Abnormal road semantic recognition method, device, equipment and storage medium

Abstract

The present application relates to an abnormal road semantics recognition method, device, equipment and storage medium. The abnormal road semantics recognition method includes: obtaining the position information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and the regional position information of other areas on the visible surface of the trajectory where the driving trajectory is located except for the semantics to be recognized; constructing a regional grid map covering the semantics to be recognized based on the position information of the semantics to be recognized and the corresponding visible surface of the trajectory; determining the first distribution of grids covered by the semantics to be recognized in the regional grid map; determining the second distribution of grids not covered by the semantics to be recognized in the regional grid map; and determining whether the semantics to be recognized is an abnormal semantic based on the first distribution and the corresponding second distribution corresponding to each target vehicle. In the above steps, the accuracy of abnormal road semantics recognition is improved, the operation is convenient and quick, the amount of calculation is small, and the recognition efficiency is high.

Description

Technical Field

[0001] The present application relates to the field of map technology, and in particular to a method, device, equipment and storage medium for abnormal road semantic recognition. Background Art

[0002] Road maps play an important role in the navigation or path planning process of mobile devices. The generation of road maps is inseparable from the collection and analysis of semantic data on the road.

[0003] In traditional technology, the semantics to be recognized collected by vehicles are usually filtered, denoised, and processed, and the processed semantics to be recognized are used as the basis for generating road maps.

[0004] However, due to the influence of semantic data collection accuracy and environmental interference, the above method still cannot accurately identify abnormal semantics even if the semantics to be recognized are filtered and denoised, which affects the accuracy of the subsequently generated road map. Summary of the invention

[0005] Based on this, it is necessary to provide an abnormal road semantic recognition method, device, equipment and storage medium that can improve the accuracy of abnormal road semantic recognition in order to address the above technical problems.

[0006] In a first aspect, the present application provides a method for identifying abnormal road semantics, comprising:

[0007] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0008] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0009] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0010] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0011] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0012] In one of the embodiments, whether the semantics to be identified is abnormal semantics is determined based on the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, including: determining the semantically occupied grid of the semantics to be identified based on the first distribution situation corresponding to each target vehicle; determining the semantically unoccupied grid of the semantics to be identified based on the second distribution situation corresponding to each target vehicle; determining whether the semantics to be identified is abnormal semantics based on the grid distribution of the semantically occupied grids in the regional grid map and the grid distribution of the semantically unoccupied grids.

[0013] In one of the embodiments, semantically unoccupied grids of the semantics to be identified are determined based on the second distribution situation corresponding to each target vehicle, including: initializing the grid weights of each grid in the regional grid map; for each target vehicle, accumulating a first preset weight for the grid weights of the grids covered by the semantics to be identified based on the first distribution situation corresponding to the target vehicle, so as to update the grid weights of the corresponding grids; and, for each target vehicle, accumulating a second preset weight for the grid weights of the grids not covered by the semantics to be identified based on the second distribution situation corresponding to the target vehicle, so as to update the grid weights of the corresponding grids; based on the updated grid weights of each grid in the regional grid map, the semantically unoccupied grids of the semantics to be identified are determined; wherein the second preset weight has an opposite numerical sign to the first preset weight.

[0014] In one of the embodiments, semantically unoccupied grids of the semantics to be identified are determined based on the updated grid weights of each grid in the regional grid map, including: determining semantically unoccupied grids of the semantics to be identified based on the size relationship between the updated grid weights of each grid in the regional grid map and a preset grid weight threshold; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0015] In one of the embodiments, whether the semantics to be identified is abnormal semantics is determined based on the grid distribution of semantically occupied grids and the grid distribution of semantically unoccupied grids in a regional grid map, including: taking the grid corresponding to the intersection of the semantically occupied grid and the semantically unoccupied grid as the abnormally occupied grid of the semantics to be identified; determining the semantic weight of the semantics to be identified based on the proportion of the abnormally occupied grids in the semantically occupied grids; and determining whether the semantics to be identified is abnormal semantics based on the relationship between the semantic weight and a preset semantic weight threshold.

[0016] In one of the embodiments, a regional grid map covering the semantics to be identified is constructed based on the position information of the semantics to be identified and the corresponding trajectory visible surface, including: obtaining trajectory point data of the trajectory of the target vehicle; determining the semantic height at which the semantics to be identified is located based on the position information of the semantics to be identified; and determining the trajectory height at which the semantics to be identified is located based on the trajectory point data corresponding to the semantics to be identified; when the semantic height and the trajectory height are at the same height level, a regional grid map covering the semantics to be identified is constructed based on the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0017] In one of the embodiments, a regional grid map covering the semantics to be identified is constructed based on the position information of the semantics to be identified and the corresponding trajectory visible surface, including: determining the visible surface boundary position of the corresponding trajectory visible surface based on the trajectory point position of the target vehicle; constructing a regional map covering the semantics to be identified and the visible surface corresponding to the visible surface boundary position in a preset positive direction based on the position information of the semantics to be identified and the visible surface boundary position; and expanding the regional map according to a preset grid size to obtain a regional grid map.

[0018] In a second aspect, the present application also provides an abnormal road semantic recognition device, comprising:

[0019] An acquisition module, used to acquire the position information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and the area position information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0020] A construction module, used to construct a regional grid map covering the semantics to be identified based on the location information of the semantics to be identified and the corresponding trajectory visible surface;

[0021] A first determination module, used to determine a first distribution of grids covered by the semantics to be identified in the regional grid map according to the location information of the semantics to be identified;

[0022] A second determination module is used to determine a second distribution of semantically uncovered grids to be identified in the regional grid map according to the regional location information;

[0023] The third determination module is used to determine whether the semantics to be identified is abnormal semantics according to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle.

