A fusion method and fusion device for traffic data

Abstract

The present application provides a method and device for fusing traffic data, the method comprising: obtaining multiple traffic data sets collected from a target road section, and rasterizing two of the traffic data sets to be fused respectively, to obtain a spatiotemporal traffic map corresponding to each traffic data set to be fused; based on the data source corresponding to each traffic data set to be fused, fusing the spatiotemporal traffic map to obtain a target spatiotemporal traffic map of the target road section, and determining at least one first target congestion coordinate set in the target spatiotemporal traffic map through morphological operations and connected domain analysis; or fusing at least one congestion coordinate set corresponding to each spatiotemporal traffic map to obtain at least one second target congestion coordinate set corresponding to the spatiotemporal traffic map of the target road section. Through the above method, the accuracy and reliability of traffic data fusion are improved, thereby improving the efficiency of traffic management.

Description

Technical Field

[0001] The present application relates to the technical field of traffic data processing, and in particular to a method and device for fusing traffic data. Background Art

[0002] Since data from different sources and ranges may have their own limitations, such as sensor failure, incomplete data collection, etc., in addition, there are certain differences between different data sources. Therefore, traffic data fusion plays a vital role in modern traffic management systems. By integrating traffic data, the accuracy and reliability of the overall traffic data can be improved, thereby improving the efficiency and effectiveness of traffic management.

[0003] At present, when fusing traffic data from multiple sources, they are generally fused at the original traffic parameter level, and then the traffic status is judged based on the fused traffic parameters. The fusion of traffic parameters depends on the road segment division method of the underlying road, and the road segment division methods used by different sources may be different, which makes it difficult to fuse data, reduces the accuracy and reliability of traffic data fusion, and thus reduces the efficiency of traffic management. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a traffic data fusion method and fusion device, which collects multiple traffic road condition data sets from multiple data sources for a target section in a highway, and obtains a spatiotemporal road condition map corresponding to each traffic road condition data set to be fused by rasterizing two of the traffic road condition data sets to be fused respectively, and based on the data sources corresponding to the two traffic road condition data sets to be fused, the spatiotemporal road condition maps are correspondingly fused to obtain a target spatiotemporal road condition map corresponding to the target section and determine at least one target congestion coordinate set corresponding to the target spatiotemporal road condition map, or for two full road condition data sets, morphological operations and connected domain analysis are performed on each spatiotemporal road condition map to obtain at least one congestion coordinate set corresponding to each spatiotemporal road condition map and further fuse the congestion coordinate sets, thereby realizing multi-source data fusion of multi-source traffic conditions and traffic congestion at the decision-making level, improving the accuracy and reliability of traffic data fusion, and thereby improving the efficiency of traffic management.

[0005] The present application provides a method for fusing traffic data, the method comprising:

[0006] Acquire multiple traffic condition data sets collected for a target section of the highway; wherein the multiple traffic condition data include at least one full traffic condition data set collected based on all network devices and at least one local traffic condition data set collected for a local section of the target section;

[0007] For two traffic condition data sets to be fused among the multiple traffic condition data sets, rasterizing each of the traffic condition data sets to be fused to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused;

[0008] The step of performing rasterization processing on each of the traffic condition data sets to be fused to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused includes:

[0009] Calculate the image range of the spatiotemporal road condition map based on the preset spatiotemporal resolution and the target start and end stake numbers and target analysis time period corresponding to the target road section;

[0010] For each of the traffic condition data sets to be integrated, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be integrated in the traffic condition data set to be integrated;

[0011] Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a plurality of spatiotemporal road condition coordinates corresponding to the traffic condition data sets to be fused are obtained using a preset coordinate conversion formula and a preset resolution;

[0012] Based on the multiple spatiotemporal road condition coordinates, the traffic state information and the image range, obtaining a spatiotemporal road condition map corresponding to each of the traffic road condition data sets to be fused;

[0013] Based on the data sources corresponding to the two traffic condition data sets to be fused, the spatiotemporal traffic condition graphs are correspondingly fused to obtain a target spatiotemporal traffic condition graph of the target road section, and morphological operations and connected domain analysis are performed on the target spatiotemporal traffic condition graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic condition graph;

[0014] Alternatively, morphological operations and connected domain analysis are performed on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused are correspondingly fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

[0015] Furthermore, after obtaining at least one target congestion coordinate set, the fusion method further includes:

[0016] Based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set, multiple coordinate set matching calculations are performed using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

[0017] Further, the method of performing multiple coordinate set matching calculations based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set includes:

[0018] Based on at least one of the target congestion coordinate sets and a preset calibrated congestion coordinate set, using a preset similarity calculation formula, a similarity matrix corresponding to the target congestion coordinate set and the calibrated congestion coordinate set is calculated;

[0019] For each of the target congestion coordinate sets, based on the similarity corresponding to the target congestion coordinate set in the similarity matrix, it is determined whether the target congestion coordinate set meets the matching condition corresponding to the calibration congestion coordinate set, and a determination result is obtained; wherein the matching condition includes: the similarity corresponding to the target congestion coordinate set is greater than a first preset similarity threshold and the similarity is a maximum value among multiple similarities corresponding to elements in the calibration congestion coordinate set;

[0020] Based on the judgment result, respectively determining a first number of target congestion coordinate sets that meet the matching condition and a second number of target congestion coordinate sets that do not meet the matching condition in at least one of the target congestion coordinate sets, as well as the number of unmatched elements in the calibration congestion coordinate set;

[0021] Based on the first number, the second number and the number of elements, the accuracy and recall rate corresponding to at least one of the target congestion coordinate sets are respectively obtained using a preset accuracy calculation formula and a recall rate calculation formula.

[0022] Furthermore, based on the data sources corresponding to the two traffic condition data sets to be fused, the spatiotemporal traffic condition graphs are correspondingly fused to obtain the target spatiotemporal traffic condition graph of the target road section, including:

[0023] When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets or one full traffic condition data set and one partial traffic condition data set, determining the weight values ​​corresponding to the traffic condition data sets to be fused respectively;

[0024] Based on the pixel values ​​and the weight values ​​respectively corresponding to the spatiotemporal road condition graphs, the spatiotemporal road condition graphs are correspondingly fused by using a mean overlap fusion method to obtain a to-be-processed spatiotemporal road condition graph of the target road section;

[0025] The overlapping boundary area of ​​the to-be-processed spatiotemporal traffic condition map is smoothed to obtain a target spatiotemporal traffic condition map of the target road section.

[0026] Further, the performing of morphological operations and connected domain analysis on the target spatiotemporal traffic condition graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic condition graph includes:

[0027] Performing threshold segmentation on the target spatiotemporal traffic map to convert the spatiotemporal traffic map into a first spatiotemporal traffic map represented as a binary image;

[0028] Using a preset convolution kernel to perform a morphological operation including an opening operation and a closing operation on the first spatiotemporal road condition map, to obtain a second spatiotemporal road condition map converted from the first spatiotemporal road condition map, so as to denoise and fill holes in the first spatiotemporal road condition map;

[0029] Performing a connected domain analysis on the second spatiotemporal traffic map using a preset multi-neighborhood definition to determine a congestion boundary box in the second spatiotemporal traffic map;

[0030] Based on the congestion boundary box, at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal road condition map is obtained.

[0031] Further, the morphological operation and connected domain analysis are performed on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused are correspondingly fused to obtain at least one second target congestion coordinate set spatiotemporal road condition graph corresponding to the target road section, including:

[0032] When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets, morphological operations and connected domain analysis are performed on each of the spatiotemporal traffic condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal traffic condition graphs;

[0033] Based on the congestion coordinate set corresponding to each of the traffic condition data sets to be fused, a preset similarity calculation formula is used to obtain the similarity between each of the congestion coordinate sets, so as to obtain a first similarity matrix between each of the congestion coordinate sets;

[0034] Sorting the similarities in descending order according to the quantity, and determining whether the first similarity in the descending order is greater than a second preset similarity threshold;

[0035] If it is greater than, performing a union calculation on the congestion coordinate sets corresponding to the first similarity to obtain a first target coordinate set, and replacing the congestion coordinate set corresponding to the first similarity with the first target coordinate set to obtain an updated second similarity matrix;

[0036] Based on the second similarity matrix, similarity calculation, threshold judgment and union calculation are repeated until there is no congested coordinate set in the congested coordinate set that can be used for union calculation, so as to obtain at least one second target congested coordinate set corresponding to the target road section.

