Perception data stream processing method, server and vehicle

Abstract

The invention provides a sensing data stream processing method, a server and a vehicle, and relates to the fields of automatic driving, unmanned driving, unmanned vehicles and data processing. The perception data stream processing method is applied to a server, and comprises the following steps: in response to a received operation perception data stream sent by a target vehicle, writing an operation perception data sequence determined based on the operation perception data stream in a preset storage area, and in the process of receiving the operation perception data stream, executing a specified detection instruction sent by the target vehicle, a detection feedback result is obtained, the detection feedback result is sent to the target vehicle, and the specified detection instruction is used for detecting the operation state of the server; and performing quality detection on the job perception data sequence, recording a quality detection result, and processing the job perception data sequence with a quality problem.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of automatic driving, unmanned driving, unmanned vehicle, and data processing, and more particularly, to a perception data stream processing method, a server and a vehicle. BACKGROUND

[0002] With the development of the automatic driving field, data loop has become an important part of the iteration of automatic driving algorithms. By playing back and analyzing the perception data collected from the vehicle driving process in real scenes, the deficiencies in the automatic driving algorithm can be found, thereby providing guidance for the optimization of the automatic driving algorithm.

[0003] However, the inventors have found that in specific scenarios such as mine operations, the perception data collected by the vehicle will occupy a large amount of device resources, resulting in a high cost burden for the scheduling of vehicles in specific scenarios. SUMMARY

[0004] Therefore, the present disclosure provides a perception data stream processing method, a server and a vehicle.

[0005] One aspect of the present disclosure provides a perception data stream processing method applied to a server, comprising: in response to receiving a job perception data stream sent by a target vehicle, writing a job perception data sequence determined based on the job perception data stream in a preset storage area, wherein during the reception of the job perception data stream, a specified detection instruction sent by the target vehicle is executed to obtain a detection feedback result, and the detection feedback result is sent to the target vehicle, and the specified detection instruction is used to detect the running state of the server; quality detection is performed on the job perception data sequence, and the quality detection result is recorded and the job perception data sequence with quality problems is processed.

[0006] According to an embodiment of the present disclosure, the job perception data sequence comprises a job perception image sequence; the quality detection on the job perception data sequence comprises: performing pixel attribute detection on at least one frame of job perception image in the job perception image sequence to obtain pixel attribute detection information; and in response to the pixel attribute detection information not satisfying a preset pixel quality condition, determining a target detection result.

[0007] According to an embodiment of the present disclosure, the quality detection on the job perception data sequence comprises: determining a target running state of the target vehicle in a specified period based on a time attribute of the job perception data sequence and running state information sent from the target vehicle, the specified period being determined based on the time attribute; and determining a target detection result based on the target running state; the target running state comprises at least one of the following: a first running state indicating that the target vehicle is at rest; and a second running state indicating that a specified lighting device related to the job perception data sequence does not perform lighting in a specified lighting period, the specified lighting device being configured on the target vehicle.

[0008] According to an embodiment of the present disclosure, the work perception data sequence comprises a work perception image sequence; wherein, the target running state of the target vehicle in the specified period is determined based on the time attribute of the work perception data sequence, comprising: determining the image brightness state in the specified lighting period based on the time attribute of the work perception image sequence and the image brightness attribute of at least one frame of work perception image; determining the second running state based on the image brightness state and the device field of the specified lighting device related to the image collection angle of the work perception image.

[0009] According to an embodiment of the present disclosure, the quality detection of the work perception data sequence comprises: performing difference detection based on the time attribute of the plurality of work perception data in the work perception data sequence to obtain a time difference detection result; and determining a frame loss defect detection result based on the comparison result between the time difference detection result and the preset difference time length threshold.

[0010] According to an embodiment of the present disclosure, the perception data stream processing method further comprises: executing a detection instruction before receiving the work perception data stream to obtain a detection feedback result; preferably, executing the detection instruction before receiving the work perception data stream to obtain the detection feedback result, comprising: executing a process state detection instruction to determine a first detection feedback result representing the amount of idle process resources; and / or executing a connectivity detection instruction to generate a feedback data packet, wherein the connectivity detection instruction is determined based on a detection data packet sent by the target vehicle; further, executing the detection instruction to obtain the detection feedback result further comprises: executing a time synchronization detection instruction to determine a second detection feedback result representing the system time attribute of the server; and / or executing a storage space detection instruction to determine a third detection feedback result representing the remaining storage space in the server.

[0011] Another aspect of the present disclosure provides a perception data stream method applied to a vehicle end, comprising: obtaining a work perception data stream; before sending the work perception data stream to a server, and in the process of sending the work perception data stream, detecting the running state of the server by sending a detection data packet carrying a specified detection instruction to the server to obtain a detection result; and in the case that the detection result represents that the running state of the server is abnormal, sending a target instruction to the server to control the server to stop recording the work perception data.

[0012] According to an embodiment of the present disclosure, the running state of the service end is detected, including: receiving a detection feedback result sent by the service end, the detection feedback result being determined by the service end executing a specified detection instruction; determining an abnormal detection result representing an abnormal running state of the service end based on the detection feedback result, wherein the detection feedback result includes at least one of: a first detection feedback result representing an idle process resource amount of the service end, the abnormal detection result corresponding to the first detection feedback result indicating that the idle process resource of the service end is abnormal; a second detection feedback result representing a system time attribute of the service end, the abnormal detection result corresponding to the second detection feedback result indicating that there is a system time synchronization abnormality between the service end and the vehicle end. Preferably, the running state of the service end is detected, further including: in response to the number of feedback data packets in the target detection period not satisfying a preset number condition, determining a communication abnormality detection result, the feedback data packet being determined by the service end based on a detection data packet.

[0013] Another aspect of the present disclosure provides a server, comprising: a first communication module configured to receive a job perception data stream sent by a target vehicle; a memory comprising a preset storage area; a first processor configured to, in response to receiving the job perception data stream sent by the target vehicle, write a job perception data sequence determined based on the job perception data stream in the preset storage area, wherein during the process of receiving the job perception data stream, the first processor is further configured to execute a specified detection instruction sent by the target vehicle to obtain a detection feedback result, the specified detection instruction being used to detect the running state of the service end; the first communication module is further configured to send the detection feedback result to the target vehicle; and the first processor is further configured to perform quality detection on the job perception data sequence, record a quality detection result and process the job perception data sequence with quality problems.

