Systems, methods, and machine-readable storage media for cloud processing of vehicle data

Abstract

The application provides a system, a method and a machine readable storage medium for cloud processing of vehicle data. The method comprises: obtaining at least part of vehicle data uploaded by an information collection vehicle, wherein the vehicle data comprises: vehicle body data, positioning trajectory data and camera collected data; mapping vehicle data occurring at similar times to the same data identifier respectively; performing semantic analysis on the vehicle data and adding a data label to establish another mapping relationship; and aggregating data identifiers and data labels within a preset range according to a preset rule to generate a main label, and establishing a further mapping relationship between the main label and the data identifiers and the data labels, thereby obtaining a multi-level mapping relationship between the main label, the data identifiers and the data labels, and the vehicle data. Through this method, the vehicle data is integrated and processed by labeling, thereby improving the execution efficiency of subsequent playback of the vehicle data.

Description

Technical Field

[0001] This invention relates to the fields of map navigation and autonomous driving, and in particular to a system, method, and machine-readable storage medium for cloud-based vehicle data processing. Background Technology

[0002] With the continuous development of autonomous driving technology, high-precision map data has become one of the key capabilities for vehicles to achieve high-level autonomous driving. Because autonomous driving algorithms have extremely high requirements for the richness, freshness, and accuracy of map data, high-precision map providers need to arrange information collection vehicles to collect high-precision map data periodically to maintain these qualities. However, as the frequency and scope of collection continue to expand, effective status monitoring and management of map collection vehicles are necessary to consistently improve their collection efficiency. Among these measures, vehicle data playback is crucial for remote vehicle monitoring and management, tracing abnormal events, analyzing the causes of map collection quality issues, and verifying mapping algorithm annotations.

[0003] Current methods for vehicle data playback typically rely on external GPS devices to upload GPS data for monitoring vehicle location and tracing the location of abnormal events. This method often only supports historical backtracking in terms of map geographic location and requires operators to conduct extensive, indiscriminate searches of vast amounts of trajectory data to identify the target trajectory. Therefore, establishing effective vehicle data playback capabilities is indispensable for the overall map data acquisition process. This invention aims to provide a fundamental guarantee for the efficiency and quality of subsequent map acquisition by proposing an effective vehicle data playback method. Summary of the Invention

[0004] One objective of this invention is to perform tagging and integration processing on the data uploaded by information collection vehicles.

[0005] A further objective of this invention is to improve the efficiency of vehicle data playback.

[0006] A further objective of this invention is to improve the comprehensiveness of vehicle data playback.

[0007] Specifically, the present invention provides a vehicle data cloud processing system, comprising:

[0008] The information collection vehicle includes at least one or more sensors among the vehicle positioning module, camera, and vehicle status module, and is capable of uploading vehicle data collected by at least one or more of the above sensors to a cloud playback device.

[0009] Cloud processing equipment, including:

[0010] The data interface is configured to receive vehicle data uploaded by information collection vehicles, and to establish a first mapping relationship between vehicle data that occur within the same preset time period and the same data identifier. The data identifier is generated according to the preset time period. The vehicle data includes: vehicle body data, positioning trajectory data and camera data.

[0011] The data analysis and processing module is configured to establish a second mapping relationship between preset vehicle body labels and vehicle body data, a third mapping relationship between preset positioning trajectory labels and positioning trajectory data, and a fourth mapping relationship between preset camera acquisition labels and camera acquisition data. The vehicle body labels are generated based on the predefined vehicle body data, the positioning trajectory labels are generated based on the predefined positioning trajectory data, and the camera acquisition labels are generated based on the predefined camera acquisition data.

[0012] The data tag integration module is configured to integrate data identifiers and data tags according to preset rules to obtain a master tag, and establish a fifth mapping relationship between the master tag and the data identifiers and data tags. The data tags include: vehicle body tags, positioning trajectory tags, and camera acquisition tags. Thus, through the first, second, third, fourth, and fifth mapping relationships, a multi-level mapping relationship between the master tag, data identifiers and data tags, and vehicle data is obtained. The master tag is used to identify data identifiers and data tags within a preset range.

[0013] Optionally, the data analysis and processing module is also configured to:

[0014] Extract feature-based positioning trajectory data that meets preset conditions from the positioning trajectory data;

[0015] Denoise the trajectory points in the feature localization trajectory data according to preset denoising rules;

[0016] The denoised feature localization trajectory data is analyzed and matched to preset trajectory labels.

[0017] Optionally, the cloud processing device also includes a data matching module and a data playback module:

[0018] The data matching module is configured to: acquire a data playback request, which includes the main label to be played back; select the main label for the corresponding time range based on the data playback request; acquire the target data label corresponding to the main label based on the multi-level mapping relationship; acquire the target vehicle data corresponding to the target data label based on the multi-level mapping relationship; extract non-feature positioning trajectory data that does not meet preset conditions from the positioning trajectory data; perform thinning processing on the non-feature positioning trajectory data according to preset thinning rules; and integrate the main label, target vehicle data, and thinned non-feature positioning trajectory data to obtain the playback data.