[0037] The embodiment of the present application further provides a traffic data fusion device, the fusion device comprising:

[0038] A data acquisition module, used to acquire multiple traffic condition data sets collected for a target section in the highway; wherein the multiple traffic condition data include at least one full traffic condition data set collected based on the whole network equipment and at least one local traffic condition data set collected for a local section in the target section;

[0039] A raster processing module, for performing rasterization processing on two traffic condition data sets to be fused among the multiple traffic condition data sets, respectively, to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused;

[0040] When the grid processing module is used to perform grid processing on each of the traffic condition data sets to be fused, and obtain the spatiotemporal traffic condition graphs corresponding to each of the traffic condition data sets to be fused, the grid processing module is used to:

[0041] Calculate the image range of the spatiotemporal road condition map based on the preset spatiotemporal resolution and the target start and end stake numbers and target analysis time period corresponding to the target road section;

[0042] For each of the traffic condition data sets to be integrated, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be integrated in the traffic condition data set to be integrated;

[0043] Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a plurality of spatiotemporal road condition coordinates corresponding to the traffic condition data sets to be fused are obtained using a preset coordinate conversion formula and a preset resolution;

[0044] Based on the multiple spatiotemporal road condition coordinates, the traffic state information and the image range, obtaining a spatiotemporal road condition map corresponding to each of the traffic road condition data sets to be fused;

[0045] A first fusion module is used to fuse the spatiotemporal traffic graphs based on the data sources corresponding to the two traffic data sets to be fused, respectively, to obtain a target spatiotemporal traffic graph of the target road section, and to perform morphological operations and connected domain analysis on the target spatiotemporal traffic graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic graph;

[0046] The second fusion module is used to perform morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and to fuse the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

[0047] Furthermore, the fusion device further includes a data verification module, which is used to:

[0048] Based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set, multiple coordinate set matching calculations are performed using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

[0049] An embodiment of the present application also provides an electronic device, including: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps of the traffic data fusion method as described above are performed.

[0050] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned traffic data fusion method are executed.

[0051] The traffic data fusion method and fusion device provided in the embodiment of the present application, compared with the method in the prior art that fuses at the original traffic parameter level and then judges the traffic state of the fused traffic parameters, the fusion method collects multiple traffic road condition data sets from multiple data sources for the target section in the expressway, and obtains the spatiotemporal road condition map corresponding to each traffic road condition data set to be fused by rasterizing two of the traffic road condition data sets to be fused respectively, based on the data sources corresponding to the two traffic road condition data sets to be fused, the spatiotemporal road condition maps are correspondingly fused to obtain the target spatiotemporal road condition map corresponding to the target section and determine at least one target congestion coordinate set corresponding to the target spatiotemporal road condition map, or for two full road condition data sets, morphological operations and connected domain analysis are performed on each spatiotemporal road condition map to obtain at least one congestion coordinate set corresponding to each spatiotemporal road condition map and further the congestion coordinate sets are fused, thereby realizing multi-source data fusion of multi-source traffic conditions and traffic congestion at the decision-making level, improving the accuracy and reliability of traffic data fusion, and thereby improving the efficiency of traffic management.

[0052] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 One of the flow charts of a method for integrating traffic data provided in an embodiment of the present application;

[0055] Figure 2 An example diagram of a spatiotemporal traffic condition diagram provided in an embodiment of the present application;

[0056] Figure 3 A schematic diagram of image processing of a congestion boundary box provided in an embodiment of the present application;

[0057] Figure 4 A schematic diagram of the fusion of a congestion coordinate set provided in an embodiment of the present application;

[0058] Figure 5 A second flowchart of a method for integrating traffic data provided in an embodiment of the present application;

[0059] Figure 6 One of the structural schematic diagrams of a traffic data fusion device provided in an embodiment of the present application;

[0060] Figure 7 A second structural diagram of a traffic data fusion device provided in an embodiment of the present application;

[0061] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0062] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0063] Research has found that, at present, when fusing traffic data from multiple sources, they are generally fused at the original traffic parameter level, and then the traffic status is judged based on the fused traffic parameters. Among them, the fusion of traffic parameters depends on the road segment division method of the underlying road, and the road segment division methods used by different sources may be different, which makes it difficult to fuse data, reduces the accuracy and reliability of traffic data fusion, and thus reduces the efficiency of traffic management.

[0064] Based on this, an embodiment of the present application provides a method for fusing traffic data, which collects multiple traffic condition data sets from multiple data sources for a target section in a highway, and obtains a spatiotemporal traffic condition map corresponding to each traffic condition data set to be fused by rasterizing two of the traffic condition data sets to be fused respectively. Based on the data sources corresponding to the two traffic condition data sets to be fused, the spatiotemporal traffic condition maps are correspondingly fused to obtain a target spatiotemporal traffic condition map corresponding to the target section and determine at least one target congestion coordinate set corresponding to the target spatiotemporal traffic condition map, or for two full traffic condition data sets, morphological operations and connected domain analysis are performed on each spatiotemporal traffic condition map to obtain at least one congestion coordinate set corresponding to each spatiotemporal traffic condition map and further the congestion coordinate sets are fused, thereby realizing multi-source data fusion at the decision-making level of multi-source traffic conditions and traffic congestion, improving the accuracy and reliability of traffic data fusion, and thereby improving the efficiency of traffic management.

[0065] See also Figure 1 , Figure 1 This is one of the flow charts of a method for integrating traffic data provided in an embodiment of the present application. Figure 1 As shown in , the method for fusing traffic data provided by the embodiment of the present application includes:

[0066] S101. Acquire multiple traffic condition data sets collected for a target section in a highway; wherein the multiple traffic condition data sets include at least one full traffic condition data set collected based on equipment in the entire network and at least one local traffic condition data set collected for a local section in the target section.

[0067] It should be noted that since there are many technologies for detecting traffic conditions on highways, in one possible implementation of the present application, depending on the data collection equipment and data sources, the multiple traffic road condition data sets collected for the target section of the highway may include at least one full road condition data set and at least one local road condition data set.

[0068] Among them, the at least one full road condition data set is collected based on the whole network equipment, and the data source of the full road condition data set includes the whole network equipment such as Internet maps and high-speed ETC toll collection. For example, the traffic conditions are calculated based on the positioning data of the Internet map data, or the road conditions are obtained by restoring the vehicle travel trajectory based on the high-speed ETC toll data. Here, the full road condition data set refers to the data set of the traffic road condition status of the whole network.

[0069] The at least one local road condition data set is collected for a local section in the target section. The data sources of the local road condition data set include videos and images, wave frequency detectors (microwave detectors, ultrasonic detectors, radars and other equipment) and command and dispatch systems, etc. Among them, the data collected by the command and dispatch system refers to the slow-moving data discovered through manual video inspections. Here, the local road condition data set can only perceive the traffic status of the local section.

[0070] In this step, at least one full road condition data set and at least one local road condition data set of the target section in the highway are collected respectively by the collection devices set up in the whole network and the local sections, and multiple traffic road condition data sets including the collected full road condition data sets and local road condition data sets are obtained.

[0071] For example, an example of a traffic condition data set is shown in the following table.

[0072]

[0073] Among them, linkid represents the sequence number of the path segment; roadNum represents the line number of the target segment; direction represents the direction of the target segment (1 represents upward movement, 2 represents downward movement); StartStake represents the starting stake number; endStake represents the ending stake number; startTime represents the starting time; endTime represents the ending time; and state represents the traffic state value.

[0074] S102: for two traffic condition data sets to be fused among the multiple traffic condition data sets, rasterize each of the traffic condition data sets to be fused to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused.