[0014] Another aspect of the present disclosure provides a vehicle, comprising: a collection module configured to obtain a job perception data stream; a second communication module configured to send the job perception data stream to a service end; the second communication module is further configured to, before sending the job perception data stream to the service end and during the process of sending the job perception data stream to the service end, detect the running state of the service end by sending a detection data packet carrying a specified detection instruction to the service end, and receive a detection result; and a second processor configured to, in the case that the detection result represents an abnormal running state of the service end, generate a target instruction for controlling the service end to stop recording the job perception data; and the second communication module is further configured to send the target instruction to the service end. BRIEF DESCRIPTION OF DRAWINGS

[0015] The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure taken in conjunction with the accompanying drawings, in which:

[0016] Figure 1An application scenario diagram of a perception data stream processing method, a server and a vehicle according to an embodiment of the present disclosure is schematically shown;

[0017] Figure 2 A flowchart of a perception data stream processing method according to an embodiment of the present disclosure is schematically shown;

[0018] Figure 3 A target vehicle and a NAS device according to an embodiment of the present disclosure are schematically shown;

[0019] Figure 4 A quality detection result according to an embodiment of the present disclosure is exemplarily shown;

[0020] Figure 5 A flowchart of a perception data stream processing method according to another embodiment of the present disclosure is schematically shown;

[0021] Figure 6 A flowchart of a perception data processing method according to an embodiment of the present disclosure is schematically shown;

[0022] Figure 7 A structural block diagram of a server according to an embodiment of the present disclosure is schematically shown;

[0023] Figure 8 A structural block diagram of a vehicle according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0024] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It is to be understood, however, that the description is merely exemplary and is intended to provide a thorough understanding of the embodiments of the present disclosure. The following detailed description and specific examples are presented to provide a thorough understanding of the embodiments of the present disclosure. However, it will be apparent that one or more embodiments can be practiced without these specific details. In addition, in the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily obscuring the concept of the present disclosure.

[0025] The terms used herein are merely used to describe specific embodiments and are not intended to limit the present disclosure. The terms "include", "comprise" and the like used herein indicate the presence of the described features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0026] All terms used herein, including technical and scientific terms, have meanings commonly understood by one of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having meanings consistent with the context of the present specification, and should not be interpreted in an idealized or overly formal manner.

[0027] In the case where expressions such as "at least one of A, B, and C, etc." are used, it generally means the same as "at least one of A, at least one of B, and at least one of C" unless clearly specified otherwise.

[0028] In embodiments of the present disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, storage, etc. of the data involved (for example, including but not limited to user personal information) comply with the relevant legal regulations, are used for legal purposes, and do not violate public order and good customs. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and to maintain user personal information security and network security.

[0029] In embodiments of the present disclosure, the authorization or consent of the user is obtained before the user's personal information is acquired or collected.

[0030] A high-quality data closed loop can significantly improve the robustness and performance of automatic driving algorithms in core modules such as perception, prediction, and decision-making, and is one of the key means to promote the continuous development of automatic driving technology. For example, vehicle operation perception data reflecting the vehicle driving and operation state can be recorded and stored to analyze and process the vehicle operation perception data, thereby realizing a data closed loop. However, due to the limited disk space of the vehicle domain controller, directly recording data to the domain controller not only cannot meet the demand of large-scale data generation, but also accelerates the disk wear and tear, further shortening the service life of the equipment. To solve this problem, in one example, a network attached storage (NAS) device can be introduced as a new data storage device on the vehicle. For example, the NAS device can be directly connected to the domain controller through a network cable, and the vehicle-end data can be transmitted to the NAS device via the network and stored on the disk.

[0031] In the process of implementing the concept of the disclosure, the inventors found that, although the introduction of the NAS device can significantly improve the data capacity of single recording, it also brings new problems, mainly including the following two aspects: 1. Recording environment problem: the connection between the NAS device and the domain controller may be disconnected due to vehicle bumping, in addition, there are problems such as insufficient storage space of the NAS device or the system time of the NAS device is not synchronized with the domain controller, which will affect the continuity and accuracy of data recording; 2. Data quality problem: if there is a problem in the recording environment, it will cause the data to be unable to be recorded or the quality of the recorded data to be poor, which cannot be used for data closed loop. In addition, the time of single data recording is usually long, usually lasting for several hours or even dozens of hours, and it is unrealistic for people to monitor whether there is a problem throughout the process, so when the data recording has a problem, the on-site test personnel often cannot find it in time. If the problem data is uploaded to the cloud, not only the network traffic and cloud storage resources will be wasted, but also the efficiency of algorithm iteration will be affected.

[0032] Therefore, the embodiments of the disclosure provide a perception data stream processing method, a server and a vehicle, the perception data stream processing method is applied to a server side, and the perception data stream processing method comprises the following steps: in response to receiving job perception data stream sent by a target vehicle, writing job perception data sequence determined based on the job perception data stream in a preset storage area, wherein, in the process of receiving the job perception data stream, a specified detection instruction sent by the target vehicle is executed to obtain a detection feedback result, and the detection feedback result is sent to the target vehicle, and the specified detection instruction is used to detect the running state of the server side; performing quality detection on the job perception data sequence, recording a quality detection result, and processing the job perception data sequence with quality problems.

[0033] According to the embodiments of the disclosure, by detecting the running state of the NAS device, the possible software and hardware environment problems in the NAS device can be found in time based on the detection feedback result, so as to ensure the stability and reliability of the recording process, find potential problems in time, and avoid data recording interruption or abnormality. By performing real-time quality detection on the job perception data sequence, the available data and problem data can be distinguished in time, and the reliable job perception data sequence is screened out for data closed loop, thereby reducing the waste of invalid data on network traffic and NAS device storage resources, avoiding the uploading of data with quality problems to the cloud, improving the reliability of data recording, reducing the workload of manual intervention and post-data cleaning, and helping to improve the iteration efficiency of the automatic driving algorithm.

[0034] Figure 1 An application scenario diagram of the perception data stream processing method, the server and the vehicle according to the embodiments of the disclosure is shown.

