Recharge method and device of test data, recharge system of test data

Abstract

The application discloses a test data backfilling method and device and a test data backfilling system, and relates to the technical field of data processing. The method comprises the following steps: receiving a data backfilling request, wherein the data backfilling request is used for requesting to backfill test data of an auxiliary driving system in a target driving scene, and the test data is used for testing the auxiliary driving performance of the auxiliary driving system in the target driving scene; in response to the data backfilling request, reference driving data collected in the target driving scene is obtained from a driving data set of the auxiliary driving system; the reference driving data is subjected to data optimization in a data optimization mode matched with a test demand of the auxiliary driving system, so that target driving data with improved matching degree with the test demand is obtained; and the target driving data is backfilled to the auxiliary driving system. By adopting the technical scheme, the problem of low test efficiency of the auxiliary driving performance of the auxiliary driving system in the related art is solved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method and apparatus for reinjecting test data, and a test data reinjection system. Background Technology

[0002] With the continuous development of the times, people have increasingly higher requirements for the safety and comfort of car driving. ADAS (Advanced Driver Assistance System) has also made great strides, with increasing demand and more sensors installed in vehicles, collecting more and more data. This data is used for obstacle recognition, distance accuracy detection, and hazard warnings. The realization of these functions requires a large amount of data to verify and evaluate their reliability.

[0003] Currently, most testing and verification of advanced driver assistance systems (ADAS) in the field are conducted through real-world road tests. This involves installing the developed ADAS equipment on vehicles and testing it in real-world environments. This is an essential step in the development and testing process, providing a direct evaluation of ADAS performance reliability. However, relying solely on reliability testing consumes significant manpower and resources and has a long testing cycle. Therefore, relying solely on this method is not feasible for rapidly optimizing ADAS performance in a short period. Furthermore, the main drawback of real-world road testing is its high manpower and resource consumption, susceptibility to weather and road conditions, long feedback time for test results, and inability to quickly pinpoint problems, which hinders algorithm performance evaluation and iterative optimization.

[0004] There is still no effective solution to the problem of low testing efficiency of the assisted driving performance of related technologies. Summary of the Invention

[0005] This application provides a method and apparatus for backfeeding test data, a storage medium, and an electronic device to at least solve the problem of low testing efficiency for the assisted driving performance of assisted driving systems in related technologies.

[0006] According to one embodiment of this application, a method for feeding back test data is provided, comprising: receiving a data feeding back request, wherein the data feeding back request is used to request the feeding back of test data of a driver assistance system to be tested in a target driving scenario, the test data being used to test the driver assistance performance of the driver assistance system in the target driving scenario; responding to the data feeding back request, obtaining reference driving data collected in the target driving scenario from the driving data set of the driver assistance system; optimizing the reference driving data according to a data optimization method matching the test requirements of the driver assistance system to obtain target driving data with increased matching degree with the test requirements; and feeding the target driving data back to the driver assistance system.

[0007] Optionally, optimizing the reference driving data according to a data optimization method that matches the test requirements of the assisted driving system to obtain target driving data with increased matching degree with the test requirements includes at least one of the following: adjusting the frame rate of the reference driving image in the reference driving data according to the frame rate requirements of the assisted driving system to obtain a first driving image matching the frame rate requirements as the target driving data, wherein the test requirements include the frame rate requirements; optimizing the image region where the target object is located in the reference driving image according to the target object involved in the assisted driving system to obtain a second driving image as the target driving data, wherein the test requirements include the target object, and the target object is a road object that affects the assisted driving process of the assisted driving system.

[0008] Optionally, adjusting the frame rate of the reference driving image in the reference driving data according to the frame rate requirement of the assisted driving system to obtain a first driving image matching the frame rate requirement as the target driving data includes: determining the timestamp of the image frame to be inserted in the reference driving image according to the frame rate requirement and the reference frame rate of the reference driving image; determining the target position of the target object in the image at the timestamp from a local motion model, wherein the local motion model is used to characterize the motion trajectory of the moving object in the reference driving image; adding the target object to the target position of the reference image frame to obtain a candidate image frame, wherein the image size of the reference image frame is the same as the image size of the reference driving image; filling the background of the candidate image frame to obtain a target image frame; and inserting the target image frame into the position corresponding to the timestamp in the reference driving image to obtain the first driving image.

