A vehicle offline scene data processing method and system

Abstract

The present invention relates to a vehicle offline scene data processing method and system, the method comprising: obtaining a data source standard format and an original log file of the vehicle offline scene data; parsing the original log file according to the data source standard format to obtain a log file to be processed; filtering the log file to be processed in at least one manner to obtain a target scene data log file; extracting target scene data from the target scene data log file, and storing the target scene data. The present invention filters the vehicle data stored in a private cloud, especially the scene data collected during driving, to obtain high-quality valid data. While meeting application requirements, the storage and calculation amount of data is reduced, and the construction and operation and maintenance costs of the data center are greatly reduced.

Description

Technical Field

[0001] The present invention relates to the field of vehicle data processing, and in particular to a vehicle offline scene data processing method and system. Background Art

[0002] As smart driving cars gradually achieve small-scale mass production or pre-mass production, the device-centric traditional data center is increasingly unable to store, analyze, machine learn, or visualize the raw data collected from smart driving cars. For L3 smart driving cars, with an average daily usage time of 8 hours, each car will generate 25TB to 30TB of data per day. The huge amount of data places higher demands on data storage and computing. If traditional data centers want to meet the high requirements of current or future data storage and computing, the economic cost of investment will be very huge, and traditional data centers with complex systems face huge difficulties in operation, maintenance, and upgrades. More importantly, the model of traditional data centers cannot respond to the current rapidly developing business needs.

[0003] In response to this situation, one approach is to replace traditional data centers with cloud data centers. Cloud data centers use a service-oriented architecture to treat the equipment, systems, and functional outputs of data centers as services and deliver them on demand at any time. Another approach is to filter or screen the data. Unlike public clouds, private clouds do not need to save invalid or useless data after data filtering. They only need to store valid data that meets application requirements, which will significantly reduce the cost of distributed storage and distributed computing. Summary of the invention

[0004] In order to filter the vehicle data of smart driving cars stored in a private cloud data center, especially the large amount of scene data collected by the vehicle during driving, to obtain high-quality effective data, the present invention combines the above two solutions and discloses a vehicle offline scene data processing method and system. The technical solution is as follows:

[0005] In a first aspect, the present invention discloses a method for processing vehicle offline scene data, the method comprising:

[0006] Obtain the data source standard format of the vehicle offline scene data and the original log file of the vehicle offline scene data;

[0007] Parsing the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed;

[0008] Performing at least one filtering process on the log file of the to-be-processed vehicle offline scene data to obtain a target scene data log file;

[0009] The target scene data is extracted from the target scene data log file, and the target scene data is stored.

[0010] Furthermore, the data source standard format for obtaining the vehicle offline scene data and the original log file of the vehicle offline scene data include:

[0011] Establishing a mapping table between vehicle-mounted sensor devices and data formats, wherein the mapping table includes a standard format of a data source;

[0012] Acquiring scene data collected by the vehicle-mounted sensor device;

[0013] The scene data is recorded by a vehicle terminal data recording device, and an original log file of the vehicle offline scene data is generated.

[0014] Furthermore, the original log file of the vehicle offline scene data is parsed according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed, including:

[0015] According to the data source of the original log file of the vehicle offline scene data, selecting a data source standard format matching each original log file in the original log files from the data source standard format of the vehicle offline scene data;

[0016] According to the data source standard format matching each of the original log files, the data content of each of the original log files is parsed to obtain a log file of the to-be-processed vehicle offline scene data that complies with a preset format.

[0017] Furthermore, the parsing of the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed also includes:

[0018] When the original log file of the vehicle offline scene data is a log file generated according to time, and / or there are at least two data sources for the original log file,

[0019] According to the data source of the data content in the original log file of the vehicle offline scene data, splitting and classifying the original log file of the vehicle offline scene data into a log file set of at least one independent data source;

[0020] Selecting a data source standard format that matches the log file set of each independent data source in the log file set of the at least one independent data source from the data source standard format of the vehicle offline scene data;

[0021] According to a data source standard format that matches the log file set of each independent data source, the log files in each log file set are parsed to obtain log files of the vehicle offline scene data to be processed that conform to a preset format.

[0022] Furthermore, the filtering process of the log file of the to-be-processed vehicle offline scene data in at least one manner to obtain the target scene data log file includes:

[0023] Setting a time filter condition for the vehicle offline scene data through a time selector, wherein the time filter condition includes a first moment and a second moment;

[0024] According to the time filtering condition, data content between the first moment and the second moment is extracted from the log file of the vehicle offline scene data to be processed to obtain a first target scene data log file.