[0019] The data playback module is configured to render the feature-based positioning trajectory data and the thinned non-feature-based positioning trajectory data in the playback data to the map trajectory playback compiler according to the data playback request. When playing back the trajectory, the corresponding playback data is obtained according to the time point for synchronous playback display.

[0020] Optionally, the data analysis and processing module is also configured to: process the data collected by the camera to generate image data; perform similarity matching and classification on the image data; select a target image in each category of image data for separate semantic analysis, and synchronize the analysis results to other images in the category of the target image; extract video data or image data with corresponding preset abnormal labels from the data collected by the camera as abnormal data; and store the abnormal data in a preset area.

[0021] The data playback module is also configured to: obtain abnormal event playback requests, which include abnormal tags that need to be played back; obtain abnormal tags based on the abnormal event playback requests; obtain vehicle data during the occurrence of the abnormal event corresponding to the abnormal tag, and perform overall playback.

[0022] Optionally, the data analysis and processing module is also configured to:

[0023] Errors occurred in the acquisition of vehicle body data, positioning trajectory data, and camera-collected data;

[0024] Analyze the causes of errors in the erroneous data;

[0025] Establish a mapping relationship between the error cause and the error data, and generate prompt information.

[0026] According to another aspect of the present invention, a vehicle data cloud processing method is also provided, comprising:

[0027] Acquire at least some of the vehicle data uploaded by the information collection vehicle. The vehicle data includes: vehicle body data, positioning trajectory data, and camera data.

[0028] The vehicle data that occurred within the same preset time period are respectively associated with the same data identifier. The data identifier is generated according to the preset time period. The vehicle data includes: vehicle body data, positioning trajectory data and camera data.

[0029] Semantic analysis is performed on vehicle data to establish a second mapping relationship between preset vehicle body tags and vehicle body data, a third mapping relationship between preset positioning trajectory tags and positioning trajectory data, and a fourth mapping relationship between preset camera acquisition tags and camera acquisition data. Among them, vehicle body tags are predefined and generated based on vehicle body data, positioning trajectory tags are predefined and generated based on positioning trajectory data, and camera acquisition tags are predefined and generated based on camera acquisition data.

[0030] Data identifiers and data tags are stored according to data type, tag type, and time. Data tags include: vehicle body tags, location trajectory tags, and camera acquisition tags.

[0031] Data identifiers and data tags within a preset range are aggregated according to preset rules to generate a master label, and a fifth mapping relationship is established between the master label and the data identifiers and data tags. Thus, through the first, second, third, fourth, and fifth mapping relationships, a multi-level mapping relationship between the master label, data identifiers and data tags, and vehicle data is obtained. The master label is used to identify data identifiers and data tags within a preset range.

[0032] Optionally, the step of establishing a third mapping relationship between preset positioning trajectory labels and positioning trajectory data includes:

[0033] Extract feature-based positioning trajectory data that meets preset conditions from the positioning trajectory data;

[0034] Denoise the trajectory points in the feature localization trajectory data according to preset denoising rules;

[0035] The denoised feature localization trajectory data is analyzed and matched to preset trajectory labels.

[0036] Optionally, the step of establishing the fifth mapping relationship between the main label and the data identifier and the data label includes:

[0037] Get the data replay request, which includes the main tag that needs to be replayed.

[0038] Select the main tag for the corresponding time range based on the data playback request;

[0039] Obtain the target data label corresponding to the main label based on the multi-level mapping relationship;

[0040] Obtain the target vehicle data corresponding to the target data label based on the multi-level mapping relationship;

[0041] Extract non-feature positioning trajectory data from the positioning trajectory data that do not meet preset conditions;

[0042] Non-feature positioning trajectory data are thinned according to preset thinning rules;

[0043] The playback data is obtained by integrating the main label, target vehicle data, and thinned non-feature positioning trajectory data.

[0044] Based on the data playback request, the feature positioning trajectory data and the thinned non-feature positioning trajectory data in the playback data are rendered to the map trajectory playback compiler. When playing back the trajectory, the corresponding playback data is obtained according to the time point for synchronous playback display.

[0045] Optionally, the step of establishing a fourth mapping relationship between preset camera acquisition labels and camera acquisition data includes: processing the camera acquisition data to generate image data; performing similarity matching and classification on the image data; selecting a target image in each category of image data for separate semantic analysis, and synchronizing the analysis results to other images in the category of the target image; extracting video data or image data corresponding to preset abnormal labels from the camera acquisition data as abnormal data; and storing the abnormal data in a preset area.

[0046] After establishing the third mapping relationship between the main label and the data identifier and data label, the process also includes: obtaining an abnormal event playback request, which includes the abnormal label that needs to be played back; obtaining the abnormal label according to the abnormal event playback request; obtaining the vehicle data during the occurrence of the abnormal event corresponding to the abnormal label, and performing overall playback.

[0047] Optionally, the step of establishing the fifth mapping relationship between the main label and the data identifier and the data label includes:

[0048] Errors occurred in the acquisition of vehicle body data, positioning trajectory data, and camera-collected data;

[0049] Analyze the causes of errors in the erroneous data;

[0050] Establish a mapping relationship between the error cause and the error data, and generate prompt information.