[0009] Optionally, before determining the target position of the target object in the image at the timestamp from the local motion model, the method further includes: extracting feature descriptors corresponding to each image frame of the reference driving image, wherein the feature descriptors are used to describe the features of corner points in the image frames; calculating affine model transformation parameters corresponding to two adjacent image frames based on the feature descriptors of adjacent image frames in the reference driving image, wherein the affine model transformation parameters are used to characterize the motion trajectory of objects in the two adjacent image frames; and constructing the local motion model using the affine model transformation parameters.

[0010] Optionally, the step of optimizing the image region where the target object is located in the reference driving image based on the target object involved in the assisted driving system to obtain a second driving image as the target driving data includes: using a target image recognition model to recognize the reference driving image to obtain a target image region output by the target image recognition model, wherein the target image region is the region where the target object is located, and the target image recognition model is obtained by training an initial image recognition model using image samples labeled with the image region corresponding to the reference object; and enlarging the target image region in the reference driving image to obtain the second driving image.

[0011] Optionally, the step of magnifying the target image region in the reference driving image to obtain the second driving image includes: segmenting the target image region from the reference driving image; and performing bicubic interpolation magnification on the image region to obtain the second driving image.

[0012] Optionally, obtaining the reference driving data collected in the target driving scenario from the driving data set of the assisted driving system includes: obtaining driving data collected from multiple collection angles in the target driving scenario from the driving data set of the assisted driving system; and aligning the timestamps of the driving data collected from the multiple collection angles to obtain the reference driving data.

[0013] According to another embodiment of this application, a test data reinjection system is also provided, comprising: a host computer and a data processing module, wherein the host computer is connected to an assisted driving system, the host computer is also connected to the data processing module, and the data processing module is connected to the assisted driving system; the host computer is configured to generate a data reinjection request, wherein the data reinjection request is used to request the reinjection of test data of the assisted driving system under a target driving scenario, the test data being used to test the assisted driving performance of the assisted driving system under the target driving scenario; the data processing module is configured to receive the data reinjection request; respond to the data reinjection request by obtaining reference driving data collected under the target driving scenario from the driving data set of the assisted driving system; optimize the reference driving data according to a data optimization method matching the test requirements of the assisted driving system to obtain target driving data with increased matching degree with the test requirements; and reinject the target driving data into the assisted driving system; the host computer is also configured to monitor the data reinjection result of the assisted driving system.

[0014] According to another embodiment of this application, a test data reinjection device is also provided, comprising: a receiving module, configured to receive a data reinjection request, wherein the data reinjection request is used to request the reinjection of test data of a driver assistance system under test in a target driving scenario, the test data being used to test the driver assistance performance of the driver assistance system in the target driving scenario; an acquisition module, configured to respond to the data reinjection request and acquire reference driving data collected in the target driving scenario from the driving data set of the driver assistance system; an optimization module, configured to optimize the reference driving data according to a data optimization method matching the test requirements of the driver assistance system, thereby obtaining target driving data with increased matching degree with the test requirements; and a reinjection module, configured to reinject the target driving data into the driver assistance system.

[0015] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to execute the above-described test data backfeeding method when running.

[0016] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the test data re-feeding method through the computer program.

[0017] In this embodiment, a data feedback request is received, wherein the data feedback request is used to request the feedback of test data of the driver assistance system to be tested in a target driving scenario, the test data being used to test the driver assistance performance of the driver assistance system in the target driving scenario; in response to the data feedback request, reference driving data collected in the target driving scenario is obtained from the driving data set of the driver assistance system; the reference driving data is optimized according to a data optimization method matching the test requirements of the driver assistance system to obtain target driving data with increased matching degree with the test requirements; and the target driving data is fed back into the driver assistance system. In essence, the driving dataset stores driving data collected in specific driving scenarios. Upon receiving a request to re-feed back test data of the assisted driving system under test in the target driving scenario, the system retrieves reference driving data collected in the target driving scenario from the driving dataset in response to the request. This reference data is then optimized according to a data optimization method that matches the testing requirements. This optimization increases the matching degree between the testing requirements and the optimized target driving data. The optimized target driving data is then used to re-feed back the assisted driving system, thereby better testing the assisted driving performance of the system in the target driving scenario. This technical solution solves the problem of low testing efficiency for the assisted driving performance of assisted driving systems in related technologies, achieving the technical effect of improving the testing efficiency of assisted driving performance. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the hardware environment for a test data refeeding method according to an embodiment of this application;