[0025] Furthermore, the filtering process of the log file of the to-be-processed vehicle offline scene data in at least one manner to obtain the target scene data log file further includes:

[0026] Get the vehicle's tag list file;

[0027] Setting a tag filtering condition for the vehicle offline scene data according to the tag list file of the vehicle, the tag filtering condition including a first tag scene and a second tag scene;

[0028] Parsing a corresponding marking moment according to the marking filtering condition, the marking moment comprising a first marking moment when the first marking scene is marked and a second marking moment when the second marking scene is marked;

[0029] According to the marking time, data content between the first marking time and the second marking time is extracted from the log file of the vehicle offline scene data to be processed to obtain a second target scene data log file.

[0030] Furthermore, the filtering process of the log file of the to-be-processed vehicle offline scene data in at least one manner to obtain the target scene data log file further includes:

[0031] Obtaining a high-precision positioning file of the vehicle, wherein the high-precision positioning file includes high-precision real-time positioning information, a high-precision map and / or a planned track of the vehicle's driving route;

[0032] Setting a positioning filter condition for the vehicle offline scene data according to the high-precision positioning file of the vehicle, wherein the positioning filter condition includes a first positioning position and a second positioning position;

[0033] Parsing the corresponding positioning time according to the positioning filtering condition, the positioning time including a first positioning time when the first positioning position is determined and a second positioning time when the second positioning position is determined;

[0034] According to the positioning moment, data content between the first positioning moment and the second positioning moment is extracted from the log file of the vehicle offline scene data to be processed to obtain a third target scene data log file.

[0035] Furthermore, the filtering process of the log file of the to-be-processed vehicle offline scene data in at least one manner to obtain the target scene data log file further includes:

[0036] Searching and querying the scene data of the first target scene data log file, the second target scene data log file and / or the third target scene data log file according to the vehicle identification number, the time filtering condition, the tag filtering condition and / or the location filtering condition to obtain a retrieval result;

[0037] When the search result indicates that the first target scene data log file, the second target scene data log file and / or the third target scene data log file do not cover all target scene data, and / or there are duplicate scene data or invalid scene data,

[0038] Performing a time-based micro-filtering process on the first target scene data log file, the second target scene data log file, and / or the third target scene data log file;

[0039] And / or repeatedly performing at least one filtering process on the first target scene data log file, the second target scene data log file and / or the third target scene data log file to obtain a final target scene data log file.

[0040] Further, the extracting the target scene data from the target scene data log file and storing the target scene data includes:

[0041] Deleting the original log file of the vehicle offline scene data, the first target scene data log file, the second target scene data log file and / or the third target scene data log file;

[0042] The target scene data in the target scene data log file is distributedly stored in a private cloud data center.

[0043] In a second aspect, the present invention discloses a vehicle offline scene data processing system, the system comprising:

[0044] An acquisition unit, used for acquiring a data source standard format of vehicle offline scene data and an original log file of the vehicle offline scene data;

[0045] A parsing unit, used for parsing the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed;

[0046] A filtering unit, comprising a first filtering subunit based on timestamp, a second filtering subunit based on active marking, a third filtering subunit based on high-precision positioning, and a backtracking unit, for filtering the log file of the to-be-processed vehicle offline scene data in at least one manner to obtain a target scene data log file;

[0047] The storage unit is used to extract the target scene data from the target scene data log file and store the target scene data.

[0048] By adopting the above technical solution, the vehicle offline scene data processing method and system described in the present invention have the following beneficial effects: the present invention filters the vehicle data stored in the private cloud, especially the scene data collected during driving, to obtain high-quality valid data. While meeting the application requirements, the storage and calculation amount of data is reduced, and the construction and operation and maintenance costs of the data center can be greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 It is a flow chart of a vehicle offline scene data processing method provided by an embodiment of the present invention;

[0051] Figure 2 It is a schematic diagram of parsing an original log file of vehicle offline scene data provided by an embodiment of the present invention;

[0052] Figure 3 It is a schematic diagram of multiple filtering methods for filtering log files of vehicle offline scene data provided by an embodiment of the present invention;

[0053] Figure 4 It is a schematic diagram of an algorithm flow of another vehicle offline scene data processing method provided by an embodiment of the present invention;

[0054] Figure 5It is a schematic diagram of an application scenario of vehicle offline scenario data processing provided by an embodiment of the present invention;

[0055] Figure 6 It is a structural schematic diagram of a vehicle offline scene data processing system provided by an embodiment of the present invention;

[0056] Figure 7 It is a structural schematic diagram of a vehicle offline scene data processing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0058] "One embodiment" or "embodiment" referred to herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. In the description of the present invention, it should be understood that the terms "first", "second", "third" and "fourth" in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units that are inherent to these processes, methods, products or devices.

