Data screening method and device and computer readable storage medium

Abstract

The embodiment of the invention discloses a data screening method and device and a computer readable storage medium. The data screening method comprises the steps that one or more pieces of track data and feature data corresponding to the track data are acquired; track points included in the track data are matched with generated road data, and road sections covered by each piece of track data are determined; based on a road section covered by the trajectory data, establishing an observation matrix for representing a coverage relationship between the trajectory data and the road section; based on feature data corresponding to the trajectory data, determining a quality evaluation matrix of the trajectory data; and screening the trajectory data according to the observation matrix and the quality evaluation matrix. According to the track data screened by the technical scheme, the data processing amount can be reduced, the coverage degree of high-precision map construction can be ensured, the quality of collected data can be ensured, the production efficiency and quality of a high-precision map can be effectively improved, and the mapping or updating requirements of the high-precision map can be met.

Description

Technical Field

[0001] The present disclosure relates to the technical field of collected data processing, and in particular to a data screening method, device and computer-readable storage medium. Background Art

[0002] With the development of technology, high-precision maps (also called high-standard or high-quality maps) with more accurate location information and richer road elements have gradually become the basic data for various location-based service systems such as autonomous driving and intelligent driving. High-precision maps are usually made based on road data collected by acquisition equipment. Common acquisition equipment includes: LiDAR, visual sensors (such as monocular cameras), positioning sensors, inertial navigation, etc. Among them, the acquisition solution based on LiDAR has limited the development of high-precision maps to a certain extent due to its high acquisition cost.

[0003] In order to reduce the cost of collection and improve the coverage and freshness of high-precision maps, more and more manufacturers have begun to develop vision-based collection solutions. The advantages of vision-based collection solutions are: visual sensors are low-cost and widely used. Vision-based collection solutions can be deployed on a large scale, which can significantly improve the frequency and coverage of data collection. However, more devices also mean that the collected data is very large (up to a massive level). If all the data is processed, it will take a lot of time, which will directly affect the efficiency of high-precision map production. In addition, vision-based collection solutions require not only visual sensors, but also positioning sensors and other sensors that can generate location points. Due to differences in sensor performance, the quality of data collected by vision-based collection solutions varies. Processing all data will also affect the quality of high-precision maps. Therefore, for vision-based collection solutions, it is necessary to provide a data screening solution that can not only reduce the amount of data processing, but also meet the quality requirements of high-precision maps. Summary of the invention

[0004] Embodiments of the present disclosure provide a data screening method, device, and computer-readable storage medium.

[0005] In a first aspect, a data screening method is provided in an embodiment of the present disclosure.

[0006] Specifically, the data screening method includes:

[0007] Obtaining one or more trajectory data and feature data corresponding to the trajectory data;

[0008] Matching the trajectory points included in the trajectory data with the generated road data to determine the road section covered by each trajectory data;

[0009] Based on the road sections covered by the trajectory data, establishing an observation matrix for characterizing the coverage relationship between the trajectory data and the road sections;

[0010] Determining a quality evaluation matrix of the trajectory data based on feature data corresponding to the trajectory data;

[0011] The trajectory data is screened according to the observation matrix and the quality evaluation matrix.

[0012] In a second aspect, a data screening device is provided in an embodiment of the present disclosure.

[0013] Specifically, the data screening device includes:

[0014] An acquisition module, configured to acquire one or more trajectory data and feature data corresponding to the trajectory data;

[0015] A determination module, configured to match the trajectory points included in the trajectory data with the generated road data to determine the road section covered by each trajectory data;

[0016] An establishing module configured to establish an observation matrix for characterizing a coverage relationship between trajectory data and road sections based on the road sections covered by the trajectory data;

[0017] An evaluation module, configured to determine a quality evaluation matrix of the trajectory data based on feature data corresponding to the trajectory data;

[0018] The screening module is configured to screen the trajectory data according to the observation matrix and the quality evaluation matrix.

[0019] In a third aspect, an embodiment of the present disclosure provides an electronic device comprising a memory and at least one processor, wherein the memory is used to store one or more computer instructions, and wherein the one or more computer instructions are executed by the at least one processor to implement the above-mentioned data screening method.

[0020] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium for storing computer instructions used by a data screening device, which includes computer instructions involved in the data screening device for executing the above-mentioned data screening method.

[0021] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, including a computer program / instructions, wherein the computer program / instructions implement the above-mentioned data screening method when executed by a processor.

[0022] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0023] The above technical solution screens the trajectory data by determining an observation matrix for characterizing the coverage relationship between the trajectory data and the road section, and a quality evaluation matrix for characterizing the quality of trajectory data collection. Since the present disclosure takes into account the coverage factor and quality factor of the trajectory data for the road section, although the number of trajectories screened out according to the technical solution is reduced, the screened trajectory data can not only ensure the coverage of the road section, but also have good quality. Therefore, it can effectively improve the efficiency and quality of high-precision map production and meet the needs of high-precision map construction or update.