[0021] Figure 2 This is a flowchart of a test data reinjection method according to an embodiment of this application;

[0022] Figure 3 This is a flowchart of an optional image optimization method according to an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of a test data reinjection system according to an embodiment of this application. Figure 1 ;

[0024] Figure 5 This is a schematic diagram of another optional test data reinjection system according to an embodiment of this application. Figure 2 ;

[0025] Figure 6 This is a schematic diagram of another optional test data reinjection system according to an embodiment of this application. Figure 3 ;

[0026] Figure 7 This is a structural block diagram of a test data reinjection device according to an embodiment of this application. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] The methods and embodiments provided in this application can be executed on a computer terminal, device terminal, or similar computing device. Taking running on a computer terminal as an example, Figure 1 This is a schematic diagram of the hardware environment for a test data refeeding method according to an embodiment of this application. Figure 1 As shown, a computer terminal may include one or more ( Figure 1Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. In one exemplary embodiment, the computer terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the computer terminal described above. For example, the computer terminal may also include components that are more complex than those described above. Figure 1 The more or fewer components shown, or having the same Figure 1 Equivalent functions or ratios shown Figure 1 The functions shown have more different configurations.

[0030] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the message push sending method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer programs stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to a computer terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0031] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the computer terminal. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0032] This embodiment provides a method for backfeeding test data, applied to the aforementioned device terminal. Figure 2 This is a flowchart of a test data reinjection method according to an embodiment of this application, such as... Figure 2 As shown, the process includes the following steps:

[0033] Step S202: Receive a data feedback request, wherein the data feedback request is used to request the feedback of test data of the driver assistance system to be tested in the target driving scenario, and the test data is used to test the driver assistance performance of the driver assistance system in the target driving scenario;

[0034] Step S204: In response to the data feedback request, obtain the reference driving data collected in the target driving scenario from the driving data set of the assisted driving system;

[0035] Step S206: Optimize the reference driving data according to the data optimization method that matches the test requirements of the assisted driving system to obtain target driving data with increased matching degree with the test requirements;

[0036] Step S208: Feed the target driving data back into the driver assistance system.

[0037] Through the above steps, the driving dataset stores driving data collected in driving scenarios. Upon receiving a request to re-feed back test data of the assisted driving system under test in the target driving scenario, the system retrieves reference driving data collected in the target driving scenario from the driving dataset in response to the request. This reference data is then optimized according to a data optimization method that matches the testing requirements. This optimization increases the matching degree between the testing requirements and the optimized target driving data. The optimized target driving data is then used to re-feed back the assisted driving system, thereby better testing the assisted driving performance of the system in the target driving scenario. This technical solution solves the problem of low testing efficiency for the assisted driving performance of assisted driving systems in related technologies, achieving the technical effect of improving the testing efficiency of assisted driving performance.

[0038] In the technical solution provided in step S202 above, the driver assistance system is used to assist the driving status of the driving equipment. The driver assistance system may include, but is not limited to, lane keeping assist system, automatic parking assist system, brake assist system, reversing assist system and driving assist system, etc. This solution does not limit this.

[0039] Optionally, in this embodiment, the target driving scenario can be used to characterize the test scenario of the assisted driving system, such as test weather, test road conditions, etc. At the same time, the target driving scenario can also be used to characterize the test requirements for testing the assisted driving system. For example, if the assisted driving system is an automatic parking assisted driving system, then the test scenario is the automatic parking scenario.

[0040] Optionally, in this embodiment, the test data may be, but is not limited to, video data, radar data, etc., wherein the radar data may be, but is not limited to, radar data collected by radar devices such as lidar, ultrasonic radar, and corner radar.

[0041] In the technical solution provided in step S204 above, the driving data set is used to store driving data collected during the road test under different driving scenarios.

[0042] Optionally, in this embodiment, driving data is used to characterize the vehicle's driving state, and may include, but is not limited to, the vehicle's driving posture, lane information during the vehicle's driving process, relative positional relationship with other vehicles or obstacles, etc. This solution does not limit this.

[0043] In the technical solution provided in step S206 above, the data optimization method may include, but is not limited to, adjusting the clarity of the image data, adjusting the format of the image data, adjusting the frame rate of the image data, highlighting objects in the image data, etc. This solution does not limit this.

