A method for determining road data errors, related methods and devices

Abstract

The present invention discloses a method for determining road data errors, as well as related methods and devices. The method comprises: obtaining a driving trajectory represented by a series of trajectory points that meets quality standards; obtaining a road to be matched within a preset distance range around the driving trajectory, wherein the road to be matched includes production attributes and virtual attributes, and the values of the virtual attributes are opposite to those of the production attributes; based on the production attributes of the road to be matched, determining a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched, thereby obtaining a first matching probability of the driving trajectory to the road to be matched; based on the virtual attributes of the road to be matched, determining a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, thereby obtaining a second matching probability of the driving trajectory to the road to be matched; and when the second matching probability is greater than the first matching probability, determining that the production attributes of the road to be matched are incorrect. This method provides a solution for accurately finding road data with errors.

Description

Technical Field

[0001] The present invention relates to the field of geographic information technology, and in particular to a method for determining road data errors, and related methods and devices. Background Art

[0002] In the field of geographic information technology, the integrity and accuracy of road data are crucial. Errors in road data can lead to errors in data-based operations. For example, incorrect road directions can lead to errors in navigation route planning. Therefore, accurately identifying erroneous road data is a technical challenge facing those skilled in the art. Summary of the Invention

[0003] In view of the above problems, the present invention is proposed to provide a method for determining road data errors, related methods and devices that overcome the above problems or at least partially solve the above problems.

[0004] In a first aspect, an embodiment of the present invention provides a method for determining road data errors, comprising the following steps:

[0005] Acquire a driving trajectory represented by a series of trajectory points that meets quality standards, wherein the quality standards include a consistency standard, a fluctuation standard, and a quantity standard of the trajectory points;

[0006] Acquire a to-be-matched road within a preset distance range around the driving trajectory, wherein the to-be-matched road includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute;

[0007] Based on the production attributes of the road to be matched, determining a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched according to a preset projection probability and transition probability algorithm, and obtaining a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability;

[0008] Determining, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtaining a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability;

[0009] When the second matching probability of the driving trajectory to the road to be matched is greater than the first matching probability, it is determined that the production attribute of the road to be matched is wrong.

[0010] In a second aspect, an embodiment of the present invention provides a method for determining a road data error type, comprising the following steps:

[0011] Determine whether there is an error in the properties of the road to be matched with the driving trajectory;

[0012] If so, determining the error type of the road to be matched according to the virtual attribute of the road to be matched;

[0013] The step of determining whether there is an error in the production attributes of the road to be matched with the driving trajectory adopts the above-mentioned method for determining road data errors.

[0014] In a third aspect, an embodiment of the present invention provides a device for determining road data errors, comprising:

[0015] a trajectory extraction module, configured to obtain a driving trajectory represented by a series of trajectory points that meets quality standards, wherein the quality standards include a consistency standard, a volatility standard, and a quantity standard of the trajectory points;

[0016] A road acquisition module is used to acquire a road to be matched within a preset distance range around the driving trajectory, wherein the road to be matched includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute;

[0017] a first matching probability determination module, configured to determine, based on the production attributes of the road to be matched and in accordance with a preset projection probability and transition probability algorithm, a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtain a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability;

[0018] a second matching probability determination module, which determines, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtains a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability;

[0019] The road data error determination module is used to compare the first matching probability obtained by the first matching probability determination module with the second matching probability obtained by the second matching probability determination module, and when the second matching probability is greater than the first matching probability, determine that the production attribute of the road to be matched is erroneous.

[0020] In a fourth aspect, an embodiment of the present invention provides a device for determining a road data error type, comprising:

[0021] A data error judgment module is used to determine whether there is an error in the properties of the road to be matched in the driving trajectory;

[0022] a type error determination module, configured to determine the error type of the to-be-matched road according to the virtual attributes of the to-be-matched road when an error occurs in the production attributes of the to-be-matched road of the driving trajectory;

[0023] The step of determining whether there is an error in the production attributes of the road to be matched with the driving trajectory adopts the above-mentioned method for determining road data errors.

[0024] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the above-mentioned road error determination method, or implement the above-mentioned road data error type determination method.

[0025] In a sixth aspect, an embodiment of the present invention provides a server comprising: a processor and a memory for storing processor executable commands; wherein the processor is configured to execute the above-mentioned road error determination method, or execute the above-mentioned road data error type determination method.