[0051] According to another aspect of the present invention, a machine-readable storage medium is also provided, on which a machine-executable program is stored, wherein the machine-executable program, when executed by a processor, implements the vehicle data cloud processing method described above.

[0052] According to another aspect of the present invention, a computer device is also provided, including a memory, a processor, and a machine-executable program stored in the memory and running on the processor, wherein the processor executes the machine-executable program to implement the vehicle data cloud processing method described above.

[0053] In this invention, after receiving vehicle data uploaded by an information collection vehicle, vehicle data occurring within the same preset time period are associated with the same data identifier to establish a first mapping relationship. Semantic analysis is performed on the vehicle data, and data tags are created, along with a mapping relationship between the data tags and the vehicle data. Data identifiers and data tags are stored according to data type, tag type, and time. Data identifiers and data tags within a preset range are aggregated according to preset rules to generate a master tag, and a mapping relationship is established between the master tag and the data identifiers and data tags, thus obtaining a multi-level mapping relationship between the master tag, data identifiers and data tags, and vehicle data. This method performs tag-based integration processing on the data uploaded by the information collection vehicle, thereby improving the efficiency of subsequent vehicle data matching.

[0054] Furthermore, in the solution of this invention, when a data playback request is received, relevant target tags are selected according to the data playback request, playback data is matched according to the target tags, and the positioning trajectory data is rendered into the map trajectory playback compiler. When playing back the positioning trajectory data, corresponding vehicle body data, tag data, and camera-collected data are obtained according to the time point and displayed synchronously. When a filter tag request is received, the vehicle data corresponding to the selected tag is obtained according to the filter tag request and played back. When an abnormal event playback request is received, the abnormal tag is obtained according to the abnormal event playback request, the vehicle data during the occurrence of the abnormal event corresponding to the abnormal tag is obtained, and the entire event is played back. This solution provides multiple vehicle data playback methods, improving the comprehensiveness of vehicle data playback.

[0055] Furthermore, in the solution of this invention, when vehicle data is acquired, erroneous data that appears in the vehicle data is extracted; the cause of the erroneous data is analyzed; a mapping relationship is established between the cause of the error and the erroneous data, and a prompt message is generated. This method enables timely detection of faulty sensors or devices, ensuring the integrity of the system.

[0056] The above and other objects, advantages and features of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description

[0057] The following sections will describe some specific embodiments of the invention in a detailed manner by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0058] Figure 1 This is a schematic diagram of a vehicle data cloud processing system according to an embodiment of the present invention;

[0059] Figure 2 This is a schematic diagram of the architecture of a vehicle data cloud processing system according to an embodiment of the present invention;

[0060] Figure 3 This is a schematic diagram of vehicle data association data identification in a vehicle data cloud processing system according to an embodiment of the present invention;

[0061] Figure 4 This is a schematic diagram of the architecture of a data analysis and processing module of a vehicle data cloud processing system according to an embodiment of the present invention;

[0062] Figure 5 This is a schematic diagram of the mapping relationship after data tag integration in a vehicle data cloud processing method according to an embodiment of the present invention;

[0063] Figure 6 This is a functional diagram of the data playback module of a vehicle data cloud processing system according to an embodiment of the present invention;

[0064] Figure 7 This is a flowchart illustrating a method for cloud-based processing of vehicle data according to an embodiment of the present invention;

[0065] Figure 8 This is a schematic diagram of a machine-readable storage medium in a method for calibrating multi-sensor parameters according to an embodiment of the present invention; and

[0066] Figure 9 This is a schematic diagram of a computer device in a method for calibrating parameters of multiple sensors according to an embodiment of the present invention. Detailed Implementation

[0067] Those skilled in the art should understand that the embodiments described below are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. These partial embodiments are intended to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. Based on the embodiments provided by the present invention, all other embodiments obtained by those skilled in the art without creative effort should still fall within the scope of protection of the present invention.

[0068] Figure 1 This is a schematic diagram of a vehicle data cloud processing system according to an embodiment of the present invention. The vehicle data cloud processing system may include an information collection vehicle 110 and a cloud playback device 120.

[0069] The information collection vehicle 110 can be a collection vehicle equipped with a camera and / or a geographic information collection device, or other mobile collection vehicles. The information collection vehicle 110 has at least one or more sensors, including a vehicle positioning module, a camera, and a vehicle status module, and can upload vehicle data collected by at least one or more of these sensors to the cloud playback device 120. Collection vehicles include both professional collection vehicles and crowdsourced collection vehicles. Typically, professional collection vehicles are equipped with lidar, high-precision cameras, and high-precision positioning equipment, and are used by professional surveyors to collect and produce geographic information data, including lane-level road topology and feature information, through surveying methods. Crowdsourced collection vehicles, including private cars and buses, can use onboard sensors to monitor environmental changes in real time and upload vehicle data to the cloud playback device 120, thereby achieving comprehensive and rapid collection of vehicle data.