[0024] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Other features, objectives and advantages of the present disclosure will become more apparent through the following detailed description of non-limiting embodiments in conjunction with the accompanying drawings. In the accompanying drawings:

[0026] Figure 1 A flow chart showing a data screening method according to an embodiment of the present disclosure;

[0027] Figure 2 A schematic diagram showing a method of determining a road section completely covered by road collection trajectory data according to an embodiment of the present disclosure;

[0028] Figure 3 A structural block diagram of a data screening device according to an embodiment of the present disclosure is shown;

[0029] Figure 4 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown;

[0030] Figure 5 It is a structural schematic diagram of a computer system suitable for implementing a data screening method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for the sake of clarity, parts not related to the description of the exemplary embodiments are omitted in the accompanying drawings.

[0032] In the present disclosure, it should be understood that terms such as "include" or "have" are intended to indicate the presence of features, numbers, steps, behaviors, components, parts, or a combination thereof disclosed in the present specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, behaviors, components, parts, or a combination thereof exist or are added.

[0033] It should also be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0034] The technical solution provided by the embodiment of the present disclosure screens the trajectory data by determining an observation matrix for characterizing the coverage relationship between the trajectory data and the road section, and a quality evaluation matrix for characterizing the quality of trajectory data collection. Since the technical solution takes into account both the coverage factor and the quality factor of the collected data, the trajectory data screened according to the technical solution can reduce the amount of data processing and ensure the coverage of high-precision map construction, and can also ensure the quality of the collected data, thereby effectively improving the efficiency and quality of high-precision map production and meeting the needs of high-precision map construction or updating.

[0035] Figure 1 A flow chart showing a data screening method according to an embodiment of the present disclosure is shown as follows: Figure 1 As shown, the data screening method includes the following steps S101-S105:

[0036] In step S101, one or more trajectory data and feature data corresponding to the trajectory data are obtained;

[0037] In step S102, the trajectory points included in the trajectory data are matched with the generated road data to determine the road section covered by each trajectory data;

[0038] In step S103, based on the road sections covered by the trajectory data, an observation matrix for characterizing the coverage relationship between the trajectory data and the road sections is established;

[0039] In step S104, a quality evaluation matrix of the trajectory data is determined based on the feature data corresponding to the trajectory data;

[0040] In step S105, the trajectory data is screened according to the observation matrix and the quality evaluation matrix.

[0041] As mentioned above, with the development of technology, high-precision maps (also called high-standard or high-quality maps) with more accurate location information and richer road elements have gradually become the basic data for various location-based service systems such as autonomous driving and intelligent driving. High-precision maps are usually produced based on road data collected by acquisition equipment. Common acquisition equipment includes: LiDAR, visual sensors (such as monocular cameras), positioning sensors, inertial navigation, etc. Among them, the acquisition solution based on LiDAR has limited the development of high-precision maps to a certain extent due to its high acquisition cost. In order to reduce the acquisition cost and improve the coverage and freshness of high-precision maps, more and more manufacturers have begun to develop vision-based acquisition solutions. The advantages of vision-based acquisition solutions are: visual sensors are low in cost and widely used. Vision-based acquisition solutions can be deployed on a large scale, which can significantly improve the frequency and coverage of data collection. However, more equipment also means that the collected data is very large (up to massive levels). If all the data is processed, it will take a lot of time, which directly affects the efficiency of high-precision map production. In addition, vision-based acquisition solutions require not only visual sensors, but also positioning sensors and other sensors that can generate location points. Due to differences in sensor performance, the quality of data collected by vision-based acquisition solutions varies greatly, and processing all data will also affect the quality of high-precision maps. Therefore, for vision-based acquisition solutions, a data screening solution is needed to reduce the amount of data processing and meet the quality requirements of high-precision maps.

[0042] In view of the above defects, this embodiment provides a data screening method, which screens the trajectory data by determining an observation matrix used to characterize the coverage relationship between trajectory data and road sections, and a quality evaluation matrix used to characterize the quality of trajectory data collection. Since this technical solution takes into account both the coverage factor and the quality factor of the collected data, the trajectory data screened according to this technical solution can reduce the amount of data processing and ensure the coverage of high-precision map construction, and can also ensure the quality of the collected data, thereby effectively improving the efficiency and quality of high-precision map production and meeting the needs of high-precision map construction or updating.

[0043] In one embodiment of the present disclosure, the data screening method may be applicable to computers, computing devices, electronic devices, servers, service clusters, etc. that screen trajectory data.

[0044] In one embodiment of the present disclosure, the trajectory data refers to the road collection trajectory data collected by a transportation vehicle equipped with a collection device during road driving, wherein each road collection trajectory data includes multiple trajectory points. Based on the trajectory data and image data collected by the collection device, one or more of the following feature data corresponding to the trajectory data can be extracted: trajectory quality, mapping vector, image features, etc., wherein the trajectory quality can be characterized by trajectory positioning reliability, and the image features can be, for example, road elements and other features appearing in the image data.