[0026] The beneficial effects of the above technical solutions provided by the embodiments of the present invention include at least:

[0027] The above-mentioned method for determining road data errors provided by an embodiment of the present invention obtains a driving trajectory represented by a series of trajectory points that meets quality standards, obtains the roads to be matched within a preset distance range around the driving trajectory, and introduces corresponding virtual attributes to the production attributes corresponding to the roads to be matched when performing trajectory matching. Based on the production attributes and the virtual attributes, a first matching probability and a second matching probability corresponding to the driving trajectory to the roads to be matched are obtained, and compared. If the second matching probability is greater than the first matching probability, it indicates that an error has occurred in the road data. The embodiment of the present invention utilizes a driving trajectory that meets quality standards to measure the correctness of the road data of the roads to be matched. During the road matching process, virtual attributes of the roads to be matched are introduced to assume various possible errors in the roads to be matched. Then, by comparing the matching results, it can be determined whether the production attributes of the roads to be matched are erroneous. This provides a solution for accurately finding erroneous road data, provides a good foundation for subsequent correction of road data, and ensures the accurate implementation of various services based on road data.

[0028] The method for determining the road data error type provided by the embodiment of the present invention can not only accurately determine whether there is an error in the road data, but also, when the production attributes of the road to be matched are erroneous, further accurately determine the road data error type corresponding to the road to be matched based on the determined values of the virtual attributes of the road to be matched. The method is simple and easy to implement, and it is convenient and efficient to correct road data errors.

[0029] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0030] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0032] Figure 1 is a flow chart of a method for determining road data errors in an embodiment of the present invention;

[0033] Figure 2 Flowchart of a method for extracting a vehicle trajectory that meets quality standards according to an embodiment of the present invention;

[0034] Figure 3 Schematic diagram of road network error types in an embodiment of the present invention;

[0035] Figure 4 Schematic diagram of trajectory matching and roads in a road network in an embodiment of the present invention;

[0036] Figure 5 is a flow chart of a method for determining a road data error type in an embodiment of the present invention;

[0037] Figure 6 2 is a schematic structural diagram of a device for determining road data errors according to an embodiment of the present invention;

[0038] Figure 7 2 is a schematic diagram of the structure of a device for determining an error type of road data in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0040] In view of the problems existing in the above-mentioned prior art, the embodiment of the present invention provides a method for determining road data errors. Figure 1 As shown, the following steps are included:

[0041] S11. Obtaining a driving trajectory represented by a series of trajectory points that meets quality standards, where the quality standards include a continuity standard, a volatility standard, and a quantity standard of the trajectory points;

[0042] S12, obtaining a road to be matched within a preset distance range around the driving trajectory; wherein the road to be matched includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute;

[0043] The roads to be matched include a point coordinate sequence of road data, a road travel direction, and a road connection relationship. The threshold value of the above-mentioned preset distance range can be set according to actual conditions, and is not limited in the embodiment of the present invention.

[0044] S13. Based on the production attributes of the road to be matched, and in accordance with a preset projection probability and transition probability algorithm, determine a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtain a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability;

[0045] S14. Determine, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtain a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability;

[0046] S15: When the second matching probability between the driving trajectory and the road to be matched is greater than the first matching probability, it is determined that the production attribute of the road to be matched is wrong.

[0047] The above-mentioned method for determining road data errors provided by an embodiment of the present invention obtains a driving trajectory represented by a series of trajectory points that meets quality standards, obtains the roads to be matched within a preset distance range around the driving trajectory, and introduces corresponding virtual attributes to the production attributes corresponding to the roads to be matched when performing trajectory matching. Based on the production attributes and virtual attributes, a first matching probability and a second matching probability corresponding to the driving trajectory to the roads to be matched are respectively obtained and compared. If the second matching probability is greater than the first matching probability, it indicates that an error has occurred in the road data.

[0048] The embodiment of the present invention utilizes driving trajectories that meet quality standards to measure the correctness of the road data of the road to be matched. In the process of road matching, virtual attributes of the road to be matched are introduced to assume various possible errors in the road to be matched. Then, by comparing the matching results, it can be determined whether the production attributes of the road to be matched are wrong. This provides a solution for accurately finding erroneous road data, provides a good foundation for subsequent correction of road data, and ensures the accurate implementation of various services based on road data.

[0049] The inventors of the present invention have discovered that existing technical solutions mainly use trajectory matching methods based on the HMM model. The premise of this method is that "the roads in the road network are correct, but the driving trajectories may be wrong." Therefore, the focus is mainly on how to provide the best matching results, rather than how to determine road data errors and repair road data. After obtaining the matching results of driving trajectories and roads, the problem of road data errors can only be inferred based on the matching results obtained. Therefore, the accuracy and efficiency of handling road data errors are not high. The inventors of the present invention process the original driving trajectory to obtain a driving trajectory represented by a series of trajectory points that meets the quality standards. The driving trajectory that meets the quality standards has good consistency, small fluctuations, and the number of trajectory points meets the trajectory matching requirements, which helps to accurately determine whether there are road data errors on the roads in the road network. Therefore, before determining road data errors, it is necessary to solve the problem of how to obtain driving trajectories that meet the quality standards.