[0070] After receiving a data collection task instruction, the information collection vehicle 110 uploads at least a portion of the collected vehicle data to the cloud playback device 120. In some embodiments of this application, the information collection vehicle 110 can upload vehicle data according to a data upload request from the cloud playback device 120 or activate the data upload function according to a trigger event. In other embodiments of this application, the information collection vehicle 110 can automatically upload vehicle data according to a pre-configured system.

[0071] The cloud-based playback device 120 performs image AI algorithm analysis, positioning trajectory data analysis, and vehicle status data analysis on the received vehicle data. This allows for semantic analysis of the vehicle data and the definition of tags. After a data matching process, valid target data is selected. Finally, a map editing and playback front-end tool is used to effectively display the multi-dimensional data playback. This helps operations personnel quickly retrieve key trajectory events, vehicle status, and surrounding environment data, saving playback time and improving efficiency. It also provides a solid foundation for subsequent in-depth data mining, data mapping and annotation, and data quality review.

[0072] Figure 2 This is a schematic diagram of the architecture of a vehicle data cloud processing system according to an embodiment of the present invention. The information collection vehicle 110 is configured to perform a data collection task and upload at least a portion of the collected data to the data interface 121 of the cloud playback device 120. The collected data includes vehicle GPS data, camera data, vehicle status and other vehicle data of different dimensions.

[0073] The cloud playback device 120 includes: a data interface 121, a data analysis and processing module 122, a data tag integration module 123, a data matching module 124, and a data playback module 125.

[0074] Data interface 121 is configured to receive vehicle data uploaded by information collection vehicle 110. The appropriate data transmission and storage formats are selected according to the data type of the vehicle data. For example, vehicle body data or positioning trajectory data is usually in JSON format, so when receiving vehicle body data or positioning trajectory data, the appropriate data transmission and storage formats are selected. Camera-collected data is usually in JPG or AVI format, so when receiving camera-collected data, the appropriate data transmission and storage formats are usually selected according to the data type when the camera-collected data was uploaded. Those skilled in the art can extend the data type of vehicle data and the data transmission and storage formats according to the actual situation.

[0075] Data interface 121 is also configured to, while storing vehicle data, establish a first mapping relationship between vehicle data occurring within the same preset time period and the same data identifier, since the upload frequency of each data point is inconsistent. The data identifier can be set by those skilled in the art. Specific examples can be found in [link to example]. Figure 3 And the corresponding instruction manual description.

[0076] The data analysis and processing module 122 is configured to perform semantic analysis on vehicle data, add tags to the vehicle data, and establish mapping relationships. Vehicle data may include vehicle-related information such as vehicle body data, positioning trajectory data, and camera-collected data. Those skilled in the art can expand the vehicle data according to actual conditions.

[0077] During semantic analysis, vehicle body data is matched to preset vehicle body tags using a first preset analysis rule, establishing a second mapping relationship between the preset vehicle body tags and the vehicle body data. The first preset analysis rule can be customized. Vehicle body tags may include: sensor status tags (camera normal, camera abnormal), acquisition time characteristic tags, vehicle configuration information tags (7V, 5V, 3V), vehicle dashboard mileage tags, and vehicle dashboard speed tags, etc., which describe the vehicle's driving status.

[0078] For positioning trajectory data, it is necessary to first extract feature positioning trajectory data that meets preset conditions. In some embodiments of this application, feature positioning trajectory data may include positioning trajectory data with features such as curves and roundabouts. Since feature positioning trajectory data may be distorted due to positioning module accuracy or signal issues, it is necessary to denoise the trajectory points relative to the trajectory distance, time, and estimated speed. Then, trajectory analysis and trajectory tag matching are performed on the denoised positioning trajectory data. The trajectory tags include positioning trajectory data tags such as network signal strength, number of satellites, and continuity. In some embodiments of this application, a speed and distance offset can be defined. Trajectory points exceeding the offset are removed. The speed data uploaded by the nearest vehicle can be used as a reference point to remove noise points exceeding the offset. The denoised feature positioning trajectory data is stored in a separate database, and then trajectory analysis is performed to establish a third mapping relationship between preset positioning trajectory tags and positioning trajectory data.

[0079] For camera-captured data, the data is first processed into image data. Then, the image data is categorized by similarity matching. In each category, a target image is selected and subjected to a separate semantic analysis using a predefined image AI analysis algorithm. The analysis results are then synchronized to other images in the same category as the target image. Subsequently, a fourth mapping relationship is established between preset camera-captured tags and the camera-captured data. Image tags include: daytime, nighttime, elevated road, ground, traffic accident, obstacle, traffic sign, etc. In some embodiments of this application, if the camera-captured data is video data, frame extraction is required to obtain image data.

[0080] During the data acquisition and processing process of the camera, there are also preset anomaly tags. When there is data with the corresponding anomaly tag in the image data, the image data is extracted and stored separately to provide data support for subsequent anomaly data analysis and reduce the workload of subsequent data screening.

[0081] In some embodiments of this application, the information collection vehicle 110 may also extract partial tags from the vehicle data at the time of collecting the vehicle data.