[0045] In one embodiment of the present disclosure, the coverage means that a certain road section is completely covered by the corresponding road collection trajectory data. The road section that is not completely covered by the road collection trajectory data may affect the quality of the high-precision map. In order to ensure the quality of the high-precision map, the present disclosure preferably considers the road section that is completely covered by the road collection trajectory data. The meaning of complete coverage is that the road collection trajectory data covers the road section from beginning to end. A road collection trajectory data may completely cover one or more road sections.

[0046] In one embodiment of the present disclosure, the observation matrix refers to a matrix used to characterize the coverage relationship between the road collection trajectory data and the road sections, that is, the observation matrix characterizes the correspondence between each road collection trajectory data and the road sections completely covered by the road collection trajectory data.

[0047] In one embodiment of the present disclosure, the quality evaluation matrix represents the corresponding relationship between the road collection trajectory data and its quality evaluation value, and the quality evaluation value of the road collection trajectory data is determined based on the feature data corresponding to the road collection trajectory data.

[0048] In the above implementation, firstly, more than one road collection trajectory data and its corresponding feature data are obtained, then the trajectory points included in the road collection trajectory data are matched with the generated road data, and the road sections completely covered by each road collection trajectory data are determined, and then based on the road sections covered by the trajectory data, an observation matrix for characterizing the coverage relationship between the road collection trajectory data and the road sections is established, wherein the generated road data can be obtained in advance from ordinary map data, which at least records the starting point, the end point and the corresponding road section identification data of each road section, and the trajectory points included in the road collection trajectory data are matched with the road data to determine whether the road collection trajectory data completely covers a certain road section; and, based on the feature data corresponding to the road collection trajectory data, the quality evaluation value of the road collection trajectory data is determined, and then a quality evaluation matrix for characterizing the road collection trajectory data and its corresponding quality evaluation value is obtained; finally, according to the observation matrix and the quality evaluation matrix, the road collection trajectory data is screened to obtain road collection trajectory data that can cover more road sections and has guaranteed quality from the road collection trajectory data.

[0049] In one embodiment of the present disclosure, if the number of the road collection trajectory data is n, and the road sections covered by the n road collection trajectory data are m, then the step S103, i.e., the step of establishing an observation matrix for characterizing the coverage relationship between the trajectory data and the road sections based on the road sections covered by the trajectory data, may include the following steps:

[0050] Construct an observation matrix A with m rows and n columns m×n ;

[0051] If the i-th road section among the m road sections is covered by the j-th track data among the n track data, then the observation matrix A m×n The element value at the i-th row and j-th column is set to 1, otherwise it is set to 0.

[0052] That is, if the number of the road collection trajectory data is n, and the n road collection trajectory data cover m road sections, when establishing an observation matrix for representing the coverage relationship between the trajectory data and the road sections, first construct an observation matrix A with m rows and n columns: m×n Then, according to the coverage relationship between the road collection trajectory data and its fully covered road section, the observation matrix A is set m×n Specifically, if the i-th road section in the m road sections is completely covered by the j-th trajectory data in the n trajectory data, then the observation matrix A m×n The element value of the i-th row and j-th column is set to 1, otherwise it is set to 0. It can be seen that the observation matrix A m×nIt represents the coverage relationship between the road collection trajectory data and the road section. Therefore, the observation matrix A m×n By filtering the road collection trajectory data as an observation element, it is possible to ensure the coverage of the road section by the filtered road collection trajectory data.

[0053] Furthermore, in one embodiment of the present disclosure, when determining a road section that is completely covered by road collection trajectory data, in order to improve processing efficiency, it can be determined based on the generated road data whether trajectory points included in the road collection trajectory data exist before the starting point of a road section and after the end point of the road section along the direction of the road collection trajectory data. If so, it can be determined that the road collection trajectory data completely covers the road section; if not, it can be determined that the road collection trajectory data does not completely cover the road section.

[0054] Figure 2 A schematic diagram of determining a road section completely covered by road collection trajectory data according to an embodiment of the present disclosure is shown. Assuming that a certain road collection trajectory data includes 10 trajectory points, the road collection trajectory data passes through 4 road sections: section 1, section 2, section 3 and section 4. Figure 2 As shown, the direction of collecting trajectory data on the road is from west to east, trajectory point 1 and trajectory point 2 fall within the range of section 1, trajectory point 3 and trajectory point 4 fall within the range of section 2, trajectory point 5 falls between section 2 and section 3, trajectory point 6 and trajectory point 7 fall within the range of section 3, trajectory point 8 and trajectory point 9 fall within the range of section 4, and trajectory point 10 falls outside section 4. Therefore, along the direction of collecting trajectory data on the road, it can be determined that there is no trajectory point before the starting point of section 1, there is trajectory point 3 after the end of section 1, and there is trajectory point 4 before the starting point of section 2. Track point 2, there is a track point 5 after the end of section 2; there is a track point 5 before the starting point of section 3, and there is a track point 8 after the end of section 3; there is a track point 7 before the starting point of section 4, and there is a track point 10 after the end of section 4. That is to say, along the direction of the road trajectory data collection, there are track points before the starting point and after the end point of section 2, section 3 and section 4, while there is a track point only after the end point of section 1. Therefore, section 2, section 3 and section 4 are sections that are completely covered by the road trajectory data collection, but section 1 is not.