[0059] Figure 1 It is a flowchart of a vehicle offline scene data processing method provided by an embodiment of the present invention. This specification provides method operation steps as described in the embodiment or flowchart, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps listed in the embodiment is only one way of executing the steps among many orders, and does not represent the only order of execution. When the actual system or server product is executed, it can be executed in sequence or in parallel according to the method shown in the embodiment or the accompanying drawings (for example, a parallel processor or a multi-threaded processing environment). Specifically, Figure 1 As shown, the vehicle offline scene data processing method may include:

[0060] S110: Acquire a data source standard format of the vehicle offline scene data and an original log file of the vehicle offline scene data.

[0061] In some feasible implementations, step S110 in the embodiment of the present invention may include the following steps:

[0062] S111: Establishing a mapping table between vehicle-mounted sensor devices and data formats, wherein the mapping table includes a standard format of a data source.

[0063] It is understandable that the vehicle-mounted sensor equipment may include but is not limited to cameras, laser radars, millimeter-wave radars, high-precision map sensors, and intelligent driving domain controllers. Taking into account the universal applicability of the application, the vehicle-mounted sensor equipment may come from different equipment providers. Different equipment providers set different data formats and / or standards, which require adaptation management, that is, establishing a mapping relationship between each vehicle-mounted sensor equipment and the data format, and generating a mapping table, in which each data source, that is, the data standard format of each vehicle-mounted sensor equipment, can be queried. The mapping table can also be modified accordingly according to the addition, deletion or upgrade of the vehicle-mounted sensor equipment.

[0064] S112: Acquire scene data collected by the vehicle-mounted sensor device.

[0065] S113: Recording the scene data through a vehicle terminal data recording device, and generating an original log file of the vehicle offline scene data.

[0066] S120: Parsing the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed.

[0067] Preferably, each data source generates its own scene data file in a predefined data file standard format, and the original log files stored by the vehicle terminal data recording device will also have different data formats. The log files stored by the vehicle terminal data recording device can be parsed in sequence from front to back according to the data source standard format.

[0068] In some feasible implementations, step S120 in the embodiment of the present invention may include:

[0069] According to the data source of the original log file of the vehicle offline scene data, selecting a data source standard format matching each original log file in the original log files from the data source standard format of the vehicle offline scene data;

[0070] According to the data source standard format matching each of the original log files, the data content of each of the original log files is parsed to obtain a log file of the to-be-processed vehicle offline scene data that complies with a preset format.

[0071] Preferably, the vehicle-mounted terminal data recording device generates original log files of different categories of vehicle offline scene data according to the data source, and then parses the original log files according to the data source standard format matching the original log files of each category to obtain the log files of the vehicle offline scene data to be processed.

[0072] In some feasible implementations, step S120 in the embodiment of the present invention may further include:

[0073] When the original log file of the vehicle offline scene data is a log file generated according to time, and / or there are at least two data sources for the original log file,

[0074] According to the data source of the data content in the original log file of the vehicle offline scene data, splitting and classifying the original log file of the vehicle offline scene data into a log file set of at least one independent data source;

[0075] Selecting a data source standard format that matches the log file set of each independent data source in the log file set of the at least one independent data source from the data source standard format of the vehicle offline scene data;

[0076] According to a data source standard format that matches the log file set of each independent data source, the log files in each log file set are parsed to obtain log files of the vehicle offline scene data to be processed that conform to a preset format.

[0077] It is understandable that the vehicle terminal data recording device generates the original log file of the vehicle offline scene data according to time. If a log file contains scene data collected by at least one data source device, it is necessary to first split the original log file, separate the data collected by each data source device, and generate at least one independent data source log file set. Then, the format of each independent data source log file set in the at least one independent data source log file set is parsed to obtain the log file of the vehicle offline scene data to be processed.

[0078] Figure 2 FIG. 1 is a schematic diagram of parsing an original log file of vehicle offline scene data provided by an embodiment of the present invention. In a specific implementation, the splitting and parsing of the file can be processed in parallel. Figure 2As shown, log file _0001 contains scene data collected by multiple sensors such as radar, camera and intelligent driving domain controller. The original log file needs to be split and parsed to generate a log file set of independent data sources after parsing. Specifically, the scene data from the radar in "log file _0001" is separated and parsed by parsing tool _1 and standard format _1 to generate "log file _0001_radar". Similarly, the scene data from the radar in "log file _0002" is separated and parsed to generate "log file _0002_radar", and the scene data from the radar in "log file _0003" is separated and parsed to generate "log file _0003_radar"... "Log file _0001_radar", "log file _0002_radar", "log file _0003_radar" and the like constitute a log file set whose independent data source after parsing is radar. Accordingly, a log file set whose independent data source after parsing is the camera and a log file set whose independent data source after parsing is the intelligent driving domain controller can be obtained, that is, the log files of the vehicle offline scene data to be processed are obtained.

[0079] S130: Perform at least one filtering process on the log file of the to-be-processed vehicle offline scene data to obtain a target scene data log file.