[0082] The data analysis and processing module 122 is also configured to extract erroneous data from the acquired vehicle data; analyze the causes of the erroneous data; establish a mapping relationship between the causes of the errors and the erroneous data; and generate a prompt message. In some embodiments of this application, when the data analysis and processing module 122 detects incomplete or fragmented vehicle data, it determines that the data transmission channel or sensor has an error, and generates a corresponding prompt message. Those skilled in the art can then locate the corresponding sensor of the vehicle that acquired the information based on the prompt message to troubleshoot the problem. This method enables timely detection of problematic sensors or devices, ensuring the integrity of the system.

[0083] The data tag integration module 123 is configured to integrate the data identifiers generated by the data interface 121 and the data tags generated by the data analysis and processing module 122 according to preset rules and establish a multi-level mapping relationship. In some embodiments of this application, the data identifiers generated by the data interface 121 and the data tags generated by the data analysis and processing module 122 after semantic analysis of vehicle data can be stored according to data type, tag, and time. Then, data tags or data identifiers within the same range in a preset time period are aggregated to obtain the main tag. A fifth mapping relationship is established between the main tag and the data identifier and the data tag. Thus, a multi-level mapping relationship between the main tag, the data identifier and the data tag, and the vehicle data is obtained through the first mapping relationship, the second mapping relationship, the third mapping relationship, the fourth mapping relationship, and the fifth mapping relationship.

[0084] The data matching module 124 is configured to acquire a data playback request, which includes information such as the main label to be played back; select the main label for the corresponding time range according to the data playback request; acquire the target data label corresponding to the main label according to the multi-level mapping relationship; acquire the target vehicle data corresponding to the target data label according to the multi-level mapping relationship; extract non-feature positioning trajectory data that does not meet the preset conditions from the positioning trajectory data; perform thinning processing on the non-feature positioning trajectory data according to the preset thinning rules; and integrate the main label, target vehicle data, and thinned non-feature positioning trajectory data to obtain playback data; in some embodiments of this application, the playback data may include: vehicle body data, positioning trajectory data, and camera acquisition data. Since location trajectory data is usually continuous and dense, the modules mentioned above typically select feature-based location trajectory data that meet preset conditions for semantic analysis and labeling. Therefore, it is necessary to further thin out the non-feature trajectory data. The thinning rules can be set according to the specificity of the location point and the time. The actual road trajectory is matched using a method of approximation. At the straight line position of the trajectory, adjacent points are removed for thinning. Finally, thinning is performed according to the time range at time intervals. After thinning is completed, the main label, target vehicle data, and thinned non-feature location trajectory data are integrated to obtain the playback data.

[0085] The data playback module 125 is configured to render the positioning trajectory data obtained by the data matching module 124 to a preset map trajectory playback compiler. During the playback of the positioning trajectory data, corresponding vehicle data is obtained based on the time point and displayed synchronously. In some embodiments of this application, when playing back the positioning trajectory data, when the playback reaches a certain moment, other vehicle data at that moment is displayed synchronously. For example, when the vehicle is driving to a certain location, the camera data and vehicle body data at that moment are displayed synchronously, thus completely displaying the vehicle's status information at a certain moment.

[0086] The data playback module 125 is also configured to filter tags according to instructions, and supports displaying the original data playback of a single tag class or multiple tag classes.

[0087] The data playback module 125 is also configured to acquire an abnormal event playback request; acquire an abnormal tag based on the abnormal event playback request; acquire vehicle data during the occurrence of the abnormal event corresponding to the abnormal tag, and perform overall playback. In some embodiments, the abnormal tag in the camera can be used to match the corresponding vehicle data, thereby fully displaying all the vehicle's status information when the abnormality occurs.

[0088] Figure 3This is a schematic diagram of vehicle data association data identifiers in a vehicle data cloud processing system according to an embodiment of the present invention. The diagram includes: vehicle data 310-314 and data identifiers 320-321.

[0089] Among them, vehicle data 310 stores location trajectory data in JSON format and records its occurrence time as 2022-11-10 11:00:00; vehicle data 311 stores camera data in JPG format and records its occurrence time as 2022-11-10 11:00:30; and vehicle data 312 stores vehicle status data in JSON format and records its occurrence time as 2022-11-10 11:01:00. In this schematic diagram, a mapping relationship is established between vehicle data within 10 minutes and the same data identifier 320. The time range can be set by those skilled in the art. Finally, a mapping relationship is established between vehicle data 310-312 and data identifier 320. Data identifier 320 records the ID of this time period on the timeline and the data type of the corresponding vehicle data 310-312. Similarly, a mapping relationship is established between vehicle data 313-314 and data identifier 321. This facilitates rapid time matching of vehicle data in the subsequent data tag storage and integration stages, as well as the data matching stage.

[0090] Figure 4 This is a schematic diagram of the architecture of a data analysis and processing module of a vehicle data cloud processing system according to an embodiment of the present invention. A predefined rule engine 420 is configured in the data analysis and processing module 122, which defines the logic for data semantic analysis. In some embodiments of this application, camera data 410, after being processed by the predefined rule engine 420, can generate camera tag data 430; vehicle status data 411, after being processed by the predefined rule engine 420, can generate vehicle status tag data 431; positioning trajectory data 412, after being processed by the predefined rule engine 420, can generate positioning trajectory tag data 432; one optional format for the tag data is TXT format. Those skilled in the art can extend the data to other dimensions and corresponding semantic analysis logic based on the schematic diagram according to actual conditions.