[0044] Optionally, in this embodiment, the test requirements can correspond to the assisted driving functions of the assisted driving system. Different assisted driving systems with different assisted driving functions correspond to different test requirements. For example, the test requirement for an automatic parking system is the impact of the parking point reference on the parking during parking, while the test requirement for a lane keeping assist system is the impact of lane lines on the driving state during driving. This solution does not limit this.

[0045] In the technical solution provided in step S208 above, data backfeed is used to instruct the transmission of test data used to test the driving performance of the assisted driving system in a certain driving scenario to the assisted driving system to be tested, so as to instruct the assisted driving system to be tested to run the assisted driving program according to the test data, thereby realizing the corresponding assisted driving function.

[0046] As an optional embodiment, optimizing the reference driving data according to a data optimization method that matches the testing requirements of the assisted driving system to obtain target driving data with increased matching degree with the testing requirements includes at least one of the following:

[0047] The frame rate of the reference driving image in the reference driving data is adjusted according to the frame rate requirement of the driving assistance system to obtain a first driving image that matches the frame rate requirement as the target driving data, wherein the test requirement includes the frame rate requirement;

[0048] Based on the target object involved in the assisted driving system, the image region where the target object is located in the reference driving image is optimized to obtain a second driving image as the target driving data. The test requirements include the target object, which is a road object that affects the assisted driving process of the assisted driving system.

[0049] Optionally, in this embodiment, different types of driver assistance systems may have fixed requirements for the frame rate of the input driving images. For example, lower-version driver assistance systems require image data with a frame rate of 20 FPS, while higher-version driver assistance systems require image data with a frame rate of 30 FPS. This solution does not impose any limitations on this.

[0050] Optionally, in this embodiment, the image data input to the same driver assistance system can be collected by different acquisition devices in the same driving scenario. The frame rate of the images collected by different acquisition devices is different. The frame rate requirement of the driver assistance system is that the frame rate of the driving data collected by different acquisition devices in the same driving scenario should be the same. For example, images collected by cameras set in different positions on the vehicle. However, due to different camera hardware conditions or camera failures, most of the images collected by multiple cameras have the same frame rate, while a small number of cameras have dropped frames. Therefore, it is necessary to unify the frame rate of the images collected by multiple cameras before inputting them into the driver assistance system.

[0051] Optionally, in this embodiment, adjusting the frame rate may include, but is not limited to, frame interpolation and frame deletion; this solution does not limit this.

[0052] Optionally, in this embodiment, the target object can be pedestrians, vehicles, lane lines, etc. in the image, and this application does not limit this.

[0053] Optionally, in this embodiment, image optimization is used to increase the recognition of the target object by the driver assistance system. Image optimization may be achieved by adjusting the sharpness of the area where the target object is located in the image, or by enlarging the area corresponding to the target object in the image; this solution does not limit this approach.

[0054] Optionally, in this embodiment, the road object may include, but is not limited to, vehicles traveling on the road, pedestrians, reference objects located around the road (such as buildings, streetlights, etc.), road signs, lane lines, etc., and this solution does not limit this.

[0055] As an optional embodiment, adjusting the frame rate of the reference driving image in the reference driving data according to the frame rate requirement of the driving assistance system to obtain a first driving image matching the frame rate requirement as the target driving data includes:

[0056] The timestamps of the image frames to be inserted in the reference driving image are determined based on the frame rate requirement and the reference frame rate of the reference driving image.

[0057] The target position of the target object in the image at the timestamp is determined from the local motion model, wherein the local motion model is used to characterize the motion trajectory of the moving object in the reference driving image;

[0058] The target object is added to the target position of the reference image frame to obtain a candidate image frame, wherein the image size of the reference image frame is the same as the image size of the reference driving image;

[0059] The candidate image frames are filled with background to obtain the target image frames;

[0060] The target image frame is inserted into the position corresponding to the timestamp in the reference driving image to obtain the first driving image.

[0061] Optionally, in this embodiment, the timestamp of the image frame to be inserted can be determined based on the frame rate difference between the reference frame rate and the target frame rate of the reference driving image. For example, the target number of image frames to be inserted can be determined based on the target difference between the reference frame rate and the target frame rate, and the image frames to be inserted at the target number of frames can be randomly inserted into the reference driving image. Alternatively, the reference driving image can be divided into the target number of segments, and the middle timestamp of each segment can be determined as the timestamp of the inserted image frame. This solution does not limit this.

