Data processing system and method

Abstract

The present invention provides a data processing system and method, which relate to the field of intelligent transportation technology. The data processing system includes: a sensor system, a data collector system and a data extraction platform; wherein the sensor system is used to obtain first data information during the automatic driving process of the vehicle to be detected; the data collector system is used to process the first data information and obtain second data information; the data extraction platform is used to mark the second data information and store it in a scene library data center. By collecting and automatically marking the real scene data in the process of automatic driving of the vehicle, a core tool system is provided for the construction of an automatic driving scene database, which can effectively reduce a lot of time and manpower costs wasted due to manual data collection and data labeling.

Description

Technical Field

[0001] The present invention relates to the field of intelligent transportation technology, and in particular to a data processing system and method. Background Art

[0002] With the development of science and technology, cars, as an indispensable means of transportation in people's lives, have also been given more functions. For example, the unmanned driving function, in order to prove that unmanned vehicles are safer than human-driven vehicles, needs to test 100 vehicles 24 hours a day in various traffic scenarios, and test them continuously for 225 years. To achieve the rapid road access of smart cars, collecting autonomous driving scene data, building a scene library, and realizing accelerated testing have become the only effective ways. Therefore, a system that can collect autonomous driving scene data is currently needed. Summary of the invention

[0003] The embodiments of the present invention provide a data processing system and method for solving the problem of how to collect data of an unmanned vehicle.

[0004] In order to solve the above technical problems, an embodiment of the present invention provides a data processing system, including:

[0005] sensor systems, data logger systems, and data extraction platforms;

[0006] Wherein, the sensor system is used to obtain first data information during the automatic driving process of the vehicle to be detected;

[0007] The data collector system is used to process the first data information to obtain second data information;

[0008] The data extraction platform is used to annotate the second data information and store it in the scene library data center.

[0009] Further, the sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module;

[0010] Wherein, the camera is used to collect video information;

[0011] The first radar is used to detect the position information, distance information and speed information of obstacles around the vehicle to be detected;

[0012] The GPS / IMU module is used to provide a calibration clock source and collect the position information of the vehicle to be detected;

[0013] The second radar is used to obtain the three-dimensional spatial structure of the road where the vehicle to be detected is located;

[0014] The CAN information acquisition module is used to obtain the real-time driving status information of the vehicle to be detected.

[0015] Further, the data collector system includes: a data reading module and a data recording module;

[0016] Wherein, the data reading module is used to read the first data information collected by the sensor system;

[0017] The data recording module is used to perform clock correction on the first data information to set a timestamp, and to package the first data information to obtain the second data information.

[0018] Further, the data extraction platform includes: a data generator and a fusion controller;

[0019] The data generator is used to restore the second data information and send it to the fusion controller in the form of UDP;

[0020] The fusion controller is used to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

[0021] The embodiment of the present invention further provides a data processing method, which is applied to the above-mentioned data processing system, and the method includes:

[0022] Acquiring first data information detected by the sensor system during the vehicle's automatic driving process;

[0023] Controlling the data acquisition system to process the first data information and obtain second data information;

[0024] The control data extraction platform marks the second data information and stores it in the scene library data center.

[0025] Furthermore, the sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module, and the first data information detected by the sensor system during the vehicle automatic driving process is acquired, including:

[0026] Obtaining video information collected by the camera;

[0027] Obtaining position information, distance information, and speed information of obstacles around the vehicle to be detected detected by the first radar;

[0028] Obtaining the location information of the vehicle to be detected collected by the GPS / IMU module;

[0029] Acquire the three-dimensional spatial structure of the road on which the vehicle to be detected is located detected by the second radar;

[0030] Acquire the real-time driving status information of the vehicle to be detected collected by the CAN information.

[0031] Furthermore, the data acquisition system includes: a data reading module and a data recording module, and the control data acquisition system processes the first data information to obtain the second data information, including:

[0032] Controlling the data reading module to read the first data information collected by the sensor system;

[0033] The data recording module is controlled to perform clock correction on the first data information and set a timestamp, and the first data information is packaged to obtain the second data information.