[0024] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0025] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0026] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0027] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0028] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0029] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0030] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0031] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0032] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0033] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0034] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0035] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0036] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0037] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0038] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0039] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0040] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0041] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0042] The above-mentioned abnormal road semantics recognition method, device, equipment and storage medium provide an original basis for abnormal recognition of the semantics to be recognized by constructing a regional grid map covering the semantics to be recognized, and introducing the first distribution of the grid covered by the semantics to be recognized in the regional grid map corresponding to each target vehicle and the second distribution of the grid not covered by the semantics to be recognized. The original basis combines the first distribution and the second distribution corresponding to each target vehicle. Since the semantics to be recognized of each target vehicle corresponds to the same road semantics in the real space, the distribution of the semantics to be recognized of each target vehicle in the regional grid map should be relatively concentrated or similar, and then it can be determined whether the semantics to be recognized is abnormal, thereby improving the accuracy and reliability of abnormal semantic recognition. In addition, in the process of abnormal semantic recognition, only the first distribution of the grid covered by the semantics to be recognized in the regional grid map corresponding to each target vehicle and the second distribution of the grid not covered by the semantics to be recognized are needed, without the need to use other data, the operation is convenient and fast, the amount of calculation is small, and the recognition efficiency is high. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A schematic diagram of a flow chart of an abnormal road semantic recognition method in one embodiment;

[0045] Figure 2 1 is a flow chart of the steps of determining abnormal semantics in one embodiment;

[0046] Figure 3 is a flow chart of a step of determining a semantically unoccupied grid in one embodiment;

[0047] Figure 4 A schematic diagram of a flow chart of steps for constructing a regional grid map in one embodiment;

[0048] Figure 5A A schematic diagram of a flow chart of an abnormal road semantic recognition method in another embodiment;

[0049] Figure 5B is a schematic diagram of an abnormal semantic elimination process in one embodiment;

[0050] Figure 6 is a structural block diagram of an abnormal road semantic recognition device in one embodiment;

[0051] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] In an exemplary embodiment, Figure 1 As shown, a method for abnormal road semantic recognition is provided. This embodiment takes the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0054] S110, obtaining position information of semantics to be recognized when at least one target vehicle is traveling on a road to be recognized, and area position information of other areas on a visible surface of the traveling trajectory except for the semantics to be recognized.

[0055] Among them, the semantics to be identified can be understood as feature information used to describe the road to be identified and / or the surrounding environment. Exemplarily, the semantics to be identified may include at least one of lane lines and guide signs. Among them, the guide signs may include guide arrows, etc. The location information of the semantics to be identified may include at least one of the longitude, latitude and altitude information corresponding to the semantics to be identified. In order to ensure the accuracy of the recognition result, the above-mentioned location information usually includes three-dimensional location information. Among them, the semantics to be identified is collected or identified by the target vehicle during its driving on the road to be identified; exemplarily, the semantics to be identified can be collected or identified by the sensor system of the target vehicle, and this application does not impose any limitation on the specific implementation of the sensor system.

[0056] The semantics to be recognized can be face-type semantics to be recognized or non-face-type semantics to be recognized. For non-face-type semantics to be recognized, such as one-dimensional semantics to be recognized corresponding to linear semantics to be recognized, in order to ensure the accuracy of the recognition result of subsequent abnormal semantic recognition, the non-face-type semantics to be recognized can also be converted into two-dimensional face-type semantics to be recognized.

[0057] The driving trajectory can be understood as the path taken by the target vehicle when driving on the road to be identified. In an optional embodiment, the driving trajectory can be generated by sorting the trajectory point data collected by the target vehicle in chronological order. The trajectory point data may include at least one of the longitude, latitude and altitude corresponding to the trajectory point. In order to ensure the accuracy of the recognition result, the above-mentioned trajectory point data usually includes three-dimensional position information.

[0058] Among them, the trajectory visible surface can be understood as the trajectory surface where the driving trajectory of the target vehicle is located. In an optional embodiment, the trajectory visible surface can be generated by expanding the surrounding area of ​​the driving trajectory. Exemplarily, the trajectory visible surface can be generated by expanding the buffer of a preset distance on both sides of the trajectory of the driving trajectory. The preset distance can be set by the technician according to needs or experience, or determined through a large number of experiments. This application does not impose any restrictions on this. For example, the preset distance can be 1m-5m, and in particular, the preset distance is 3m. It is worth noting that during the expansion process, the preset distances extended in different directions may be the same or different. It is not difficult to understand that the trajectory visible surface includes areas where there are semantics to be identified and other areas except for the semantics to be identified. The regional location information of other areas refers to at least one of the corresponding longitude, latitude and altitude information of other areas. In order to ensure the accuracy of the recognition result, the above-mentioned regional location information usually includes three-dimensional location information.

[0059] S120 , constructing a regional grid map covering the semantics to be recognized based on the location information of the semantics to be recognized and the corresponding track visible surface.

[0060] The grid size of the regional grid map can be set by technicians according to needs or experience, or determined repeatedly through a large number of experiments, and this application does not impose any limitation on this. In a specific example, the grid size can be 0.2m×0.2m.

[0061] In an optional embodiment, based on the position information of the semantics to be identified and the corresponding trajectory visible surface, constructing a regional grid map covering the semantics to be identified may include the following steps: determining the visible surface boundary position of the corresponding trajectory visible surface based on the trajectory point position of the target vehicle; constructing a regional map covering the semantics to be identified and the visible surface corresponding to the visible surface boundary position in a preset positive direction based on the position information of the semantics to be identified and the visible surface boundary position; and expanding the regional map according to a preset grid size to obtain a regional grid map.

[0062] The position of the visible surface boundary can be understood as the position of each boundary point of the visible surface boundary. The preset positive direction can be set according to actual needs, such as the true north direction.

[0063] It is understandable that in the process of dividing the regional map according to the preset grid size, there may be some areas in the regional map that cannot be allocated grids of the preset grid size. At this time, the regional map is expanded and the grids that are less than the preset grid size are filled according to the preset grid size, so that all grids can meet the preset grid size requirements, which is convenient for the subsequent determination of the distribution of the semantics to be identified.

[0064] In another optional implementation, the location information of the semantics to be identified and the trajectory point data corresponding to the semantics to be identified can be grouped according to the height level; the location information of the semantics to be identified and the corresponding trajectory visible surface at the same height level are used to construct a regional grid map covering the semantics to be identified. By grouping according to the height level, when constructing the regional grid map, the interference of the location information and corresponding trajectory point data of the semantics to be identified with large height differences (i.e., other height levels) (such as the influence of data of roads at different height levels in elevated roads) is avoided, and the regional grid maps are constructed at different height levels, which is convenient for the subsequent identification of abnormal semantics.

