intersection identification. However, for some special areas, such as a park, a factory, and the like, it is relatively expensive to establish intersection data by using a conventional map construction method.

In the prior art, one mode is to install a camera on a vehicle and identify an intersection in an image/video identification mode, but the mode needs complex training, occupies more computing resources, and is relatively high in cost.

Therefore, there is a need for a low-cost intersection identification method to obtain intersection information in a certain area.

Disclosure of Invention

In order to reduce the cost of constructing intersection identification, the invention provides an intersection identification method, a computer readable storage medium, a computer program product, an intersection data model, an intersection reminding method and intersection reminding equipment.

The technical scheme adopted by the invention is as follows:

in one aspect, a method for identifying an intersection is provided, which includes the following steps:

obtaining trajectory data of a vehicle within a target area;

analyzing the track data to obtain one or more position data sets { Li | i ═ 1,2,3 … } indicating that the driving direction has left-turn action, or/and to obtain one or more position data sets { Ri | i ═ 1,2,3 … } indicating that the driving direction has right-turn action;

calculating an intersection center position set { LOi | i ═ 1,2,3, … } of the position data set { Li | i ═ 1,2,3 … } and constructing a data set for determining an intersection, or/and calculating an intersection center position set { ROi | i ═ 1,2,3, … } of the position data set { Ri | i ═ 1,2,3 … } and constructing a data set for an intersection to be determined; wherein:

li denotes a location data set, which comprises several items of data, for example: the obtained result has three position data sets which are expressed as { L1, L2, L3 };

ri represents a position data set comprising several items of data, for example: the obtained result has three position data sets which are expressed as { R1, R2, R3 };

one intersection center position LOi corresponds to data of one determined intersection, and one intersection center position ROi corresponds to data of one intersection to be determined.

Further, the method also comprises the steps of identifying and updating the newly added intersection, and specifically comprises the following steps:

updating trajectory data of the vehicle within the target area;

analyzing the updated track data to obtain one or more position data sets { Lxi | i ═ 1,2,3 … } indicating that the driving direction has left-turning action, or/and obtaining one or more position data sets { Rxi | i ═ 1,2,3 … } indicating that the driving direction has right-turning action;

calculating an intersection center position set { LxOi | i ═ 1,2,3, … } of the position data set { Lxi | i ═ 1,2,3, 3626 }, or/and calculating an intersection center position set { RxOi | i ═ 1,2,3, … } of the position data set { Rxi ═ 1,2,3 … };

judging whether each intersection center position LxOi exists in a data set of a determined intersection or not, if so, judging that the intersection center position LxOi is not a newly added intersection, if not, judging that the intersection center position LxOi is the newly added intersection and is expressed as an intersection center position LOi, and adding the intersection center position LOi into the data set of the determined intersection; and/or the first and/or second light sources,

judging whether each intersection center position LxOi exists in a data set of an intersection to be determined, if so, judging that the intersection center position ROi of the same intersection described by the intersection center position LxOi exists in the data set of the intersection to be determined, removing the intersection center position ROi from the data set of the intersection to be determined, representing the intersection center position LxOi as an intersection center position LOi, and adding the intersection center position LOi into the data set of the determined intersection; and/or the first and/or second light sources,

judging whether each intersection center position RxOi exists in a data set of the intersection to be determined, if so, judging that the intersection center position RxOi is not a newly added intersection, if not, judging that the intersection center position RxOi is the newly added intersection and is expressed as an intersection center position ROi, and adding the intersection center position ROi into the data set of the intersection to be determined; wherein:

lx, Rx have the same physical meaning as L, R, and their definition may correspond to L, R.

Further, a first intersection distance threshold value for judging the position repetition is defined, and the method further comprises the following steps:

judging whether the distance between each intersection center position ROi in the data set of the intersection to be determined and each intersection center position LOi in the data set of the intersection to be determined is smaller than the first intersection distance threshold value or not, and if so, moving out the intersection center position ROi from the data set of the intersection to be determined.

In another aspect, a computer readable storage medium is provided, on which a computer program/instructions are stored, which when executed by a processor implement the steps of the above intersection identification method.

In another aspect, a computer program product is provided, comprising computer programs/instructions which, when executed by a processor, implement the steps of the above intersection identification method.

In another aspect, an intersection data model is provided, which is constructed according to the intersection-determining data set and/or the intersection-to-be-determined data set obtained by the intersection identification method.

In another aspect, an intersection reminding method is provided, defining an intersection area, and the intersection reminding method includes the steps of:

obtaining a vehicle position;

and judging whether the vehicle position is in the intersection area or not by means of the intersection data model, and sending out reminding information if the vehicle enters the intersection area.

In another aspect, an electronic device is provided, which includes a memory, a processor and a computer program stored in the memory, and is configured to receive the reminding information of the intersection reminding method and send out a reminder to a driver by voice, vibration or the like, or the processor executes the computer program to implement the steps of the intersection reminding method.

Compared with the prior art, the invention has the beneficial effects that: the intersection position is obtained by analyzing the track data of the vehicle, the cost of intersection identification is reduced, and the track data of the vehicle has the attribute which is easy to obtain. In addition, the left-turn analysis result and the right-turn analysis result are distinguished, the left-turn analysis result is used as the determined intersection, the accuracy is high, meanwhile, intersection data to be determined can be used temporarily under the condition that no left-turn track exists, and the intersection data can be used under the condition that the relative accuracy requirement is low, such as reminding and the like. Aiming at the newly added vehicle track data, the newly added updating of the intersection can be realized. In addition to the above advantages, other advantages of the present invention can be seen in the details of the examples.

Drawings

Fig. 1 is a flowchart of an intersection identification method according to an embodiment.

FIG. 2 is a block diagram of an embodiment of a sliding window algorithm for analyzing a window analysis set to obtain a commutation data set.

FIG. 3 is a diagram illustrating a sliding step of the sliding window shifting algorithm in the embodiment.

FIG. 4 is a diagram illustrating a section of turning trajectory in the embodiment.

Fig. 5 is a flowchart of an intersection identification updating method according to an embodiment.

Fig. 6 is a flowchart illustrating the steps of determining and processing the intersection center position LxOi in the embodiment.

Fig. 7 is a flowchart illustrating the steps of determining and processing the intersection center position RxOi in the embodiment.