[0075] In an embodiment of the present application, based on the user's indication of the data set desired to be fused in the data fusion requirement or the business requirement indication of the traffic management system, two traffic condition data sets to be fused are determined from multiple traffic condition data sets, so as to perform subsequent data fusion on the two traffic condition data sets to be fused.

[0076] Here, since the original traffic status recognition results are usually based on the traffic conditions of the underlying road sections, the traffic status of each basic section in each update cycle is generally recorded. In addition, since traffic conditions are structured data, traffic congestion directly identified by traffic conditions is very discrete in spatial and temporal dimensions, which is not conducive to traffic control; therefore, in order to extract unstructured data information from structured traffic condition data, the traffic condition data set to be fused can be converted into a spatiotemporal traffic condition map through rasterization.

[0077] Among them, the spatiotemporal traffic condition map refers to a visualization tool used to describe traffic conditions. It integrates information in two dimensions, time and space. The horizontal axis direction (i.e., the length of the image) of the spatiotemporal traffic condition map reflects spatial information, the vertical axis direction (i.e., the width of the image) reflects time information, and the pixel value reflects the road condition information, so that observers can quickly understand the changing trend of traffic conditions in a specific area.

[0078] In one possible implementation of the present application, during specific implementation, step S102 may include:

[0079] S1021. Calculate the image range of the spatiotemporal road condition map based on a preset spatiotemporal resolution, the target start and end pile numbers corresponding to the target road section, and the target analysis time period.

[0080] It should be noted that when determining the target section on the highway, the corresponding target start and end pile numbers and the target analysis time period are set in advance for the target section; wherein the target start and end pile numbers include the target starting pile number and the target ending pile number; the target analysis time period includes the target start time and the target end time.

[0081] In this step, determine the target starting stake number of the target section ( ) and the target end station number ( ); Here, when on the uplink route, the target starting pile number is smaller than the target ending pile number, and when on the downlink route, the target starting pile number is larger than the target ending pile number.

[0082] The image range of the spatiotemporal road condition map includes image length and image width.

[0083] Specifically, the mileage length of the target section is calculated based on the target starting pile number and the target ending pile number, and then combined with the preset spatial resolution ( ), the image length of the corresponding spatiotemporal traffic map can be calculated by the following expression .

[0084] .

[0085] in, Indicates the preset spatial resolution in pixels / km; Indicates the target starting pile number; Indicates the target end stake number; here, the length of one kilometer in the spatial dimension is represented by how many pixels. The larger the spatial resolution value, the finer the spatial information is depicted.

[0086] Further, according to the target analysis time period of the target road section, the image width of the spatiotemporal road condition map is calculated; wherein the target analysis time period includes the target start time ( ) and target end time ( ), the time is represented by a timestamp, accurate to the second, and then the preset time resolution ( ), the image width of the spatiotemporal road condition map can be calculated by the following expression .

[0087] .

[0088] in, Indicates the preset spatial resolution in pixel / h; Indicates the target start time; Indicates the target end time; here, the length of one hour in the time dimension is represented by how many pixels. The larger the time resolution value, the finer the time dimension is depicted.

[0089] S1022. For each of the traffic condition data sets to be merged, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be merged in the traffic condition data set to be merged.

[0090] In an embodiment of the present application, each traffic condition data set to be integrated may include multiple traffic condition data to be integrated; each traffic condition data to be integrated includes at least the start and end pile numbers, start and end time periods and traffic status information corresponding to the target section.

[0091] Among them, the starting and ending pile numbers include the starting pile number and the ending pile number; the starting and ending time periods include the start time and the end time; and the traffic status information is not limited to the traffic status value.

[0092] S1023. Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a preset coordinate conversion formula and a preset resolution are used to obtain a plurality of spatiotemporal traffic condition coordinates corresponding to the traffic condition data sets to be fused.

[0093] Here, the starting and ending pile numbers include a starting pile number and an ending pile number; the starting and ending time periods include a start time and an end time.

[0094] In the embodiment of the present application, the expression of the preset coordinate transformation formula is as follows.

[0095] , (Formula 1).

[0096] , (Formula 2).

[0097] in, Indicates the starting and ending pile numbers; Indicates the start and end time period; Represents the horizontal axis coordinate of the time-space road condition diagram, and is composed of the start and end pile numbers ( ) is converted; Indicates the target starting pile number among the corresponding target starting and ending pile numbers of the target road section; Indicates the preset spatial resolution; Represents the vertical axis coordinate of the spatiotemporal traffic map, and is composed of the start and end time periods ( ) is converted; Indicates the target start time in the target analysis time period corresponding to the target road segment; Indicates the time resolution.

[0098] For example, each traffic condition data to be fused has a starting stake number (startStake), an end stake number (endStake), a start time (startTime) and an end time (endTime). Substitute startStake into Formula 1 to obtain the horizontal coordinate corresponding to the starting stake number, recorded as x1, and substitute endStake into Formula 1 to obtain the horizontal coordinate corresponding to the end stake number, recorded as x2; substitute startTime into Formula 2 to obtain the vertical coordinate corresponding to the start time, recorded as y1, and substitute endTime into Formula 2 to obtain the vertical coordinate corresponding to the end time, recorded as y2; that is, the coordinates of the traffic condition data in the spatiotemporal road condition diagram can be obtained (x1, x2, y1, y2).

[0099] S1024. Based on the multiple spatiotemporal traffic condition coordinates, the traffic state information and the image range, obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused.

[0100] In this step, based on the corresponding multiple spatiotemporal road condition coordinates (x1, x2, y1, y2) in the spatiotemporal road condition map obtained from each piece of traffic road condition data to be fused, and the traffic state information, the traffic state value of the corresponding area is determined, and the pixel values ​​corresponding to the multiple spatiotemporal road condition coordinates in the piece of traffic road condition data to be fused are assigned; then, the multiple spatiotemporal road condition coordinates, the pixel values ​​and the image range are constructed to obtain the spatiotemporal road condition map corresponding to each piece of traffic road condition data to be fused, that is, the rasterization of the piece of traffic road condition data to be fused is completed.

[0101] Furthermore, all the traffic condition data of the traffic condition data set to be fused are subjected to the same transformation, so as to obtain a spatiotemporal traffic condition graph corresponding to the traffic condition data set to be fused.

[0102] For example, see Figure 2 , Figure 2 This is an example diagram of a spatiotemporal traffic condition diagram provided in an embodiment of the present application.

[0103] like Figure 2 As shown in the figure, the traffic conditions on the upward direction of G11 Expressway from milepost K1416 (target starting pile number) to milepost K1430 (target ending pile number) from 12:00 (target starting time) to 19:00 (target ending time) on October 24, 2023 are shown. The horizontal axis represents the mileage (starting and ending pile numbers) of the spatial dimension, and the vertical axis represents the time (starting and ending time periods) of the time dimension. The preset spatial resolution is 100 pixels / km, and the preset time resolution is 60 pixels / h. The pixel value (state) is derived from the traffic status value in the traffic status information of the target section.

[0104] S103. Based on the data sources corresponding to the two traffic condition data sets to be merged, the spatiotemporal traffic condition graphs are correspondingly merged to obtain a target spatiotemporal traffic condition graph of the target road section, and morphological operations and connected domain analysis are performed on the target spatiotemporal traffic condition graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic condition graph.

[0105] In an embodiment of the present application, when the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets or a full traffic condition data set and a local traffic condition data set, the spatiotemporal road condition maps corresponding to the two traffic condition data sets to be fused are correspondingly fused. Specifically, based on the pixel values ​​and the weight values ​​respectively corresponding to the spatiotemporal road condition maps, the fusion is performed using the mean overlapping fusion method, and after further smoothing, the fused target spatiotemporal road condition map is obtained.

[0106] Furthermore, by performing morphological operations and connected domain analysis on the target spatiotemporal traffic condition graph, a congestion boundary box in the target spatiotemporal traffic condition graph is determined to obtain at least one first target congestion coordinate set in the spatiotemporal traffic condition graph.