[0055] In one embodiment of the present disclosure, continuing with the previous example, if there are n pieces of trajectory data, then the step S104, i.e., the step of determining the quality evaluation matrix of the trajectory data based on the feature data corresponding to the trajectory data, may include the following steps:

[0056] Determine a quality evaluation value of the trajectory data according to one or more of trajectory quality, mapping vector, and image features included in feature data corresponding to the trajectory data, where the quality evaluation value is between 0 and 1;

[0057] Construct a quality evaluation matrix Q with n rows and 1 column n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data.

[0058] As mentioned above, the feature data corresponding to the road collection trajectory data may include one or more of trajectory quality, mapping vector, and image features. Therefore, in this embodiment, when determining the quality evaluation matrix of the trajectory data based on the feature data corresponding to the trajectory data, first, the quality evaluation value of the trajectory data is determined according to one or more of the trajectory quality, mapping vector, and image features included in the feature data corresponding to the trajectory data. In order to facilitate the subsequent screening of the road collection trajectory data, the quality evaluation value may be set between 0 and 1; then, a quality evaluation matrix Q having n rows and 1 column is constructed. n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data. It can be seen that the quality evaluation matrix Q n×1 It can characterize the collection quality of road collection trajectory data. Therefore, the quality evaluation matrix Q n×1 As an evaluation element, screening the road collection trajectory data can ensure the quality of the screened road collection trajectory data.

[0059] Among them, when evaluating the collection quality of the road collection trajectory data according to the feature data corresponding to the road collection trajectory data, one or more of the feature data corresponding to the road collection trajectory data can be used according to the needs of the actual application to evaluate the collection quality of the road collection trajectory data, or multiple evaluation methods can be used to evaluate the collection quality of the road collection trajectory data. The present disclosure does not specifically limit the specific method of collection quality evaluation. For example, the collection quality of the road collection trajectory data can be evaluated based on any one of the feature data corresponding to the road collection trajectory data; or the collection quality of the road collection trajectory data can be comprehensively evaluated based on two or more of the feature data corresponding to the road collection trajectory data. Specifically, corresponding weights can be set for two or more feature data, and then the collection quality of the road collection trajectory data can be evaluated based on the feature data to obtain the quality evaluation value corresponding to the feature data. Finally, the quality evaluation values ​​corresponding to different feature data are weighted and summed to obtain the comprehensive collection quality evaluation value of the road collection trajectory data.

[0060] In one embodiment of the present disclosure, continuing with the previous example, if there are n pieces of trajectory data and the number of covered road sections is m, then the step S105, i.e., the step of screening the trajectory data according to the observation matrix and the quality evaluation matrix, may include the following steps:

[0061] Construct a trajectory data screening matrix X with n rows and 1 column n×1 ;

[0062] According to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , construct the screening objective function;

[0063] Solve the screening objective function to obtain the trajectory data screening matrix X n×1 The value of each row element, where the trajectory data screening matrix X n×1 The element value of the i-th row element of is 1, indicating that the trajectory data is selected, and the element value is 0, indicating that the trajectory data is not selected.

[0064] In this embodiment, in order to achieve the purpose of screening the road collection trajectory data, firstly, a road collection trajectory data screening matrix X with n rows and 1 column is constructed. n×1 , where the trajectory data screening matrix X n×1 The i-th row element of represents whether the i-th road collection trajectory data among the n road collection trajectory data is selected. If so, the element value is set to 1, otherwise the element value is set to 0. It can be seen that the trajectory data screening matrix X n×1 is the quantity to be determined, i.e. the unknown quantity; according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , construct a screening objective function; then, solve the screening objective function to obtain the trajectory data screening matrix X that meets the requirements of the objective function n×1 , and the screening results of the road collection trajectory data are obtained. Since the above screening of the road collection trajectory data takes into account both the coverage of the trajectory data on the road sections and the quality of the trajectory data, the screened road collection trajectory data can cover as many road sections as possible and have guaranteed quality, which can effectively improve the efficiency and quality of high-precision map production and meet the needs of high-precision map construction or update.

[0065] Furthermore, according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , the step of constructing the screening objective function may include the following steps:

[0066] According to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 ;

[0067] According to the observation matrix A m×n , quality evaluation matrix Q n×1 , trajectory data screening matrix X n×1 , the minimum expected number b and the relaxation matrix ξ m×1 , construct the screening objective function.

[0068] According to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , when constructing the screening objective function, in order to avoid the situation where the number of times some sections are completely covered does not reach the predetermined value and the screening objective function has no solution, in this embodiment, a relaxation matrix corresponding to the completely covered sections is also set to compensate for the gap between the number of times the sections are completely covered and the predetermined value. That is, according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 When constructing the screening objective function, first obtain the minimum expected number of times b that each road section is completely covered, wherein the minimum expected number b can be set according to the needs of the actual application, for example, it can be set to 4; then according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the fully covered section m×1 ; Finally, according to the observation matrix A m×n , quality evaluation element matrix Q n×1 , trajectory data screening matrix X n×1 , the minimum expected number b and the relaxation matrix ξ m×1 , and jointly construct the screening objective function.