[0050] In one embodiment, in the above step S11, the driving trajectory represented by a series of trajectory points that meets the quality standard can be manually screened out from the original driving trajectory by an operator according to the above quality standard, or can be obtained by machine processing the original driving trajectory according to the above quality standard through set standard values.

[0051] As a specific implementation of the embodiment of the present invention, refer to Figure 2 As shown, obtaining a driving trajectory represented by a series of trajectory points that meets the quality standard can be achieved through the following steps:

[0052] S21, splitting the original driving trajectory at two adjacent trajectory points that do not meet the continuity criterion to obtain at least one segmented trajectory that meets the continuity criterion;

[0053] Specifically, whether two adjacent trajectory points in the original driving trajectory meet the preset continuity standard can be determined, for example, by the following method:

[0054] Determine, based on the coordinates of the track points in the original driving track, whether the distance between two adjacent track points is greater than a set distance threshold; and / or

[0055] According to the observation time of the trajectory points in the original driving trajectory, it is determined whether the observation time of two adjacent trajectory points is greater than the set time threshold;

[0056] If so, it is determined that the two adjacent trajectory points do not meet the continuity criterion.

[0057] Because each trajectory point in the original driving trajectory has four-dimensional information of longitude, latitude, speed and driving direction, as well as the observation time information of the trajectory point, the distance between two adjacent trajectory points can be determined based on the four-dimensional information of the trajectory point, and the observation time interval between two adjacent trajectory points can be determined based on the observation time information of the trajectory point. When the trajectory is split, a distance threshold or time threshold is set. If the distance between two adjacent trajectory points is less than or equal to the distance threshold or the observation time interval is less than or equal to the time threshold, it indicates that the two trajectory points meet the continuity standard.

[0058] S22, determining whether the number of track points in each segmented track meets a quantity standard, and discarding the segmented track if the number of track points does not meet the quantity standard;

[0059] If the number of track points in the segmented track is greater than or equal to the preset number threshold, it indicates that the number of track points meets the quantity standard; if the number of track points in the segmented track is less than the preset number threshold, it indicates that the number of track points is too small, and the information contained in the track is small. Such a track has little effect on the subsequent road data error determination and cannot meet the quantity standard. Therefore, it is necessary to discard the track with too few track points.

[0060] S23. For each segmented trajectory, filter out the trajectory segments that do not meet the volatility criterion to obtain at least one filtered trajectory;

[0061] Whether the above segmented trajectory meets the volatility standard can be determined, for example, by the following method:

[0062] In the trained Convolutional Neural Network (CNN) model, a preset number of point sequence windows are used to traverse each segmented trajectory, and the probability that the volatility of each segmented trajectory segment composed of a preset number of trajectory points in the segmented trajectory meets the quality requirements is determined.

[0063] Determine whether the probability is greater than a set probability threshold. If so, determine that the trajectory segment composed of the preset number of trajectory points is a trajectory segment that meets the volatility standard. If not, determine that the trajectory segment composed of the preset number of trajectory points is a trajectory segment that does not meet the volatility standard, and therefore the trajectory segment that does not meet the quality standard needs to be discarded.

[0064] For example, a segmented trajectory has 50 points. A sliding window of a 10-by-4 point sequence (i.e., 10 points containing four-dimensional information about longitude, latitude, speed, and direction of travel) is used to traverse the segmented trajectory. The first sliding window is applied to points 1-10, the second sliding window is applied to points 2-11, and so on. The probability that the volatility of a segment consisting of 10 adjacent points meets the volatility standard is output, and a determination is made as to whether the probability is greater than a set probability threshold. Assuming that the probability that the volatility of the segment consisting of points 21-30 meets the volatility standard is less than the set threshold, the segment is determined to be of substandard quality and is eliminated, resulting in two trajectories: one consisting of points 1-20 and the other consisting of points 31-50.

[0065] S24: Determine at least one selected trajectory whose number of trajectory points meets the quantity standard as a driving trajectory that meets the quality standard.

[0066] In step S24, whether the number of track points in a filtered track exceeds a preset threshold value can be determined to determine whether the number of track points meets the quantity standard. If the judgment result is yes, the filtered track is determined to be a driving track that meets the quality standard. If not, the filtered track is discarded.

[0067] After obtaining a driving trajectory that meets the quality standards, it can be guaranteed that the driving trajectory that meets the quality standards is correct during the matching process with the road, thereby determining whether the roads in the road network have at least one similar road data error. Based on the relationship between the driving trajectory that meets the quality standards and the roads in the road network, it is assumed that the roads in the road network have at least one type of road data error. The road data is introduced to represent real attributes (also called fabricated attributes, i.e., various attributes of existing roads in the road network) and virtual attributes (i.e., virtual attributes corresponding to the data errors assumed to exist on the roads) of the road to be matched.