[0091] Figure 5This diagram illustrates the mapping relationship after data tag integration in a vehicle data cloud processing method according to an embodiment of the present invention. The diagram includes: main tag data 510, which aggregates data tags of different types within the same range according to the principle of time proximity. Through the main tag data 510, corresponding camera tag data 430, vehicle body status tag data 431, and positioning trajectory tag data 432 can be found. Simultaneously, camera data 410 can be obtained from the camera tag data 430, vehicle body status data 411 can be obtained from the vehicle body status tag data 431, and positioning trajectory data 412 can be obtained from the positioning trajectory tag data 432. This mapping relationship improves the efficiency of data matching during data playback. Those skilled in the art can extend the data types according to actual conditions.

[0092] Figure 6 This is a functional diagram of the data playback module of a vehicle data cloud processing system according to an embodiment of the present invention. The data playback module 125 is configured to support synchronous data playback 610, filtered tag playback 611, and abnormal tag event playback 612. In some embodiments of this application, synchronous data playback 610 can render positioning trajectory data to a preset map trajectory playback compiler, and synchronously play back and display the corresponding vehicle data according to the time point when playing back the positioning trajectory data; filtered tag playback 611 can filter tags according to instructions, thereby displaying the vehicle data playback corresponding to a single tag class or multiple tags; abnormal tag event playback 612 can match the vehicle data at the time of the abnormal event according to the abnormal tag selected by the instructions, and perform overall playback.

[0093] Figure 7 This is a flowchart illustrating a method for cloud-based processing of vehicle data according to an embodiment of the present invention.

[0094] This process includes:

[0095] Step S701: Obtain at least a portion of the vehicle data uploaded by the information collection vehicle. This vehicle data may include positioning trajectory data, vehicle body data, and camera data. This step includes: Since the data structures and data types of the positioning trajectory data, vehicle body data, and camera data uploaded by the information collection vehicle may differ, it is necessary to store these different types of data in different databases. In this process, standard data transmission formats and storage formats need to be defined for different data types. In some embodiments of this application, the data format for vehicle body data may be JSON, the data format for positioning trajectory data may be JSON, and the data format for camera data may be JPG or AVI. Those skilled in the art can extend the data types and formats according to actual circumstances.

[0096] Step S702: Establish a first mapping relationship between vehicle data that occurred at similar times and the same data identifier. In this step, since the upload frequency of vehicle data varies, a mapping relationship is established between vehicle data that occurred at similar times and the same data identifier according to preset rules. The data identifier can be set by professionals in the field according to the actual situation. This facilitates rapid time matching of vehicle data in the subsequent data tag storage and integration stage and data matching stage.

[0097] Step S703: Perform semantic analysis on the vehicle data and establish a second mapping relationship between the preset vehicle body label and the vehicle body data, a third mapping relationship between the preset positioning trajectory label and the positioning trajectory data, and a fourth mapping relationship between the preset camera acquisition label and the camera acquisition data.

[0098] Step S704: Store the data identifier and data tag according to data type, tag type and time.

[0099] Step S705: Aggregate data identifiers and data tags within a preset range according to preset rules to generate a master tag, and establish a fifth mapping relationship between the master tag and the data identifiers and data tags, thereby obtaining a multi-level mapping relationship between the master tag, data identifiers and data tags, and vehicle data.

[0100] After the data is processed by the above method, the cloud device can obtain the corresponding playback data according to the data tag corresponding to the instruction and the multi-level mapping relationship after receiving the data playback instruction, thereby realizing the function of uploading and playing back vehicle data.

[0101] This embodiment also provides a machine-readable storage medium and a computer device. Figure 8 This is a schematic diagram of a machine-readable storage medium 801 according to an embodiment of the present invention. Figure 9 This is a schematic diagram of a computer device 903 according to an embodiment of the present invention.

[0102] A machine-readable storage medium 801 stores a machine-executable program 802 thereon, which, when executed by a processor, implements the vehicle data cloud processing method of any of the above embodiments.

[0103] Computer device 903 may include memory 901, processor 902 and machine-executable program 802 stored on memory 901 and running on processor 902, and processor 902 executes machine-executable program 802 to implement the vehicle data cloud processing method of any of the above embodiments.

[0104] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, such as semantic analysis, may be specifically implemented in any machine-readable storage medium for use by, or in conjunction with, instruction execution systems, apparatuses, or devices (such as computer-based systems, processor-based systems, or other systems that can fetch and execute instructions from, an instruction execution system, apparatus, or device).

[0105] For the purposes of this embodiment, the machine-readable storage medium 801 can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the machine-readable storage medium 801 include: an electrical connection (electronic device) having one or more wires, a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, the machine-readable storage medium 801 can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0106] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system.