[0069] Furthermore, according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 The steps may include the following steps:

[0070] Construct a relaxation matrix ξ with m rows and 1 column m×1 , where the relaxation matrix ξ m×1 The i-th row element of represents the relaxation amount corresponding to the i-th road segment among all covered road segments;

[0071] When the actual number of times c that the i-th road section is covered is less than the minimum expected number b, the relaxation amount corresponding to the road section is set to the minimum expected number b minus the actual number of times c that the road section is covered, where the actual number of times c that the i-th road section is covered is calculated based on the observation matrix A. m×n Sure;

[0072] When the actual number of times c that the i-th road section is covered is greater than or equal to the minimum expected number b, the relaxation amount corresponding to the road section is set to 0.

[0073] As mentioned above, the relaxation matrix is ​​used to compensate for the gap between the number of times a road section is completely covered and a predetermined value. Therefore, in this embodiment, according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the fully covered section m×1 When , we first construct a relaxation matrix ξ with m rows and 1 column. m×1 , where the relaxation matrix ξ m×1 The i-th row element of represents the relaxation amount corresponding to the i-th road segment among all covered road segments. Then, according to the observation matrix A m×n Determine the actual number of times c that each road section among all the completely covered road sections is completely covered; then determine whether the actual number of times c that a certain road section is completely covered is less than the minimum expected number b. If it is less than, it means that the gap between the number of times the road section is completely covered and the predetermined value needs to be compensated, and the value obtained by subtracting the actual number of times c that the road section is completely covered from the minimum expected number b can be used as the relaxation amount corresponding to the road section. If it is greater than or equal to, it means that there is no need to compensate for the gap between the number of times the road section is completely covered and the predetermined value, and the relaxation amount corresponding to the road section can be set to 0; finally, fill the relaxation matrix ξ according to the relaxation amounts corresponding to all the completely covered road sections obtained by the above method. m×1 The corresponding elements of the relaxation matrix ξ are obtained m×1 .

[0074] Furthermore, the screening objective function can be expressed as:

[0075] Minimize(1-Q) T X,

[0076] So that AX+ξ≥b·1,

[0077] Among them, Minimize is the minimum function; 1 is a matrix of size n×1, with all element values ​​1; Q is the quality evaluation matrix, which represents the evaluation element; (·) T represents the transpose operation; X is the screening matrix; A is the observation matrix, representing the observation element; ξ is the relaxation matrix; b is the minimum expected number of times each road section is covered.

[0078] Among them, the constraint condition of AX+ξ≥b·1 can ensure that each road section is fully covered a sufficient number of times, thereby ensuring the coverage of the road section by the selected trajectory data; Minimize(1-Q) T X can ensure that the quality of the selected road collection trajectory data is high enough, thereby ensuring the quality of the selected trajectory data.

[0079] The following are embodiments of the apparatus of the present disclosure, which can be used to execute embodiments of the method of the present disclosure.

[0080] Figure 3 FIG. 1 is a block diagram showing a data screening device according to an embodiment of the present disclosure. The device can be implemented as part or all of an electronic device through software, hardware, or a combination of both. Figure 3 As shown, the data screening device comprises:

[0081] An acquisition module 301 is configured to acquire one or more trajectory data and feature data corresponding to the trajectory data;

[0082] A determination module 302 is configured to match the trajectory points included in the trajectory data with the generated road data to determine the road section covered by each trajectory data;

[0083] An establishing module 303 is configured to establish an observation matrix for characterizing a coverage relationship between trajectory data and road sections based on the road sections covered by the trajectory data;

[0084] An evaluation module 304 is configured to determine a quality evaluation matrix of the trajectory data based on feature data corresponding to the trajectory data;

[0085] The screening module 305 is configured to screen the trajectory data according to the observation matrix and the quality evaluation matrix.

[0086] In one embodiment of the present disclosure, if the number of the road collection trajectory data is n, and the n road collection trajectory data cover m road sections, the establishment module 303 may be configured as follows:

[0087] Construct an observation matrix A with m rows and n columns m×n ;

[0088] If the i-th road section among the m road sections is covered by the j-th track data among the n track data, then the observation matrix A m×n The element value at the i-th row and j-th column is set to 1, otherwise it is set to 0.

[0089] That is, if the number of the road collection trajectory data is n, and the n road collection trajectory data cover m road sections, when establishing an observation matrix for representing the coverage relationship between the trajectory data and the road sections, first construct an observation matrix A with m rows and n columns: m×n Then, according to the coverage relationship between the road collection trajectory data and its fully covered road section, the observation matrix A is set m×n Specifically, if the i-th road section in the m road sections is completely covered by the j-th trajectory data in the n trajectory data, then the observation matrix A m×n The element value of the i-th row and j-th column is set to 1, otherwise it is set to 0. It can be seen that the observation matrix A m×n It represents the coverage relationship between the road collection trajectory data and the road section. Therefore, the observation matrix A m×n By filtering the road collection trajectory data as an observation element, it is possible to ensure the coverage of the road section by the filtered road collection trajectory data.