[0068] Generally speaking, road data in a road network may contain one or more of the following errors: incorrect road direction, incorrect road topology connection, and road missing errors. A road direction error, for example, could occur when a road is actually bidirectional, but the road data is configured as a one-way road. A road topology connection error, for example, could occur when two adjacent roads are actually connected, but the road data topology is configured as disconnected. A road missing error, for example, could occur when a road exists in the actual road network, but is not configured in the road data.

[0069] To implement the method for determining road data errors provided in the embodiment of the present invention, in the embodiment of the present invention, the production attributes and virtual attributes of the roads to be matched may include any one or a combination of the following attributes: road direction, road topological connection relationship, and road missingness; wherein:

[0070] If the value of the road direction in the production attribute indicates that the road direction is one-way, then the value of the road direction in the virtual attribute indicates that the road direction is two-way; or, if the value of the road direction in the production attribute indicates that the road direction is two-way, then the value of the road direction in the virtual attribute indicates that the road direction is one-way;

[0071] If the value of the road topology connection relationship in the production attribute indicates that the road topology is not connected, then the value of the road topology connection relationship in the virtual attribute indicates that the road topology is connected; or, if the value of the road topology connection relationship in the production attribute indicates that the road topology is connected, then the value of the road topology connection relationship in the virtual attribute indicates that the road topology is not connected;

[0072] If the value of missing road in the production attribute indicates that there is no road, then the value of missing road in the virtual attribute indicates that there is a road; or, if the value of missing road in the production attribute indicates that there is a road, then the value of missing road in the virtual attribute indicates that there is no road.

[0073] Of course, the embodiments of the present invention are not limited to the several methods listed above.

[0074] In one embodiment, the preset projection probability and transition probability algorithm in step S13 may be based on, for example, a Viterbi algorithm of a Hidden Markov Model (HMM).

[0075] The driving trajectory that meets the quality standard is matched with the roads in the road network that are adjacent to the driving trajectory. According to the production attributes of the roads to be matched, the first matching probability is obtained to determine which roads to be matched are located on which the trajectory points of the driving trajectory that meets the quality standard fall. In other words, the driving trajectory is matched to each road to be matched in the road network. For example, the first projection probability of a trajectory point and the road to be matched can be obtained by the following formula (1):

[0076]

[0077] Among them, Z t For the trajectory points of the driving trajectory that meet the quality standards, r i The trajectory points of the driving trajectory that meet the quality standards are vertically projected to the projection points on the road to be matched, x t,i is the road to be matched, |z t-x t,i | great circle Represents the trajectory point Z t The distance between the road to be matched, σ z is a constant;

[0078] The first transition probability between the to-be-matched road corresponding to one trajectory point and the to-be-matched road corresponding to the next trajectory point is obtained by the following formula (2):

[0079]

[0080] Among them, d t is the trajectory point Z t With the next trajectory point Z t+1 The distance between the trajectory point Z t The corresponding road to be matched x t,i With the next trajectory point Z t+1 The corresponding road to be matched x t+1,j The difference in distance between

[0081] In the above formula (2), β is a constant.

[0082] In one embodiment, in step S13, the first matching probability of the driving trajectory to the road to be matched is obtained based on the first projection probability and the first transition probability of each trajectory point in the driving trajectory to the road to be matched. This can be achieved, for example, by separately determining the product of the first projection probability of each trajectory point in the driving trajectory to the road to be matched, and the product of the first transition probability of each trajectory point in the driving trajectory to the road to be matched, and then fusing the obtained product of the first projection probability and the product of the first transition probability corresponding to the driving trajectory to finally obtain the first matching probability of the driving trajectory to the road to be matched. There are various ways to fuse this, for example, the product of the first projection probability and the product of the first transition probability corresponding to the driving trajectory can be weighted to obtain the maximum weighted probability, i.e., the first matching probability. The weights of the product of the first projection probability and the product of the first transition probability corresponding to the driving trajectory can be selected, for example, based on experience. This is not limited in the present embodiment.

[0083] In one embodiment, in step S14, based on the virtual attributes of the road to be matched, determining the second projection probability and the second transition probability of each trajectory point in the driving trajectory to the road to be matched is specifically achieved by the following process:

[0084] For the driving trajectory, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability are determined for the first trajectory point and the road to be matched. The second projection probability and second transition probability for each trajectory point ordered after the first trajectory point and the road to be matched are determined according to a preset projection probability and transition probability algorithm. The algorithm for calculating the second projection probability and second transition probability for each trajectory point can be the same or similar to the algorithm for calculating the first projection probability and second transition probability in step S13 above. For details, refer to the description of step S13.