[0080] Figure 3 FIG. 1 is a schematic diagram of multiple filtering methods for filtering a log file of vehicle offline scene data provided by an embodiment of the present invention. In some feasible implementations, such as Figure 3 As shown, step S130 in the embodiment of the present invention may include the following steps:

[0081] S131: Setting a time filter condition for the vehicle offline scene data through a time selector, wherein the time filter condition includes a first moment and a second moment.

[0082] It is understandable that when the vehicle terminal data recording device records each frame of scene data collected by the vehicle sensor device, it will add timestamp information that complies with the unified time reference standard to each frame of scene data. The unified time reference standard is compatible with the hardware technology of the vehicle terminal data recording device. Preferably, the accuracy of the timestamp information is 1 microsecond.

[0083] It is understandable that a timestamp is a complete and verifiable data that indicates that a piece of data existed before a certain time. It is usually a character sequence that can uniquely identify a certain moment in time. Specifically, a timestamp refers to the total number of seconds from 00:00:00 Greenwich Mean Time, January 1, 1970 (08:00:00 Beijing Time, January 1, 1970) to the current moment. It is proposed mainly to provide users with an electronic evidence to prove the generation time of certain data of the user.

[0084] S132: According to the time filtering condition, extract the data content between the first moment and the second moment from the log file of the vehicle offline scene data to be processed to obtain a first target scene data log file.

[0085] In some feasible implementation manners, step S130 in the embodiment of the present invention may further include the following steps:

[0086] S133: Obtain the vehicle's mark list file.

[0087] It is understandable that preset or specific types of scenes and / or locations can trigger the active marking behavior of the vehicle terminal data recording device, adding marking points to the scene data of the scene and / or location. The scenes may include but are not limited to sunny days, rainy days, and a certain visibility range.

[0088] Preferably, the active marking behavior can be triggered multiple times, and the marking data can be saved separately in a binary file as a marking list file of the vehicle, so that a query of the marking points can be provided to filter the scene data according to the marking points.

[0089] S134: Setting a tag filtering condition for the vehicle offline scene data according to the tag list file of the vehicle, wherein the tag filtering condition includes a first tag scene and a second tag scene.

[0090] In some feasible implementations, the vehicle offline scene data processing system provides a marking list file of the vehicle, so that two marking points that meet application requirements are selected according to the marking list file as the first marking scene and the second marking scene.

[0091] S135: Parsing corresponding marking moments according to the marking filtering condition, where the marking moments include a first marking moment when the first marking scene is marked and a second marking moment when the second marking scene is marked.

[0092] S136: According to the marking time, extract the data content between the first marking time and the second marking time from the log file of the vehicle offline scene data to be processed to obtain a second target scene data log file.

[0093] In some feasible implementation manners, step S130 in the embodiment of the present invention may further include the following steps:

[0094] S137: Obtain a high-precision positioning file of the vehicle, wherein the high-precision positioning file includes high-precision real-time positioning information, a high-precision map and / or a planned track of the vehicle's driving route.

[0095] It is understandable that when the vehicle-mounted terminal data recording device records each frame of scene data collected by the vehicle-mounted sensor device, it will write the vehicle's high-precision real-time positioning information. Generally, the error of the vehicle's high-precision real-time positioning information is at the centimeter level.

[0096] In some feasible implementations, high-precision real-time positioning information can be determined by multi-sensor data fusion positioning. Specifically, the vehicle is initially positioned using a differential positioning system (Global Positioning System, GPS) and an inertial measurement unit (Inertial Measurement Unit, IMU) in an inertial navigation system, and a vehicle-mounted laser radar or binocular vision camera combined with high-precision map data can be used for further precise local positioning.

[0097] S138: Setting a positioning filter condition for the vehicle offline scene data according to the high-precision positioning file of the vehicle, wherein the positioning filter condition includes a first positioning position and a second positioning position.

[0098] In some feasible implementations, the vehicle offline scene data processing system can also provide a high-precision map and a route planning track for the vehicle's actual travel mission, which is used to select two location points in the vehicle's driving process in the route planning track, including the first positioning position and the second positioning position.

[0099] S139: parsing a corresponding positioning moment according to the positioning filtering condition, wherein the positioning moment includes a first positioning moment when the first positioning position is determined and a second positioning moment when the second positioning position is determined.

[0100] S1310: According to the positioning time, extract the data content between the first positioning time and the second positioning time from the log file of the vehicle offline scene data to be processed to obtain a third target scene data log file.

[0101] In some feasible implementation manners, step S130 in the embodiment of the present invention may further include the following steps:

[0102] S1311: Based on the vehicle identification number, the time filter condition, the tag filter condition and / or the positioning filter condition, search and query the scene data of the first target scene data log file, the second target scene data log file and / or the third target scene data log file to obtain retrieval results.