[0107] In an embodiment of the present application, in order to extract congestion information from the target spatiotemporal traffic map, it is necessary to perform morphological operations and connected domain analysis on the target spatiotemporal traffic map to obtain at least one first target congestion coordinate set corresponding to the target spatiotemporal traffic map.

[0108] The at least one first target congestion coordinate set is represented in the target spatiotemporal traffic condition map in the form of a congestion boundary box, so as to intuitively represent the congestion situation of the traffic condition data set to be fused in two dimensions of time and space.

[0109] See also Figure 3 , Figure 3 A schematic diagram of image processing of a congestion boundary box provided in an embodiment of the present application. Figure 3 As shown in , after threshold segmentation is performed on the exemplary target spatiotemporal traffic condition map, morphological operations and connected domain analysis are performed to obtain a congestion boundary box to obtain at least one congestion coordinate set in the spatiotemporal traffic condition map.

[0110] In one possible implementation of the present application, in a specific implementation, the step of correspondingly fusing the spatiotemporal traffic condition graph based on the data sources corresponding to the two traffic condition data sets to be fused in step S103 to obtain the target spatiotemporal traffic condition graph of the target road section may include:

[0111] S1031. When the data sources corresponding to the two traffic condition data sets to be merged are two full traffic condition data sets or one full traffic condition data set and one partial traffic condition data set, determine the weight values ​​corresponding to the traffic condition data sets to be merged.

[0112] In this step, when the data sources corresponding to the two traffic condition data sets to be merged are two full traffic condition data sets or one full traffic condition data set and one partial traffic condition data set, the credibility of each traffic condition data set to be merged is determined, and based on the credibility, the weight values ​​corresponding to each traffic condition data set to be merged are determined.

[0113] S1032. Based on the pixel values ​​and the weight values ​​respectively corresponding to the spatiotemporal traffic map, the spatiotemporal traffic map is correspondingly fused by using a mean overlap fusion method to obtain a to-be-processed spatiotemporal traffic map of the target road section.

[0114] In this step, the overlapping areas of the two spatiotemporal traffic condition maps are fused, and the pixel values ​​and the weight values ​​corresponding to the two spatiotemporal traffic condition maps are correspondingly fused in the form of weighted average values ​​to obtain the spatiotemporal traffic condition map to be processed of the target section.

[0115] Here, the pixel value corresponding to the spatiotemporal road condition graph is, that is, the traffic state value corresponding to each piece of traffic data to be fused in the traffic data set to be fused corresponding to the spatiotemporal road condition graph.

[0116] It should be noted that, for the selected pixel values ​​(traffic status values ​​in the traffic status information) corresponding to multiple spatiotemporal road condition coordinates of the local road condition dataset, there are the following situations.

[0117] If the local road condition dataset can obtain the original detection results, for example, traffic parameters such as flow density and speed, the traffic condition evaluation index used by the full road condition dataset that is expected to be fused with the local road condition dataset is used to evaluate its traffic status, that is, the value range of the local road condition dataset is consistent with the evaluation standard and range of the traffic status value of the full road condition dataset to be fused.

[0118] If the local road condition dataset can only obtain the final traffic status evaluation result, and the traffic status evaluation index adopted by the local road condition dataset is inconsistent with that adopted by the full road condition dataset to be fused, for example, in the first road condition dataset, the road condition level is divided into levels 1-6; in the second road condition dataset, the road condition level is divided into levels 1-3, then traffic status mapping is required to map the traffic status evaluation index with a large number of levels to the traffic status evaluation index with a small number of levels.

[0119] If the local traffic condition dataset includes slow-moving data discovered by manual video inspection, the local traffic condition dataset is mapped to the slow-moving level and above in the traffic state evaluation index adopted by the full traffic condition dataset to be fused.

[0120] S1033: Smoothing the overlapping boundary areas of the to-be-processed spatiotemporal traffic condition graph to obtain a target spatiotemporal traffic condition graph of the target road section.

[0121] In this step, a preset Gaussian filter processing tool is used to smooth the overlapping boundary areas of the time-space traffic map to be processed, so as to obtain a target time-space traffic map of the target road section after filtering and smoothing.

[0122] In one possible implementation of the present application, in a specific implementation, the step of performing morphological operations and connected domain analysis on the target spatiotemporal traffic graph in step S103 to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic graph may include:

[0123] S1034: Perform threshold segmentation on the target spatiotemporal road condition map to convert the spatiotemporal road condition map into a first spatiotemporal road condition map represented as a binary image.

[0124] In this step, the target spatiotemporal road condition map is threshold segmented using a preset segmentation threshold, the target spatiotemporal road condition map is converted into a binary image, and the binary image is determined as the first spatiotemporal road condition map corresponding to the target spatiotemporal road condition map.

[0125] Specifically, the preset segmentation threshold is used to distinguish whether there is congestion; the segmentation threshold is applied to the original target spatiotemporal traffic map, pixels above the segmentation threshold are marked as foregrounds, and pixels below the segmentation threshold are marked as backgrounds, to obtain a first spatiotemporal traffic map represented as a binary image.

[0126] In the embodiment of the present application, the binary image is one of the simplest types of images, in which each pixel contains only two possible states or colors, for example, black and white or 0 and 1.

[0127] For example, assuming that Figure 2 If the state (pixel value, i.e., traffic state value) shown in is greater than or equal to 4, it is congested, then we can get the following Figure 3 The first spatiotemporal traffic map is shown in .

[0128] S1035. Use a preset convolution kernel to perform morphological operations including opening and closing operations on the first spatiotemporal road condition map to obtain a second spatiotemporal road condition map converted from the first spatiotemporal road condition map, so as to denoise and fill holes in the first spatiotemporal road condition map.

[0129] In the embodiment of the present application, the opening operation refers to a combination operation of first erosion and then expansion, which is used to eliminate noise existing in the image; and the closing operation refers to a combination operation of first expansion and then erosion, which is used to fill small holes in the front object in the image.

[0130] Among them, erosion is an operation used to remove boundary points in an image, which shrinks the boundary inward and makes the objects in the image smaller; dilation is the opposite process of erosion, which increases boundary points and makes the objects in the image larger by expanding the boundary outward.

[0131] In this step, a combination of opening and closing operations is used to denoise and fill holes in the first spatiotemporal road condition map using a preset convolution kernel to obtain a second spatiotemporal road condition map converted from the first spatiotemporal road condition map; wherein the size of the preset convolution kernel depends on the size of the congestion to be identified and the spatiotemporal resolution.

[0132] For example, if you want to identify congestion with a distance greater than 400m and a duration greater than 10min, when the spatial resolution is 100pixel / km and the temporal resolution is 60pixel / h, the size of the convolution kernel should be 10x40; where 10 corresponds to the vertical axis in the spatiotemporal traffic map and 40 corresponds to the horizontal axis in the spatiotemporal traffic map. The second spatiotemporal traffic map (for example, Figure 3 The spatiotemporal traffic map shown in ).

[0133] S1036: Perform a connected domain analysis on the second spatiotemporal traffic map using a preset multi-neighborhood definition to determine a congestion boundary box in the second spatiotemporal traffic map.

[0134] In the embodiment of the present application, the connected domain analysis refers to defining a pixel region with the same pixel value and adjacent position as a connected domain, that is, using a connected domain to represent an existing congestion.

[0135] Here, the preset multi-neighborhood definition refers to the judgment of positional proximity, which may include a 4-neighborhood definition and an 8-neighborhood definition, etc.; wherein the 4-neighborhood definition indicates that 4 pixel points (upper, lower, left, and right) are adjacent to the central pixel, that is, they are the same in space and adjacent in time, or the same in time and adjacent in space, and can be regarded as a connected domain; and the 8-neighborhood definition indicates 8 pixel points (upper, lower, left, right, and 4 points in the diagonal), that is, they are adjacent in space and time and can be regarded as a connected domain.

[0136] Exemplarily, after analyzing the connected domain defined by 8 neighborhoods on the second spatiotemporal traffic graph, Figure 3 As shown in , two connected domains, namely, two congestions, are marked, and then a congestion bounding box (bounding box) of the congested connected domain in the second spatiotemporal traffic map is determined.