[0034] Furthermore, the data extraction platform includes: a data generator and a fusion controller, and the control data extraction platform marks the second data information and stores it in the scene library data center, including:

[0035] Control the data generator to restore the second data information and send it to the fusion controller in the form of UDP;

[0036] The fusion controller is controlled to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

[0037] An embodiment of the present invention further provides a data processing device, the device comprising:

[0038] An acquisition module, used to acquire first data information detected by a sensor system during the vehicle automatic driving process;

[0039] A first control module, used for controlling the data acquisition system to process the first data information and obtain second data information;

[0040] The second control module is used to control the data extraction platform to mark the second data information and store it in the scene library data center.

[0041] The beneficial effects of the present invention are:

[0042] The above solution provides a core tool system for building an autonomous driving scenario database by collecting and automatically labeling real scene data during the vehicle's autonomous driving process, which can effectively reduce the huge amount of time and manpower costs wasted due to manual data collection and data labeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram showing the structure of a data processing system according to an embodiment of the present invention;

[0044] Figure 2 A schematic diagram showing a flow chart of a data processing method according to an embodiment of the present invention;

[0045] Figure 3 A second flowchart diagram showing a data processing method according to an embodiment of the present invention;

[0046] Figure 4 A third flowchart of a data processing method according to an embodiment of the present invention;

[0047] Figure 5 A schematic diagram showing the structure of a data processing device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] The present invention aims at the problem of how to collect data of unmanned vehicles and provides a data processing system and method.

[0050] like Figure 1 As shown, an embodiment of the present invention provides a data processing system, including:

[0051] sensor systems, data logger systems, and data extraction platforms;

[0052] Wherein, the sensor system is used to obtain first data information during the automatic driving process of the vehicle to be detected;

[0053] The data collector system is used to process the first data information to obtain second data information;

[0054] The data extraction platform is used to annotate the second data information and store it in the scene library data center.

[0055] The embodiments of the present invention collect and automatically label real scene data during the vehicle's autonomous driving process, thereby providing a core tool system for building an autonomous driving scene database, which can effectively reduce a large amount of time and manpower costs wasted due to manual data collection and data labeling.

[0056] Specifically, the sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module;

[0057] The camera is used to collect video information and cooperate with the laser radar to mark features in the image, such as lane lines, signs, stop lines, etc.

[0058] The first radar is used to detect the position information, distance information and speed information of the obstacles around the vehicle to be detected, and the first radar may be a millimeter wave radar.

[0059] The GPS / IMU module is used to provide a calibration clock source and collect the position information and posture information of the vehicle to be detected.

[0060] The second radar is used to obtain the three-dimensional spatial structure of the road where the vehicle to be detected is located, and the second radar can be a laser radar.

[0061] The CAN information acquisition module is used to acquire the real-time driving status information of the vehicle to be detected, and the real-time driving status information includes: steering wheel angle information, throttle opening information, brake pedal opening information, etc.

[0062] Specifically, the data collector system includes: a data reading module and a data recording module;

[0063] Wherein, the data reading module is used to read the first data information collected by the sensor system;

[0064] The data recording module is used to perform clock correction on the first data information to set a timestamp, and to package the first data information to obtain the second data information.

[0065] It should be noted that the data reading module is also used to initialize the CAN interface, Ethernet interface and UART interface, obtain the initialization success handle, and cyclically read the first data information sent by the sensor system, that is, the original detection data according to the handle, and classify and analyze the first data information. The data recording module is also used to calibrate the real-time clock through the GPS time.