[0062] As an optional embodiment, before determining the target position of the target object in the image at the timestamp from the local motion model, the method further includes:

[0063] Extract feature descriptors corresponding to each image frame of the reference driving image, wherein the feature descriptors are used to describe the features of corner points in the image frames;

[0064] The affine model transformation parameters corresponding to two adjacent image frames are calculated based on the feature descriptors of adjacent image frames in the reference driving image, wherein the affine model transformation parameters are used to characterize the motion trajectory of objects in two adjacent image frames.

[0065] The local motion model is constructed using the affine model transformation parameters.

[0066] Optionally, in this embodiment, corner points are points used to characterize the salient attributes of an image.

[0067] Optionally, in this embodiment, the detected corner points are described using mathematical features, such as gradient histograms and local random binary features. Extraction algorithms include neighborhood template matching, SIFT feature descriptors, and ORB feature descriptors.

[0068] As an optional embodiment, the step of optimizing the image region where the target object is located in the reference driving image based on the target object involved in the assisted driving system to obtain a second driving image as the target driving data includes:

[0069] The target image recognition model is used to recognize the reference driving image to obtain the target image region output by the target image recognition model. The target image region is the region where the target object is located. The target image recognition model is obtained by training an initial image recognition model using image samples with image regions corresponding to the reference object labeled.

[0070] The target image region in the reference driving image is magnified to obtain the second driving image.

[0071] Optionally, in this embodiment, the image recognition model can be used to identify the target image region where the target object is located in the image to be recognized. For example, it can output the image with the target image region selected, or it can output the coordinates of the target image region in the image. This solution does not limit this.

[0072] Optionally, in this embodiment, the image recognition model can also be used to segment the target image region from the image for output, that is, to output an image containing only the target image region.

[0073] Optionally, in this embodiment, the target region can be magnified using various interpolation algorithms, such as bicubic interpolation.

[0074] Figure 3 This is a flowchart of an optional image optimization method according to an embodiment of this application, such as... Figure 3 As shown, this image optimization method includes at least the following steps:

[0075] S301, the image is labeled by pre-determining the category of the object in the image or the location of the object in the image. The labeled image is used as a training sample to train the initial image recognition model to obtain the target image recognition model. The initial image recognition model can be, but is not limited to, R-CNN (a model that applies deep learning to image detection). This model can abstract the detection into two processes: first, based on the image, several regions that may contain objects are proposed; second, a classification network (AlexNet) is run on these proposed regions to obtain the category of the object in each region.

[0076] S302, the reference driving image is input into the target image recognition model to obtain the category information of the object in the image output by the target image recognition model and the location information of the area where the object is located.

[0077] 303, in practical scenarios, assigns different categories of meaning to pixels to segment the target image region corresponding to the target object.

[0078] S304 performs intra-frame interpolation on the segmented target image region to achieve resolution adaptation. It can process the image resolution by performing bicubic interpolation to enlarge the image resolution according to the required image region size.

[0079] As an optional embodiment, the step of magnifying the target image region in the reference driving image to obtain the second driving image includes:

[0080] The target image region is segmented from the reference driving image;

[0081] The image region is magnified by bicubic interpolation to obtain the second driving image.

[0082] Optionally, in this embodiment, the method of segmenting the target image region from the reference driving image may include, but is not limited to, determining the position coordinates of the target image region in each image frame of the reference driving image (the target image recognition model outputs the position coordinates of the target region in the reference driving image), cropping each frame of the image according to the position coordinates, cropping the target image region in each frame of the image, and sorting the cropped image frames containing the reference image region according to the order of the image frames in the reference driving image, thereby obtaining the target driving image, wherein the target driving data includes the target driving image, thereby completing the operation of segmenting the target image region from the reference driving image.

[0083] Optionally, in this embodiment, after performing bicubic interpolation magnification, image deblurring, noise reduction, edge enhancement, and other processing can be performed to improve image quality. Whether further processing is needed depends on actual requirements.

[0084] As an optional embodiment, obtaining the reference driving data collected in the target driving scenario from the driving data set of the assisted driving system includes:

[0085] The driving data collected from multiple angles in the target driving scenario is obtained from the driving data set of the assisted driving system.

[0086] The driving data collected from the multiple acquisition angles is timestamped to obtain the reference driving data.