[0065] S130 . Determine, according to the location information of the semantics to be recognized, a first distribution situation of grids covered by the semantics to be recognized in the regional grid map.

[0066] The first distribution situation may include distribution information of grids covered by the semantics to be identified, and the distribution information may include at least one of information such as the numbers of the covered grids and the positions of the covered grids.

[0067] In an optional embodiment, the grid containing the position information of the semantics to be recognized can be used as the grid covered by the semantics to be recognized by data processing, and the first distribution information of the grid covered by the semantics to be recognized can be recorded. In another optional embodiment, the grid covered by the image of the semantics to be recognized can be identified by image processing, and the first distribution information of the grid covered by the semantics to be recognized can be recorded.

[0068] In another optional embodiment, the coverage area of ​​the semantics to be identified covered by each grid in the regional grid map can be determined, and the grids with a coverage area greater than a first preset area threshold can be used as grids covered by the semantics to be identified. The first preset area threshold can be set according to actual needs. Exemplarily, the first preset area threshold can be a preset value, for example, an area value (such as 0.02m 2 ), or the grid area percentage (e.g. 5%).

[0069] It is not difficult to understand that, taking the preset coefficient as 0 as an example, when the coverage area is greater than 0, the grid is considered to be covered by the semantics to be identified, and the first distribution of grids covered by the semantics to be identified can be obtained by recording grids with coverage areas greater than 0. In order to reduce the amount of subsequent calculations, the preset coefficient can be increased in combination with the actual situation. For example, the preset coefficient is set to 3%. When the coverage area ratio is greater than 3%, the grid is considered to be covered by the semantics to be identified, and when the coverage area ratio is ≤3%, it is considered that the semantics to be identified is not covered. This can reduce the amount of calculation for subsequent abnormal identification while meeting the recognition accuracy.

[0070] In yet another optional embodiment, grids not covered by the semantics to be recognized in the regional grid map may be determined, and grids other than the grids not covered by the semantics to be recognized in the regional grid map may be used as grids covered by the semantics to be recognized.

[0071] S140: Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map.

[0072] The second distribution situation may include distribution information of grids not covered by the semantics to be identified, and the distribution information may include at least one of grid numbers of grids not covered by the semantics to be identified and positions of covered grids.

[0073] In an optional embodiment, the grid that does not contain the position information of the semantics to be recognized can be regarded as the grid not covered by the semantics to be recognized by data processing, and the second distribution information of the grid not covered by the semantics to be recognized can be recorded. Alternatively, the grid that does not cover the image of the semantics to be recognized can be identified by image processing, and the second distribution information of the grid not covered by the semantics to be recognized can be recorded.

[0074] In another optional embodiment, the coverage area of ​​the semantics to be identified that is not covered by each grid in the regional grid map can be determined, and the grids whose coverage area is not greater than the second preset area threshold are used as the semantics not covered by the semantics to be identified. The second preset area threshold can be the same as or different from the first preset area threshold, and this application does not make any limitation on this. For example, it can be set according to actual needs. Exemplarily, the second preset area threshold can be a preset value, for example, it can be an area value (such as 0.02m 2 ), or the grid area percentage (e.g. 5%).

[0075] In an optional embodiment, grids covered by the semantics to be recognized in the regional grid map may be determined, and grids other than the grids covered by the semantics to be recognized in the regional grid map may be regarded as grids not covered by the semantics to be recognized.

[0076] S150. Determine whether the semantics to be identified is abnormal semantics according to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle.

[0077] It can be understood that since the semantics to be identified of each target vehicle corresponds to the same road semantics in the real space, the distribution of the semantics to be identified of each target vehicle in the regional grid map should be relatively concentrated or similar, so as to determine whether the semantics to be identified are abnormal.

[0078] The embodiment of the present application provides an original basis for abnormal recognition of the semantics to be recognized by constructing a regional grid map covering the semantics to be recognized, and introducing the first distribution of the grids covered by the semantics to be recognized and the second distribution of the grids not covered by the semantics to be recognized in the regional grid map corresponding to each target vehicle. The original basis combines the first distribution and the second distribution corresponding to each target vehicle. Since the semantics to be recognized of each target vehicle corresponds to the same road semantics in the real space, the distribution of the semantics to be recognized of each target vehicle in the regional grid map should be relatively concentrated or similar, so that it can be determined whether the semantics to be recognized is abnormal, thereby improving the accuracy and reliability of abnormal semantic recognition. In addition, in the process of abnormal semantic recognition, it is only necessary to use the first distribution of the grids covered by the semantics to be recognized and the second distribution of the grids not covered by the semantics to be recognized in the regional grid map corresponding to each target vehicle, without the need to use other data, and the operation is convenient and fast, the amount of calculation is small, and the recognition efficiency is high.

[0079] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment. In this optional embodiment, the step of determining the abnormal semantics of step S150 is refined.

[0080] See also Figure 2 The steps for determining the semantics of the exception shown include:

[0081] S210 . Determine a semantic occupancy grid of the semantics to be recognized according to the first distribution situation corresponding to each target vehicle.

[0082] The semantic occupancy grid represents the grid covered by the semantics to be identified.

[0083] S220 . Determine semantically unoccupied grids of the semantics to be recognized according to the second distribution conditions corresponding to each target vehicle.

[0084] In an optional embodiment, the semantically unoccupied grid can be obtained by the following steps: determining the uncovered times of each target grid in the regional grid map according to the second distribution situation corresponding to the target vehicle; and determining the semantically unoccupied grid according to the uncovered times of each target grid. The uncovered times refers to the number of semantically unoccupied grids to be identified whose grid attributes correspond to the target grids in each target vehicle.

[0085] Exemplarily, a times threshold may be set, and when the number of times the target grid is not covered is greater than the set times threshold, the target grid is determined as a semantically unoccupied grid.