[0137] The congestion boundary box may reflect the congestion duration and maximum congestion length reflected in the spatiotemporal traffic condition diagram, for example, Figure 3 As shown in , the borders in the third spatiotemporal traffic condition map are the two congestion boundary boxes determined in the second spatiotemporal traffic condition map.

[0138] Exemplarily, assuming that [x, y, w, h] is used to describe a congested bounding box, an example of border information corresponding to the congested bounding box is shown in the following table.

[0139]

[0140] S1037. Based on the congestion boundary box, obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal road condition map.

[0141] In this step, based on the position of the congestion boundary box in the corresponding space-time traffic condition map, the pixel points included in the space-time traffic condition map where the congestion boundary box is located are converted into the corresponding multiple congestion coordinates to obtain at least one first target congestion coordinate set corresponding to the target road section in the target space-time traffic condition map.

[0142] Furthermore, if the first target congestion coordinate set needs to be converted into a spatiotemporal range, the spatiotemporal resolution may be used to restore the congestion coordinate set into a vector space.

[0143] Specifically, any spatiotemporal road condition coordinate (x, y) in the first target spatiotemporal road condition map can be converted into a spatiotemporal coordinate by the following formula ( , ).

[0144] .

[0145] .

[0146] Where, (x, y) is any space-time traffic coordinate in the target space-time traffic map; ( , ) is the space-time coordinate; Indicates the preset spatial resolution; Indicates the preset time resolution; Indicates the target starting point pile number among the target starting and ending pile numbers corresponding to the target road section; Indicates the target start time in the target analysis time period corresponding to the target segment.

[0147] S104, or, performing morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and correspondingly merging the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

[0148] In an embodiment of the application, when the data sources corresponding to the two traffic condition data sets to be merged are two full road condition data sets, at least one second target congestion coordinate set corresponding to the target road section can be obtained through another implementation method, that is, based on the spatiotemporal road condition map corresponding to each full road condition data set, at least one congestion coordinate set corresponding to each full road condition data set is obtained; then, the congestion coordinate set corresponding to each full road condition data set is correspondingly merged to obtain at least one second target congestion coordinate set corresponding to the target road section.

[0149] In one implementation of the present application, during specific implementation, step S104 may include:

[0150] S1041. When the data sources corresponding to the two traffic condition data sets to be merged are two full traffic condition data sets, morphological operations and connected domain analysis are performed on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be merged in the spatiotemporal road condition graphs.

[0151] In this step, when the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets, first, threshold segmentation is performed on each spatiotemporal road condition map to convert the spatiotemporal road condition map into a fourth spatiotemporal road condition map represented as a binary image; then, a morphological operation including an opening operation and a closing operation is performed on the fourth spatiotemporal road condition map using a preset convolution kernel to obtain a fifth spatiotemporal road condition map converted from the fourth spatiotemporal road condition map, so as to denoise and fill holes in the fourth spatiotemporal road condition map; finally, a connected domain analysis is performed on the fifth spatiotemporal road condition map using a preset multi-neighborhood definition to determine a congestion boundary box in the fifth spatiotemporal road condition map, so as to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition map.

[0152] S1042. Based on the congestion coordinate set corresponding to each of the traffic condition data sets to be merged, a preset similarity calculation formula is used to obtain the similarity between each of the congestion coordinate sets, so as to obtain a first similarity matrix between each of the congestion coordinate sets.

[0153] In this step, at least one congestion coordinate set corresponding to each of the traffic condition data sets to be merged is input into a preset similarity calculation formula, and the similarity corresponding to each of the congestion coordinate sets is calculated by the formula to obtain a first similarity matrix corresponding to each of the congestion coordinate sets.

[0154] For example, assuming that the first traffic condition data set to be fused corresponds to 2 congestion coordinate sets, and the second traffic condition data set to be fused corresponds to 3 congestion coordinate sets, then the first similarity matrix is ​​a 5*5 matrix.

[0155] In the embodiment of the present application, the expression of the preset similarity calculation formula is as follows.

[0156] .

[0157] Among them, IoU (Intersection over Union) represents the corresponding similarity between each of the congestion coordinate sets; A represents at least one congestion coordinate set corresponding to a traffic condition data set to be fused; B represents at least one congestion coordinate set corresponding to another traffic condition data set to be fused.

[0158] Here, the first similarity matrix includes the congestion coordinate set corresponding to each of the traffic condition data sets to be fused, and the similarities corresponding to each of the congestion coordinate sets.

[0159] S1043: sort the similarities in descending order according to the quantity, and determine whether the first similarity in the descending order is greater than a second preset similarity threshold.

[0160] In this step, the multiple similarity values ​​are sorted in descending order according to the obtained multiple similarity values, and the first similarity with the largest number is determined to determine whether the first similarity is greater than a second preset similarity threshold.

[0161] Here, the first similarities with the largest number are selected to determine two congestion coordinate sets with the largest overlap between one traffic condition data set to be fused and another data set to be fused.

[0162] Furthermore, it is determined whether the first similarity is greater than a second preset similarity threshold to ensure the reliability of the fused data; wherein the second preset similarity threshold can be set according to the specific data conditions and specific road section of the traffic road condition data set.

[0163] S1044: If it is greater than, performing a union calculation on the congestion coordinate sets corresponding to the first similarity to obtain a first target coordinate set, and replacing the congestion coordinate set corresponding to the first similarity with the first target coordinate set to obtain an updated second similarity matrix.

[0164] In this step, if the first similarity is greater than the second preset similarity threshold, it indicates that the reliability of the two congestion coordinate sets corresponding to the first similarity is high, and the congestion coordinate sets corresponding to the first similarity can be unioned, that is, the two congestion boundary boxes corresponding to the first similarity are fused to obtain the first target coordinate set.

[0165] Furthermore, the two congested coordinate sets corresponding to the first similarity are deleted, and the obtained first target coordinate set is added to the congested coordinate set corresponding to the two congested coordinate sets corresponding to the first similarity in the spatio-temporal road condition map, and in the first similarity matrix, the first similarity and the two congested coordinate sets corresponding to the first similarity are respectively updated based on the first target coordinate set to obtain an updated second similarity matrix, that is, the two congested bounding boxes are replaced by the target congested bounding box obtained by fusion of the two congested bounding boxes.

[0166] S1045. Repeat similarity calculation, threshold determination and union calculation based on the second similarity matrix until there is no congested coordinate set on which union calculation can be performed in the congested coordinate set, so as to obtain at least one second target congested coordinate set corresponding to the target road section.

[0167] In an embodiment of the present application, the core idea of ​​fusing the congestion coordinate sets corresponding to each traffic condition data set to be fused includes: merging the congestion borders corresponding to the congestion coordinate sets with high overlap. If there is no overlap between the congestion border and other congestion borders, it means that the congestion border (congestion coordinate set) is unreliable, so the congestion border (congestion coordinate set) is discarded.

[0168] Here, since fusing the congestion coordinate sets corresponding to each traffic condition data set to be fused is a continuous cycle, it is necessary to re-add the merged congestion coordinate sets to the data set of the congestion coordinate sets.

[0169] In this step, after obtaining the first target coordinate set and the second similarity matrix, a second similarity is selected in the second similarity matrix and the second similarity is compared with a second preset similarity threshold, and a union calculation is performed on two congestion coordinate sets corresponding to the second similarity greater than the second preset similarity threshold to obtain a second target coordinate set, and so on, until there are no two congestion coordinate sets that meet the judgment condition of the second preset similarity threshold, that is, there is no congestion coordinate set that can be union calculated among all congestion coordinate sets, and then at least one second target congestion coordinate set corresponding to the target section spatiotemporal road condition map can be obtained.