Fig. 8 is a schematic diagram of an intersection and two driving tracks at the intersection in the embodiment.

Fig. 9 is a schematic diagram of a t-junction and two driving tracks at the junction in the embodiment.

Fig. 10 is a schematic diagram of an irregular intersection and two driving tracks at the intersection in the embodiment.

Fig. 11 is a schematic device structure diagram of an electronic apparatus according to an embodiment.

Detailed Description

The invention is further described below with reference to the accompanying drawings and specific embodiments.

In the present invention, the sequence numbers and serial numbers of the steps are easy-to-read labels, and these labels should not be understood as constraints on the sequence or logic sequence, and the sequence of the steps can be properly adjusted under the condition of realizing the corresponding functions or purposes.

Example 1

Referring to fig. 1, the intersection identification method of the present embodiment includes the following steps:

s10, acquiring data: trajectory data of the vehicle within the target area is obtained.

The target area is an area of intersection information to be identified, and may be a relatively independent area such as a factory area, or a municipal public area, which is not limited in the present invention.

The trajectory data is a set of data including a plurality of types, which are generated when the vehicle travels within the target area. The trajectory data specifically includes position data of the vehicle, a speed of the vehicle, and a traveling direction of the vehicle, and may be obtained by using vehicle-mounted equipment such as a GPS and a compass that are positioned by a satellite.

In the present invention, when processing and analyzing the traveling direction are described, the traveling direction can be understood as data corresponding thereto. The direction of travel can also be understood as the direction of speed of the vehicle.

As a preferred embodiment, the data is typically collected periodically at certain intervals, for example, one data per 10 s. Thus, for a vehicle's travel track, a time-including data set can be constructed based on the time axis. In the present embodiment, as an example, the data set describing the travel locus is expressed as follows:

{ { Tim [ i ], Loc [ i ], Dir [ i ], Vol [ i ] } | i ═ 1,2,3, … }, in which:

tim represents a time scale, Loc represents the position of the vehicle, Dir represents the traveling direction of the vehicle, and Vol represents the speed of the vehicle;

{ Tim [ i ], Loc [ i ], Dir [ i ], Vol [ i ] } denotes a piece of data at Tim [ i ].

And S20, analyzing the track data.

In the invention, the data mainly processed by intersection identification are position data and driving direction. An upper speed threshold is defined, for example 60 yards. Preferably, the speed data is mainly used as a characteristic quantity of data validity, and the position data and the driving direction when the speed exceeds the upper limit threshold are regarded as invalid data, and the invalid data in the trajectory data is removed.

For a driving trajectory, there may be several transition diversions, so that it is first necessary to find all diversions data sets in the driving trajectory, wherein a diversions data set is a data describing the vehicle from the beginning to the end of the diversion process.

In this embodiment, as a manner, the step of obtaining the commutation data set by using the sliding window algorithm specifically includes:

referring to fig. 2, the size of the sliding window is defined as N, and the minimum step size of the window sliding is M +1 before the current commutation data set is found, and obviously N, M is a natural number. Taking the above 10s as the acquisition period, the value of N is preferably 30-40, and at this time, the window may cover the turning trajectory of one intersection, and in this embodiment, N is preferably 35. N, M can be adjusted as appropriate according to other information such as the acquisition period. When efficiency and accuracy are both considered, M is preferably 5.

S211, selecting 35 pieces of data corresponding to Tim [ i ] -Tim [ i +35] from the data set { { Tim [ i ], Loc [ i ], Dir [ i ], Vol [ i ] } | i ═ 1,2,3, … } as a window analysis set, referring to table 1. The Dir [ i ] is expressed in terms of an angle.

TABLE 1 analysis set of windows

In this step: initially, i is typically 1; the angular difference (Tim [ i + k +1] -Tim [ i + k ]) represents the angular difference between two pieces of data adjacent to each other in the time axis.

In addition, in the present invention, when the description is made with respect to data, the time tag Tim [ i ] is regarded as a primary key, which can be understood as referring to 1 piece of data corresponding thereto, corresponding to one row of data in reference table 1.

And defining a first angle difference threshold value for judging that the steering occurring in the track belongs to the steering passing through the intersection or the small-angle steering belonging to other situations. In the present embodiment, the first angle difference threshold is set to 10 °.

S212, on the time axis, judging whether the angle difference between two adjacent data Tim [ i ] and Tim [ i +1] exceeds the first angle threshold one by one, and when two adjacent data with the angle difference exceeding the first angle threshold are found, regarding as the starting point of the vehicle steering.

Referring to the data in Table 1, Tim [ i ] -Tim [ i +3] is a small change in the angle of the trajectory, and it is not possible to determine whether the vehicle is turning in a general sense, and possible situations include, but are not limited to, lane changing, or simply adjusting the vehicle position. And when Tim [ i +4] -Tim [ i +5], the angle change of the track is 11 degrees, and the first angle difference threshold value is exceeded, and the Tim [ i +4] can be taken as a starting point ST 1.

It is easily understood that in the above steps, Tim [ i ] is selected as the starting point, but the present invention is not limited to Tim [ i ] and Tim [ i +1], Tim [ i ] being selected as the starting point, or Tim [ i +1] being selected as the starting point, where the angle difference exceeds the first angle threshold.

If the data corresponding to the Tim [ I +25] is analyzed as the starting point, if (I +25) - (I) ═ 25> M, the current window analysis set does not cover a complete intersection turning track, at this time, the data corresponding to the Tim [ I +25] -Tim [ I +59] is selected from the Tim [ I +25], and the window analysis set is updated, that is, in the data set of the driving track, the step length of the window analysis set sliding along the time axis is 25.

Referring to fig. 3, if the data corresponding to Tim [ i + k ] is analyzed as a starting point, (i + k) - (i) ═ k > M, the step size of the window analysis set sliding along the time axis is k. For example, if the data set of the driving track includes 200 data, none of the first 5 data Tim [1] -Tim [5] finds the starting point, the value of j in FIG. 3 is 0, and the starting point ST is found at Tim [8], then newly selecting the data corresponding to Tim [8] -Tim [42] from Tim [8] as the window analysis set.