[0086] Exemplarily, based on the first preset function, the uncovered confidence weight of the corresponding target grid can be determined according to the number of uncovered times; and the semantically unoccupied grid can be determined according to the uncovered confidence weight of each target grid. The first preset function is an increasing function with the number of uncovered times as the independent variable and the uncovered confidence weight as the dependent variable. The uncovered confidence weight of the target grid can be understood as the numerical quantification result of the credibility of the uncovered semantics to be identified in the target grid, which is used to characterize the probability that the actual spatial area corresponding to the target grid does not contain the semantics to be identified.

[0087] In another optional embodiment, weights may be set for grids not covered by the semantics to be identified based on the second distribution corresponding to the target vehicle, and the semantically unoccupied grids of the semantics to be identified may be determined based on the accumulated weights of each grid after update in the regional grid map.

[0088] S230 , determining whether the semantics to be identified is abnormal semantics according to the grid distribution of semantically occupied grids and the grid distribution of semantically unoccupied grids in the regional grid map.

[0089] In an optional embodiment, abnormal semantics can be determined by the following steps: taking the grid corresponding to the intersection of the semantic occupied grid and the semantic unoccupied grid as the abnormal occupied grid of the semantics to be identified; determining the semantic weight of the semantics to be identified based on the proportion of the abnormal occupied grid in the semantic occupied grid; and determining whether the semantics to be identified is abnormal semantics based on the relationship between the semantic weight and a preset semantic weight threshold.

[0090] Exemplarily, the semantic weight of the semantics to be identified can be calculated by the following formula:

[0091] ;

[0092] Among them, w represents the semantic weight of the semantics to be recognized; Represents a collection of semantic occupancy grids; Represents a collection of semantically unoccupied grids; represents the number of semantic occupancy grids in the semantic occupancy grid set; Indicates the number of abnormal occupancy grids in the abnormal occupancy grid collection.

[0093] The preset semantic weight threshold can be set according to actual needs. For example, the preset semantic weight threshold can be set to any value between 0.5 and 0.7. For example, the preset semantic weight threshold can be specifically set to 0.5. For example, when the semantic weight is less than the preset semantic weight threshold, the semantics to be identified is determined to be abnormal semantics; when the semantic weight is not less than the preset semantic weight threshold, the semantics to be identified is determined to be normal semantics.

[0094] The embodiment of the present application determines the semantically occupied grid of the semantics to be identified and the semantically unoccupied grid of the semantics to be identified, and introduces the grid distribution of the semantically occupied grid and the grid distribution of the semantically unoccupied grid in the regional grid map, so as to determine whether the semantics to be identified is an abnormal semantics.

[0095] On the basis of the technical solutions of the above-mentioned embodiments, the present application further provides an optional embodiment. In this optional embodiment, the step of determining the semantically unoccupied grid in step S220 is refined.

[0096] See also Figure 3 The steps for determining the semantically unoccupied grid include:

[0097] S310: Initialize the grid weight of each grid in the regional grid map.

[0098] The grid weight is the numerical quantification result of the coverage credibility of the semantics to be identified in the grid, which is used to characterize the probability of whether the actual spatial area corresponding to the grid contains the semantics to be identified. Initializing the grid weight means resetting the grid weight to the initial value, which can be set according to actual needs, for example, setting the initial value to 0.

[0099] S320. For each target vehicle, according to the first distribution situation corresponding to the target vehicle, a first preset weight is accumulated for the grid weight of the grid covered by the semantics to be identified, so as to update the grid weight of the corresponding grid.

[0100] Exemplarily, for each target vehicle, the covered grid is determined according to the first distribution corresponding to the target vehicle; for any covered grid, the cumulative number of grids covered by the semantics to be identified is counted; and the grid weight of the corresponding grid is updated according to the product of the cumulative number and the first preset weight.

[0101] S330. For each target vehicle, according to the second distribution situation corresponding to the target vehicle, a second preset weight is accumulated for the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid.

[0102] Exemplarily, for each target vehicle, the uncovered grids are determined according to the second distribution corresponding to the target vehicle; for any uncovered grid, the cumulative number of semantically uncovered grids to be identified is counted; and the grid weight of the corresponding grid is updated according to the product of the cumulative number and the second preset weight.

[0103] S340 determines the semantically unoccupied grids of the semantics to be identified according to the updated grid weights of the grids in the regional grid map.

[0104] The second preset weight has a numerical sign opposite to that of the first preset weight. The numerical values ​​of the first preset weight and the second preset weight can be set according to actual needs.

[0105] In an optional embodiment, the semantically unoccupied grids of the semantics to be identified are determined based on the size relationship between the updated grid weights of each grid in the regional grid map and a preset grid weight threshold; wherein the preset grid weight threshold has the same numerical sign as the second preset weight, or the preset grid weight threshold is set to 0.

[0106] Among them, when the first preset weight is a positive value, the second preset weight is a negative value, and the preset grid weight threshold is a negative value or 0, the grid whose updated grid weight is less than the preset grid weight threshold is determined as a semantically unoccupied grid.

[0107] Among them, when the first preset weight is a negative value, the second preset weight is a positive value, and the preset grid weight threshold is a positive value or 0, the grid whose updated grid weight is greater than the preset grid weight threshold is determined to be a semantically unoccupied grid.

[0108] For the sake of convenience, the constructed 3*3 regional grid map is taken as an example for explanation, and the first preset weight is set to +0.7, the second preset weight is set to -0.4, and the preset grid weight threshold is 0. The first distribution and the second distribution corresponding to the target vehicle A1-target vehicle A3 are shown in Table-1:

[0109] Table-1 The first distribution and the second distribution

[0110] Target vehicle The first distribution Second distribution <![CDATA[A1]]> <![CDATA[G 11 、G 12 、G 13 ]]> <![CDATA[G 21 、G 22 、G 23 、G 31 、G 32 、G 33 ]]> <![CDATA[A2]]> <![CDATA[G 11 、G 12 ]]> <![CDATA[G 13 、G 21 、G 22 、G 23 、G 31 、G 32 、G 33 ]]> <![CDATA[A3]]> <![CDATA[G 11 、G 12 ]]> <![CDATA[G 13 、G 21 、G 22 、G 23 、G 31 、G 32 、G 33 ]]>

[0111] Each grid G ​​in the regional grid map is pre-set ij The grid weight of the grid G ​​(where 1≤i≤3, 1≤j≤3, and i and j are integers) is set to 0. As shown in Table 1, the grid G 11 , G 12 The grid covered by the grid to be identified semantics is accumulated three times with the first preset weight, and the grid weight is updated to "+2.1"; grid G 13 The grid covered by the semantics to be identified accumulates the first preset weight once, and the grid not covered by the semantics to be identified accumulates the second preset weight twice, and the grid weight is updated to "-0.1". Similarly, the grid G 21 , G 22 , G 23 , G 31 , G 32 , G 33 The grid weight of G is updated to "-1.2". Since the preset grid weight threshold is 0, and -1.2<0, -0.1<0, the grid G 13 , G 21 , G 22 , G 23 , G 31 , G 32 , G 33 Determined as semantically unoccupied raster.