[0035] As Figure 1As shown, the application scenario 100 according to this embodiment can include vehicles 101, 102, 103, a network 104, and a server 105. The network 104 is a medium for providing a communication link between the vehicles 101, 102, 103 and the server 105. The network 104 can include various connection types, such as wired, wireless communication links, or fiber optic cables, and the like.

[0036] A user can use the vehicles 101, 102, 103 to interact with the server 105 through the network 104 to receive or send messages, and the like. The vehicles 101, 102, 103 can be unmanned vehicles, or can also include vehicles driven based on a driver user.

[0037] The vehicles 101, 102, 103 can be any type of vehicle with unmanned driving function or automatic auxiliary driving function, such as a truck, a car, an unmanned mining truck, and the like. Exemplarily, the vehicles 101, 102, 103 can be deployed with a domain controller (DCU). The domain controller can integrate a plurality of electronic control units (ECUs) scattered on one or several powerful processors to realize centralized control of a specific function domain (such as automatic driving, power system, body control, and the like). It can realize intelligent decision and control of the vehicle by integrating multi-source sensor data, running algorithms, and coordinating actuators.

[0038] The server 105 can be a server providing various services, for example, can be a data storage server providing network storage services for data (for example, work perception data) representing the motion state, work state, and the like of the vehicles 101, 102, 103 (only as an example). The data storage server can analyze, store, and the like process the data transmitted by the vehicles received, and feed back the processing result to any one of the vehicles.

[0039] In the field of automatic driving technology, the work perception data can be understood as the vehicle surrounding environment information collected by the vehicle in real time through multi-modal sensors (such as a camera, a laser radar, a millimeter wave radar, and the like), for example, can include visual data (such as image data, video data collected by a camera), laser radar point cloud, millimeter wave radar data, ultrasonic data, fusion perception data (such as target trajectory, semantic segmentation map), and the like.

[0040] It should be noted that the perception data stream processing method provided by the embodiments of the present disclosure can generally be executed by the server 105. The perception data stream processing method provided by the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the vehicles 101, 102, 103 and / or the server 105.

[0041] Alternatively, the perception data stream processing method provided by the embodiments of the present disclosure can also be generally executed by any one or more of the vehicles 101, 102, and 103.

[0042] It should be understood that Figure 1 The number of vehicles, networks, and servers in the above description is only illustrative. Any number of vehicles, networks, and servers can be provided according to the implementation needs.

[0043] Figure 2 A flowchart of a perception data stream processing method according to an embodiment of the present disclosure is schematically shown.

[0044] As shown in Figure 2 The method 200 includes operations S210-S220.

[0045] At operation S210, in response to receiving a job perception data stream sent by a target vehicle, writing a job perception data sequence determined based on the job perception data stream in a preset storage area, wherein during the reception of the job perception data stream, a specified detection instruction sent by the target vehicle is executed to obtain a detection feedback result, and the detection feedback result is sent to the target vehicle, and the specified detection instruction is used to detect the running state of the server.

[0046] At operation S220, quality detection is performed on the job perception data sequence, the quality detection result is recorded, and the job perception data sequence with quality problems is processed.

[0047] According to the embodiments of the present disclosure, the method 200 is applied to a server, which can be, for example, a NAS device. The target vehicle can be any type of vehicle with a self-driving function or an automatic auxiliary driving function, and the domain controller of the target vehicle can be connected to the NAS device through a network.

[0048] Figure 3 A target vehicle and a NAS device according to an embodiment of the present disclosure are schematically shown.

[0049] As shown in Figure 3 The target vehicle can include a domain controller 310, which can include, for example, an automatic driving control module 311 and an environmental problem detection module 312. The NAS device 320 can include, for example, a data recording module 321, a data storage module 322, and a data quality detection module 323. The domain controller 310 can be connected to the NAS device 320 through a network cable, for example. In an embodiment, the data recording module and the data quality detection module of the NAS device 320 can be remotely started by a program in the domain controller.

[0050] Exemplarily, the autonomous driving control module can be used to process raw data collected by in-vehicle sensors, laser radars, etc., to generate a job perception data stream. The environment problem detection module can be used to detect the running environment of the NAS device, and is responsible for real-time monitoring and detecting potential problems in the data recording environment, to ensure the stability of the recording process. The data recording module can record the received data (e.g., the job perception data stream), the data storage module can store the recorded data, and the data quality detection module can detect the quality of the stored data, to ensure that the recorded data meets the required quality standards.

[0051] According to one embodiment of the present disclosure, the domain controller of the target vehicle can continuously send a job perception data stream to the NAS device through a network. In response to receiving the job perception data stream, the NAS device can use a network packet capturing tool to capture data packets flowing through the network interface, and continuously record the data transmitted from the domain controller through the network cable based on a data recording module. For example, one file can be recorded and saved every 30 seconds, to obtain a job perception data sequence. The job perception data sequence can be named by the creation time, and written to a preset storage area in the NAS device. The job perception data sequence can include at least one frame of job perception data, for example.

[0052] As an example, during the process of receiving the job perception data stream by the NAS device, the domain controller can send a specified detection instruction to the NAS device based on the environment problem detection module, which can be used to detect whether there is a problem in the running state of the NAS device. The NAS device can execute the specified detection instruction to obtain a detection feedback result (e.g., normal running state, abnormal running state, etc.), and send the detection feedback result to the domain controller of the target vehicle.

[0053] For example, in the case where the detection feedback result represents that the running state of the NAS device is normal, the NAS device can continue to perform the operations of receiving the job perception data stream and recording the job perception data sequence. For example, in the case where the detection feedback result represents that the running state of the NAS device is abnormal, the NAS device can stop performing the operations of receiving the job perception data stream and recording the job perception data sequence.

[0054] According to one embodiment of the present disclosure, the NAS device can perform quality detection on the job perception data sequence based on the data quality detection module, record the quality detection result, and process the job perception data sequence with quality problems. For example, whenever a job perception data sequence is written to a preset storage area, quality detection can be performed on the job perception data sequence, and a corresponding quality detection result can be obtained, which can indicate the quality problems of the job perception data sequence. According to the real-time quality detection result, the available data and the problem data can be automatically classified to realize real-time quality detection and screening processing of the job perception data sequence, thereby reducing the workload of subsequent manual data cleaning and effectively reducing the labor cost.