[0107] Computer device 903 can be, for example, a server, desktop computer, laptop computer, tablet computer, or smartphone. In some examples, computer device 903 can be a cloud computing node. Computer device 903 can be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. Typically, program modules can include routines, programs, object programs, components, logic, data structures, etc., that perform specific tasks or implement specific abstract data types. Computer device 903 can be implemented in a distributed cloud computing environment where tasks are performed by remote processing devices linked via a communication network. In a distributed cloud computing environment, program modules can reside on local or remote computing system storage media, including storage devices.

[0108] Computer device 903 may include a processor 902 adapted to execute stored instructions and a memory 901 that provides temporary storage space for the operation of said instructions during operation. The processor 902 may be a single-core processor, a multi-core processor, a computing cluster, or any other configuration. The memory 901 may include random access memory (RAM), read-only memory, flash memory, or any other suitable storage system.

[0109] The processor 902 can be connected via a system interconnect (e.g., PCI, PCI-Express, etc.) to an I / O interface (input / output interface) suitable for connecting the computer device 903 to one or more I / O devices (input / output devices). I / O devices may include, for example, a keyboard and indicating devices, where indicating devices may include a touchpad or touchscreen, etc. I / O devices may be built into the computer device 903 or may be external devices connected to the computing device.

[0110] The processor 902 may also be linked via a system interconnect to a display interface suitable for connecting the computer device 903 to a display device. The display device may include a display screen as a built-in component of the computer device 903. The display device may also include an external computer monitor, television, or projector connected to the computer device 903. Furthermore, a network interface controller (NIC) may be adapted to connect the computer device 903 to a network via a system interconnect. In some embodiments, the NIC may use any suitable interface or protocol (such as an Internet Minicomputer System Interface) to transmit data. The network may be a cellular network, a radio network, a wide area network (WAN), a local area network (LAN), or the Internet, etc. Remote devices may connect to the computing device via the network.

[0111] The flowchart provided in this embodiment is not intended to indicate that the operations of the method will be performed in any particular order, or that all operations of the method are included in every case. Furthermore, the method may include additional operations. Within the scope of the technical concept provided by the method in this embodiment, additional variations can be made to the above method.

[0112] Therefore, those skilled in the art should recognize that although numerous exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications conforming to the principles of the present invention can be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the invention. Thus, the scope of the present invention should be understood and construed as covering all such other variations or modifications.

Claims

1. A vehicle data cloud processing system, comprising: an information collection vehicle comprising at least one or more sensors of a vehicle positioning module, a camera and a vehicle body state module, and capable of uploading vehicle data collected by the at least one or more sensors to a cloud playback device; a cloud processing device comprising: a data interface configured to receive the vehicle data uploaded by the information collection vehicle, and to establish a first mapping relationship between the vehicle data occurring at a same preset time period and a same data identifier respectively, the data identifier being generated according to the preset time period, the vehicle data comprising: vehicle body data, positioning trajectory data and camera collected data; a data analysis processing module configured to establish a second mapping relationship of a preset vehicle body label corresponding to the vehicle body data, a third mapping relationship of a preset positioning trajectory label corresponding to the positioning trajectory data, and a fourth mapping relationship of a preset camera collected label corresponding to the camera collected data, wherein the vehicle body label is generated by predefinition according to the vehicle body data, the positioning trajectory label is generated by predefinition according to the positioning trajectory data, and the camera collected label is generated by predefinition according to the camera collected data; a data label integration module configured to integrate the data identifier and the data label according to a preset rule to obtain a main label, and to establish a fifth mapping relationship of the main label corresponding to the data identifier and the data label, the data label comprising: the vehicle body label, the positioning trajectory label and the camera collected label, so as to obtain a multi-level mapping relationship of the main label, the data identifier and the data label, and the vehicle data through the first mapping relationship, the second mapping relationship, the third mapping relationship, the fourth mapping relationship and the fifth mapping relationship, the main label being used to identify the data identifier and the data label within a preset range.

2. The vehicle data cloud processing system of claim 1, wherein, The data analysis processing module is further configured to: extract feature positioning trajectory data in the positioning trajectory data that meet a preset condition; perform trajectory point denoising on the feature positioning trajectory data according to a preset denoising rule; perform trajectory analysis on the denoised feature positioning trajectory data and match to a preset trajectory label. 3.The vehicle data cloud processing system of claim 2, wherein the cloud processing device further comprises a data matching module and a data playback module: the data matching module is configured to: obtain a data playback request, the data playback request comprising a main label that needs to be played back; select the main label of a corresponding time range according to the data playback request; obtain a target data label corresponding to the main label according to the multi-level mapping relationship; obtain target vehicle data corresponding to the target data label according to the multi-level mapping relationship; extract non-feature positioning trajectory data in the positioning trajectory data that do not meet a preset condition; perform thinning processing on the non-feature positioning trajectory data according to a preset thinning rule; integrate the main label, the target vehicle data and the non-feature positioning trajectory data after the thinning processing to obtain playback data. The data playback module is configured to: render the feature positioning track data in the playback data and the non-feature positioning track data after the thinning processing into a map track playback compiler according to the data playback request, and synchronize and playback the corresponding playback data according to a time point when playing back a track.