[0112] The embodiment of the present application introduces a first preset weight and a second preset weight respectively based on two dimensions, namely, the grid covered by the semantics to be identified and the grid not covered by the semantics to be identified, so as to quantify the carrying conditions of the semantics to be identified corresponding to different target vehicles and improve the accuracy of the quantification results. Thus, according to the updated grid weight, the semantically unoccupied grid of the semantics to be identified is determined, and the determination process is more convenient and has a small amount of data calculation, thereby improving the determination efficiency of the semantically unoccupied grid and the accuracy of the determination results.

[0113] In another optional embodiment, the grid weight of the corresponding grid can also be determined only by the second distribution. The step of determining the semantically unoccupied grid can include: initializing the grid weight of each grid in the regional grid map; for each target vehicle, according to the second distribution corresponding to the target vehicle, accumulating the second preset weight for the grid weight of the semantically uncovered grid to be identified, so as to determine the grid weight of the corresponding grid; according to the updated grid weights of each grid in the regional grid map, determining the semantically unoccupied grid of the semantics to be identified.

[0114] On the basis of the technical solutions of the above-mentioned embodiments, the present application further provides an optional embodiment. In this optional embodiment, the step of constructing the regional grid map in step S120 is refined.

[0115] See also Figure 4 The steps for constructing the raster map of the area shown include:

[0116] S410: Acquire the track point data of the target vehicle's travel track.

[0117] S420: Determine the semantic height of the semantics to be recognized according to the position information of the semantics to be recognized.

[0118] In an optional embodiment, the semantic height of the semantics to be recognized can be the height values ​​of each position point corresponding to the semantics to be recognized. In another optional embodiment, the height value of each position point can be obtained from the position information of the semantics to be recognized, and the average height of the height values ​​of each position point can be calculated and used as the semantic height of the semantics to be recognized. In yet another optional embodiment, the polygon vertices of the semantics to be recognized can be obtained from the position information of the semantics to be recognized, and the average height of each polygon vertices can be used as the semantic height of the semantics to be recognized.

[0119] Exemplarily, the average height of each polygon vertex is calculated by the following formula:

[0120] ;

[0121] in, Represents the average height of each polygon vertex; Represents the set of vertices of each polygon; represents the number of polygon vertices; pz represents the height of the polygon vertex p corresponding to the semantics to be recognized.

[0122] S430 , determining the trajectory height of the semantics to be recognized according to the trajectory point data corresponding to the semantics to be recognized.

[0123] In an optional embodiment, the track height of the semantics to be recognized may be the median of the height values ​​of the track points corresponding to the semantics to be recognized. In another optional embodiment, the track height of the semantics to be recognized may be the average of the height values ​​of the track points corresponding to the semantics to be recognized.

[0124] S440: When the semantic height and the trajectory height are at the same height level, a regional grid map covering the semantics to be identified is constructed according to the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0125] Optionally, the height level refers to different height intervals divided from the height dimension according to certain standards.

[0126] Exemplarily, the height level of the semantic height can be obtained by dividing the semantic height by a preset value and rounding down to obtain the height level of the semantic height. Similarly, the height level of the track height can also be obtained by dividing the track height by a preset value and rounding down to obtain the height level of the track height. The preset value can be set according to actual needs, for example, the preset value is set to 5.

[0127] Or optionally, the height level may also be a height interval within a neighborhood range to which a preset height belongs.

[0128] Exemplarily, when the difference between the semantic height and the trajectory height is less than a height threshold, it can be determined that the semantic height and the trajectory height are at the same height level.

[0129] In an optional embodiment, when the semantic height and the trajectory height are at different height levels, the position information of the semantics to be identified and the trajectory point data corresponding to the semantics to be identified are grouped according to the height levels they belong to, and a regional grid map covering the semantics to be identified is constructed only based on the position information of the semantics to be identified and the corresponding trajectory visible surface at the same height level.

[0130] In the embodiment of the present application, by introducing height levels, when constructing a regional grid map, interference from the location information and corresponding trajectory point data of the semantics to be identified with large height differences (i.e., other height levels) can be avoided. At the same time, regional grid maps are constructed at different height levels to facilitate subsequent identification of abnormal semantics.

[0131] Based on the technical solutions of the above embodiments, the present application also provides an optional embodiment, in which the abnormal road semantic recognition method is described in detail.

[0132] See also Figure 5A The abnormal road semantic recognition method shown includes:

[0133] S501, obtaining position information of semantics to be recognized when at least one target vehicle is traveling on a road to be recognized, and area position information of other areas on a visible surface of the traveling trajectory except for the semantics to be recognized.

[0134] S502: Acquire the track point data of the target vehicle's travel track.

[0135] S503: Determine the semantic height of the semantics to be recognized according to the position information of the semantics to be recognized.

[0136] S504: Determine the trajectory height of the semantics to be recognized according to the trajectory point data corresponding to the semantics to be recognized.

[0137] Among them, S503 can be determined before S504, or after S504, or both can be determined together or crosswise, and this application does not impose any limitation on the order of execution of the two processes.

[0138] S505 . When the semantic height and the trajectory height are at the same height level, a regional grid map covering the semantics to be identified is constructed according to the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0139] S506 . Determine a semantic occupancy grid of the semantics to be recognized according to the first distribution situation corresponding to each target vehicle.

[0140] S507: Initialize the grid weight of each grid in the regional grid map.