[0087] Optionally, in this embodiment, when performing data feedback for the assisted driving system, in order to realistically recreate the driving scenario, it is usually necessary to use driving data collected from different collection angles on the vehicle. The driving data collected from different collection angles is collected during the road test by collection devices installed at different locations on the road test vehicle. Then, during the testing phase, the driving scenario is realistically recreated by retrieving the driving data collected from different angles.

[0088] This embodiment provides a test data reinjection system. Figure 4 This is a schematic diagram of a test data reinjection system according to an embodiment of this application. Figure 1 ,like Figure 4 As shown, the test data reinjection system includes at least: a host computer 42 and a data processing module 44, wherein the host computer 42 is connected to the driver assistance system, the host computer 42 is also connected to the data processing module 44, and the data processing module 44 is connected to the driver assistance system;

[0089] The host computer 42 is used to generate a data feedback request, wherein the data feedback request is used to request the feedback of test data of the assisted driving system to be tested in the target driving scenario, and the test data is used to test the assisted driving performance of the assisted driving system in the target driving scenario.

[0090] The data processing module 44 is configured to receive the data feedback request; respond to the data feedback request by obtaining reference driving data collected in the target driving scenario from the driving data set of the assisted driving system; optimize the reference driving data according to the data optimization method that matches the test requirements of the assisted driving system to obtain target driving data with increased matching degree with the test requirements; and feed the target driving data back to the assisted driving system.

[0091] The host computer 42 is also used to monitor the data feedback results of the driver assistance system.

[0092] Based on the above, the driving dataset stores driving data collected in driving scenarios. Upon receiving a request to re-feed back test data of the assisted driving system under test in the target driving scenario, the system retrieves reference driving data collected in the target driving scenario from the driving dataset in response to the request. This reference data is then optimized according to a data optimization method that matches the testing requirements. This optimization increases the matching degree between the testing requirements and the optimized target driving data. The optimized target driving data is then used to re-feed back the assisted driving system, thereby better testing the assisted driving performance of the system in the target driving scenario. This technical solution solves the problem of low testing efficiency for the assisted driving performance of assisted driving systems in related technologies, achieving the technical effect of improving the testing efficiency of assisted driving performance.

[0093] Optionally, in this embodiment, the host computer can also be used to obtain the data source of the re-feedback data. The data source may include, but is not limited to, multiple video data, various radar data, and vehicle body data, etc. The data source can be stored on a local hard drive or in a local area network.

[0094] Optionally, in this embodiment, the host computer can also be used to parse the file format and assemble corresponding reference driving data. For example, it can parse the format of video data. If the parsing reveals that the video data format is inconsistent or that the video data format is inconsistent with the data format requirements of the driver assistance system under test, the video data can be converted to a format that meets the data format requirements of the driver assistance system. Similarly, the host computer can also parse radar data. If the parsing reveals that the radar data format is inconsistent or that the radar data format is inconsistent with the data format requirements of the driver assistance system under test, the radar data can be converted to a format that meets the data format requirements of the driver assistance system. This solution does not limit this aspect.

[0095] Optionally, in this embodiment, the data processing module may be, but is not limited to, an FPGA chip packaged according to the test data feedback requirements. Different functional areas are packaged in the FPGA chip to execute the test data feedback method described in this application. For example, the FPGA is used to optimize the reference real-time data, and different functional areas are packaged in the FPGA chip according to different optimization methods. For example, the first functional area is used to adjust the frame rate of the reference driving image of the reference driving data, and the second functional area is used to optimize the image area where the target object in the reference driving image is located. The target object is a road object that affects the assisted driving process of the assisted driving system. Figure 5This is a schematic diagram of another optional test data reinjection system according to an embodiment of this application. Figure 2 ,like Figure 5 As shown, the test data refeeding system includes a host computer and a data processing module obtained by integrating functions from an FPGA. The FPGA is configured with general-purpose hardware interfaces, such as Flexray, CAN, CANFD, Ethernet, etc., for data transmission, thus adapting to different product projects. The FPGA reads the refeeding data source through PCIe, performs data synchronization and time synchronization processing on the FPGA, and sends the data to the assisted driving system through the hardware interface. It supports CAN, CANFD, Flexray, and LAN. The video stream uses the Fakra interface to support the MIPI protocol, enabling high-definition multi-channel video refeeding. The video data link can be configured through IIC, supporting not only image data of different resolutions but also different image formats, such as RAW12, YUV, H265, etc., supporting high-speed data streams and strong image compatibility. For video display requirements, the refeeding system provides an HDMI interface for connecting an external monitor for video display. If storage is required, a memory can also be connected for corresponding data transmission.