[0085] In one embodiment, in steps S13 and S14 above, the values of the production attributes and corresponding virtual attributes of the to-be-matched roads can be continuously determined in real time and incorporated into the corresponding road data during the process of matching the trajectory that meets the quality criteria to the surrounding roads. For example, to obtain the to-be-matched roads within a preset distance range around the driving trajectory, during the road matching process, the values of the introduced production attributes and virtual attributes can be determined in the following possible ways:

[0086] 1. Based on the four-dimensional information and observation time information of each track point of the driving trajectory that meets the quality standards, as well as the road data of the road to be matched, if the track point of the driving trajectory cannot be matched to the road to be matched, a value indicating that there is no road is added to the road data in the production attribute, and a value indicating that there is a road is introduced in the virtual attribute;

[0087] 2. Based on the four-dimensional information and observation time information of each track point of the driving trajectory that meets the quality standards, as well as the road data of the road to be matched, if the track point of the driving trajectory needs to jump to a topologically disconnected road, a value indicating the topological disconnection of the road is added to the road data, and a value indicating the topological connectivity of the road is introduced in the virtual attribute;

[0088] 3. Based on the four-dimensional information and observation time information of each trajectory point of the driving trajectory that meets the quality standards, as well as the road data of the road to be matched, if the trajectory point of the driving trajectory needs to be matched to a reverse road that does not exist in the road network, a value indicating that the road travel direction is one-way is added to the road data in the production attribute, and a value indicating that the road travel direction is two-way is introduced in the virtual attribute.

[0089] Assume that at least one type of road data error exists on the roads in the road network. This means that at least one type of fabricated attribute and virtual attribute is introduced into the roads to be matched. Regardless of the number of fabricated attributes introduced, the road data for each road to be matched is unique within the road network. Therefore, using the acquired driving trajectories that meet the quality criteria and the Viterbi algorithm based on the fabricated attributes of the roads to be matched, the first match probability corresponding to the optimal match result is often unique.

[0090] As for the virtual attributes of the roads to be matched, since the values of the virtual attributes are opposite to the values of the production attributes, and one or more possible virtual attributes may be introduced during the calculation process, it is equivalent to making one or more assumptions about the road attributes of the road data of the roads to be matched in the road network. In this way, one or more possible hypothetical roads corresponding to the roads to be matched under the virtual attributes will be obtained. For convenience, these one or more possible hypothetical roads are referred to as virtual roads to be matched.

[0091] For example: Refer to Figure 3 As shown, the driving trajectory is represented by a shape composed of continuous arrows, and the direction of each arrow is the driving direction of the trajectory point of the driving trajectory; the black solid line is used to represent the road data to be matched in the road network, and the black arrow is used to represent the road travel direction. Assuming that the driving trajectory is correct, it can be assumed that the production attributes of the road data in the road network have one or more of the following errors: road shape error, road direction error, road topology connection error, and road missing error. Figure 3 In the process of road matching, the driving direction of the trajectory points of the driving trajectory is inconsistent with the road traffic direction. Therefore, in the process of road matching, a production attribute value indicating that the road traffic direction is one-way and a virtual attribute value indicating that the road traffic direction is two-way can be introduced for the road data of the road to be matched in the road network; the driving trajectory can turn right at the intersection, and the topological connection relationship in the road data is that the road is not connected. Therefore, in the process of road matching, a production attribute value indicating that the road topology is not connected and a virtual attribute value indicating that the road topology is connected can be introduced for the road data of the road to be matched in the road network; the trajectory points of the driving trajectory cannot be matched to the roads in the road network. Therefore, in the process of road matching, a production attribute value indicating that there is no road and a virtual attribute value indicating that there is a road can be introduced for the road data of the road to be matched in the road network.

[0092] During the road matching process, each virtual attribute value introduced for roads in the road network is equivalent to processing the road data in the road network, resulting in a virtual road corresponding to the road to be matched in the road network. This shows that by introducing one or more virtual attribute values during the road matching process, at least one virtual road corresponding to the road to be matched can be obtained under the virtual attribute conditions. In other words, one or more virtual roads with corrected errors can be hypothesized. Subsequently, the Hidden Markov Model and Viterbi algorithm can be used to infer whether the previous assumptions hold true. If so, this confirms that the original virtual attributes should be correct, and the original road attributes are incorrect.

[0093] Similar to the calculation process of the first matching probability, in the above step S14, in order to obtain the second matching probability between the driving trajectory and the road to be matched under the virtual attributes, the projection probability and transition probability of each trajectory point in the driving trajectory to the at least one virtual road to be matched can be determined respectively, and at least one second projection probability and second transition probability of the driving trajectory to the road to be matched under the virtual attributes can be obtained; then, based on the obtained at least one second projection probability and second transition probability, at least one second matching probability of the driving trajectory to the road to be matched can be obtained.