It should be understood that the step size of the window sliding is uncertain, and the parameter M is to confirm that the data in the current window is effectively covered, so as to avoid that when the 32 th and 33 th data in the current window are found, for example, only the 34 th and 35 th data in the current window cannot cover a complete turn track. And sliding the window once when the initial point is found by exceeding M pieces of data from the 1 st piece of data in the window, wherein the step length is the data amount from the 1 st piece of data to the initial point, so that the track is ensured to be continuously analyzed, and meanwhile, when the initial point is found in the window, the initial point is positioned in the front M pieces of data in the window, and N-M pieces of data are arranged behind the initial point and are enough to cover a section of complete turning track.

Alternatively, the distance between the starting point and the item 1 data in the current window is obtained as N1, and the size of the current window is temporarily expanded backwards by N1. For example, the value of j in FIG. 3 takes 0, and the start point ST is found at Tim [8 ]. The distance N1 between the starting point ST and the item 1 ST data in the current window is 8, and it should be understood that the distance referred to herein is the number of items 1 to the starting point ST, not the actual length distance. Thus, the original 27 items of data after the starting point ST in the window temporarily expand the size of the current window backwards by 8 items, so that 35 items of data still exist after the starting point ST, and the data can cover a section of complete turning track.

As described above, after the start point ST1 is found in 5 pieces of data from Tim [ i ]; or, the starting point ST1 is found by more than 5 pieces of data, a certain step length is slid backwards, and the window analysis set is updated; step S213 is performed.

And S213, in the current window analysis set, comparing the current window analysis set with the current window analysis set in step S212 one by one, searching data of which the angle change of the last pair of tracks exceeds the first angle difference threshold value, and regarding the data as the end point of turning.

Referring to the data in table 1, the two data sets, i.e., Tim [ i +30] and Tim [ i +31], can be sequentially traversed from front to back, and then the last angle difference exceeding the first angle difference threshold is found to be 12 °. Or, the last piece of data Tim [ i +34] can be searched forward, and the first angle difference exceeding the first angle difference threshold corresponds to two pieces of data Tim [ i +30] and Tim [ i +31 ].

The same choice as the starting point ST1 is made, and the invention does not limit the two data of the specific Tim [ i + k ], Tim [ i + k +1], where k is a positive integer. In this embodiment, Tim [ i + k +1], i.e., Tim [ i +31] is preferably the termination point EN 1.

Through steps S212, S213, the start point ST1 and the end point EN1, ST 1-EN 1 of a turn on the trajectory and the data therebetween are found as a commutation data set describing a turn trajectory. In the invention, the starting point ST and the end point EN are two end points of the corresponding turning track, and are not limited to find the starting point first and then find the end point.

If there is still data on the trajectory that has not been analyzed, there may be additional turns. At this time, further, for the window analysis set for obtaining the commutation data set, from the next piece of data in the window analysis set, if the step size is the window size, then Tim [ i +35] -Tim [ i +69] is selected backward from Tim [ i +35] as a new window analysis set, and analysis is continued, and the above steps S211, S212, and S213 are repeated, correspondingly, the second commutation data set includes the start point ST2 and the end point EN2 and data therebetween, the third commutation data set includes the start point ST3 and the end point EN3, and so on until the analysis of the trajectory is completed and the commutation data sets at all turns are found.

In the above step, one condition of the window sliding operation is whether the position of the data corresponding to the currently found starting point relative to the 1 st data in the current window analysis set exceeds the parameter M, another condition is that the analysis processing of the current window analysis set is completed and the reversing data set corresponding to the turning track is found, and if any one of the two conditions is satisfied, the window is slid.

In addition, a trajectory may also be analyzed by a clustering algorithm, a regression algorithm, or an SVM (support vector machine), and the like, to distinguish and identify a straight trajectory and a turning trajectory in the trajectory, which are all the prior art, and are not described herein again. After the turning track is identified, the starting point and the ending point of the turning can be found for subsequent analysis and processing.

And S220, judging the type of the reversing data set, including left turning or right turning. Setting: the positive north direction is 0 degrees, the anticlockwise direction is a negative direction, and the clockwise direction is a positive direction. Comparing the angle change conditions of the starting point and the ending point of each reversing set, wherein if the angle is increased, the reversing set rotates clockwise, and the reversing set can be understood as turning to the right; the angle is reduced, and the rotation is in the anticlockwise direction, and the left steering can be understood.

Referring to the data of table 1, the angle is reduced by 62 ° from 85 ° of the start point ST1 to 13 ° of the end point EN1, and therefore, the left turn is judged as a corresponding steering data set as the position data set L1 where the left turn motion occurs.

Continuing to judge a second reversing data set, and if the steering is left-turning, taking the second reversing data set as a position data set L2 for left-turning; if the direction is the right direction, the direction data set is used as a position data set R1 for right direction action;

judging a third reversing data set, and if the steering is left steering, taking the third reversing data set as a position data set L3 for left steering; in the case of a right turn, the steering data set is used as the position data set R2 for the right turn maneuver.

Until the steering judgment of all the steering data sets is completed. One or more position data sets { Li | i ═ 1,2,3 … } indicating that a left turn motion occurs in the traveling direction are obtained, and one or more position data sets { Ri | i ═ 1,2,3 … } indicating that a right turn motion occurs in the traveling direction are obtained.

In the above steps, the left steering and the right steering are opposite directions, and there are differences in selecting different reference frames and different description manners, and those skilled in the art can also make adaptive adjustments based on conditions and needs, therefore, how to judge the change of the driving direction is not limited to the above method in the present invention.

Through the screening of the first angle threshold value, the angles of the two positions of the starting point and the ending point are respectively changed by at least 10 degrees, so that small-angle steering of the vehicle under the conditions of normal attitude adjustment, lane changing and the like is removed.

A second angle threshold is defined, which is approximately 150 ° and 160 °, and in this embodiment, the second angle threshold is preferably 155 °.

A third angle threshold is defined, and the value range of the third angle threshold is approximately 30 to 45 °, and in this embodiment, the third angle threshold is preferably 35 °.

Further, in order to screen out turning tracks of the similar situations such as turning around or changing lanes continuously, the method also comprises the following steps:

and S221, judging whether the angle change from the starting point to the end point of each position data set Li exceeds the second angle threshold, and if so, removing the position data set Li or removing the corresponding reversing data set.