[0090] Furthermore, in one embodiment of the present disclosure, when determining a road section that is completely covered by road collection trajectory data, the determination module 302 may be configured to: based on the generated road data, determine whether trajectory points included in the road collection trajectory data exist before the starting point of a road section and after the end point of the road section along the direction of the road collection trajectory data; if so, it can be determined that the road collection trajectory data completely covers the road section; if not, it can be determined that the road collection trajectory data does not completely cover the road section.

[0091] In one embodiment of the present disclosure, continuing with the previous example, if there are n pieces of trajectory data, the evaluation module 304 may be configured as follows:

[0092] Determine a quality evaluation value of the trajectory data according to one or more of trajectory quality, mapping vector, and image features included in feature data corresponding to the trajectory data, where the quality evaluation value is between 0 and 1;

[0093] Construct a quality evaluation matrix Q with n rows and 1 column n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data.

[0094] As mentioned above, the feature data corresponding to the road collection trajectory data may include one or more of trajectory quality, mapping vector, and image features. Therefore, in this embodiment, when determining the quality evaluation matrix of the trajectory data based on the feature data corresponding to the trajectory data, first, the quality evaluation value of the trajectory data is determined according to one or more of the trajectory quality, mapping vector, and image features included in the feature data corresponding to the trajectory data. In order to facilitate the subsequent screening of the road collection trajectory data, the quality evaluation value may be set between 0 and 1; then, a quality evaluation matrix Q having n rows and 1 column is constructed. n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data. It can be seen that the quality evaluation matrix Q n×1 It can characterize the collection quality of road collection trajectory data. Therefore, the quality evaluation matrix Q n×1 As an evaluation element, screening the road collection trajectory data can ensure the quality of the screened road collection trajectory data.

[0095] Among them, when evaluating the collection quality of the road collection trajectory data according to the feature data corresponding to the road collection trajectory data, one or more of the feature data corresponding to the road collection trajectory data can be used according to the needs of the actual application to evaluate the collection quality of the road collection trajectory data, or multiple evaluation methods can be used to evaluate the collection quality of the road collection trajectory data. The present disclosure does not specifically limit the specific method of collection quality evaluation. For example, the collection quality of the road collection trajectory data can be evaluated based on any one of the feature data corresponding to the road collection trajectory data; or the collection quality of the road collection trajectory data can be comprehensively evaluated based on two or more of the feature data corresponding to the road collection trajectory data. Specifically, corresponding weights can be set for two or more feature data, and then the collection quality of the road collection trajectory data can be evaluated based on the feature data to obtain the quality evaluation value corresponding to the feature data. Finally, the quality evaluation values ​​corresponding to different feature data are weighted summed to obtain the comprehensive collection quality evaluation value of the road collection trajectory data.

[0096] In one embodiment of the present disclosure, continuing with the previous example, if there are n pieces of trajectory data and m covered road sections, the screening module 305 may be configured as follows:

[0097] Construct a trajectory data screening matrix X with n rows and 1 column n×1 ;

[0098] According to the observation matrix A m×n , quality evaluation matrix Q n×1And trajectory data screening matrix X n×1 , construct the screening objective function;

[0099] Solve the screening objective function to obtain the trajectory data screening matrix X n×1 The value of each row element, where the trajectory data screening matrix X n×1 The element value of the i-th row element of is 1, indicating that the trajectory data is selected, and the element value is 0, indicating that the trajectory data is not selected.

[0100] In this embodiment, in order to achieve the purpose of screening the road collection trajectory data, firstly, a road collection trajectory data screening matrix X with n rows and 1 column is constructed. n×1 , where the trajectory data screening matrix X n×1 The i-th row element of represents whether the i-th road collection trajectory data among the n road collection trajectory data is selected. If so, the element value is set to 1, otherwise the element value is set to 0. It can be seen that the trajectory data screening matrix X n×1 is the quantity to be determined, i.e. the unknown quantity; according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , construct a screening objective function; then, solve the screening objective function to obtain the trajectory data screening matrix X that meets the requirements of the objective function n×1 , and the screening results of the road collection trajectory data are obtained. Since the above screening of the road collection trajectory data takes into account both the coverage of the trajectory data on the road sections and the quality of the trajectory data, the screened road collection trajectory data can cover as many road sections as possible and have guaranteed quality, which can effectively improve the efficiency and quality of high-precision map production and meet the needs of high-precision map construction or update.

[0101] Furthermore, according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , the part that builds the screening objective function can be configured as:

[0102] According to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 ;

[0103] According to the observation matrix A m×n , quality evaluation matrix Q n×1 , trajectory data screening matrix X n×1 , the minimum expected number b and the relaxation matrix ξ m×1 , construct the screening objective function.