[0094] In a specific embodiment, since the value of the road to be matched under the virtual attribute condition may be one or more, the road to be matched is processed based on the value of the virtual attribute of each road to be matched, and at least one virtual road to be matched can be obtained. When performing road matching, for each of the at least one virtual roads to be matched, the projection probability and transition probability between the first trajectory point of the driving trajectory and the virtual road to be matched are first determined, and then the projection probability and transition probability of other trajectory points ranked after the first trajectory point and the virtual road to be matched are determined according to a preset projection probability and transition probability algorithm. The projection probability and transition probability of each trajectory point in the driving trajectory to the at least one virtual road to be matched are statistically obtained, which are at least one second projection probability and second transition probability of the driving trajectory to the road to be matched under the virtual attribute condition.

[0095] In a specific embodiment, referring to Figure 4As shown, assume these continuous arrows form a driving trajectory that meets quality standards. Each arrow represents a trajectory point, and the direction of each arrow indicates the driving direction of the trajectory point. During the road matching process, each trajectory point can obtain at least one matching road within a preset distance range. In the prior art, the degree of matching between the driving trajectory and the matching road is reflected by the size of the projection probability and transition probability. The larger the projection probability and transition probability, the smaller the distance between the trajectory point and the road. If the road has no road network errors, the resulting projection probability and transition probability will be larger.

[0096] Assuming that there is no road data error in the road network, there will be many options for matching driving trajectories with roads. Figure 4 It can be seen that when the driving trajectory is finally matched to the road marked with black lines in the road network, the projection probability and transition probability between the driving trajectory and the road to be matched are the largest. The projection probability and transition probability at this time are the first projection probability and the first transition probability, and the road with missing roads that is closest to each trajectory point cannot be matched. This is because, without considering the road errors in the road network, the trajectory is matched to the road marked with black lines. Although the values of the first projection probability and the first transition probability will be relatively small, the fused first matching probability is the maximum probability value obtained by the Viterbi algorithm. That is to say, although it is the optimal matching result, there is still a deviation between the road marked with black lines in the road network and the actual road passed by the driving trajectory, indicating that there is an error in the road data in the road network.

[0097] In an embodiment of the present invention, during the road matching process, if a trajectory point cannot be matched to a road in the road network, it is assumed that the trajectory point corresponds to a missing road. Specifically, a virtual road to be matched is added to the missing location of the road in the map, a hard jump is made for the disconnected road in the road network, the road data is actively repaired, and a value of a production attribute and a virtual attribute is introduced into the road data. At this time, based on the virtual attributes of the road to be matched, the second projection probability and the second transition probability of the trajectory when matched to the virtual road to be matched closest to each trajectory point are both greater than the first projection probability and the first transition probability when matched to a road marked with black lines in the road network. Therefore, the second matching probability obtained by fusing the second projection probability and the second transition probability is greater than the first matching probability, indicating that the matching result when the driving trajectory is matched to the repaired virtual road is better than the matching result when it is matched to a road marked with black lines in the road network, thereby indicating that the production attributes of the roads in the road network are incorrect.

[0098] Based on the same inventive concept, Figure 5 As shown, an embodiment of the present invention further provides a method for determining a road data error type, comprising:

[0099] S51: Determine whether there is an error in the properties of the road to be matched with the driving trajectory;

[0100] If yes, go to step S52, if no, go to step S53;

[0101] S52: Determine the error type of the road to be matched according to the virtual attribute of the road to be matched;

[0102] S53. Exit the current process.

[0103] In the above step S51, the step of determining whether there is an error in the properties of the road to be matched with the driving trajectory can be implemented by using the above method for determining road data errors.

[0104] In the above step S52, the error type of the road to be matched can be determined by the following method:

[0105] comparing and determining a maximum value of at least one second matching probability;

[0106] Determining a value of the virtual attribute of the to-be-matched road corresponding to the maximum value of the second matching probability;

[0107] The error type of the road to be matched is determined according to the determined value of the virtual attribute of the road to be matched.

[0108] During the road matching process, virtual attributes can be introduced into the road data of each road to be matched. When at least one second matching probability is determined to be greater than the first matching probability, it indicates that the attributes of the road to be matched are incorrect. According to the description of the road data error determination method described above, the greater the projection probability and transition probability, the better the match between the driving trajectory and the road. Therefore, the maximum value of the at least one second matching probability determined by comparison corresponds to the optimal matching degree of the road to be matched. Therefore, based on the value of the virtual attribute of the road to be matched corresponding to the maximum second matching probability, the error type of the road to be matched can be determined.

[0109] The method for determining the road data error type provided by the embodiment of the present invention can not only accurately determine whether there is an error in the road data, but also, when the production attributes of the road to be matched are erroneous, further accurately determine the road data error type corresponding to the road to be matched based on the determined values of the virtual attributes of the road to be matched. The method is simple and easy to implement, and it is convenient and efficient to correct road data errors.

[0110] In one embodiment, the method for determining the road data error type further includes: correcting the road network road data based on the determined values of the virtual attributes of the roads to be matched. For example, reverse iteration may be performed based on the determined values of the virtual attributes of the roads to be matched to correct the point coordinate sequence, road travel directions, and road connectivity of the road network road data.