[0055] For example, according to the quality detection result, the job perception data sequence and / or at least one frame of job perception data with quality problems can be identified, and the job perception data sequence and / or at least one frame of job perception data with quality problems can be deleted or removed from the preset storage area, thereby realizing timely cleaning of the perception data that does not meet the quality requirements, screening reliable job perception data for data closed loop, thereby saving the storage space of the NAS device and improving the reliability of the data.

[0056] According to an embodiment of the present disclosure, by detecting the running state of the NAS device, the possible software and hardware environment problems in the NAS device can be found in time based on the detection feedback result, so as to ensure the stability and reliability of the recording process, find potential problems in time, and avoid data recording interruption or abnormality. By performing real-time quality detection on the job perception data sequence, the available data and the problem data can be distinguished in time, and reliable job perception data sequence is screened for data closed loop, thereby reducing the waste of network traffic and NAS device storage resources by invalid data, avoiding uploading of data with quality problems to the cloud, thereby improving the reliability of data recording, reducing the workload of manual intervention and post-data cleaning, and helping to improve the iteration efficiency of the automatic driving algorithm.

[0057] According to an embodiment of the present disclosure, the perception data stream processing method further comprises: determining a quality alarm packet based on the quality detection result; and sending the quality alarm packet to a specified terminal, so that the specified terminal controls the target vehicle to adjust the execution state of the perception data stream acquisition task.

[0058] In one embodiment, in the case that the quality detection result indicates that the job perception data sequence has quality problems, a quality alarm packet can be generated based on the quality detection result. The NAS device can send the quality alarm packet to a specified terminal (such as a terminal of an operating personnel, a terminal of a vehicle infotainment system of a target vehicle, etc.), so that the specified terminal controls the target vehicle to adjust the execution state of the perception data stream acquisition task.

[0059] For example, in response to receiving the quality alarm packet, the designated terminal can control the NAS device to stop performing the operation of receiving the job-aware data stream and the job-aware data sequence.

[0060] According to an embodiment of the present disclosure, the job-aware data sequence includes a job-aware image sequence; the quality detection on the job-aware data sequence includes: performing pixel attribute detection on at least one frame of job-aware image in the job-aware image sequence to obtain pixel attribute detection information; and in response to the pixel attribute detection information not satisfying a preset pixel quality condition, determining a target detection result.

[0061] According to an embodiment of the present disclosure, the job-aware image sequence may, for example, include perception video data obtained based on a vehicle-mounted camera, and the job-aware image sequence may include at least one frame of job-aware image.

[0062] As an example, the pixel attribute detection may, for example, include image brightness detection, image resolution detection, etc. The preset pixel quality condition may, for example, include but is not limited to image brightness requirement, image resolution requirement, etc. For example, the job-aware image is a monitoring image captured in a scene without turning on the light at night, and the image brightness detection can be performed on at least one frame of job-aware image to obtain pixel attribute detection information representing the image brightness. In response to the pixel attribute detection information representing the image brightness not satisfying the image brightness requirement (for example, the image brightness of the job-aware image is lower than a preset image brightness threshold), a target detection result can be determined, which indicates that the corresponding job-aware image has a quality problem. Thus, in the case that the job-aware image in the perception data collected by the vehicle does not satisfy the preset pixel quality condition, the corresponding job-aware image sequence can be marked or deleted, and the job-aware data in the low-quality job-aware data stream can be processed in time to avoid the low-quality job-aware image sequence occupying storage space, causing redundant occupation of device storage resources, and to reduce the waste of computing overhead generated by performing data processing tasks based on low-quality job-aware data.

[0063] According to an embodiment of the present disclosure, the quality detection on the job-aware data sequence includes: determining a target running state of the target vehicle in a specified time period based on a time attribute of the job-aware data sequence and running state information sent from the target vehicle, the specified time period being determined based on the time attribute; and determining a target detection result based on the target running state; the target running state includes at least one of the following: a first running state indicating that the target vehicle is stationary; and a second running state indicating that a specified lighting device related to the job-aware data sequence does not perform lighting in a specified lighting time period, the specified lighting device being configured on the target vehicle.

[0064] According to one embodiment of the present disclosure, the job perception data sequence can have a time attribute, which can represent, for example, a job period corresponding to the job perception data sequence. The running state information of the target vehicle can include, for example, but not limited to, the running state of the vehicle device (such as the running state of the vehicle light, etc.), the job execution state of the vehicle (such as the stop job state, etc.).

[0065] As an example, the time attribute of the job perception data sequence is 22:00-22:10, and it can be determined that the job time corresponding to the job perception data sequence is 22:00-22:10 at night. It can be understood that the natural light at 22:00-22:10 at night is insufficient, which belongs to the lighting period, and the target vehicle should turn on the lighting device (such as the vehicle light) for lighting when performing the job task in the lighting period.

[0066] Based on the foregoing time attribute and the running state information sent by the target vehicle, the target running state of the target vehicle in the specified period of 22:00-22:10 at night can be determined. The target running state can include at least one of the following: a first running state indicating that the target vehicle is stationary; a second running state indicating that the specified lighting device related to the job perception data sequence in the specified lighting period is not illuminated. The foregoing specified lighting device is configured on the target vehicle, and the specified lighting device can include a lighting device (such as a vehicle light) with a lighting direction similar to the perception direction of the perception device (such as the camera direction of the vehicle-mounted camera). The first running state indicates that the target vehicle is in a stop job state at 22:00-22:10 at night, and the corresponding job perception data sequence has a lower value for data closed loop and automatic driving algorithm iteration. The second running state indicates that the specified lighting device of the target vehicle is not illuminated at 22:00-22:10 at night, and the pixel brightness of the corresponding job perception data sequence can not meet the preset image brightness requirement.

[0067] Exemplarily, in the case where it is determined that the target vehicle is in the target running state in the specified period, the target detection result can be determined based on the target running state, which indicates that the job perception data sequence has a quality problem.