[0103] S1312: When the search result indicates that the first target scene data log file, the second target scene data log file, and / or the third target scene data log file do not cover all target scene data, and / or there are duplicate scene data or invalid scene data,

[0104] Performing a time-based micro-filtering process on the first target scene data log file, the second target scene data log file, and / or the third target scene data log file;

[0105] And / or repeatedly performing at least one filtering process on the first target scene data log file, the second target scene data log file and / or the third target scene data log file to obtain a final target scene data log file.

[0106] It is understandable that if Figure 3 As shown, the three scene data filtering methods can be arranged and combined in priority order according to actual application requirements, and the embodiment of the present invention does not make specific limitations.

[0107] S140: extracting target scene data from the target scene data log file, and storing the target scene data.

[0108] In some feasible implementations, step S140 in the embodiment of the present invention may include:

[0109] Deleting the original log file of the vehicle offline scene data, the first target scene data log file, the second target scene data log file and / or the third target scene data log file;

[0110] The target scene data in the target scene data log file is distributedly stored in a private cloud data center.

[0111] It can be understood that in the embodiment of the present invention, data is stored in a private cloud data center. Since the private cloud data center stores target scenario data that meets actual application needs, there is no need to save a large amount of invalid and useless data after filtering and processing. Therefore, the original log files and temporary log files generated in the intermediate process can be deleted, thereby significantly reducing the cost of distributed storage and distributed computing.

[0112] Preferably, the target scene data obtained by a vehicle offline scene data processing method provided by an embodiment of the present invention can be used for training models in the field of big data and / or artificial intelligence. At present, there is a large amount of invalid and useless data in the data collected by intelligent driving vehicles for model training, which undoubtedly increases the cost of model training and reduces the efficiency of model training. The method provided by the embodiment of the present invention can greatly reduce unnecessary data storage and calculation while ensuring the acquisition of effective scene data, greatly reducing the construction cost of the data center.

[0113] Figure 4 It is an algorithm flow chart of another vehicle offline scene data processing method provided by an embodiment of the present invention. This specification provides method operation steps as described in the embodiment or the flow chart, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps listed in the embodiment is only one way of executing the steps among many orders, and does not represent the only order of execution. When the actual system or server product is executed, it can be executed in sequence or in parallel according to the method shown in the embodiment or the accompanying drawings (for example, a parallel processor or a multi-threaded processing environment). Specifically, Figure 4 As shown, the vehicle offline scene data processing method is implemented by a software system, and the implementation algorithm may include the following steps:

[0114] S410: When the process starts, the data source standard format is loaded.

[0115] In some feasible implementations, the software system loads standard formats of data from different equipment suppliers and different data sources, including but not limited to standard formats of on-board Advanced Driving Assistant System (ADAS) sensors, intelligent driving domain controllers, controller area networks (CAN), etc.

[0116] S420: Read the original log file.

[0117] S430: Parse the original log file according to the data source standard format.

[0118] In some feasible implementations, before parsing, the original log files need to be split and classified according to the sources of the data contents.

[0119] S440: Load filtering conditions.

[0120] Specifically, the filtering conditions may include but are not limited to a start time, an end time, a marking point, and a positioning position.

[0121] S450: Determine the scene data filtering method.

[0122] like Figure 4 As shown, step S450 may include the following steps:

[0123] S451: Read the vehicle's tag list file for the marked scene data filtering method. Select two tag points in the tag list file that meet the application requirements. Parse the two corresponding timestamps based on the two tag points. Then filter the scene data based on the timestamp method and generate a new scene data log file.

[0124] S452: For the scene data filtering method of positioning, read the high-precision positioning file of the vehicle. Select two positioning positions that meet the application requirements in the high-precision positioning file. Parse the two corresponding timestamps according to the two positioning positions. Then filter the scene data based on the timestamp method and generate a new log file of the scene data.

[0125] S453: In the scene data filtering method based on timestamp, data in the log file is extracted according to the timestamps of the start time and the end time, and a log file of the target scene data is generated.

[0126] In some feasible implementations, the order of steps S451-S453 listed in this embodiment is only one of many execution orders of steps, and does not represent the only execution order. In actual applications, one or more filtering methods can also be selected according to application requirements.

[0127] S460: Persistently save the log file of the newly generated target scene data after filtering.

[0128] S470: Delete the original log file and the log file temporarily generated in the intermediate step, and the process ends.

[0129] In some feasible implementations, the software system also has a backtracking unit that can playback and view the newly generated target scene data log file. If the playback and viewing target scene data log file does not completely cover the target scene, some or all of steps S410-S460 can be repeated until the target scene data log file obtained meets the scene requirements of the actual application, and finally step S470 is executed.