[0141] S508. For each target vehicle, according to the first distribution situation corresponding to the target vehicle, the first preset weight is accumulated to the grid weight of the grid covered by the semantics to be identified, so as to update the grid weight of the corresponding grid.

[0142] S509 . For each target vehicle, according to the second distribution situation corresponding to the target vehicle, the second preset weight is accumulated for the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid.

[0143] Among them, S508 can be determined before S509, or after S509, or both can be determined together or crosswise, and this application does not impose any limitation on the order of execution of the two processes.

[0144] S510, determining the semantically unoccupied grids of the semantics to be identified based on the size relationship between the updated grid weights of each grid in the regional grid map and the preset grid weight threshold; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0145] S511 , taking a grid corresponding to the intersection of a semantically occupied grid and a semantically unoccupied grid as an abnormally occupied grid of the semantics to be identified.

[0146] S512: Determine the semantic weight of the semantics to be identified according to the proportion of the abnormal occupancy grid in the semantic occupancy grid.

[0147] S513: Determine whether the semantics to be identified is abnormal semantics according to the size relationship between the semantic weight and a preset semantic weight threshold.

[0148] The detailed technical contents of the above steps S501-S513 have been recorded in the above examples and will not be repeated here.

[0149] In an optional embodiment, when the semantics to be identified are abnormal semantics, the abnormal semantics can be removed. Figure 5B The schematic diagram of the abnormal semantic elimination process shown in FIG. 1 shows that the abnormal road semantic recognition method provided by the present application can accurately identify the abnormal road semantics such as Figure 5B The abnormal semantics in the dotted line frame are removed to avoid the interference of abnormal semantics, which is beneficial to improving the accuracy of the road map.

[0150] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0151] Based on the same inventive concept, the embodiment of the present application also provides an abnormal road semantic recognition device for implementing the abnormal road semantic recognition method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more abnormal road semantic recognition device embodiments provided below can refer to the limitations of the abnormal road semantic recognition method above, and will not be repeated here.

[0152] In an exemplary embodiment, Figure 6 As shown, an abnormal road semantic recognition device is provided, including: an acquisition module 610, a construction module 620, a first determination module 630, a second determination module 640 and a third determination module 650, wherein:

[0153] An acquisition module 610 is used to acquire the position information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and the area position information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0154] A construction module 620, for constructing a regional grid map covering the semantics to be identified according to the location information of the semantics to be identified and the corresponding track visible surface;

[0155] A first determination module 630 is used to determine a first distribution of grids covered by the semantics to be identified in the regional grid map according to the location information of the semantics to be identified;

[0156] The second determination module 640 is used to determine the second distribution of the semantically uncovered grids to be identified in the regional grid map according to the regional location information;

[0157] The third determination module 650 is used to determine whether the semantics to be identified is abnormal semantics according to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle.

[0158] In one embodiment, the third determination module 650 includes: a first determination unit, used to determine the semantically occupied grid of the semantics to be identified according to the first distribution corresponding to each target vehicle; a second determination unit, used to determine the semantically unoccupied grid of the semantics to be identified according to the second distribution corresponding to each target vehicle; a third determination unit, used to determine whether the semantics to be identified is an abnormal semantics based on the grid distribution of the semantically occupied grids in the regional grid map and the grid distribution of the semantically unoccupied grids.

[0159] In one embodiment, the second determination unit includes: an initialization subunit, which is used to initialize the grid weights of each grid in the regional grid map; a first updating subunit, which is used to accumulate a first preset weight for the grid weights of the grids covered by the semantics to be identified for each target vehicle according to a first distribution situation corresponding to the target vehicle, so as to update the grid weights of the corresponding grids; a second updating subunit, which is used to accumulate a second preset weight for the grid weights of the grids not covered by the semantics to be identified for each target vehicle according to a second distribution situation corresponding to the target vehicle, so as to update the grid weights of the corresponding grids; the first determination subunit is used to determine the semantically unoccupied grids of the semantics to be identified according to the updated grid weights of each grid in the regional grid map; wherein the second preset weight has an opposite numerical sign to the first preset weight.

[0160] In one embodiment, the first determination subunit is specifically used to: determine the semantically unoccupied grids of the semantics to be identified based on the size relationship between the updated grid weights of each grid in the regional grid map and a preset grid weight threshold; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0161] In one embodiment, the third determination unit includes: a first processing subunit, used to take the grid corresponding to the intersection of the semantic occupied grid and the semantic unoccupied grid as the abnormal occupied grid of the semantics to be identified; a second determination subunit, used to determine the semantic weight of the semantics to be identified according to the proportion of the abnormal occupied grid in the semantic occupied grid; and a third determination subunit, used to determine whether the semantics to be identified is an abnormal semantics according to the size relationship between the semantic weight and a preset semantic weight threshold.

[0162] In one embodiment, the construction module 620 includes: a first acquisition unit, used to acquire the trajectory point data of the target vehicle's travel trajectory; a fourth determination unit, used to determine the semantic height of the semantics to be identified based on the position information of the semantics to be identified; a fifth determination unit, used to determine the trajectory height of the semantics to be identified based on the trajectory point data corresponding to the semantics to be identified; a first construction unit, used to construct a regional grid map covering the semantics to be identified based on the position information of the semantics to be identified and the corresponding trajectory visible surface when the semantic height and the trajectory height are at the same height level.

[0163] In one embodiment, the construction module 620 includes: a sixth determination unit, used to determine the visible surface boundary position of the corresponding track visible surface according to the track point position of the target vehicle; a second construction unit, used to construct an area map covering the visible surface corresponding to the semantics to be identified and the visible surface boundary position in a preset positive direction according to the position information of the semantics to be identified and the visible surface boundary position; an expansion unit, used to expand the area map according to a preset grid size to obtain a regional grid map.

[0164] Each module in the abnormal road semantic recognition device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0165] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 7As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (Near Field Communication, NFC) or other technologies. When the computer program is executed by the processor, an XXX method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0166] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0167] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0168] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0169] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0170] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0171] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0172] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0173] In one embodiment, when the processor executes the computer program, the following steps are also implemented: determining the semantically occupied grid of the semantics to be identified based on the first distribution corresponding to each target vehicle; determining the semantically unoccupied grid of the semantics to be identified based on the second distribution corresponding to each target vehicle; and determining whether the semantics to be identified is an abnormal semantics based on the grid distribution of the semantically occupied grids and the grid distribution of the semantically unoccupied grids in the regional grid map.