[0066] Specifically, the data extraction platform includes: a data generator and a fusion controller;

[0067] The data generator is used to restore the second data information and send it to the fusion controller in the form of UDP;

[0068] The fusion controller is used to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

[0069] It should be noted that the data generator includes: a data receiving module, a data timing processing module, and a data sending module. The data generator is used to restore the second data information stored in the data recording module and simulate the generation of each sensor signal to send to the fusion controller. Among them, the data receiving module receives the original data information of each sensor collected by the display platform through the TCP / IP protocol, and splits and stores the data information based on the data information collection time and data type; the data timing processing module sends the data to be sent stored in the buffer queue to the data sending module for data transmission according to the time interrupt (minimum period 50μs) generated by the time series generator and the time interval difference of each data in the data buffer queue; the data sending module simulates the signal generation of each sensor under actual working conditions, and sends the data information to the fusion controller in the form of UDP.

[0070] It should be further explained that the fusion controller includes: a data receiving and parsing module, a fusion algorithm processing module and a fusion data sending module. The data receiving and parsing module takes out the data within a cycle of different sensors, parses them into fusion input types and stores them in a buffer queue; the fusion algorithm processing module performs fusion processing on the detection data information of different sensors to obtain fusion output data; the fusion data sending module sends the fusion output data to the display platform and the scene library data center through the network.

[0071] The data processing system of the embodiment of the present invention can extract time information, scene information, environmental information, and vehicle driving status information in the driving scene, and perform semantic-level annotation on the data corresponding to this information, which becomes the storage label of the data set and is finally stored in the scene library data center.

[0072] like Figure 2 As shown, an embodiment of the present invention further provides a data processing method, which is applied to the above-mentioned data processing system, and the method includes:

[0073] Step 21, obtaining first data information during the vehicle automatic driving process detected by the sensor system;

[0074] Step 22, controlling the data acquisition system to process the first data information to obtain second data information;

[0075] Step 23, controlling the data extraction platform to mark the second data information and store it in the scene library data center.

[0076] The embodiments of the present invention collect and automatically label real scene data during the vehicle's autonomous driving process, thereby providing a core tool system for building an autonomous driving scene database, which can effectively reduce a large amount of time and manpower costs wasted due to manual data collection and data labeling.

[0077] Specifically, the sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module. The step 21 acquires the first data information detected by the sensor system during the vehicle automatic driving process, including:

[0078] Obtaining video information collected by the camera;

[0079] Obtaining position information, distance information, and speed information of obstacles around the vehicle to be detected detected by the first radar;

[0080] Obtaining the location information of the vehicle to be detected collected by the GPS / IMU module;

[0081] Acquire the three-dimensional spatial structure of the road on which the vehicle to be detected is located detected by the second radar;

[0082] Acquire the real-time driving status information of the vehicle to be detected collected by the CAN information.

[0083] Specifically, the data acquisition system includes: a data reading module and a data recording module. The step 22 controls the data acquisition system to process the first data information and obtain the second data information, including:

[0084] Controlling the data reading module to read the first data information collected by the sensor system;

[0085] The data recording module is controlled to perform clock correction on the first data information and set a timestamp, and the first data information is packaged to obtain the second data information.

[0086] It should be noted that the data reading module is also used to initialize the CAN interface, Ethernet interface and UART interface, obtain the initialization success handle, and cyclically read the first data information sent by the sensor system, that is, the original detection data according to the handle, and classify and analyze the first data information. The data recording module is also used to calibrate the real-time clock through the GPS time.

[0087] Specifically, the data extraction platform includes: a data generator and a fusion controller. The step 23 controls the data extraction platform to mark the second data information and store it in the scene library data center, including:

[0088] Control the data generator to restore the second data information and send it to the fusion controller in the form of UDP;

[0089] The fusion controller is controlled to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

[0090] It should be noted that the data generator includes: a data receiving module, a data timing processing module, and a data sending module. The data generator is used to restore the second data information stored in the data recording module and simulate the generation of each sensor signal to send to the fusion controller. Figure 3 As shown, the data receiving module receives the raw data information of each sensor collected by the display platform through the TCP / IP protocol, and splits and stores the data information based on the data collection time and data type; the data timing processing module sends the waiting data stored in the buffer queue to the data sending module for data transmission according to the time interrupt (minimum period 50μs) generated by the time series generator and the time interval difference of each data in the data buffer queue; the data sending module simulates the signal generation of each sensor under actual working conditions and sends the data information to the fusion controller in the form of UDP.