[0104] According to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , when constructing the screening objective function, in order to avoid the situation where the number of times some sections are completely covered does not reach the predetermined value and the screening objective function has no solution, in this embodiment, a relaxation matrix corresponding to the completely covered sections is also set to compensate for the gap between the number of times the sections are completely covered and the predetermined value. That is, according to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 When constructing the screening objective function, first obtain the minimum expected number of times b that each road section is completely covered, wherein the minimum expected number b can be set according to the needs of the actual application, for example, it can be set to 4; then according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the fully covered section m×1 ; Finally, according to the observation matrix A m×n , quality evaluation element matrix Q n×1 , trajectory data screening matrix X n×1 , the minimum expected number b and the relaxation matrix ξ m×1 , and jointly construct the screening objective function.

[0105] Furthermore, according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 The section can be configured as:

[0106] Construct a relaxation matrix ξ with m rows and 1 column m×1 , where the relaxation matrix ξ m×1 The i-th row element of represents the relaxation amount corresponding to the i-th road segment among all covered road segments;

[0107] When the actual number of times c that the i-th road section is covered is less than the minimum expected number b, the relaxation amount corresponding to the road section is set to the minimum expected number b minus the actual number of times c that the road section is covered, where the actual number of times c that the i-th road section is covered is calculated based on the observation matrix A. m×n Sure;

[0108] When the actual number of times c that the i-th road section is covered is greater than or equal to the minimum expected number b, the relaxation amount corresponding to the road section is set to 0.

[0109] As mentioned above, the relaxation matrix is ​​used to compensate for the gap between the number of times a road section is completely covered and a predetermined value. Therefore, in this embodiment, according to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the fully covered section m×1 When , we first construct a relaxation matrix ξ with m rows and 1 column. m×1 , where the relaxation matrix ξ m×1 The i-th row element of represents the relaxation amount corresponding to the i-th road segment among all covered road segments. Then, according to the observation matrix A m×n Determine the actual number of times c that each road section among all the completely covered road sections is completely covered; then determine whether the actual number of times c that a certain road section is completely covered is less than the minimum expected number b. If it is less than, it means that the gap between the number of times the road section is completely covered and the predetermined value needs to be compensated, and the value obtained by subtracting the actual number of times c that the road section is completely covered from the minimum expected number b can be used as the relaxation amount corresponding to the road section. If it is greater than or equal to, it means that there is no need to compensate for the gap between the number of times the road section is completely covered and the predetermined value, and the relaxation amount corresponding to the road section can be set to 0; finally, fill the relaxation matrix ξ according to the relaxation amounts corresponding to all the completely covered road sections obtained in the above manner m×1 The corresponding elements of the relaxation matrix ξ are obtained m×1 .

[0110] Furthermore, the screening objective function can be expressed as:

[0111] Minimize(1-Q) T X,

[0112] So that AX+ξ≥b·1,

[0113] Among them, Minimize is the minimum function; 1 is a matrix of size n×1, with all element values ​​1; Q is the quality evaluation matrix, which represents the evaluation element; (·) T represents the transpose operation; X is the screening matrix; A is the observation matrix, representing the observation element; ξ is the relaxation matrix; b is the minimum expected number of times each road section is covered.

[0114] Among them, the constraint condition of AX+ξ≥b·1 can ensure that each road section is fully covered a sufficient number of times, thereby ensuring the coverage of the road section by the selected trajectory data; Minimize(1-Q) T X can ensure that the quality of the selected road collection trajectory data is high enough, thereby ensuring the quality of the selected trajectory data.

[0115] The present disclosure also discloses an electronic device, Figure 4A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown. Figure 4 As shown, the electronic device 400 includes a memory 401 and a processor 402; wherein,

[0116] The memory 401 is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor 402 to implement the above method steps.

[0117] Figure 5 It is a structural schematic diagram of a computer system suitable for implementing a data screening method according to an embodiment of the present disclosure.

[0118] like Figure 5 As shown, the computer system 500 includes a processing unit 501, which can perform various processes in the above-mentioned embodiments according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the computer system 500 are also stored. The processing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0119] The following components are connected to the I / O interface 505: an input part 506 including a keyboard, a mouse, etc.; an output part 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage part 508 including a hard disk, etc.; and a communication part 509 including a network interface card such as a LAN card, a modem, etc. The communication part 509 performs communication processing via a network such as the Internet. The drive 510 is also connected to the I / O interface 505 as needed. Removable media 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., are installed on the drive 510 as needed, so that the computer program read therefrom is installed into the storage part 508 as needed. Among them, the processing unit 501 can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, an NPU, etc.

[0120] In particular, according to an embodiment of the present disclosure, the method described above can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program tangibly contained on a readable medium thereof, and the computer program includes a program code for executing the method. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 509, and / or installed from a removable medium 511.

[0121] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the road map or block diagram can represent a part of a module, program segment or code, and the part of the module, program segment or code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a different order from the order marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0122] The units or modules involved in the embodiments described in the present disclosure may be implemented by software or hardware. The units or modules described may also be set in a processor, and the names of these units or modules do not constitute limitations on the units or modules themselves in some cases.