[0174] In one embodiment, the processor further implements the following steps when executing the computer program: initializing the grid weight of each grid in the regional grid map; for each target vehicle, according to the first distribution situation corresponding to the target vehicle, accumulating the first preset weight to the grid weight of the grid covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; and, for each target vehicle, according to the second distribution situation corresponding to the target vehicle, accumulating the second preset weight to the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; according to the updated grid weights of each grid in the regional grid map, determining the semantically unoccupied grid of the semantics to be identified; wherein the second preset weight has an opposite numerical sign to the first preset weight.

[0175] In one embodiment, when the processor executes the computer program, the following steps are also implemented: based on the size relationship between the updated grid weights of each grid in the regional grid map and the preset grid weight threshold, the semantically unoccupied grids of the semantics to be identified are determined; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0176] In one embodiment, when the processor executes the computer program, the following steps are also implemented: the grid corresponding to the intersection of the semantic occupied grid and the semantic unoccupied grid is used as the abnormal occupied grid of the semantics to be identified; the semantic weight of the semantics to be identified is determined according to the proportion of the abnormal occupied grid in the semantic occupied grid; and whether the semantics to be identified is abnormal semantics is determined according to the relationship between the semantic weight and a preset semantic weight threshold.

[0177] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining trajectory point data of the target vehicle's travel trajectory; determining the semantic height of the semantics to be identified based on the position information of the semantics to be identified; and determining the trajectory height of the semantics to be identified based on the trajectory point data corresponding to the semantics to be identified; when the semantic height and the trajectory height are at the same height level, constructing a regional grid map covering the semantics to be identified based on the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0178] In one embodiment, when the processor executes the computer program, the following steps are also implemented: according to the trajectory point position of the target vehicle's travel trajectory, the visible surface boundary position of the corresponding trajectory visible surface is determined; according to the position information of the semantics to be identified and the visible surface boundary position, an area map covering the visible surface corresponding to the semantics to be identified and the visible surface boundary position in a preset positive direction is constructed; according to the preset grid size, the area map is expanded to obtain a regional grid map.

[0179] In an optional embodiment, the computer device may be a vehicle-mounted terminal.

[0180] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0181] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0182] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0183] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0184] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0185] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0186] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: determining the semantically occupied grid of the semantics to be identified based on the first distribution corresponding to each target vehicle; determining the semantically unoccupied grid of the semantics to be identified based on the second distribution corresponding to each target vehicle; and determining whether the semantics to be identified is an abnormal semantics based on the grid distribution of the semantically occupied grids and the grid distribution of the semantically unoccupied grids in the regional grid map.

[0187] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: initializing the grid weight of each grid in the regional grid map; for each target vehicle, according to the first distribution situation corresponding to the target vehicle, accumulating the first preset weight for the grid weight of the grid covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; and, for each target vehicle, according to the second distribution situation corresponding to the target vehicle, accumulating the second preset weight for the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; according to the updated grid weights of each grid in the regional grid map, determining the semantically unoccupied grid of the semantics to be identified; wherein the second preset weight has an opposite numerical sign to the first preset weight.

[0188] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: based on the size relationship between the updated grid weights of each grid in the regional grid map and the preset grid weight threshold, the semantically unoccupied grids of the semantics to be identified are determined; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0189] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: the grid corresponding to the intersection of the semantic occupied grid and the semantic unoccupied grid is used as the abnormal occupied grid of the semantics to be identified; the semantic weight of the semantics to be identified is determined according to the proportion of the abnormal occupied grid in the semantic occupied grid; and whether the semantics to be identified is abnormal semantics is determined according to the relationship between the semantic weight and a preset semantic weight threshold.

[0190] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining trajectory point data of the target vehicle's travel trajectory; determining the semantic height of the semantics to be identified based on the position information of the semantics to be identified; and determining the trajectory height of the semantics to be identified based on the trajectory point data corresponding to the semantics to be identified; when the semantic height and the trajectory height are at the same height level, constructing a regional grid map covering the semantics to be identified based on the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0191] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: according to the trajectory point position of the target vehicle's travel trajectory, the visible surface boundary position of the corresponding trajectory visible surface is determined; according to the position information of the semantics to be identified and the visible surface boundary position, an area map covering the visible surface corresponding to the semantics to be identified and the visible surface boundary position in a preset positive direction is constructed; according to the preset grid size, the area map is expanded to obtain an area grid map.

[0192] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0193] Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized;

[0194] According to the location information of the semantics to be identified and the corresponding trajectory visible surface, a regional grid map covering the semantics to be identified is constructed;

[0195] Determine, according to the location information of the semantics to be identified, a first distribution of grids covered by the semantics to be identified in the regional grid map;

[0196] Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map;

[0197] According to the first distribution situation and the corresponding second distribution situation corresponding to each target vehicle, it is determined whether the semantics to be identified is abnormal semantics.

[0198] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: determining the semantically occupied grid of the semantics to be identified based on the first distribution corresponding to each target vehicle; determining the semantically unoccupied grid of the semantics to be identified based on the second distribution corresponding to each target vehicle; and determining whether the semantics to be identified is an abnormal semantics based on the grid distribution of the semantically occupied grids and the grid distribution of the semantically unoccupied grids in the regional grid map.

[0199] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: initializing the grid weight of each grid in the regional grid map; for each target vehicle, according to the first distribution situation corresponding to the target vehicle, accumulating the first preset weight for the grid weight of the grid covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; and, for each target vehicle, according to the second distribution situation corresponding to the target vehicle, accumulating the second preset weight for the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; according to the updated grid weights of each grid in the regional grid map, determining the semantically unoccupied grid of the semantics to be identified; wherein the second preset weight has an opposite numerical sign to the first preset weight.

[0200] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: based on the size relationship between the updated grid weights of each grid in the regional grid map and the preset grid weight threshold, the semantically unoccupied grids of the semantics to be identified are determined; wherein the preset grid weight threshold has the same numerical sign as the second preset weight.