S222, judging whether the angle change from the starting point to the end point of each position data set Li exceeds the third angle threshold, if not, removing the position data set Li, or removing the corresponding reversing data set.

The angle change from the starting point to the ending point is the total angle change of the vehicle from the beginning of turning to the completion of turning. By means of the second angle threshold the U-turn data is removed, and by means of the first angle threshold the trajectories of actions, such as lane changes, which would form a large angle turn are removed.

An intersection reversing distance threshold is defined, and the value range of the intersection reversing distance threshold is approximately within 120m, and in the embodiment, the intersection reversing distance threshold is preferably 100 m.

Further, the method also comprises the following steps:

s223, calculating the distance from the starting point to the end point of each position data set Li, judging whether the distance exceeds the intersection reversing distance threshold value, and if not, describing a turning track at an intersection by the position data set Li; if the position data set Li does not describe the turning track, the turning track is not taken at the position of an intersection, and the position data set Li is removed.

The above screening procedure is also applicable to the position data set Ri.

Therefore, the turning track of the intersection position can be more accurately described through screening.

S30, calculating an intersection center position set { LOi | i ═ 1,2,3, … } of the position data set { Li | i ═ 1,2,3 … }, and constructing a data set for determining the intersection based on the intersection center position set { LOi | i ═ 1,2,3, … }; and calculating an intersection center position set { ROi | i ═ 1,2,3, … } of the position data set { Ri | i ═ 1,2,3 … }, and constructing a data set of the intersection to be determined according to the intersection center position set { ROi | i ═ 1,2,3, … }.

The calculation process of the position data set Li and the position data set Ri is the same, and the position data set L1 is taken as an example.

S311, referring to fig. 4, there are 29 pieces of data from the start point ST1 to the end point EN1, corresponding to the position of 29 points. Sequentially calculating the distance delta di between two adjacent points according to the position data of each piece of data, wherein i is 1,2,3, …, P, and accumulating to obtain the track length of the whole turning track

For the position data set L1, P28. When other parameters such as the vehicle speed are different, the value of P also changes.

S312, calculating the sum of delta di and the first n distances in the P distances

n starts at 1 and has an upper limit of P.

S313, comparison D_nAnd D/2 size:

if D is_n<D/2, adding 1 to the value of n, and repeating the step S312;

if D is_nIf D/2, then take Tim [ i +3+ n +1]Position data of the corresponding data as an intersection center position LO1 of the position data set L1;

if D is_n>D/2, then | D is further compared_n-1-D/2| and | D_nThe magnitude of D/2 |:

if | D_n-1-D/2|<|D_nD/2, then take Tim [ i +3+ n-1+1 |)]Position data of the corresponding data as an intersection center position LO1 of the position data set L1;

if | D_n-1-D/2|>|D_nD/2, then take Tim [ i +3+ n +1 |)]The position data of the corresponding data is the intersection center position LO1 of the position data set L1.

By analogy, the intersection center position LO2, the intersection center position LO3 … …, the intersection center position RO1 and the intersection center position RO2 … … are obtained

The intersection center position LO1, the intersection center position LO2 and the intersection center position LO3 … … of the left turning track form a data set for determining the intersection; the intersection center position RO1 and the intersection center position RO2 … … of the right turning trajectory constitute a data set of the intersection to be determined.

In the above steps, the position data set may also be analyzed and calculated by means of a clustering algorithm, etc., to obtain the center position of the intersection.

When there are a plurality of pieces of trajectory data within the target area, there may be duplication in the identified intersection data.

A first intersection distance threshold is defined, and is typically about 20 m.

And defining a second intersection distance threshold, wherein the value range of the second intersection distance threshold is about 50m under the normal condition.

Further, the method also comprises the steps of judging and processing repeated intersection data:

judging whether the distance between each intersection center position ROi in the data set of the intersection to be determined and each intersection center position LOi in the data set of the intersection to be determined is smaller than the first intersection distance threshold value or not, and if so, moving out the intersection center position ROi from the data set of the intersection to be determined;

and judging whether the distance between any two intersection center positions LOi in the data set of the intersection is smaller than the second intersection distance threshold value or not, if so, regarding the two intersection center positions LOi as repetition, and combining the two intersection center positions LOi. Wherein, one of the repeated intersection center positions LOi can be removed from the data set of the determined intersection, or the repeated intersection center positions LOi are subjected to statistical processing, such as averaging;

similarly, whether the distance between any two intersection center positions ROi in the data set of the intersection to be determined is smaller than the second intersection distance threshold value or not is judged, if so, the two intersection center positions ROi are regarded as repetition, and the two intersection center positions ROi are combined.

The accuracy of the intersection to be determined is relatively poor, so that the reference value of the intersection to be determined is improved. In this embodiment, the position of the intersection to be determined is corrected.

As an embodiment: defining an adjusting parameter of the center position of the intersection, adjusting the center position ROi of the intersection by means of the adjusting parameter, and adding a data set of the intersection to be determined. For example, the adjustment parameter is set to 10m, and the position of the intersection center position ROi is moved 10m to the left front direction, and the specific moving direction and amplitude can be adjusted according to the road condition and other factors.

As another embodiment: and fitting and correcting the central position ROi of the intersection by means of a regression algorithm and the like, and then adding a data set of the intersection to be determined.

Example 2

In this embodiment, the intersection identification method of embodiment 1 is applied to trajectory data of a certain range of a certain intersection to identify the position of the intersection.

The shape of the intersection is shown in fig. 8, in which (a) is a right-turn trajectory and (b) is a left-turn trajectory.

Table 2 is track data of a vehicle turning to the right when passing through the intersection. In which part of the data is omitted.