[0091] It should be further explained that the fusion controller includes: a data receiving and analyzing module, a fusion algorithm processing module and a fusion data sending module. Figure 4 As shown, the data receiving and parsing module takes out the data within a cycle of different sensors, parses it into fusion input type and stores it in the buffer queue; the fusion algorithm processing module fuses the detection data information of different sensors to obtain fusion output data; the fusion data sending module sends the fusion output data to the display platform and scene library data center through the network.

[0092] The data processing system of the embodiment of the present invention can extract time information, scene information, environmental information, and vehicle driving status information in the driving scene, and perform semantic-level annotation on the data corresponding to this information, which becomes the storage label of the data set and is finally stored in the scene library data center.

[0093] like Figure 5 As shown, an embodiment of the present invention further provides a data processing device, the device comprising:

[0094] An acquisition module 51 is used to acquire first data information during the vehicle automatic driving process detected by the sensor system;

[0095] A first control module 52, used to control the data acquisition system to process the first data information and obtain second data information;

[0096] The second control module 53 is used to control the data extraction platform to mark the second data information and store it in the scene library data center.

[0097] Specifically, the sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module, and the acquisition module 51 detects the first data information of the vehicle in the automatic driving process detected by the sensor system, and includes:

[0098] A first acquisition unit, used to acquire video information collected by the camera;

[0099] A second acquisition unit, used to acquire position information, distance information and speed information of obstacles around the vehicle to be detected detected by the first radar;

[0100] A third acquisition unit is used to acquire the position information of the vehicle to be detected collected by the GPS / IMU module;

[0101] a fourth acquisition unit, configured to acquire the three-dimensional spatial structure of the road on which the vehicle to be detected is located, detected by the second radar;

[0102] The fifth acquisition unit is used to acquire the real-time driving status information of the vehicle to be detected collected by the CAN information.

[0103] Specifically, the data acquisition system includes: a data reading module and a data recording module, and the first control module 52 includes:

[0104] A first control unit, used for controlling the data reading module to read the first data information collected by the sensor system;

[0105] The second control unit is used to control the data recording module to perform clock correction on the first data information to set a timestamp, and to package the first data information to obtain the second data information.

[0106] Specifically, the data extraction platform includes: a data generator and a fusion controller, and the second control module 53 includes:

[0107] A third control unit, used for controlling the data generator to restore the second data information and send it to the fusion controller in the form of UDP;

[0108] The fourth control unit is used to control the fusion controller to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

[0109] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary personnel in this technical field, several improvements and modifications can be made without departing from the principles described in the present invention. These improvements and modifications are also within the scope of protection of the present invention.