4. The vehicle data cloud processing system of claim 3, wherein, The data analysis processing module is further configured to: process the camera collected data to generate picture data; and perform similarity matching classification on the picture data. A target picture is selected from each type of picture data for individual semantic analysis, and the analysis result is synchronized to other pictures in the classification where the target picture is located. Video data or picture data corresponding to a preset abnormality label in the camera collected data is extracted as abnormal data, and the abnormal data is stored in a preset area. The data playback module is further configured to: obtain an abnormal event playback request, the abnormal event playback request including an abnormal label that needs to be played back; and obtain the abnormal label according to the abnormal event playback request. The vehicle data in the abnormal event corresponding to the abnormal label is obtained and played back as a whole.

5. The vehicle data cloud processing system of claim 1, wherein, The data analysis processing module is further configured to: Obtain error data in the vehicle body data, the positioning track data, and the camera collected data; Analyze the error cause of the error data; Map the error cause and the error data, and generate a prompt information.

6. A vehicle data cloud processing method, comprising: Obtaining at least part of vehicle data uploaded by an information collection vehicle, the vehicle data including: vehicle body data, positioning track data, and camera collected data; Establishing a first mapping relationship between the vehicle data occurring at a same preset time period and a same data identifier, the data identifier being generated according to the preset time period, the vehicle data including: vehicle body data, positioning track data, and camera collected data; Performing semantic analysis on the vehicle data to establish a second mapping relationship between a preset vehicle body label and the vehicle body data, a third mapping relationship between a preset positioning track label and the positioning track data, and a fourth mapping relationship between a preset camera collected label and the camera collected data, wherein the vehicle body label is generated by predefinition according to the vehicle body data, the positioning track label is generated by predefinition according to the positioning track data, and the camera collected label is generated by predefinition according to the camera collected data; Storing the data identifier and data label according to data type, label type, and time, the data label including: the vehicle body label, the positioning track label, and the camera collected label; and According to a preset rule, the data identifiers and the data labels in a preset range are aggregated to generate a main label, and a fifth mapping relationship that the main label corresponds to the data identifiers and the data labels is established, so that the multi-level mapping relationship of the main label, the data identifiers and the data labels, and the vehicle data is obtained through the first mapping relationship, the second mapping relationship, the third mapping relationship, the fourth mapping relationship, and the fifth mapping relationship, and the main label is used to identify the data identifiers and the data labels in the preset range.

7. The vehicle data cloud processing method of claim 6, wherein, The step of establishing a third mapping relationship that a preset positioning track label corresponds to the positioning track data comprises: extracting characteristic positioning track data in the positioning track data that meets a preset condition; performing track point denoising on the characteristic positioning track data according to a preset denoising rule; performing track analysis on the denoised characteristic positioning track data and matching to a preset track label.

8. The vehicle data cloud processing method of claim 7, wherein, After the step of establishing the fifth mapping relationship that the main label corresponds to the data identifiers and the data labels, the method further comprises: obtaining a data playback request, wherein the data playback request includes a main label that needs to be played back; selecting the main label corresponding to a time range according to the data playback request; obtaining a target data label corresponding to the main label according to the multi-level mapping relationship; obtaining target vehicle data corresponding to the target data label according to the multi-level mapping relationship; extracting non-characteristic positioning track data in the positioning track data that does not meet the preset condition; performing thinning processing on the non-characteristic positioning track data according to a preset thinning rule; integrating the main label, the target vehicle data, and the non-characteristic positioning track data after the thinning processing to obtain playback data; rendering the characteristic positioning track data in the playback data and the non-characteristic positioning track data after the thinning processing to a map track playback compiler according to the data playback request, and synchronously playing back and displaying the corresponding playback data according to a time point when playing back a track.

9. The vehicle data cloud processing method of claim 8, wherein, The step of establishing a fourth mapping relationship that a preset camera collection label corresponds to the camera collection data comprises: processing the camera collection data to generate picture data; performing similarity matching classification on the picture data; selecting a target picture in each type of the picture data for separate semantic analysis, and synchronously transmitting the analysis result to other pictures in the classification of the target picture; extracting video data or picture data corresponding to a preset abnormal label in the camera collection data as abnormal data; and storing the abnormal data to a preset area. The step of establishing the third mapping relationship of the main label corresponding to the data identifier and the data label further comprises: obtaining an abnormal event playback request, the abnormal event playback request including an abnormal label that needs to be played back; obtaining the abnormal label according to the abnormal event playback request; obtaining the vehicle data in the process of the abnormal event corresponding to the abnormal label, and playing back the data as a whole.

10. The vehicle data cloud processing method of claim 6, wherein, The step of establishing the fifth mapping relationship of the main label corresponding to the data identifier and the data label further comprises: obtaining error data in the vehicle body data, the positioning trajectory data, and the camera acquisition data; analyzing the error cause of the error data; establishing a mapping relationship between the error cause and the error data, and generating a prompt information.

11. A machine readable storage medium having stored thereon a machine executable program which, when executed by a processor, implements the vehicle data cloud processing method according to any one of claims 6 to 10.

12. A computer device comprising a memory, a processor, and a machine executable program stored on the memory and running on the processor, and the processor implements the vehicle data cloud processing method according to any one of claims 6 to 10 when executing the machine executable program.

Images