TABLE 2A Window analysis set of Right turn trajectories at crossroads

Label (R)	Right turn track (longitude and latitude)	Angle/° (north direction is 0 °)	Angular difference/° of adjacent points
				T1	121.434619,31.388805	351	/
T2	121.434619,31.388839	350	-1
				T3	121.434592,31.388934	350	0
T4	121.434556,31.389019	350	0
				T5	121.434538,31.389076	350	0
T6	121.434525,31.38913	2	12
				T7	121.434516,31.389215	11	9
T8	121.434498,31.389288	18	7
				T9	121.434507,31.3893	28	10
T10	121.434511,31.389365	36	8
				T11	121.434529,31.389396	45	9
T12	121.434605,31.389427	54	9
				T13	121.434673,31.389454	65	11
T14	121.434727,31.389485	79	14
				T15	121.434835,31.389527	79	0
T16	121.43497,31.389594	80	1
				T17	121.435163,31.389652	80	0
T18	121.435392,31.389756	81	1
				T19	121.43567,31.389872	80	-1
T20	121.435769,31.389906	80	0
				T21	121.436295,31.390115	80	0
T22	121.436865,31.39033	79	-1
				…	…	…	…
T35	…	…	…

In table 2: in calculating the angular difference, for the data at T5 and T6, the angular difference is-348 °, for convenience of comparison the steering direction is clockwise, and 360 ° is added on the basis thereof, which is denoted as 12 °. Similarly, if there is counterclockwise rotation, for example, the angle of one piece of data is 2 °, the next piece of data is 350 °, the angle difference between the two is 348 °, for convenience of comparison, the direction of rotation is clockwise, and 360 ° is subtracted from the clockwise direction, which is expressed as-12 °.

The intersection identification method of the embodiment 1 is adopted to process the right turn track in the table 2, and the main steps are as follows:

the angle between the data T5 and the data T6 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T5 is the 5 th data in the window, conforming to the definition, the start point ST2 is found and is T5.

The angle between the data T13 and the data T14 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T14 EN 2.

Thus, data T5-data T14 are a commutation data set. The angle change of 77 ° is accumulated from the data T5 to the data T14 and is increased, and therefore, the corresponding trajectory is a right-turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position RO can be obtained as the position of the data T10 [121.434511,31.389365 ].

The intersection center position corresponding to the table 2 obtained by the Baidu map data is [121.43434,31.389381 ].

Through longitude and latitude distance conversion, the distance between the center position RO [121.434511,31.389365] of the intersection and the center position [121.43434,31.389381] of the intersection obtained by the hundredth map data is 16.3 m. The latitude distance is converted into the prior art, and the invention does not limit this and is not described in detail.

Table 3 is track data of a vehicle turning left when passing through the intersection. In which part of the data is omitted.

TABLE 3A Window analysis set of left turn trajectories at crossroads

The intersection identification method of the embodiment 1 is adopted to process the left-turn track in the table 3, and the main steps are as follows:

the angle between the data T5 and the data T6 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T5 is the 5 th data in the window, conforming to the definition, the start point ST3 is found and is T5.

The angle between the data T13 and the data T14 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T14 EN 3.

Thus, data T5-data T14 are a commutation data set. The angle change is accumulated by 84 ° from the data T5 to the data T14, and is reduced, and therefore, the corresponding trajectory is a left turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position LO can be obtained as the position [121.43434,31.389373] of the data T10.

The intersection center position corresponding to the table 3 obtained by the Baidu map data is [121.43434,31.389381 ]. The trajectory corresponding to table 2 and the trajectory corresponding to table 3 occur at the same intersection.

Through longitude and latitude distance conversion, the distance between the intersection center position LO [121.43434,31.389373] and the intersection center position [121.43434,31.389381] obtained by the hundredth map data is 0.88 m.

Therefore, under the condition of neglecting the error of the hundred-degree map data, the accuracy of the intersection position determined by the left-turn track to the intersection position determined by the right-turn track is greatly improved, and the accuracy is about 1 m.

Example 3

In this embodiment, the intersection identification method of embodiment 1 is applied to the trajectory data of a certain range of a certain t-junction to identify the position of the intersection.

The shape of the t-junction is shown in fig. 9, in which (a) is a right turn trajectory and (b) is a left turn trajectory.

Table 4 is track data for a vehicle turning to the right when passing through the t-junction. In which part of the data is omitted.

TABLE 4 Window analysis set of Right turn trajectory for a T-junction

The intersection identification method of the embodiment 1 is adopted to process the right-turn track in the table 4, and the main steps are as follows:

the angle between the data T4 and the data T5 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T4 is the 4 th data in the window, conforming to the definition, the start point ST4 is found and is T4.

The angle between the data T12 and the data T13 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T13 EN 4.

Thus, data T4-data T13 are a commutation data set. The angle change is accumulated to 98 ° from the data T4 to the data T13, and is increased, and thus, the corresponding trajectory is a right-turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position RO can be obtained as the position of the data T19 [114.118334,22.56586 ].

The intersection center position corresponding to the table 4 obtained by the Baidu map data is [114.118401,22.565989 ].

Through longitude and latitude distance conversion, the distance between the center position RO [114.118334,22.56586] of the intersection and the center position [114.118401,22.565989] of the intersection obtained by the hundredth map data is 15.89 m.

Table 5 is track data of a vehicle turning left when passing through the intersection. In which part of the data is omitted.

TABLE 5 Window analysis set of left turn trajectory for a T-junction

Label (R)	Left turn track (longitude and latitude)	Angle/° (north direction is 0 °)	Angular difference/° of adjacent points
				T1	114.117705,22.565776	78	/
T2	114.117777,22.565801	78	0
				T3	114.117871,22.565822	78	0
T4	114.117952,22.565843	78	0
				T5	114.118019,22.565864	68	-10
T6	114.118136,22.565893	61	-7
				T7	114.118217,22.565914	54	-7
T8	114.118334,22.565951	46	-8
				T9	114.118401,22.565981	38	-8
T10	114.118428,22.566026	30	-8
				T11	114.118441,22.566114	21	-9
T12	114.118441,22.566202	11	-10
				T13	114.118446,22.566227	358	-13
T14	114.118446,22.566298	358	0
				T15	114.118442,22.566479	358	0
T16	114.118446,22.566805	359	1
				T17	114.118428,22.567351	358	-1
…	…	…	…
				T35	…	…	…

In table 5: in calculating the angular difference, for the data at T12 and T13, the angular difference was 347 °, for convenience of comparison the steering direction was counter-clockwise, and on that basis 360 ° was subtracted, which is denoted-13 °.

The intersection identification method of the embodiment 1 is adopted to process the left-turn track in the table 5, and the main steps are as follows:

the angle between the data T4 and the data T5 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T4 is the 4 th data in the window, conforming to the definition, the start point ST5 is found and is T4.

The angle between the data T12 and the data T13 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T13 EN 5.