[0201] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: the grid corresponding to the intersection of the semantic occupied grid and the semantic unoccupied grid is used as the abnormal occupied grid of the semantics to be identified; the semantic weight of the semantics to be identified is determined according to the proportion of the abnormal occupied grid in the semantic occupied grid; and whether the semantics to be identified is abnormal semantics is determined according to the relationship between the semantic weight and a preset semantic weight threshold.

[0202] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining trajectory point data of the target vehicle's travel trajectory; determining the semantic height of the semantics to be identified based on the position information of the semantics to be identified; and determining the trajectory height of the semantics to be identified based on the trajectory point data corresponding to the semantics to be identified; when the semantic height and the trajectory height are at the same height level, constructing a regional grid map covering the semantics to be identified based on the position information of the semantics to be identified and the corresponding trajectory visible surface.

[0203] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: according to the trajectory point position of the target vehicle's travel trajectory, the visible surface boundary position of the corresponding trajectory visible surface is determined; according to the position information of the semantics to be identified and the visible surface boundary position, an area map covering the visible surface corresponding to the semantics to be identified and the visible surface boundary position in a preset positive direction is constructed; according to the preset grid size, the area map is expanded to obtain an area grid map.

[0204] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0205] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0206] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for identifying abnormal road semantics, characterized in that: include: Acquire location information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and area location information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized; Constructing a regional grid map covering the semantics to be identified according to the location information of the semantics to be identified and the corresponding track visible surface; Determining, according to the location information of the semantics to be identified, a first distribution situation of grids covered by the semantics to be identified in the regional grid map; Determine, according to the regional location information, a second distribution of the semantically uncovered grids to be identified in the regional grid map; According to the first distribution situation and the corresponding second distribution situation corresponding to each of the target vehicles, it is determined whether the semantics to be identified is abnormal semantics.

2. The method according to claim 1, characterized in that The determining whether the semantics to be identified is abnormal semantics according to the first distribution situation and the corresponding second distribution situation corresponding to each of the target vehicles includes: Determining a semantic occupancy grid of the semantics to be identified according to the first distribution conditions corresponding to each of the target vehicles; Determining, according to the second distribution conditions corresponding to each of the target vehicles, a semantically unoccupied grid of the semantics to be recognized; According to the grid distribution of the semantically occupied grids and the grid distribution of the semantically unoccupied grids in the regional grid map, it is determined whether the semantics to be identified is an abnormal semantics.

3. The method according to claim 2, characterized in that The determining, according to the second distribution conditions corresponding to each of the target vehicles, the semantically unoccupied grid of the semantics to be recognized includes: Initializing the grid weight of each grid in the regional grid map; For each target vehicle, according to the first distribution situation corresponding to the target vehicle, the grid weight of the grid covered by the semantics to be identified is accumulated with a first preset weight to update the grid weight of the corresponding grid; and For each target vehicle, according to the second distribution situation corresponding to the target vehicle, the second preset weight is accumulated for the grid weight of the grid not covered by the semantics to be identified, so as to update the grid weight of the corresponding grid; Determining the semantically unoccupied grids of the semantics to be identified according to the updated grid weights of each grid in the regional grid map; The second preset weight has a numerical sign opposite to that of the first preset weight.

4. The method according to claim 3, characterized in that The determining, according to the updated grid weights of each grid in the regional grid map, the semantically unoccupied grids of the semantics to be identified comprises: Determining the semantically unoccupied grids of the semantics to be identified according to the magnitude relationship between the updated grid weights of each grid in the regional grid map and a preset grid weight threshold; The preset grid weight threshold has the same numerical sign as the second preset weight.

5. The method according to any one of claims 2 to 4, characterized in that: The determining whether the semantics to be identified is an abnormal semantics according to the grid distribution of the semantically occupied grids and the grid distribution of the semantically unoccupied grids in the regional grid map includes: The grid corresponding to the intersection of the semantically occupied grid and the semantically unoccupied grid is used as the abnormally occupied grid of the semantics to be identified; Determining the semantic weight of the semantics to be identified according to the proportion of the abnormal occupancy grid in the semantic occupancy grid; Whether the semantics to be identified is abnormal semantics is determined according to the magnitude relationship between the semantic weight and a preset semantic weight threshold.

6. The method according to any one of claims 1 to 4, characterized in that: The step of constructing a regional grid map covering the semantics to be identified based on the location information of the semantics to be identified and the corresponding track visible surface includes: Acquiring track point data of the target vehicle's travel track; Determining the semantic height of the semantic to be recognized according to the position information of the semantic to be recognized; and Determining the trajectory height of the semantics to be recognized according to the trajectory point data corresponding to the semantics to be recognized; In the case that the semantic height and the trajectory height are at the same height level, a regional grid map covering the semantics to be identified is constructed according to the position information of the semantics to be identified and the corresponding trajectory visible surface.

7. The method according to any one of claims 1 to 4, characterized in that: The step of constructing a regional grid map covering the semantics to be identified based on the location information of the semantics to be identified and the corresponding track visible surface includes: Determining the visible surface boundary position of the visible surface of the corresponding track according to the track point position of the target vehicle; According to the position information of the semantics to be recognized and the position of the boundary of the visible surface, constructing an area map covering the semantics to be recognized and the visible surface corresponding to the position of the boundary of the visible surface in a preset positive direction; The regional map is expanded according to a preset grid size to obtain the regional grid map.

8. An abnormal road semantic recognition device, characterized in that: include: An acquisition module, used to acquire the position information of the semantics to be recognized when at least one target vehicle is traveling on the road to be recognized, and the area position information of other areas on the visible surface of the trajectory where the traveling trajectory is located except for the semantics to be recognized; A construction module, used to construct a regional grid map covering the semantics to be identified according to the location information of the semantics to be identified and the corresponding track visible surface; A first determination module, configured to determine a first distribution of grids covered by the semantics to be identified in the regional grid map according to the location information of the semantics to be identified; A second determination module is used to determine a second distribution of the semantically uncovered grids to be identified in the regional grid map according to the regional location information; The third determination module is used to determine whether the semantics to be identified is abnormal semantics according to the first distribution situation and the corresponding second distribution situation corresponding to each of the target vehicles.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images