[0130] It is understandable that the implementation process of the entire algorithm is completed on the private cloud data center, which can fully utilize the advantages of distributed storage and distributed computing. For example, the reading and parsing of the original log files of the vehicle offline scene data can be based on the idea of ​​MapReduce (a programming model for parallel computing of large-scale data sets) distributed computing, and the storage of the target scene data log files can use the distributed object storage service.

[0131] Figure 5 Schematic diagram of an application scenario of vehicle offline scene data processing provided by an embodiment of the present invention. Figure 5 As shown, the vehicle starts at the starting point O and arrives at the highway gate entrance point A after 40 minutes of driving, then drives on the highway for 15 minutes to reach the gate exit point B, then exits the highway from point B, drives for another 50 minutes to reach the destination T, and finally stops and turns off the engine, completing a vehicle travel mission.

[0132] exist Figure 5 In the application scenario shown, the target scenario database is to store the scenario data of the vehicle driving between point A and point B more accurately, so as to serve as a training data set for the subsequent intelligent driving model on the private cloud data center.

[0133] During the data collection and recording stage, from the start of vehicle ignition to the parking and shutdown process, the vehicle terminal data recording device will continue to record the original driving logs of different sensor devices in the vehicle in real time. The sensor devices may include GPS, IMU, CAN bus controller, camera, LiDAR, millimeter wave radar, ultrasonic radar, high-precision map sensor, intelligent driving domain controller, etc. The data collected by all sensor devices can be collectively referred to as scene data. Because the vehicle terminal data recording device will temporarily store the scene data locally, it can also be called vehicle offline scene data. In some feasible implementations, the data can be uploaded to the server in real time after parsing the original log file while meeting the communication quality. In addition, when the vehicle arrives near point A and point B, the vehicle's active marking behavior can be triggered in a timely manner to generate marking points corresponding to point A and point B.

[0134] After the vehicle travel mission is completed, the original log data recorded by the vehicle terminal data recording device needs to be imported and uploaded to the private cloud data center in the form of offline files, and the distributed object storage service is used to save these big data files.

[0135] In a specific implementation, after replaying and visualizing the data of the vehicle travel mission, it is possible to verify whether the route and data of the vehicle travel are correct, and then filter the scene data according to the steps mentioned in a vehicle offline scene data method provided in an embodiment of the present invention.

[0136] First, the original log file can be split, and then the effective vehicle scene data of the highway section can be extracted through a variety of scene data filtering methods. Specifically, the approximate time of the vehicle's arrival at point A and point B can be estimated based on information such as the distance between the location OABT and the vehicle's driving speed, and scene data filtering and playback can be performed based on timestamps and vehicle identification numbers; or the list of all actively marked points of the vehicle's current travel mission can be queried first, and the marked points corresponding to points A and B can be selected from them for scene data filtering and playback; or, based on high-precision maps and the high-precision real-time positioning information of the vehicle's current travel mission, point A and point B can be selected for scene data filtering and playback.

[0137] The above three methods may not fully cover all highway scene data between points AB, or may include some scene data outside points AB. At this time, you can further adjust based on the timestamp according to the target scene data filtering results, that is, by fine-tuning the timestamps before or after the two result points, to shrink or expand the scene filtering data. Through the combination of the above various scene filtering methods, a more accurate target scene data segment can be obtained. Pay attention to the query and search of scene data, and consider using distributed full-text search engine technology to achieve it.

[0138] Finally, only valid highway scene data is filtered out, and invalid data before the vehicle enters the highway and after the vehicle leaves the highway is deleted. When this step is completed, only valid scene data is retained in the private cloud storage.

[0139] The embodiment of the present invention also provides a vehicle offline scene data processing system, such as Figure 6 As shown, the vehicle offline scene data processing system includes:

[0140] The acquisition unit 610 is used to acquire the data source standard format of the vehicle offline scene data and the original log file of the vehicle offline scene data.

[0141] The parsing unit 620 is used to parse the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed.

[0142] The filtering unit 630 includes a first filtering subunit 631 based on timestamp, a second filtering subunit 632 based on active marking, a third filtering subunit 633 based on high-precision positioning, and a backtracking unit 634, which is used to filter the log file of the vehicle offline scene data to be processed in at least one way to obtain the target scene data log file.

[0143] The storage unit 640 is used to extract the target scene data from the target scene data log file and store the target scene data.

[0144] The vehicle offline scene data processing system and method embodiment described in the embodiment of the present invention are based on the same inventive concept. Please refer to the method embodiment for details, which will not be repeated here.

[0145] An embodiment of the present invention also provides a computer device, comprising: a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement a vehicle offline scene data processing method as in an embodiment of the present invention.

[0146] The memory can be used to store software programs and modules, and the processor executes various functional applications by running the software programs and modules stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, application programs required for functions, etc.; the data storage area can store data created according to the use of the device, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory can also include a memory controller to provide the processor with access to the memory.