Thus, data T4-data T13 are a commutation data set. The angle change is accumulated by 80 ° from the data T4 to the data T13, and is reduced, and therefore, the corresponding trajectory is a left turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position LO can be obtained as the position [114.118401,22.565981] of the data T9.

The intersection center position corresponding to the table 5 obtained by the Baidu map data is [114.118401,22.565989 ]. The trajectory corresponding to table 4 occurs at the same intersection as the trajectory corresponding to table 5.

Through longitude and latitude distance conversion, the distance between the intersection center position LO [114.118401,22.565981] and the intersection center position [114.118401,22.565989] obtained by the hundredth map data is 1.47 m.

Example 4

In this embodiment, the intersection identification method of embodiment 1 is applied to the trajectory data of a certain range of an irregular intersection to identify the position of the irregular intersection.

The shape of the t-junction is shown in fig. 10, in which (a) is a right turn trajectory and (b) is a left turn trajectory.

Table 6 is track data of a vehicle turning to the right when passing through the irregular intersection. In which part of the data is omitted.

TABLE 6 Window analysis set of Right turn trajectory for irregular intersections

Label (R)	Right turn track (longitude and latitude)	Angle/° (north direction is 0 °)	Angular difference/° of adjacent points
				T1	121.519544,31.328684	43	/
T2	121.519598,31.328727	43	0
				T3	121.519679,31.328796	43	0
T4	121.519814,31.328919	43	0
				T5	121.519904,31.329004	45	2
T6	121.520003,31.329093	57	12
				T7	121.520043,31.32912	67	10
T8	121.520079,31.329143	78	11
				T9	121.520128,31.329139	90	12
T10	121.520151,31.329128	100	10
				T11	121.520196,31.329108	109	9
T12	121.520218,31.329097	120	11
				T13	121.520263,31.329081	122	2
T14	121.52033,31.329064	122	0
				T15	121.520438,31.329017	123	1
T16	121.520438,31.329017	123	0
				T17	121.520811,31.328848	124	1
…	…	…	…
				T35	…	…	…

The intersection identification method of the embodiment 1 is adopted to process the right-turn track in the table 6, and the main steps are as follows:

the angle between the data T5 and the data T6 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T5 is the 5 th data in the window, conforming to the definition, the start point ST5 is found and is T5.

The angle between the data T11 and the data T12 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T12 EN 5.

Thus, data T5-data T12 are a commutation data set. The angle change is accumulated by 75 ° from the data T5 to the data T12, and is increased, and therefore, the corresponding trajectory is a right-turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position RO can be obtained as the position of the data T8 [121.520079,31.329143 ].

The intersection center position corresponding to the table 6 obtained by the Baidu map data is [121.520079,31.329209 ].

Through longitude and latitude distance conversion, the distance between the center position RO [121.520079,31.329143] of the intersection and the center position [121.520079,31.329209] of the intersection obtained by the hundredth map data is 12.22 m.

Table 7 is track data of a vehicle turning to the left when passing through the intersection. In which part of the data is omitted.

TABLE 7 Window analysis set of left turn trajectories at irregular intersections

Label (R)	Left turn track (longitude and latitude)	Angle (true north direction is 0 degree)
				T1	121.520132,31.329638	195	/
T2	121.520128,31.329568	195	0
				T3	121.520106,31.32946	196	1
T4	121.520092,31.329368	196	0
				T5	121.520092,31.329368	195	-1
T6	121.520083,31.329295	184	-11
				T7	121.52007,31.329202	174	-10
T8	121.520061,31.329199	163	-11
				T9	121.520128,31.32916	152	-11
T10	121.520186,31.329129	141	-11
				T11	121.520245,31.329098	131	-10
T12	121.520362,31.329048	121	-10
				T13	121.520438,31.329009	122	1
T14	121.520658,31.328909	123	1
				T15	121.520847,31.32882	124	1
T16	121.520972,31.328763	124	0
				T17	121.52104,31.328728	122	-2
…	…	…	…
				T35	…	…	…

The intersection identification method of the embodiment 1 is adopted to process the left-turn track in the table 7, and the main steps are as follows:

the angle between the data T5 and the data T6 is two adjacent data with the 1 st angle difference being larger than 10 degrees in the window. Since T5 is the 5 th data in the window, conforming to the definition, the start point ST6 is found and is T5.

The angle between the data T11 and the data T12 is two adjacent data with the reciprocal 1 st angle difference in the window being larger than 10 degrees. Thus found and terminated by T12 EN 6.

Thus, data T5-data T12 are a commutation data set. The angle change is accumulated by 74 ° from the data T5 to the data T12, and is reduced, and therefore, the corresponding trajectory is a left turn trajectory. Through the calculation of step S30 in embodiment 1, the corresponding intersection center position LO can be obtained as the position [121.520061,31.329199] of the data T8.

The intersection center position corresponding to the table 7 obtained by the Baidu map data is [121.520079,31.329209 ]. The trajectory corresponding to table 6 occurs at the same intersection as the trajectory corresponding to table 7.

Through longitude and latitude distance conversion, the distance between the intersection center position LO [121.520061,31.329199] and the intersection center position [121.520079,31.329209] obtained by hundred-degree map data is 3.39 m.

It can be seen that, in combination with

embodiments

2,3, and 4, the accuracy of the intersection position determined by the left-turn trajectory to the intersection position determined by the right-turn trajectory is greatly improved, which is approximately by one order of magnitude.

Example 5

The method for identifying the intersection based on the vehicle driving track depends on track data, on one hand, the method depends on track data and accumulation, and on the other hand, along with road construction or reconstruction, the intersection itself may change and needs to be corrected or updated on the basis of the existing intersection data.

In this embodiment, the intersection data may be obtained based on the intersection identification method in embodiment 1, or may be obtained based on other methods, which is not limited in the present invention.

Referring to fig. 5, the intersection identification updating method of the present embodiment includes the following steps:

and S40, updating the track data of the vehicle in the target area or acquiring new track data in the target area. Refer to step S10 in example 1.

And S50, analyzing the updated track data. Referring to step S20 in example 1, one or more position data sets { Lxi ═ 1,2,3 … } indicating that a left turn motion occurs in the traveling direction and one or more position data sets { Rxi ═ 1,2,3 … } indicating that a right turn motion occurs in the traveling direction are obtained.