[0170] For example, see Figure 4 , Figure 4 This is a schematic diagram of a fusion of a congestion coordinate set provided in an embodiment of the present application. Figure 4As shown in , each traffic condition data set to be fused corresponds to at least one congestion coordinate set, wherein the congestion coordinate sets A and B correspond to one traffic condition data set to be fused; the congestion coordinate sets C and D correspond to another traffic condition data set to be fused. Here, the two congestion coordinate sets corresponding to the first similarity are the congestion coordinate set A and the congestion coordinate set C, and the congestion boundary boxes corresponding to the congestion coordinate sets A and C spatiotemporal road condition maps are fused to obtain the second target congestion coordinate set.

[0171] In another possible implementation of the application, in response to receiving an external congestion dataset inputted from outside, for a traffic condition dataset to be fused whose data source is a full road condition dataset, at least one congestion coordinate set corresponding to the full road condition dataset can be fused based on the external congestion dataset to obtain at least one third target congestion coordinate set corresponding to the target road section.

[0172] Specifically, first, the external congestion data set is rasterized using the formula and parameters described in step S1022 to obtain at least one external congestion coordinate set corresponding to the external congestion data set; finally, based on at least one congestion coordinate set corresponding to the full road condition data set and at least one external congestion coordinate set, the at least one congestion coordinate set and at least one external congestion coordinate set are merged in the form of a congestion border through the above-mentioned congestion coordinate set fusion method to obtain at least one third target congestion coordinate set corresponding to the target road section.

[0173] See also Figure 5 , Figure 5 This is a second flowchart of a method for fusing traffic data provided in an embodiment of the present application.

[0174] like Figure 5 As shown in, optionally, in addition to the traffic data fusion method described in steps S101 to S104, step S105 is also included. Specifically, step S105 is used to illustrate a method for verifying at least one target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

[0175] S105. Based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set, multiple coordinate set matching calculations are performed using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

[0176] In an embodiment of the present application, multiple coordinate set matching calculations are performed on at least one target congestion coordinate set using at least one preset calibrated congestion coordinate set and a first preset similarity threshold, so that the precision and recall rate corresponding to the target congestion coordinate set can be obtained, which is used to verify the accuracy of the target congestion coordinate set.

[0177] Here, the data source and data basis of the preset at least one calibrated congestion coordinate set may include but are not limited to public transportation data or manually reported data with high reliability, and rasterization processing is performed based on the data to obtain at least one calibrated congestion coordinate set corresponding to the data. In one embodiment of the present application, in specific implementation, step S105 may include:

[0178] S1051. Based on at least one of the target congestion coordinate sets and a preset calibrated congestion coordinate set, a preset similarity calculation formula is used to calculate a similarity matrix corresponding to the target congestion coordinate set and the calibrated congestion coordinate set.

[0179] In this step, a similarity matrix corresponding to each target congestion coordinate set in at least one target congestion coordinate set and a preset calibration congestion coordinate set is calculated using a preset similarity calculation formula.

[0180] The similarity matrix includes the similarities between each target congestion coordinate set and a preset calibrated congestion coordinate set; the calibrated congestion coordinate set includes a plurality of elements.

[0181] S1052. For each of the target congestion coordinate sets, based on the similarity corresponding to the target congestion coordinate set in the similarity matrix, determine whether the target congestion coordinate set meets the matching condition corresponding to the calibrated congestion coordinate set, and obtain a judgment result; wherein the matching condition includes: the similarity corresponding to the target congestion coordinate set is greater than a first preset similarity threshold and the similarity is a maximum value among multiple similarities corresponding to the elements in the calibrated congestion coordinate set.

[0182] In this step, whether the similarity corresponding to each target congestion coordinate set in the similarity matrix is ​​greater than a first preset similarity threshold and whether the similarity is a maximum value among multiple similarities corresponding to the elements in the calibrated congestion coordinate set is determined to determine whether the target congestion coordinate set meets the matching conditions corresponding to the calibrated congestion coordinate set, and a judgment result is obtained.

[0183] S1053. Based on the judgment result, determine respectively a first number of target congestion coordinate sets that meet the matching condition and a second number of target congestion coordinate sets that do not meet the matching condition in at least one of the target congestion coordinate sets, as well as the number of unmatched elements in the calibration congestion coordinate set.

[0184] In this step, based on the multiple groups of judgment results obtained, the first number of target congested coordinate sets corresponding to the at least one target congested coordinate set whose similarity is greater than the first preset similarity threshold and the similarity is a maximum value among multiple similarities corresponding to the elements in the calibrated congested coordinate set, and the second number of target congested coordinate sets corresponding to the at least one target congested coordinate set whose similarity is less than or equal to the first preset similarity threshold or the similarity is not a maximum value among multiple similarities corresponding to the elements in the calibrated congested coordinate set are determined respectively.

[0185] Further, after matching of at least one target congestion coordinate set is completed, the number of elements in the calibration congestion coordinate set that are not successfully matched with the target congestion coordinate set is determined.

[0186] Exemplarily, for each target congestion coordinate set y in the set Y, each target congestion coordinate set y is matched to an element x in the calibrated congestion coordinate set X whose similarity (IoU) is greater than a first preset similarity threshold (iouThrs) and whose similarity (IoU) is the largest, and the elements in the calibrated congestion coordinate set X cannot be matched repeatedly; if a match is not possible, the target congestion coordinate set y has no corresponding matching element.

[0187] S1054: Based on the first number, the second number, and the number of elements, use a preset precision calculation formula and a recall calculation formula to respectively obtain a precision and a recall rate corresponding to at least one of the target congestion coordinate sets.

[0188] In this step, the obtained first number, second number and number of elements are respectively input into the preset precision calculation formula and recall calculation formula to obtain the corresponding calculation results, and the calculation results are further averaged to obtain the precision and recall corresponding to the at least one target congestion coordinate set.

[0189] In the embodiment of the present application, the expression of the preset accuracy calculation formula is as follows.

[0190] .

[0191] in, Indicates the accuracy of the target congestion coordinate set; Indicates a first number of target congestion coordinate sets in the target congestion coordinate set that meet the matching condition; Indicates the second number corresponding to the target congestion coordinate set that does not meet the matching condition.

[0192] The expression of the preset recovery rate calculation formula is as follows.

[0193] .

[0194] in, represents the recall rate corresponding to the target congestion coordinate set; Indicates a first number of target congestion coordinate sets in the target congestion coordinate set that meet the matching condition; Indicates the number of unmatched elements in the calibrated congestion coordinate set.

[0195] To summarize, the traffic data fusion method provided in the embodiment of the present application fuses traffic data from two decision-making levels: traffic conditions and traffic congestion. It uses a rasterized spatiotemporal road condition map to achieve the fusion of multi-source road condition data, and uses rasterized congestion information to achieve the fusion of multi-source congestion data. It targets data sources collected in various forms and is not restricted by the underlying road section division when fusing traffic data.

[0196] The traffic data fusion method provided in the embodiment of the present application collects multiple traffic road condition data sets from multiple data sources for a target section in a highway, and obtains a spatiotemporal road condition map corresponding to each traffic road condition data set to be fused by rasterizing two of the traffic road condition data sets to be fused respectively, and the spatiotemporal road condition maps are correspondingly fused based on the data sources corresponding to the two traffic road condition data sets to be fused to obtain a target spatiotemporal road condition map corresponding to the target section and determine at least one target congestion coordinate set corresponding to the target spatiotemporal road condition map, or for two full road condition data sets, morphological operations and connected domain analysis are performed on each spatiotemporal road condition map to obtain at least one congestion coordinate set corresponding to each spatiotemporal road condition map and further the congestion coordinate sets are fused, thereby realizing multi-source data fusion at the decision-making level of multi-source traffic conditions and traffic congestion, improving the accuracy and reliability of traffic data fusion, and thereby improving the efficiency of traffic management.

[0197] See also Figure 6 , Figure 7 , Figure 6 This is one of the structural schematic diagrams of a traffic data fusion device provided in an embodiment of the present application. Figure 7 The second structural diagram of a traffic data fusion device provided in an embodiment of the present application.