Claims

1. A data processing system, characterized in that: include: sensor systems, data logger systems, and data extraction platforms; Wherein, the sensor system is used to obtain first data information during the automatic driving process of the vehicle to be detected; The data collector system is used to process the first data information to obtain second data information; The data extraction platform is used to annotate the second data information and store it in the scene library data center; The data extraction platform includes: a data generator and a fusion controller; The data generator is used to restore the second data information and send it to the fusion controller in the form of UDP; The data generator includes: a data receiving module, a data timing processing module, and a data sending module; The data generator is used to restore the second data information stored in the data recording module and simulate the generation of each sensor signal to send to the fusion controller; Among them, the data receiving module receives the raw data information of each sensor collected by the display platform through the TCP / IP protocol, and splits and stores the data information based on the data information collection time and data type; the data timing processing module sends the waiting data stored in the buffer queue to the data sending module for data transmission according to the time interrupt generated by the time series generator and the time interval difference of each data in the data buffer queue; the data sending module simulates the signal generation of each sensor under actual working conditions and sends the data information to the fusion controller in the form of UDP; The fusion controller includes: a data receiving and analyzing module, a fusion algorithm processing module and a fusion data sending module; The data receiving and parsing module takes out the data of different sensors within one cycle, parses it into fusion input type and stores it in the buffer queue; the fusion algorithm processing module fuses the detection data information of different sensors to obtain fusion output data; the fusion data sending module sends the fusion output data to the display platform and scene library data center through the network; The fusion controller is used to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to the scene library data center; it is also used to extract the time information, scene information, environmental information, and vehicle driving status information in the driving scene, and perform semantic level annotation on the data corresponding to these information, which becomes the storage label of the data within one cycle taken out by the data receiving and parsing module from different sensors, and finally stores it in the scene library data center; The data acquisition system comprises: a data reading module and a data recording module; The data reading module is used to read the first data information collected by the sensor system; it is also used to initialize the CAN interface, Ethernet interface and UART interface, obtain a successful initialization handle, and cyclically read the first data information sent by the sensor system according to the handle, and classify and analyze the first data information; The data recording module is used to perform clock correction on the first data information, set a timestamp, and package the first data information to obtain the second data information; and is also used to correct the real-time clock through GPS time.

2. The data processing system according to claim 1, characterized in that The sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module; Wherein, the camera is used to collect video information; The first radar is used to detect the position information, distance information and speed information of obstacles around the vehicle to be detected; The GPS / IMU module is used to provide a calibration clock source and collect the position information of the vehicle to be detected; The second radar is used to obtain the three-dimensional spatial structure of the road where the vehicle to be detected is located; The CAN information acquisition module is used to obtain the real-time driving status information of the vehicle to be detected.

3. A data processing method, applied to the data processing system according to any one of claims 1 to 2, characterized in that: The method comprises: Acquiring first data information detected by the sensor system during the vehicle's automatic driving process; Controlling the data acquisition system to process the first data information and obtain second data information; Controlling the data extraction platform to mark the second data information and store it in the scene library data center; The data extraction platform includes: a data generator and a fusion controller, and the control data extraction platform marks the second data information and stores it in the scene library data center, including: Control the data generator to restore the second data information and send it to the fusion controller in the form of UDP; The fusion controller is controlled to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

4. The data processing method according to claim 3, characterized in that: The sensor system includes: a camera, a first radar, a GPS / IMU module, a second radar and a CAN information acquisition module. The first data information detected by the sensor system during the vehicle automatic driving process includes: Obtaining video information collected by the camera; Obtaining position information, distance information, and speed information of obstacles around the vehicle to be detected detected by the first radar; Obtaining the location information of the vehicle to be detected collected by the GPS / IMU module; Acquire the three-dimensional spatial structure of the road on which the vehicle to be detected is located detected by the second radar; Acquire the real-time driving status information of the vehicle to be detected collected by the CAN information acquisition module.

5. The data processing method according to claim 3, characterized in that: The data acquisition system includes: a data reading module and a data recording module. The control data acquisition system processes the first data information to obtain the second data information, including: Controlling the data reading module to read the first data information collected by the sensor system; The data recording module is controlled to perform clock correction on the first data information and set a timestamp, and the first data information is packaged to obtain the second data information.

6. A data processing device, applied to the data processing system according to any one of claims 1 to 2, characterized in that: The device comprises: An acquisition module, used to acquire first data information detected by a sensor system during the vehicle automatic driving process; A first control module, used for controlling the data acquisition system to process the first data information and obtain second data information; A second control module, used to control the data extraction platform to mark the second data information and store it in a scene library data center; The data extraction platform includes: a data generator and a fusion controller, and the second control module includes: A third control unit, used for controlling the data generator to restore the second data information and send it to the fusion controller in the form of UDP; The fourth control unit is used to control the fusion controller to receive the data sent by the data generator, store the data into a buffer queue according to a preset period, and send the data to a scene library data center.

Images