S60, referring to step S30 in example 1, the intersection center position set { LxOi | i ═ 1,2,3, … } of the position data set { Lxi | i ═ 1,2,3 … } is calculated, and the intersection center position set { RxOi | i ═ 1,2,3, … } of the position data set { Rxi ═ 1,2,3 … } is calculated.

In this embodiment, the specific steps and implementation manners of the steps S40, S50, and S60 are substantially the same as those of the related steps in embodiment 1, and reference may be made to the specific contents in embodiment 1, which is not repeated herein.

The intersection identification updating method of the embodiment further comprises the following steps:

and S70, judging the newly added intersection and updating the data set of the determined/to-be-determined intersection.

The first intersection distance threshold value and the second intersection distance threshold value defined in example 1 are referred to.

As shown in fig. 6, the steps of determining and processing the intersection center position LxOi are as follows.

S701, judging whether the intersection described by each intersection center position LxOi exists in a data set for determining the intersection or not.

When the distance between the intersection center position LxOi and an existing intersection center position LOi does not exceed the second intersection distance threshold, the intersection center position LxOi and the intersection center position LOi are regarded as the same intersection, namely the intersection described by the intersection center position LxOi already exists in the data set of the intersection is determined, and the intersection center position LxOi is not a newly added intersection.

If the distance between the intersection center position LxOi and any existing intersection center position LOi exceeds the second intersection distance threshold, the next step S702 is performed.

S702, judging whether the intersection described by the center position LxOi of each intersection exists in a data set of the intersection to be determined.

When the distance between the intersection center position LxOi and an existing intersection center position ROi does not exceed the first intersection distance threshold, the intersection center position LxOi and the intersection center position ROi are regarded as the same intersection, the intersection described by the intersection center position LxOi exists in the data set of the intersection to be determined, at the moment, the intersection center position ROi is moved out of the data set of the intersection to be determined, meanwhile, the intersection center position LxOi is expressed as an intersection center position LOi, and the intersection center position LOi is added into the data set of the determined intersection.

When the distance between the intersection center position LxOi and any one existing intersection center position ROi exceeds the first intersection distance threshold, the intersection center position LxOi and any one existing intersection center position LOi or the intersection center position ROi do not represent the same intersection, the intersection described by the intersection center position LxOi does not exist in the intersection data set or the intersection data set, the intersection center position LxOi is a newly-added intersection, the intersection center position LxOi is represented as the intersection center position LOi, and the intersection data set is added into the intersection data set.

As shown in fig. 7, the steps of determining and processing the intersection center position RxOi are as follows.

S703, judging whether the intersection described by each intersection center position RxOi exists in the data set for determining the intersection or not.

When the distance between the intersection center position RxOi and an existing intersection center position LOi does not exceed the first intersection distance threshold, the intersection center position RxOi and the intersection center position LOi are regarded as the same intersection, namely, the intersection described by the intersection center position RxOi already exists in the data set of the intersection is determined, and the intersection center position RxOi is not a newly added intersection.

If the distance between the intersection center position RxOi and any one existing intersection center position LOi exceeds the first intersection distance threshold, the next step S704 is entered.

S704, judging whether the intersection described by each intersection center position RxOi exists in the data set of the intersection to be determined.

When the distance between the intersection center position RxOi and any one existing intersection center position ROi exceeds the second intersection distance threshold, the intersection center position RxOi and any one existing intersection center position ROi do not represent the same intersection, and the intersection described by the intersection center position RxOi does not exist in the data set of the intersection to be determined. The intersection center position RxOi is a newly added intersection, is expressed as an intersection center position ROi, and is added into a data set of the intersection to be determined.

Similarly, when the distance between the central position RxOi of the intersection and the central position ROi of an existing intersection does not exceed the distance threshold of the second intersection, the central position RxOi of the intersection and the central position ROi of the existing intersection represent the same intersection, that is, the intersection described by the central position RxOi of the intersection exists in the data set of the intersection to be determined.

In the above steps, S702 and S703 are optional steps, and the steps related to determining and processing the repeated intersection data in embodiment 1 may be adopted instead.

Example 6

The computer-readable storage medium of the present embodiment has stored thereon a computer program/instructions which, when executed by a processor, implement the steps of the intersection identification method of the embodiment 1 described above, or implement the steps of the intersection identification update method of the embodiment 2 described above.

Example 7

The computer program product of this embodiment includes a computer program/instruction, which when executed by a processor implements the steps of the intersection identification method of embodiment 1 or implements the steps of the intersection identification update method of embodiment 2.

Example 8

The intersection data model of the embodiment is constructed according to the intersection identification method of the embodiment 1 or the intersection identification updating method of the embodiment 2 to obtain the data set of the determined intersection and/or the data set of the intersection to be determined.

Specifically, the data of the center position of each intersection in the data set of the determined intersection and the data set of the intersection to be determined are usually longitude and latitude coordinates. The position coordinates in the target area and the center positions of all intersections can be processed by adopting a GeoHash algorithm to generate corresponding GeoHash codes, and the center positions of the intersections are marked to construct intersection data models.

Example 9

And defining an intersection area, for example, setting a certain radius parameter by taking the central position of each intersection as the center of a circle, wherein the area covered by the radius is the intersection area. And when the distance between the vehicle and the intersection position is less than the radius parameter, the vehicle is considered to enter the intersection area.

The intersection reminding method of the embodiment comprises the following steps:

obtaining a vehicle position;

and judging whether the vehicle position is in the intersection area or not according to the intersection data model, and sending out reminding information if the vehicle enters the intersection area.

Alternatively, the vehicle position is also GeoHash processed and compared with the data in the intersection data model of the embodiment 8 to determine that the vehicle enters the intersection area and approaches the intersection.

Example 10

The electronic device of this embodiment includes a memory, a processor, and a computer program stored on the memory, the processor and the memory are connected through a system bus, the electronic device is configured to receive the reminding information of the intersection reminding method of embodiment 9 and issue a reminder to a driver through voice, vibration, and the like, and at this time, the reminding information may be understood as electronic information or an instruction transmitted through a network; alternatively, the processor executes the computer program to implement the steps of the intersection reminding method of the embodiment 9, and at this time, the reminding information can be understood as an alarm such as voice, vibration, and the like.