[0068] According to an embodiment of the present disclosure, the job perception data sequence includes a job perception image sequence; wherein, based on the time attribute of the job perception data sequence, the target running state of the target vehicle in the specified period is determined, including: based on the time attribute of the job perception image sequence and the image brightness attribute of at least one frame of job perception image, the image brightness state in the specified lighting period is determined; based on the image brightness state and the device field of the specified lighting device related to the image acquisition angle of the job perception image, the second running state is determined.

[0069] According to one embodiment of the present disclosure, the job awareness data sequence can include a job awareness image sequence, and the job awareness image sequence can include at least one frame of job awareness image. The specified lighting device related to the image acquisition angle of the job awareness image can include, for example but not limited to, the lighting device around the camera, or a plurality of specified lighting devices related to the image acquisition angle range of the camera.

[0070] As an example, the job awareness image sequence is the image sequence captured by the camera installed beside the vehicle side marker. The specified lighting device related to the image acquisition angle of the job awareness image sequence can include, for example, the vehicle side marker beside the camera and the searchlight installed on the roof of the vehicle.

[0071] In one embodiment, the time attribute of the job awareness image sequence is 22:00-22:10, and it can be determined that the job time corresponding to the job awareness image sequence is 22:00-22:10 at night. For example, if the image brightness attribute of at least one frame of job awareness image does not meet the image brightness requirement, the image brightness state of the specified lighting period (22:00-22:10 at night) can be determined to be insufficient image brightness based on the time attribute of the job awareness image sequence and the image brightness attribute of at least one frame of job awareness image. For example, based on the running state information of the target vehicle, it is detected that the searchlight installed on the roof of the vehicle is not on during the specified lighting period, which causes the image captured by the camera installed beside the vehicle side marker to also not meet the image brightness requirement. Based on the image brightness state (insufficient image brightness) and the device field of the specified lighting device in the specified lighting period, a second running state indicating that the specified lighting device (such as the lighting installed at the rear of the vehicle, etc.) corresponding to the device field among the plurality of lighting devices in the vehicle has an abnormal lighting state can be determined, so that the target vehicle can be determined to not perform normal lighting work for the specified lighting device related to the job awareness data sequence during the specified lighting period based on the second running state. Therefore, a prompt data packet can be sent to the vehicle to control the specified lighting device to work, or a prompt data packet can be sent to the specified control end to prompt the relevant personnel to perform maintenance. At the same time, the job awareness image sequence with quality problems in the range of the specified lighting device indicating the abnormal lighting state can be marked or deleted, etc., to reduce the storage space occupation of the equipment resources.

[0072] According to an embodiment of the present disclosure, the quality detection of the job awareness data sequence includes: performing difference detection based on the time attributes of the plurality of job awareness data in the job awareness data sequence to obtain a time difference detection result; and determining a frame loss defect detection result based on the comparison result between the time difference detection result and a preset difference time threshold.

[0073] According to one embodiment of the present disclosure, the job awareness data sequence can include multiple frames of job awareness data, and the time attribute of the job awareness data can be, for example, a timestamp. Illustratively, the timestamp difference (ΔT) between adjacent two frames of data can be calculated according to the respective timestamps of the multiple frames of job awareness data, to obtain a time difference detection result. The timestamp difference ΔT can be compared with a preset difference time threshold, to determine a frame loss defect detection result, which represents that the job awareness data sequence has a quality problem (frame loss problem). For example, if it is detected that ΔT is significantly greater than an expected frame interval (i.e., the preset difference time threshold, which is, for example, 1 / 30 second), it is determined that there is frame loss. If ΔT>1.5×expected frame interval (tolerating network jitter), the number of frame loss is increased by 1.

[0074] According to one embodiment of the present disclosure, the data quality detection module of the NAS device can continuously perform quality detection on the recorded data (job awareness data sequence), and the quality detection content can include, for example, pixel attribute detection, target running state detection, and frame loss defect detection. Based on the target detection result and / or the frame loss defect detection result described above, a quality detection result can be determined, so that the job awareness data sequence with a quality problem can be processed according to the quality detection result.

[0075] Figure 4 An example of a quality detection result according to an embodiment of the present disclosure is shown.

[0076] In one embodiment, the data quality checking module at the NAS device end continuously performs quality detection on the data files written locally. When a new data file (job awareness data sequence) is completed, each frame of udp data in the file is automatically parsed, and a detailed data quality checking report in json format is generated, and a brief checking result is recorded in the log (as shown in Figure 4 The data that fails the quality check is moved to a folder named error_data, so as to be distinguished from other data. When a relevant person wants to view the data on the vehicle, the checking result of each piece of data can be quickly determined according to the log and the folder where the data is stored. If the specific reason why the data fails the check is wanted to be known, the json format report can be consulted.

[0077] More preferably, since the vehicle job awareness data can be recorded in batches and then copied out of the NAS device at one time, in order to avoid repeated checking, a database can also be maintained to record the file names of the data that have been checked. Each time the data quality checking module is started, it can query the database to confirm whether the files already existing in the NAS device have been checked. Optionally, the checking records in the database, the log, and the detailed checking report of the local data are automatically cleaned up according to a preset storage period (for example, 15 days), and the expired records are automatically deleted.

[0078] According to an embodiment of the present disclosure, the perception data stream processing method further comprises: executing a detection instruction before receiving the job perception data stream to obtain a detection feedback result; preferably, the detection instruction is executed before receiving the job perception data stream to obtain a detection feedback result, comprising: executing a process state detection instruction to determine a first detection feedback result representing an amount of idle process resources; and / or executing a connectivity detection instruction to generate a feedback data packet, wherein the connectivity detection instruction is determined based on a detection data packet sent by the target vehicle; further, the detection instruction is executed to obtain a detection feedback result, further comprising: executing a time synchronization detection instruction to determine a second detection feedback result representing a system time attribute of the server; and / or executing a storage space detection instruction to determine a third detection feedback result representing a remaining storage space in the server.

[0079] According to an embodiment of the present disclosure, after starting the environment problem detection module in the domain controller, the environment problem detection module in the domain controller can send a detection instruction to the NAS device, and the NAS device can execute the detection instruction before receiving the job perception data stream to obtain a detection feedback result.