[0111] The road error can be corrected manually, automatically by a machine, or by a combination of manual and automatic methods, which is not limited in the embodiment of the present invention.

[0112] In an embodiment of the present invention, in the process of matching a driving trajectory with a road to be matched within a preset distance range around the driving trajectory, one or more production attributes and virtual attributes can be introduced into each road to be matched to obtain a first matching probability and at least one second matching probability corresponding to the matching result between the trajectory point and the road to be matched. In each matching result, the values of the production attributes and virtual attributes introduced into the road data of the road to be matched are determined. Therefore, based on the production attributes and virtual attributes of the road data, it can be determined whether the road data of the road to be matched is erroneous, and the type of error when an error occurs. When it is determined that the road data is erroneous, the road data in the road network is continuously corrected based on the road data of the road to be matched corresponding to the maximum value of the second matching probability, so that the roads in the corrected road network are closer to the roads in the actual road network in reality.

[0113] Based on the same inventive concept, an embodiment of the present invention also provides a device for determining road data errors, a device for determining road data error types, related storage media and a server. Since the principles of the problems solved by these devices, related storage media and servers are similar to those of the aforementioned road network error determination method, the implementation of the device, related storage media and server can refer to the implementation of the aforementioned method, and the repeated parts will not be repeated.

[0114] Reference Figure 6 As shown, an embodiment of the present invention provides a road network error determination device, comprising:

[0115] a trajectory extraction module 61 for obtaining a driving trajectory represented by a series of trajectory points that meets quality standards, wherein the quality standards include a consistency standard, a volatility standard, and a quantity standard of the trajectory points;

[0116] A road acquisition module 62 is configured to acquire a road to be matched within a preset distance range around the driving trajectory, wherein the road to be matched includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute;

[0117] a first matching probability determination module 63 for determining, based on the production attributes of the road to be matched and in accordance with a preset projection probability and transition probability algorithm, a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtaining a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability;

[0118] a second matching probability determination module 64, configured to determine a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched based on the virtual attributes of the road to be matched, and to obtain a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability;

[0119] The road data error determination module 65 is used to compare the first matching probability obtained by the first matching probability determination module with the second matching probability obtained by the second matching probability determination module, and when the second matching probability is greater than the first matching probability, determine that the production attribute of the road to be matched is erroneous.

[0120] In one embodiment, the second matching probability determination module 64 is specifically used to determine, for the driving trajectory, based on the virtual attributes of the road to be matched, the second projection probability and the second transition probability between the first trajectory point and the road to be matched, and determine the second projection probability and the second transition probability between the trajectory points sorted after the first trajectory point and the road to be matched according to a preset projection probability and transition probability algorithm.

[0121] In one embodiment, the second matching probability determination module 64 is specifically configured to determine, based on the virtual attributes of the road to be matched, at least one virtual road to be matched corresponding to the road to be matched under the virtual attributes;

[0122] Determining respectively a projection probability and a transition probability of each trajectory point in the driving trajectory to the at least one virtual road to be matched, and obtaining at least one second projection probability and at least one second transition probability of each trajectory point in the driving trajectory to the road to be matched under the virtual attribute condition;

[0123] At least one second matching probability of the driving trajectory to the road to be matched is obtained based on the obtained at least one second projection probability and the at least one second transition probability.

[0124] In one embodiment, the second matching probability determination module 64 is specifically configured to compare the at least one second matching probability with the first matching probability;

[0125] If any one of the at least one second matching probability is greater than the first matching probability, it is determined that the production attribute of the road to be matched is erroneous.

[0126] Reference Figure 7 As shown, an embodiment of the present invention provides a device for determining a road data error type, comprising:

[0127] A data error judgment module 71 is used to determine whether there is an error in the properties of the road to be matched in the driving trajectory;

[0128] A type error determination module 72 is configured to determine the error type of the to-be-matched road according to the virtual attributes of the to-be-matched road when an error occurs in the properties of the to-be-matched road in the driving trajectory;

[0129] The data error judgment module 71 uses the above-mentioned road data error determination method to determine whether there is an error in the production attributes of the to-be-matched road of the driving trajectory.

[0130] An embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon. When the instructions are executed by a processor, the above-mentioned road network error determination method or the above-mentioned road data error type determination method is implemented.

[0131] An embodiment of the present invention provides a server comprising: a processor and a memory for storing processor-executable commands; wherein the processor is configured to execute the above-mentioned road network error determination method, or execute the above-mentioned road data error type determination method.