Any reference to memory, storage, or other media used in various embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.

The embodiments of the present invention are merely illustrative, and not restrictive, of the scope of the claims, and other substantially equivalent alternatives may occur to those skilled in the art and are within the scope of the present invention.

Claims

1. An intersection identification method, characterized by comprising the steps of:

obtaining trajectory data of a vehicle within a target area;

li denotes a position data set, Ri denotes a position data set;

2. The intersection identification method according to claim 1, wherein a first intersection distance threshold value for judging the position repetition is defined, further comprising the steps of:

3. The intersection identification method according to claim 1, further comprising the steps of identifying and updating a newly added intersection:

updating trajectory data of the vehicle within the target area;

lxi denotes one position data set, and Rxi denotes one position data set.

4. The intersection recognition method according to claim 1 or 3, characterized in that trajectory data of the vehicle is represented by a data set { { Tim [ i ], Loc [ i ], Dir [ i ], Vol [ i ] } | i ═ 1,2,3, … } describing a travel trajectory, in which:

tim represents a time scale, Loc represents the position of the vehicle, Dir represents the driving direction of the vehicle, Vol represents the speed of the vehicle, { Tim [ i ], Loc [ i ], Dir [ i ], Vol [ i ] } represents a piece of data acquired at a time corresponding to Tim [ i ];

defining an upper speed threshold value, and eliminating data with the speed exceeding the upper threshold value;

the analyzing step of the trajectory data comprises:

obtaining a commutation data set describing a turn trajectory, the commutation data set including a first end point and a second end point of the turn trajectory and data therebetween;

determining a type of turn of the commutation data set, including left turn and right turn, for constructing as corresponding one or more position data sets { Li or Lxi ═ 1,2,3 … } where a left turn action occurs or/and one or more position data sets { Ri or Rxi ═ 1,2,3 … } where a right turn action occurs;

obtaining a reversing data set describing each turning track in the track data through a sliding window algorithm or a clustering algorithm or a regression algorithm or an SVM;

defining the size of a sliding window as N, defining a conditional parameter M of window sliding, defining a first angle threshold,

the sliding window algorithm comprises the steps of:

A. comparing the angles of the two adjacent data one by one, and finding out the two adjacent data with the angle difference exceeding the first angle threshold value to obtain a first endpoint serving as a turning track;

B. if the position of the first endpoint in the current window exceeds the condition parameter M, sliding the window until the first endpoint is the item 1 data in the window, or obtaining the distance between the first endpoint and the item 1 data in the current window as N1, and temporarily expanding the size of the current window backwards by N1;

C. if the position of the first endpoint in the current window does not exceed the condition parameter M, finding the last two adjacent data of which the angle difference exceeds the first angle difference threshold value in the current window to obtain a second endpoint serving as a turning track;

D. the first end point, the second end point and data between the first end point and the second end point are a reversing data set;

E. and if a reversing data set is found, sliding the window backwards, wherein the step length is the size N of the sliding window, and repeating the steps A to D.

5. The intersection identification method according to claim 4, wherein a second angle threshold and a third angle threshold are defined, and if the angle difference between the first end point and the second end point in the direction-changing data set exceeds the second angle threshold or does not exceed the third angle threshold, the direction-changing data set is removed; and/or the first and/or second light sources,

and defining an intersection reversing distance threshold, and removing the reversing data set if the distance between the first end point and the second end point in the reversing data set exceeds the intersection reversing distance threshold.

6. The intersection recognition method according to claim 1 or 3, wherein the position data set is represented by { T }₁,T₂,…,T_i,…,T_P+1}，T_iFor a piece of data in the corresponding location data set, T₁Corresponding to a first endpoint, T_P+1Corresponding to the second end point, the step of calculating the intersection center position of each position data set comprises the following steps:

a. sequentially calculating the distance delta di between two adjacent points according to the position data of each piece of data, wherein i is 1,2,3, … and P;

b. track length of the entire turning track

c. Calculating the sum of the first n distances

n begins to take a value from 1;

d. if D is_n<D/2, adding 1 to the value of n, and repeating the step c; or, if D_nD/2 or | D_n-1-D/2|>|D_n-D/2, then take T_n+1The position data of the corresponding data is used as the intersection center position of the position data set; or, if | D_n-1-D/2|<|D_nD/2, then take T_nThe position data of the corresponding data is used as the intersection center position of the position data set;

defining a first intersection distance threshold value and a second intersection distance threshold value;

if the distance between each intersection center position ROi in the data set of the intersection to be determined and each intersection center position LOi in the data set of the intersection to be determined is smaller than the first intersection distance threshold, the intersection center position ROi is moved out of the data set of the intersection to be determined; or,

if the distance between the central positions LOi of the two intersections in the data set of the intersections is smaller than the second intersection distance threshold, combining the central positions LOi of the two intersections; or,

and if the distance between the central positions ROi of the two intersections in the data set of the intersection to be determined is smaller than the second intersection distance threshold, combining the central positions ROi of the two intersections.

7. The intersection identification method according to claim 1 or 3, characterized in that adjustment parameters of the center position of the intersection are defined, and the center position ROi of the intersection is adjusted by means of the adjustment parameters and then added with the data set of the intersection to be determined; or fitting or correcting the central position ROi of the intersection and then adding the data set of the intersection to be determined.

8. A computer-readable storage medium, on which a computer program/instructions are stored, which, when being executed by a processor, carry out the steps of the intersection identification method according to any one of claims 1 to 7.

9. A computer program product comprising computer program/instructions, characterized in that the computer program/instructions, when executed by a processor, implement the steps of the intersection identification method according to any of claims 1-7.

10. An intersection data model, characterized in that, the intersection data model is constructed by the intersection-determining data set and/or the intersection-to-be-determined data set obtained by the intersection identification method according to any one of claims 1 to 7.

11. An intersection reminding method is characterized by comprising the following steps:

obtaining a vehicle position;

the intersection data model of claim 10 determining if the vehicle location is within the intersection area and sending a reminder if the vehicle is entering the intersection area.

12. An electronic device comprising a memory, a processor and a computer program stored on the memory, wherein the electronic device is configured to receive the reminding information of the intersection reminding method of claim 11 and issue a reminder to a driver, or wherein the processor executes the computer program to implement the steps of the intersection reminding method of claim 11.