[0080] As an example, the detection instruction is used to check the possible environment problems in the NAS device. The scope of the check and the checking method are as follows:

[0081] 1. NAS device connectivity check

[0082] Problem description: If the NAS device is not connected, data cannot be transmitted to the NAS device, and recording cannot be started.

[0083] Checking method: the domain controller sends a detection data packet to the IP address of the NAS device through a ping instruction (i.e. a connectivity detection instruction), and the NAS device executes the connectivity detection instruction to generate a feedback data packet.

[0084] 2. NAS device remaining space check

[0085] Problem description: If the remaining space of the NAS device is insufficient, sufficient data cannot be recorded.

[0086] Checking method: the domain controller remotely connects to the NAS device through ssh (Secure Shell, a protocol for secure remote login and command execution), and the NAS device checks the remaining space of the NAS device by executing a storage space detection instruction to determine a third detection feedback result. For example, the domain controller can remotely connect to the NAS device through ssh and execute a space detection instruction, or directly check the remaining space of the NAS device in the file system to determine the third detection feedback result.

[0087] 3. NAS device time synchronization check

[0088] Problem description: If the NAS device time is not synchronized, the time of the recorded data will have errors from the real time, which will interfere with the filtering of data in a specific time period.

[0089] Checking method: Remote connection to the NAS device through ssh, and the NAS device end checks whether the system time of the NAS device is consistent with the domain controller and whether the time synchronization service of the NAS device is running normally by executing the time synchronization detection instruction, to determine the second detection feedback result.

[0090] 4. NAS device process check

[0091] Problem description: The data recording program will occupy a large amount of disk I / O, and if there are other high I / O occupying processes running in the NAS device, it may interfere with data recording.

[0092] Checking method: Remote connection to the NAS device through ssh by the domain controller, and the NAS device end checks whether there are processes running that interfere with data recording by executing the process state detection instruction, to determine the first detection feedback result. For example, the ssh + iostat (I / O statistics) instruction can be used to check whether there are processes on the NAS device that occupy too much I / O resource (hard disk read / write). For example, the ssh + top instruction (the top instruction can dynamically view the resource occupation of each process in real time) can be used to check whether there are processes on the NAS device that occupy too much CPU resource.

[0093] 5. Other device connection check

[0094] Problem description: If important devices (such as cameras, radars) are not connected, important information will be missing in the recorded data, which may not be able to be put into use.

[0095] Checking method: The NAS device end listens to the port of the important device on the vehicle to detect whether there is data, to obtain the fourth detection feedback result.

[0096] Figure 5 A flowchart of a perception data stream processing method according to another embodiment of the present disclosure is schematically shown.

[0097] As Figure 5 shown, the method 500 includes operations S510-S530.

[0098] At operation S510, a job perception data stream is obtained.

[0099] In operation S520, before sending the job awareness data stream to the service end and in the process of sending the job awareness data stream, the running state of the service end is detected by sending a detection data packet carrying a specified detection instruction to the service end, and a detection result is obtained.

[0100] In operation S530, if the detection result indicates that the running state of the service end is abnormal, a target instruction is sent to the service end to control the service end to stop recording the job awareness data.

[0101] According to an embodiment of the present disclosure, the method 500 is applied to a vehicle end, which may, for example, include a domain controller of a target vehicle. The target vehicle can be any type of vehicle with a self-driving function or an automatic auxiliary driving function, and the domain controller of the target vehicle can be connected to a service end (such as a NAS device) through a network. The job awareness data stream may, for example, include vehicle surrounding environment information collected by the vehicle in real time through multi-modal sensors (such as a camera, a laser radar, a millimeter wave radar, etc.).

[0102] According to an embodiment of the present disclosure, before sending the job awareness data stream to the NAS device end and in the process of sending the job awareness data stream, the domain controller end can send a detection data packet carrying a specified detection instruction to the NAS device end to detect the running state of the NAS device end and obtain a detection result. The detection result can represent the running state of the NAS device end, so as to determine whether there is a problem in the running environment of the NAS device end.

[0103] As an example, before sending the job awareness data stream to the NAS device end, the domain controller end can send a detection data packet carrying a specified detection instruction to the NAS device end based on the environment problem detection module to detect the running state of the NAS device end. If the detection result indicates that no environment problem is detected, the environment problem detection module will automatically start the data recording module and the data quality detection module in the NAS device; if the detection result indicates that an environment problem is detected, the environment problem detection module will output a warning on a specified terminal and exit.

[0104] As an example, in the process of sending the job awareness data stream to the NAS device end, the environment problem detection module in the domain controller can continuously check the environment problem of the NAS device end at a frequency of, for example, once every 10 minutes. If the detection result indicates that an environment problem is detected in the middle of the process, the environment problem detection module will directly remotely close the data recording module and the data quality detection module in the NAS device, and then exit.

[0105] According to an embodiment of the present disclosure, the running state of the service end is detected, including: receiving a detection feedback result sent by the service end, the detection feedback result being determined by the service end executing a specified detection instruction; determining an abnormal detection result representing an abnormal running state of the service end based on the detection feedback result, wherein the detection feedback result includes at least one of: a first detection feedback result representing an idle process resource amount of the service end, the abnormal detection result corresponding to the first detection feedback result indicating that the idle process resource of the service end is abnormal; a second detection feedback result representing a system time attribute of the service end, the abnormal detection result corresponding to the second detection feedback result indicating that there is a system time synchronization abnormality between the service end and the vehicle end. Preferably, the running state of the service end is detected, further including: in response to the number of feedback data packets in the target detection period not satisfying the preset number condition, determining a communication abnormality detection result, the feedback data packet being determined by the service end based on the detection data packet.

[0106] According to an embodiment of the present disclosure, the domain controller end can receive a detection feedback result sent by the NAS device end, and can determine an abnormal detection result representing an abnormal running state of the NAS device end based on the detection feedback result.

[0107] In one embodiment, the abnormal detection result representing the abnormal running state of the service end can be determined based on the corresponding detection feedback result of at least one of the following. Exemplarily:

[0108] 1. NAS device connectivity check

[0109] For example, if the domain controller does not receive the feedback data packet sent by the NAS device end, it means that the detection data packet sending fails, indicating that the NAS device is not connected. For example, if the number of feedback data packets received by the domain controller in the target detection period is less than a preset threshold (for example, 5), it means that the detection data packet sending fails, indicating that the NAS device is not connected. If the NAS device is not connected, it is determined that the running state of the service end is abnormal, and the environmental problem detection module can output a NAS device non-connection warning and then directly exit without entering the subsequent checking process.