[0198] like Figure 6 As shown in , the fusion device 600 comprises:

[0199] The data acquisition module 610 is used to acquire multiple traffic condition data sets collected for a target road section in the highway; wherein the multiple traffic condition data include at least one full traffic condition data set collected based on the whole network equipment and at least one local traffic condition data set collected for a local road section in the target road section;

[0200] A raster processing module 620 is used for performing rasterization processing on two traffic condition data sets to be fused among the multiple traffic condition data sets, respectively, to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused;

[0201] When the grid processing module 620 is used to perform grid processing on each of the traffic condition data sets to be fused, and obtain the spatiotemporal traffic condition graph corresponding to each of the traffic condition data sets to be fused, the grid processing module 620 is used to:

[0202] Calculate the image range of the spatiotemporal road condition map based on the preset spatiotemporal resolution and the target start and end stake numbers and target analysis time period corresponding to the target road section;

[0203] For each of the traffic condition data sets to be integrated, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be integrated in the traffic condition data set to be integrated;

[0204] Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a plurality of spatiotemporal road condition coordinates corresponding to the traffic condition data sets to be fused are obtained using a preset coordinate conversion formula and a preset resolution;

[0205] Based on the multiple spatiotemporal road condition coordinates, the traffic state information and the image range, obtaining a spatiotemporal road condition map corresponding to each of the traffic road condition data sets to be fused;

[0206] The first fusion module 630 is used to fuse the spatiotemporal traffic graphs based on the data sources corresponding to the two traffic data sets to be fused, respectively, to obtain a target spatiotemporal traffic graph of the target road section, and to perform morphological operations and connected domain analysis on the target spatiotemporal traffic graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic graph;

[0207] The second fusion module 640 is used to perform morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and to fuse the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

[0208] Furthermore, the fusion device 600 further includes a data verification module 650, and the data verification module 650 is used to:

[0209] Based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set, multiple coordinate set matching calculations are performed using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

[0210] Further, when the data verification module 650 is used to perform multiple coordinate set matching calculations based on at least one of the target congestion coordinate sets and at least one preset calibration congestion coordinate set using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set, the data verification module 650 is used to:

[0211] Based on at least one of the target congestion coordinate sets and a preset calibrated congestion coordinate set, using a preset similarity calculation formula, a similarity matrix corresponding to the target congestion coordinate set and the calibrated congestion coordinate set is calculated;

[0212] For each of the target congestion coordinate sets, based on the similarity corresponding to the target congestion coordinate set in the similarity matrix, it is determined whether the target congestion coordinate set meets the matching condition corresponding to the calibration congestion coordinate set, and a determination result is obtained; wherein the matching condition includes: the similarity corresponding to the target congestion coordinate set is greater than a first preset similarity threshold and the similarity is a maximum value among multiple similarities corresponding to elements in the calibration congestion coordinate set;

[0213] Based on the judgment result, respectively determining a first number of target congestion coordinate sets that meet the matching condition and a second number of target congestion coordinate sets that do not meet the matching condition in at least one of the target congestion coordinate sets, as well as the number of unmatched elements in the calibration congestion coordinate set;

[0214] Based on the first number, the second number and the number of elements, the accuracy and recall rate corresponding to at least one of the target congestion coordinate sets are respectively obtained using a preset accuracy calculation formula and a recall rate calculation formula.

[0215] Furthermore, when the first fusion module 630 is used to fuse the spatiotemporal traffic graphs based on the data sources corresponding to the two traffic condition data sets to be fused to obtain the target spatiotemporal traffic graph of the target road section, the first fusion module 630 is used to:

[0216] When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets or one full traffic condition data set and one partial traffic condition data set, determining the weight values ​​corresponding to the traffic condition data sets to be fused respectively;

[0217] Based on the pixel values ​​and the weight values ​​respectively corresponding to the spatiotemporal road condition graphs, the spatiotemporal road condition graphs are correspondingly fused by using a mean overlap fusion method to obtain a to-be-processed spatiotemporal road condition graph of the target road section;

[0218] The overlapping boundary area of ​​the to-be-processed spatiotemporal traffic condition map is smoothed to obtain a target spatiotemporal traffic condition map of the target road section.

[0219] Furthermore, when the first fusion module 630 is used to perform morphological operations and connected domain analysis on the target spatiotemporal traffic map to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic map, the first fusion module 630 is used to:

[0220] Performing threshold segmentation on the target spatiotemporal traffic map to convert the spatiotemporal traffic map into a first spatiotemporal traffic map represented as a binary image;

[0221] Using a preset convolution kernel to perform a morphological operation including an opening operation and a closing operation on the first spatiotemporal road condition map, to obtain a second spatiotemporal road condition map converted from the first spatiotemporal road condition map, so as to denoise and fill holes in the first spatiotemporal road condition map;

[0222] Performing a connected domain analysis on the second spatiotemporal traffic map using a preset multi-neighborhood definition to determine a congestion boundary box in the second spatiotemporal traffic map;

[0223] Based on the congestion boundary box, at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal road condition map is obtained.

[0224] Further, when the second fusion module 640 is used to perform morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and to correspondingly fuse the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set spatiotemporal road condition graph corresponding to the target road section, the second fusion module 640 is used to:

[0225] When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets, morphological operations and connected domain analysis are performed on each of the spatiotemporal traffic condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal traffic condition graphs;

[0226] Based on the congestion coordinate set corresponding to each of the traffic condition data sets to be fused, a preset similarity calculation formula is used to obtain the similarity between each of the congestion coordinate sets, so as to obtain a first similarity matrix between each of the congestion coordinate sets;

[0227] Sorting the similarities in descending order according to the quantity, and determining whether the first similarity in the descending order is greater than a second preset similarity threshold;

[0228] If it is greater than, performing a union calculation on the congestion coordinate sets corresponding to the first similarity to obtain a first target coordinate set, and replacing the congestion coordinate set corresponding to the first similarity with the first target coordinate set to obtain an updated second similarity matrix;

[0229] Based on the second similarity matrix, similarity calculation, threshold judgment and union calculation are repeated until there is no congested coordinate set in the congested coordinate set that can be used for union calculation, so as to obtain at least one second target congested coordinate set corresponding to the target road section.

[0230] The traffic data fusion device provided in the embodiment of the present application collects multiple traffic road condition data sets from multiple data sources for a target section in a highway, and obtains a spatiotemporal road condition map corresponding to each traffic road condition data set to be fused by rasterizing two of the traffic road condition data sets to be fused respectively, and based on the data sources corresponding to the two traffic road condition data sets to be fused respectively, the spatiotemporal road condition maps are correspondingly fused to obtain a target spatiotemporal road condition map corresponding to the target section and determine at least one target congestion coordinate set corresponding to the target spatiotemporal road condition map, or for two full road condition data sets, morphological operations and connected domain analysis are performed on each spatiotemporal road condition map to obtain at least one congestion coordinate set corresponding to each spatiotemporal road condition map and further the congestion coordinate sets are fused, thereby realizing multi-source data fusion of multi-source traffic conditions and traffic congestion at the decision-making level, improving the accuracy and reliability of traffic data fusion, and thereby improving the efficiency of traffic management.

[0231] See also Figure 8 , Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 8 As shown in , the electronic device 800 includes a processor 810 , a memory 820 and a bus 830 .

[0232] The memory 820 stores machine-readable instructions executable by the processor 810. When the electronic device 800 is running, the processor 810 communicates with the memory 820 via the bus 830. When the machine-readable instructions are executed by the processor 810, the above-mentioned Figure 1 as well as Figure 5 The specific implementation of the steps of the traffic data fusion method in the method embodiment shown can be found in the method embodiment, which will not be repeated here.

[0233] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 as well as Figure 5 The specific implementation of the steps of the traffic data fusion method in the method embodiment shown can be found in the method embodiment, which will not be repeated here.