[0147] The method embodiments provided in the embodiments of the present invention may be executed in a computer terminal, a server or a similar computing device, that is, the above-mentioned computer device may include a computer terminal, a server or a similar computing device. Figure 7 is a hardware structure block diagram of a computer device for running a vehicle offline scene data processing method provided by an embodiment of the present invention, such as Figure 7As shown, the internal structure of the computer device may include but is not limited to: a processor, a network interface and a memory. The processor, network interface and memory in the computer device may be connected via a bus or other means. Figure 7 The computer device can be a virtualized device in a private cloud data, and the internal architecture is not limited to Figure 7 shown.

[0148] Among them, the processor (or CPU (Central Processing Unit)) is the computing core and control core of the computer device. The network interface may optionally include a standard wired interface, a wireless interface (such as WI-FI, a mobile communication interface, etc.). Memory is a memory device in a computer device for storing programs and data. It is understandable that the memory here can be a high-speed RAM storage device or a non-volatile memory device (non-volatile memory), such as at least one disk storage device; optionally, it can also be at least one storage device located away from the aforementioned processor. The memory provides a storage space, which stores the operating system of the electronic device, which may include but is not limited to: Windows system (an operating system), Linux (an operating system), Android (Android, a mobile operating system) system, IOS (a mobile operating system) system, etc., and the present invention is not limited to this; and, in the storage space, one or more instructions suitable for being loaded and executed by the processor are also stored, and these instructions may be one or more computer programs (including program codes). In the embodiment of this specification, the processor loads and executes one or more instructions stored in the memory to implement the vehicle offline scene data processing method provided in the above method embodiment.

[0149] An embodiment of the present invention also provides a computer storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored. The at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor as described in an embodiment of the present invention. A vehicle offline scene data processing method.

[0150] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store program codes.

[0151] It should be noted that the sequence of the embodiments of the present invention described above is for description only and does not represent the advantages and disadvantages of the embodiments. The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0152] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, system and server embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0153] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0154] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A vehicle offline scene data processing method, characterized in that: The method comprises: Acquire a data source standard format of vehicle offline scene data and an original log file of the vehicle offline scene data; the data source standard format represents a mapping relationship between the data source and the data format; Parsing the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed, wherein the log file of the vehicle offline scene data to be processed is divided according to the data source; According to the time filtering condition, the mark point filtering condition and the positioning filtering condition corresponding to the target scene, the log file of the vehicle offline scene data to be processed is filtered to obtain the target scene data log file, wherein the target scene is the vehicle driving scene; Extracting target scene data from the target scene data log file and storing the target scene data; The target scene data log file includes a second target scene data log file, and the second target scene data log file can be determined in the following manner: Acquire a marking list file of the vehicle; the marking points in the marking list file represent scenes and / or locations actively marked by the vehicle-mounted terminal; Setting the marking point filtering condition of the vehicle offline scene data according to the marking list file of the vehicle, wherein the marking point filtering condition includes a first marking scene and a second marking scene; Parsing the corresponding marking moments according to the marking point filtering condition, the marking moments comprising a first marking moment when the first marking scene is marked and a second marking moment when the second marking scene is marked; According to the marking time, data content between the first marking time and the second marking time is extracted from the log file of the vehicle offline scene data to be processed to obtain the second target scene data log file.

2. The vehicle offline scene data processing method according to claim 1, characterized in that: The data source standard format for obtaining the vehicle offline scene data and the original log file of the vehicle offline scene data include: Establishing a mapping table between vehicle-mounted sensor devices and data formats, wherein the mapping table includes a standard format of a data source; Acquiring scene data collected by the vehicle-mounted sensor device; The scene data is recorded by a vehicle terminal data recording device, and an original log file of the vehicle offline scene data is generated.

3. The vehicle offline scene data processing method according to claim 1, characterized in that: The parsing of the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed includes: According to the data source of the original log file of the vehicle offline scene data, selecting a data source standard format matching each original log file in the original log files from the data source standard format of the vehicle offline scene data; According to the data source standard format matching each of the original log files, the data content of each of the original log files is parsed to obtain a log file of the to-be-processed vehicle offline scene data that complies with a preset format.

4. The vehicle offline scene data processing method according to claim 3, characterized in that: The parsing of the original log file of the vehicle offline scene data according to the data source standard format of the vehicle offline scene data to obtain the log file of the vehicle offline scene data to be processed also includes: When the original log file of the vehicle offline scene data is a log file generated according to time, and / or there are at least two data sources for the original log file, According to the data source of the data content in the original log file of the vehicle offline scene data, splitting and classifying the original log file of the vehicle offline scene data into a log file set of at least one independent data source; Selecting a data source standard format that matches the log file set of each independent data source in the log file set of the at least one independent data source from the data source standard format of the vehicle offline scene data; According to a data source standard format that matches the log file set of each independent data source, the log files in each log file set are parsed to obtain log files of the vehicle offline scene data to be processed that conform to a preset format.