[0110] 2. NAS device remaining space check

[0111] For example, if the third detection feedback result represents that the remaining space of the NAS device is less than a given threshold, it is determined that the running state of the service end is abnormal, and the environmental problem detection module can output a NAS device insufficient space warning.

[0112] 3. NAS device time synchronization check

[0113] For example, if the second detection feedback result represents that there is a time difference or a time synchronization service abnormality, it is determined that the running state of the service end is abnormal, and the environmental problem detection module can output a NAS device time synchronization abnormality warning.

[0114] 4. NAS device process check

[0115] For example, if the first detection feedback result represents that the proportion of idle process resources on the NAS device side is less than 50%, it is determined that the service side running state is abnormal, and the environment problem detection module can output a NAS device process interference warning.

[0116] 5. Other device connection check

[0117] For example, if the fourth detection feedback result represents that no data is detected from the port of the important device, it is determined that the service side running state is abnormal, and the environment problem detection module can output a warning of abnormal connection of the vehicle side device.

[0118] Figure 6 The flowchart of the perception data processing method according to the embodiment of the present disclosure is schematically shown.

[0119] As shown in Figure 6 , the domain control process on the domain controller side includes an environment problem detection module, and the NAS device process on the NAS device side includes a data quality detection module and a data recording module.

[0120] As shown in Figure 6 , after starting the environment problem detection module in the domain control process, the environment problem detection module can automatically perform the operation of detecting the environment problem. The environment problem detection module can check the environment problem that may exist on the NAS device side, and determine whether there is a running environment problem on the NAS device side.

[0121] For example, if it is detected that there is an environment problem on the NAS device side, the environment problem detection module can print a warning on the terminal and exit. For example, if it is detected that there is no environment problem on the NAS device side, the environment problem detection module can detect whether the NAS device process (including the data recording module and the data quality detection module) has been started. If it is detected that the NAS device process has not been started, the environment problem detection module can remotely start the data recording module and the data quality detection module in the NAS device process; if it is detected that the NAS device process has been started, the environment problem detection module can sleep, and then continuously check the environment problem on the NAS device side at a frequency of 10 minutes. If an environment problem is detected in the middle, the environment problem detection module will directly remotely close the data recording module and the data quality detection module in the NAS device, and then exit.

[0122] As shown in Figure 6As shown, the environmental problem detection module can remotely start the data recording module in the NAS device process. The data recording module can record data in a loop until the environmental problem detection module remotely shuts down the data recording module in the NAS device, at which point the data recording module stops recording data.

[0123] like Figure 6 As shown, the environmental problem detection module can remotely start the data quality detection module within the NAS device's process. The data quality detection module can normally be in a dormant state. Whenever it detects that new data (job-aware data sequence) has been written to a preset storage area, the data quality detection module exits dormancy and performs a quality check on the new data, outputting the result. If the quality check result indicates that the new data passes the quality check, the data quality detection module can remain dormant until the next new data is detected. If the quality check result indicates that the new data fails the quality check, it indicates that the new data has a quality problem, and the data quality detection module can move the data with quality problems to the folder error_data / .

[0124] This disclosure also provides a server and a vehicle.

[0125] Figure 7 A schematic block diagram of a server according to an embodiment of the present disclosure is shown.

[0126] like Figure 7 As shown, the server 700 includes a first communication module 710, a memory 720, and a first processor 730.

[0127] The first communication module 710 is used to receive the operation perception data stream sent by the target vehicle. The first communication module is also used to send detection feedback results to the target vehicle.

[0128] The memory 720 includes a preset storage area.

[0129] The first processor 730 is configured to, in response to receiving a task perception data stream sent by a target vehicle, write a task perception data sequence determined based on the task perception data stream into a preset storage area. During the process of receiving the task perception data stream, the first processor is also configured to execute a specified detection instruction sent by the target vehicle to obtain detection feedback results. The specified detection instruction is used to detect the operating status of the server. The first processor is also configured to perform quality detection on the task perception data sequence, record the quality detection results, and process task perception data sequences with quality problems.

[0130] Figure 8 A schematic block diagram of a vehicle according to an embodiment of the present disclosure is shown.

[0131] like Figure 8As shown, the vehicle 800 can include a collection module 810, a second communication module 820, and a second processor 830. The second communication module is further configured to send the target instruction to the server.

[0132] The collection module 810 is configured to acquire the job perception data stream.

[0133] The second communication module 820 is configured to send the job perception data stream to the server. The second communication module is further configured to, before sending the job perception data stream to the server and during the process of sending the job perception data stream, detect the running state of the server by sending a detection data packet carrying a specified detection instruction to the server, and receive a detection result.

[0134] The second processor 830 is configured to, in a case where the detection result represents that the running state of the server is abnormal, generate a target instruction for controlling the server to stop recording the job perception data.

[0135] Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure, or at least part of the functions of any one or more of the modules, sub-modules, units, sub-units, can be implemented in one module. Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be split into multiple modules for implementation. Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or any other reasonable way of hardware or firmware through integration or packaging of circuits, or in any one of software, hardware, and firmware, or in an appropriate combination of any one or more of them. Alternatively, one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be at least partially implemented as computer program modules that can perform corresponding functions when executed.

[0136] For example, any of the first communication module 710, the memory 720, and the first processor 730 can be combined into one module / unit / sub-unit, or any of them can be split into multiple modules / units / sub-units. Alternatively, at least part of the functions of one or more of these modules / units / sub-units can be combined with at least part of the functions of the other modules / units / sub-units, and implemented in one module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the first communication module 710, the memory 720, and the first processor 730 can be implemented at least in part as a hardware circuit, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on board, a system on package, an application-specific integrated circuit (ASIC), or any other reasonable manner of integrating or packaging a circuit, etc. in hardware or firmware, or in any one of or in a proper combination of software, hardware, and firmware. Alternatively, at least one of the first communication module 710, the memory 720, and the first processor 730 can be implemented at least in part as a computer program module that can perform corresponding functions when the computer program module is run.