[0234] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0235] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0236] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0237] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0238] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for fusing traffic data, characterized in that: The fusion method comprises: Acquire multiple traffic condition data sets collected for a target section of the highway; wherein the multiple traffic condition data include at least one full traffic condition data set collected based on all network devices and at least one local traffic condition data set collected for a local section of the target section; For two traffic condition data sets to be fused among the multiple traffic condition data sets, rasterizing each of the traffic condition data sets to be fused to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused; The step of performing rasterization processing on each of the traffic condition data sets to be fused to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused includes: Calculate the image range of the spatiotemporal road condition map based on the preset spatiotemporal resolution and the target start and end stake numbers and target analysis time period corresponding to the target road section; For each of the traffic condition data sets to be integrated, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be integrated in the traffic condition data set to be integrated; Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a plurality of spatiotemporal road condition coordinates corresponding to the traffic condition data sets to be fused are obtained using a preset coordinate conversion formula and a preset resolution; Based on the multiple spatiotemporal road condition coordinates, the traffic state information and the image range, obtaining a spatiotemporal road condition map corresponding to each of the traffic road condition data sets to be fused; Based on the data sources corresponding to the two traffic condition data sets to be fused, the spatiotemporal traffic condition graphs are correspondingly fused to obtain a target spatiotemporal traffic condition graph of the target road section, and morphological operations and connected domain analysis are performed on the target spatiotemporal traffic condition graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic condition graph; Alternatively, morphological operations and connected domain analysis are performed on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused are correspondingly fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

2. The method according to claim 1, characterized in that After obtaining at least one target congestion coordinate set, the fusion method further includes: Based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set, multiple coordinate set matching calculations are performed using a first preset similarity threshold to obtain the precision and recall rate corresponding to the target congestion coordinate set to verify the accuracy of the target congestion coordinate set.

3. The method according to claim 2, characterized in that The method of performing multiple coordinate set matching calculations based on at least one of the target congestion coordinate sets and at least one preset calibrated congestion coordinate set using a first preset similarity threshold to obtain the accuracy and recall rate corresponding to the target congestion coordinate set includes: Based on at least one of the target congestion coordinate sets and a preset calibrated congestion coordinate set, using a preset similarity calculation formula, a similarity matrix corresponding to the target congestion coordinate set and the calibrated congestion coordinate set is calculated; For each of the target congestion coordinate sets, based on the similarity corresponding to the target congestion coordinate set in the similarity matrix, it is determined whether the target congestion coordinate set meets the matching condition corresponding to the calibration congestion coordinate set, and a determination result is obtained; wherein the matching condition includes: the similarity corresponding to the target congestion coordinate set is greater than a first preset similarity threshold and the similarity is a maximum value among multiple similarities corresponding to elements in the calibration congestion coordinate set; Based on the judgment result, respectively determining a first number of target congestion coordinate sets that meet the matching condition and a second number of target congestion coordinate sets that do not meet the matching condition in at least one of the target congestion coordinate sets, as well as the number of unmatched elements in the calibration congestion coordinate set; Based on the first number, the second number and the number of elements, the accuracy and recall rate corresponding to at least one of the target congestion coordinate sets are respectively obtained using a preset accuracy calculation formula and a recall rate calculation formula.

4. The method according to claim 1, characterized in that: The step of fusing the spatiotemporal traffic condition graphs based on the data sources respectively corresponding to the two traffic condition data sets to be fused to obtain the target spatiotemporal traffic condition graph of the target road section includes: When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets or one full traffic condition data set and one partial traffic condition data set, determining the weight values ​​corresponding to the traffic condition data sets to be fused respectively; Based on the pixel values ​​and the weight values ​​respectively corresponding to the spatiotemporal road condition graphs, the spatiotemporal road condition graphs are correspondingly fused by using a mean overlap fusion method to obtain a to-be-processed spatiotemporal road condition graph of the target road section; The overlapping boundary area of ​​the to-be-processed spatiotemporal traffic condition map is smoothed to obtain a target spatiotemporal traffic condition map of the target road section.

5. The method according to claim 1, characterized in that The step of performing morphological operations and connected domain analysis on the target spatiotemporal traffic condition graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic condition graph includes: Performing threshold segmentation on the target spatiotemporal traffic map to convert the spatiotemporal traffic map into a first spatiotemporal traffic map represented as a binary image; Using a preset convolution kernel to perform a morphological operation including an opening operation and a closing operation on the first spatiotemporal road condition map, to obtain a second spatiotemporal road condition map converted from the first spatiotemporal road condition map, so as to denoise and fill holes in the first spatiotemporal road condition map; Performing a connected domain analysis on the second spatiotemporal traffic map using a preset multi-neighborhood definition to determine a congestion boundary box in the second spatiotemporal traffic map; Based on the congestion boundary box, at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal road condition map is obtained.

6. The method according to claim 1, characterized in that The step of performing morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and correspondingly fusing the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set spatiotemporal road condition graph corresponding to the target road section includes: When the data sources corresponding to the two traffic condition data sets to be fused are two full traffic condition data sets, morphological operations and connected domain analysis are performed on each of the spatiotemporal traffic condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal traffic condition graphs; Based on the congestion coordinate set corresponding to each of the traffic condition data sets to be fused, a preset similarity calculation formula is used to obtain the similarity between each of the congestion coordinate sets, so as to obtain a first similarity matrix between each of the congestion coordinate sets; Sorting the similarities in descending order according to the quantity, and determining whether the first similarity in the descending order is greater than a second preset similarity threshold; If it is greater than, performing a union calculation on the congestion coordinate sets corresponding to the first similarity to obtain a first target coordinate set, and replacing the congestion coordinate set corresponding to the first similarity with the first target coordinate set to obtain an updated second similarity matrix; Based on the second similarity matrix, similarity calculation, threshold judgment and union calculation are repeated until there is no congested coordinate set in the congested coordinate set that can be used for union calculation, so as to obtain at least one second target congested coordinate set corresponding to the target road section.

7. A traffic data fusion device, characterized in that: The fusion device comprises: A data acquisition module, used to acquire multiple traffic condition data sets collected for a target section in the highway; wherein the multiple traffic condition data include at least one full traffic condition data set collected based on the whole network equipment and at least one local traffic condition data set collected for a local section in the target section; A raster processing module, for performing rasterization processing on two traffic condition data sets to be fused among the multiple traffic condition data sets, respectively, to obtain a spatiotemporal traffic condition map corresponding to each of the traffic condition data sets to be fused; When the grid processing module is used to perform grid processing on each of the traffic condition data sets to be fused, and obtain the spatiotemporal traffic condition graphs corresponding to each of the traffic condition data sets to be fused, the grid processing module is used to: Calculate the image range of the spatiotemporal road condition map based on the preset spatiotemporal resolution and the target start and end stake numbers and target analysis time period corresponding to the target road section; For each of the traffic condition data sets to be integrated, determine the start and end pile numbers, start and end time periods, and traffic status information corresponding to each piece of traffic condition data to be integrated in the traffic condition data set to be integrated; Based on the start and end pile numbers and the start and end time periods corresponding to each piece of traffic condition data to be fused, a plurality of spatiotemporal road condition coordinates corresponding to the traffic condition data sets to be fused are obtained using a preset coordinate conversion formula and a preset resolution; Based on the multiple spatiotemporal road condition coordinates, the traffic state information and the image range, obtaining a spatiotemporal road condition map corresponding to each of the traffic road condition data sets to be fused; A first fusion module is used to fuse the spatiotemporal traffic graphs based on the data sources corresponding to the two traffic data sets to be fused, respectively, to obtain a target spatiotemporal traffic graph of the target road section, and to perform morphological operations and connected domain analysis on the target spatiotemporal traffic graph to obtain at least one first target congestion coordinate set corresponding to the target road section in the target spatiotemporal traffic graph; The second fusion module is used to perform morphological operations and connected domain analysis on each of the spatiotemporal road condition graphs to obtain at least one congestion coordinate set corresponding to each of the traffic condition data sets to be fused in the spatiotemporal road condition graph, and to fuse the congestion coordinate sets corresponding to each of the traffic condition data sets to be fused to obtain at least one second target congestion coordinate set corresponding to the target road section.

8. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the traffic data fusion method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the traffic data fusion method according to any one of claims 1 to 6 are executed.

Images