[0123] As another aspect, the present disclosure further provides a computer-readable storage medium, which may be a computer-readable storage medium included in the device described in the above embodiment; or a computer-readable storage medium that exists independently and is not assembled into the device. The computer-readable storage medium stores one or more programs, and the programs are used by one or more processors to execute the method described in the present disclosure.

[0124] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other.

Claims

1. A data screening method, wherein: include: Obtaining one or more trajectory data and feature data corresponding to the trajectory data; Matching the trajectory points included in the trajectory data with the generated road data to determine the road section covered by each trajectory data; Based on the road sections covered by the trajectory data, establishing an observation matrix for characterizing the coverage relationship between the trajectory data and the road sections; Determining a quality evaluation matrix of the trajectory data based on feature data corresponding to the trajectory data; The trajectory data is screened according to the observation matrix and the quality evaluation matrix.

2. The method according to claim 1, wherein: If the number of trajectory data is n and the number of road sections covered by the n trajectory data is m, then based on the road sections covered by the trajectory data, an observation matrix for characterizing the coverage relationship between the trajectory data and the road sections is established, including: Construct an observation matrix A with m rows and n columns m×n ; If the i-th road section among the m road sections is covered by the j-th track data among the n track data, then the observation matrix A m×n The element value at the i-th row and j-th column is set to 1, otherwise it is set to 0.

3. The method according to claim 2, wherein: The determining, based on the feature data corresponding to the trajectory data, a quality evaluation matrix of the trajectory data comprises: Determine a quality evaluation value of the trajectory data according to one or more of trajectory quality, mapping vector, and image features included in feature data corresponding to the trajectory data, where the quality evaluation value is between 0 and 1; Construct a quality evaluation matrix Q with n rows and 1 column n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data.

4. The method according to claim 3, wherein: The screening of the trajectory data according to the observation matrix and the quality evaluation matrix includes: Construct a trajectory data screening matrix X with n rows and 1 column n×1 ; According to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , construct the screening objective function; Solve the screening objective function to obtain the trajectory data screening matrix X n×1 The value of each row element, where the trajectory data screening matrix X n×1 The element value of the i-th row element of is 1, indicating that the trajectory data is selected, and the element value is 0, indicating that the trajectory data is not selected.

5. The method according to claim 4, wherein: According to the observation matrix A m×n , quality evaluation matrix Q n×1 And trajectory data screening matrix X n×1 , construct the screening objective function, including: According to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 ; According to the observation matrix A m×n , quality evaluation matrix Q n×1 , trajectory data screening matrix X n×1 , the minimum expected number b and the relaxation matrix ξ m×1 , construct the screening objective function.

6. The method according to claim 5, wherein: According to the observation matrix A m×n and the preset minimum expected number b, determine the relaxation matrix ξ corresponding to the covered section m×1 ,include: Construct a relaxation matrix ξ with m rows and 1 column m×1 , where the relaxation matrix ξ m×1 The i-th row element of represents the relaxation amount corresponding to the i-th road segment among all covered road segments; When the actual number of times c that the i-th road section is covered is less than the minimum expected number b, the relaxation amount corresponding to the road section is set to the minimum expected number b minus the actual number of times c that the road section is covered, where the actual number of times c that the i-th road section is covered is calculated based on the observation matrix A. m×n Sure; When the actual number of times c that the i-th road section is covered is greater than or equal to the minimum expected number b, the relaxation amount corresponding to the road section is set to 0.

7. A data screening device, comprising: An acquisition module, configured to acquire one or more trajectory data and feature data corresponding to the trajectory data; A determination module, configured to match the trajectory points included in the trajectory data with the generated road data to determine the road section covered by each trajectory data; An establishing module configured to establish an observation matrix for characterizing a coverage relationship between trajectory data and road sections based on the road sections covered by the trajectory data; An evaluation module, configured to determine a quality evaluation matrix of the trajectory data based on feature data corresponding to the trajectory data; The screening module is configured to screen the trajectory data according to the observation matrix and the quality evaluation matrix.

8. The device according to claim 7, wherein: If the number of trajectory data is n and the number of road sections covered by the n trajectory data is m, the establishment module is configured as follows: Construct an observation matrix A with m rows and n columns m×n ; If the i-th road section among the m road sections is covered by the j-th track data among the n track data, then the observation matrix A m×n The element value at the i-th row and j-th column is set to 1, otherwise it is set to 0.

9. The device according to claim 8, wherein: The evaluation module is configured to: Determine a quality evaluation value of the trajectory data according to one or more of trajectory quality, mapping vector, and image features included in feature data corresponding to the trajectory data, where the quality evaluation value is between 0 and 1; Construct a quality evaluation matrix Q with n rows and 1 column n×1 , where the quality evaluation matrix Q n×1 The i-th row element of represents the quality evaluation value of the i-th trajectory data among the n trajectory data.

10. A computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions, when executed by a processor, implement the method steps described in any one of claims 1 to 6.

Images