[0137] For example, any of the first communication module 710, the memory 720, and the first processor 730 can be combined into one module / unit / sub-unit, or any of them can be split into multiple modules / units / sub-units. Alternatively, at least part of the functions of one or more of these modules / units / sub-units can be combined with at least part of the functions of the other modules / units / sub-units, and implemented in one module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the first communication module 710, the memory 720, and the first processor 730 can be implemented at least in part as a hardware circuit, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on board, a system on package, an application-specific integrated circuit (ASIC), or any other reasonable manner of integrating or packaging a circuit, etc. in hardware or firmware, or in any one of or in a proper combination of software, hardware, and firmware. Alternatively, at least one of the first communication module 710, the memory 720, and the first processor 730 can be implemented at least in part as a computer program module that can perform corresponding functions when the computer program module is run.

[0138] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on a user's computing device, partially on a user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0139] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features described in the various embodiments of the present disclosure can be combined and / or combined in various ways, even if such combinations are not explicitly described in the present disclosure. In particular, the features described in the various embodiments of this disclosure may be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0140] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A method for processing a perception data stream, applied to a server, comprising: writing, in response to receiving a job perception data stream sent by a target vehicle, a job perception data sequence determined based on the job perception data stream in a preset storage area, wherein during the receiving of the job perception data stream, a specified detection instruction sent by the target vehicle is executed to obtain a detection feedback result, and the detection feedback result is sent to the target vehicle, the specified detection instruction being used to detect a running state of the server; performing quality detection on the job perception data sequence, recording a quality detection result, and processing a job perception data sequence having a quality problem.

2. The method of claim 1, wherein, The job perception data sequence comprises a job perception image sequence; the quality detection on the job perception data sequence comprises: performing pixel attribute detection on at least one frame of job perception image in the job perception image sequence to obtain pixel attribute detection information; and determining a target detection result in response to the pixel attribute detection information not satisfying a preset pixel quality condition.

3. The method of claim 1, wherein, The quality detection on the job perception data sequence comprises: determining a target running state of the target vehicle in a specified time period based on a time attribute of the job perception data sequence and running state information sent by the target vehicle, the specified time period being determined based on the time attribute; and determining the target detection result based on the target running state; The target running state comprises at least one of: a first running state indicating that the target vehicle is at rest; and a second running state indicating that a specified lighting device associated with the job perception data sequence does not perform lighting in a specified lighting time period, the specified lighting device being configured on the target vehicle.

4. The method of claim 3, wherein, The job perception data sequence comprises a job perception image sequence; The determination of the target running state of the target vehicle in the specified time period based on the time attribute of the job perception data sequence comprises: determining an image brightness state in the specified lighting time period based on a time attribute of the job perception image sequence and an image brightness attribute of at least one frame of job perception image; and determining the second running state based on the image brightness state and a device field of the specified lighting device associated with an image acquisition view angle of the job perception image.

5. The method of claim 1, wherein, The quality detection on the job perception data sequence comprises: performing difference detection based on time attributes of a plurality of job perception data in the job perception data sequence to obtain a time difference detection result; and determining a frame loss defect detection result based on a comparison result between the time difference detection result and a preset difference time threshold.

6. The method of claim 1, wherein, The method further comprises: executing a detection instruction before receiving the job perception data stream to obtain a detection feedback result; Preferably, the execution of the detection instruction before receiving the job perception data stream to obtain the detection feedback result comprises: executing a process state detection instruction to determine a first detection feedback result representing an idle process resource amount; and / or executing a connectivity detection instruction to generate a feedback data packet, wherein the connectivity detection instruction is determined based on a detection data packet sent by the target vehicle. Further, the detection instruction is executed to obtain a detection feedback result, and the detection feedback result further comprises: a second detection feedback result representing a system time attribute of the server; and / or a third detection feedback result representing a remaining storage space in the server.

7. A perception data stream processing method applied to a vehicle, comprising: obtaining a job perception data stream; before sending the job perception data stream to a server, and during the sending process, detecting a running state of the server by sending a detection data packet carrying a specified detection instruction to the server to obtain a detection result; and in a case where the detection result represents an abnormal running state of the server, sending a target instruction to the server to control the server to stop recording the job perception data. The detection of the running state of the server comprises:

8. The method of claim 7, wherein, receiving a detection feedback result sent by the server, the detection feedback result being determined by the server executing the specified detection instruction; determining an abnormal detection result representing an abnormal running state of the server based on the detection feedback result, wherein the detection feedback result comprises at least one of the following: a first detection feedback result representing an idle process resource amount of the server, and an abnormal detection result corresponding to the first detection feedback result indicating that the idle process resource of the server is abnormal; a second detection feedback result representing a system time attribute of the server, and an abnormal detection result corresponding to the second detection feedback result indicating that there is a system time synchronization abnormality between the server and the vehicle. Preferably, the detection of the running state of the server further comprises: in response to a number of feedback data packets in a target detection period not satisfying a preset number condition, determining a communication abnormality detection result, the feedback data packets being determined by the server based on the detection data packet.

9. A server, comprising: a first communication module configured to receive a job perception data stream sent by a target vehicle; a storage comprising a preset storage area; a first processor configured to, in response to receiving the job perception data stream sent by the target vehicle, write a job perception data sequence determined based on the job perception data stream in the preset storage area, wherein during the receiving of the job perception data stream, the first processor is further configured to execute a specified detection instruction sent by the target vehicle to obtain a detection feedback result, the specified detection instruction being used to detect a running state of the server; the first communication module is further configured to send the detection feedback result to the target vehicle; the first processor is further configured to perform quality detection on the job perception data sequence, record the quality detection result, and process the job perception data sequence having a quality problem.

10. A vehicle, comprising: a collection module configured to obtain a job perception data stream; a second communication module configured to send the job perception data stream to a server. ​ The second communication module is further configured to, before sending the job awareness data stream to the server and during sending the job awareness data stream, send a detection data packet carrying a specified detection instruction to the server to detect a running state of the server, and receive a detection result; And The second processor is configured to, in a case where the detection result represents that the running state of the server is abnormal, generate a target instruction for controlling the server to stop recording the job awareness data. The second communication module is further configured to send the target instruction to the server.

Images