5. The vehicle offline scene data processing method according to claim 1, characterized in that: The filtering process is performed on the log file of the to-be-processed vehicle offline scene data according to the time filtering condition, the mark point filtering condition and the positioning filtering condition corresponding to the target scene, and the target scene data log file is obtained, including: Setting a time filter condition for the vehicle offline scene data through a time selector, wherein the time filter condition includes a first moment and a second moment; According to the time filtering condition, data content between the first moment and the second moment is extracted from the log file of the vehicle offline scene data to be processed to obtain a first target scene data log file.

6. The vehicle offline scene data processing method according to claim 1, characterized in that: The filtering process of the log file of the to-be-processed vehicle offline scene data according to the time filtering condition, the mark point filtering condition and the positioning filtering condition corresponding to the target scene to obtain the target scene data log file also includes: Obtaining a high-precision positioning file of the vehicle, wherein the high-precision positioning file includes high-precision real-time positioning information, a high-precision map and / or a planned track of the vehicle's driving route; Setting a positioning filter condition for the vehicle offline scene data according to the high-precision positioning file of the vehicle, wherein the positioning filter condition includes a first positioning position and a second positioning position; Parsing the corresponding positioning time according to the positioning filtering condition, the positioning time including a first positioning time when the first positioning position is determined and a second positioning time when the second positioning position is determined; According to the positioning moment, data content between the first positioning moment and the second positioning moment is extracted from the log file of the vehicle offline scene data to be processed to obtain a third target scene data log file.

7. The vehicle offline scene data processing method according to claim 1, characterized in that: The target scene data log file also includes a first target scene data log file determined based on the time filtering condition and a third target scene data log file determined based on the positioning filtering condition. The target scene data log file is filtered based on the time filtering condition, the mark point filtering condition and the positioning filtering condition corresponding to the target scene. The target scene data log file also includes: Searching and querying the scene data of the first target scene data log file, the second target scene data log file and / or the third target scene data log file according to the vehicle identification number, the time filtering condition, the mark point filtering condition and / or the positioning filtering condition to obtain a retrieval result; When the search result indicates that the first target scene data log file, the second target scene data log file and / or the third target scene data log file do not cover all target scene data, and / or there are duplicate scene data or invalid scene data, Performing a time-based micro-filtering process on the first target scene data log file, the second target scene data log file, and / or the third target scene data log file; And / or repeatedly perform filtering processing based on the time filtering condition, the mark point filtering condition and / or the positioning filtering condition on the first target scene data log file, the second target scene data log file and / or the third target scene data log file to obtain the final target scene data log file.

8. The vehicle offline scene data processing method according to claim 7, characterized in that: The extracting target scene data from the target scene data log file and storing the target scene data comprises: Deleting the original log file of the vehicle offline scene data, the first target scene data log file, the second target scene data log file and / or the third target scene data log file; The target scene data in the target scene data log file is distributedly stored in a private cloud data center.

9. A vehicle offline scene data processing system, characterized in that: The system comprises: An acquisition unit, used to acquire a data source standard format of vehicle offline scene data and an original log file of the vehicle offline scene data; the data source standard format represents a mapping relationship between the data source and the data format; A parsing unit, configured to parse an original log file of the vehicle offline scene data according to a standard format of a data source of the vehicle offline scene data to obtain a log file of the vehicle offline scene data to be processed, wherein the log file of the vehicle offline scene data to be processed is divided according to the data source; A filtering unit, comprising a first filtering subunit based on timestamp, a second filtering subunit based on active marking, a third filtering subunit based on high-precision positioning, and a backtracking unit, for filtering the log file of the to-be-processed vehicle offline scene data according to a time filtering condition, a marking point filtering condition, and a positioning filtering condition corresponding to a target scene, to obtain a target scene data log file, wherein the target scene is a vehicle driving scene; A storage unit, used for extracting target scene data from the target scene data log file and storing the target scene data; The target scene data log file includes a second target scene data log file, and the filtering unit includes: a second filtering subunit based on active marking, used to obtain a vehicle marking list file; the marking points in the marking list file represent the scenes and / or locations actively marked by the vehicle terminal; the marking point filtering conditions of the vehicle offline scene data are set according to the vehicle marking list file, and the marking point filtering conditions include a first marking scene and a second marking scene; the corresponding marking moment is parsed according to the marking point filtering conditions, and the marking moment includes a first marking moment when marking the first marking scene and a second marking moment when marking the second marking scene; according to the marking moment, the data content between the first marking moment and the second marking moment is extracted from the log file of the vehicle offline scene data to be processed to obtain the second target scene data log file.

Images