[0132] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0133] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0136] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for determining road data errors, comprising: Acquire a driving trajectory represented by a series of trajectory points that meets quality standards, wherein the quality standards include a consistency standard, a fluctuation standard, and a quantity standard of the trajectory points; Acquire a to-be-matched road within a preset distance range around the driving trajectory, wherein the to-be-matched road includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute; Based on the production attributes of the road to be matched, determining a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched according to a preset projection probability and transition probability algorithm, and obtaining a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability; Determining, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtaining a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability; When the second matching probability of the driving trajectory to the road to be matched is greater than the first matching probability, it is determined that the production attribute of the road to be matched is wrong.

2. The method according to claim 1, wherein determining the second projection probability and the second transition probability of each trajectory point in the driving trajectory to the road to be matched based on the virtual attribute of the road to be matched comprises: For the driving trajectory, based on the virtual attributes of the road to be matched, the second projection probability and the second transition probability of the first trajectory point and the road to be matched are determined, and the second projection probability and the second transition probability of the trajectory points sorted after the first trajectory point and the road to be matched are determined according to a preset projection probability and transition probability algorithm.

3. The method according to claim 1 or 2, wherein the production attributes and virtual attributes include: A combination of one or more of road traffic direction, road topological connection relationship, and road missing.

4. The method of claim 3, further comprising determining a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched based on the virtual attributes of the road to be matched, and obtaining a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability, comprising: Determining, based on the virtual attributes of the road to be matched, at least one virtual road to be matched corresponding to the road to be matched under the virtual attributes; Determining respectively a projection probability and a transition probability of each trajectory point in the driving trajectory to the at least one virtual road to be matched, and obtaining at least one second projection probability and a second transition probability of the driving trajectory to the road to be matched under the virtual attributes; At least one second matching probability of the vehicle trajectory to the road to be matched is obtained according to the obtained at least one second projection probability and the second transition probability.

5. The method according to claim 4, wherein when the second matching probability of the driving trajectory to the road to be matched is greater than the first matching probability, determining that the production attribute of the road to be matched is incorrect comprises: comparing the at least one second match probability to the first match probability; If any one of the at least one second matching probability is greater than the first matching probability, it is determined that the production attribute of the road to be matched is erroneous.

6. A method for determining a road data error type, comprising: Determine whether there is an error in the properties of the road to be matched with the driving trajectory; If so, determining the error type of the road to be matched according to the virtual attribute of the road to be matched; The step of determining whether there is an error in the production attributes of the road to be matched with the driving trajectory adopts the road data error determination method according to any one of claims 1 to 5.

7. The method according to claim 6, wherein determining the error type of the road to be matched according to the virtual attribute of the road to be matched comprises: comparing and determining a maximum value of at least one second matching probability; Determining a value of the virtual attribute of the to-be-matched road corresponding to the maximum value of the second matching probability; The error type of the road to be matched is determined according to the determined value of the virtual attribute of the road to be matched.

8. A device for determining road data errors, comprising: a trajectory extraction module, configured to obtain a driving trajectory represented by a series of trajectory points that meets quality standards, wherein the quality standards include a consistency standard, a volatility standard, and a quantity standard of the trajectory points; A road acquisition module is used to acquire a road to be matched within a preset distance range around the driving trajectory, wherein the road to be matched includes a production attribute and a virtual attribute, and the value of the virtual attribute is opposite to the value of the production attribute; a first matching probability determination module, configured to determine, based on the production attributes of the road to be matched and in accordance with a preset projection probability and transition probability algorithm, a first projection probability and a first transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtain a first matching probability of the driving trajectory to the road to be matched based on the first projection probability and the first transition probability; a second matching probability determination module, which determines, based on the virtual attributes of the road to be matched, a second projection probability and a second transition probability of each trajectory point in the driving trajectory to the road to be matched, and obtains a second matching probability of the driving trajectory to the road to be matched based on the second projection probability and the second transition probability; The road data error determination module is used to compare the first matching probability obtained by the first matching probability determination module with the second matching probability obtained by the second matching probability determination module, and when the second matching probability is greater than the first matching probability, determine that the production attribute of the road to be matched is erroneous.

9. A device for determining a road data error type, comprising: A data error judgment module is used to determine whether there is an error in the properties of the road to be matched in the driving trajectory; a type error determination module, configured to determine the error type of the to-be-matched road according to the virtual attributes of the to-be-matched road when an error occurs in the production attributes of the to-be-matched road of the driving trajectory; The step of determining whether there is an error in the production attributes of the road to be matched with the driving trajectory adopts the road data error determination method according to any one of claims 1 to 5.

10. A computer-readable storage medium having computer instructions stored thereon, wherein when the instructions are executed by a processor, the method for determining a road data error according to any one of claims 1 to 5 is implemented, or the method for determining a road data error type according to claim 6 or 7 is implemented.

11. A server comprising: a processor, a memory for storing instructions executable by the processor; The processor is configured to execute the method for determining a road data error according to any one of claims 1 to 5, or to execute the method for determining a road data error type according to claim 6 or 7.

Images