[0096] Figure 6 This is a schematic diagram of another optional test data reinjection system according to an embodiment of this application. Figure 3 ,like Figure 6As shown, the test data refeeding system includes at least a data processing module obtained by integrating the FPGA. Depending on the module's function, it can be divided into a serialization module, a deserialization module, and a data optimization module. The FPGA is connected to the driver assistance system. The FPGA is used to transmit and optimize the test data corresponding to the test scenario according to the test requirements of the driver assistance system. For data transmission, after power-on, the driver assistance system accesses the configuration registers of the serialization / deserialization chips 9295 and 9296 on the FPGA via IIC, and then configures the camera parameters via IIC to achieve multi-channel camera data link configuration on the complete link. The serialization / deserialization chips are used to transmit high-speed video data between boards, supporting multiple high-definition cameras of different resolutions. For a multi-channel high-definition video real-time refeeding system, as shown in the figure, images from multiple cameras deployed at different locations on the test vehicle are acquired. A multi-channel high-definition video refeeding system has been proposed and implemented. For example, it currently supports a six-channel high-definition video acquisition system, adapting to 8MP, 3MP, and 2MP cameras, and also supports different image data formats such as raw12, raw10, and RGB. The video data is compressed using H.265 and stored. During data recovery, the data needs to be decompressed and restored on the FPGA, and then synchronized with the corresponding radar data and vehicle body data according to the timestamp alignment principle before being sent to the assisted driving system under test. Data optimization is achieved through a data optimization module, which has functions for image frame rate adjustment and local image optimization. The image frame rate adjustment function includes adjusting the frame rate of the reference driving image in the reference driving data according to the frame rate requirements of the assisted driving system to obtain a first driving image that matches the frame rate requirements as the target driving data. The local image optimization function includes optimizing the image area where the target object in the reference driving image is located according to the target object involved in the assisted driving system to obtain a second driving image as the target driving data. The test requirements include the target object, which is a road object that affects the assisted driving process of the assisted driving system.

[0097] The above-mentioned test data reinjection system has at least the following functions: the reinjection data source comes from actual road-collected data in the real environment, ensuring the data scenario is authentic and reliable; it can quickly restore the required scenario data to facilitate rapid problem location and improve development efficiency; it can repeat a large number of experiments to intuitively and quantitatively evaluate the function triggering of the assisted driving system and its performance standards, improving development efficiency and system performance; this reinjection system has low environmental dependencies, high encapsulation, and provides various hardware interfaces for domain control, requiring no extensive modifications to the assisted driving system, exhibiting strong compatibility and easy portability to different projects, improving development efficiency and reducing development cycle and labor costs; this reinjection system supports various function triggers of the assisted driving system, supporting high-speed multi-channel video streams and various radar data, including but not limited to lidar, ultrasonic radar, and corner radar. Image data formats support raw12, YUV, and H265. Modifications to the software code of the reinjection system are required, without adaptation to the ADAS system under test.

[0098] To better understand the above process, the following description will be based on optional embodiments, but these are not intended to limit the technical solutions of the embodiments of this application.

[0099] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0100] Figure 7 This is a structural block diagram of a test data reinjection device according to an embodiment of this application; as shown below. Figure 7 As shown, it includes: a receiving module 72, used to receive a data feedback request, wherein the data feedback request is used to request the feedback of test data of the driver assistance system to be tested in the target driving scenario, and the test data is used to test the driver assistance performance of the driver assistance system in the target driving scenario;

[0101] The acquisition module 74 is used to respond to the data feedback request and acquire the reference driving data collected in the target driving scenario from the driving data set of the assisted driving system;

[0102] Optimization module 76 is used to optimize the reference driving data according to the data optimization method that matches the test requirements of the assisted driving system, so as to obtain target driving data with increased matching degree with the test requirements;

[0103] The feedback module 78 is used to feed the target driving data back to the driver assistance system.

[0104] Based on the above, the driving dataset stores driving data collected in driving scenarios. Upon receiving a request to re-feed back test data of the assisted driving system under test in the target driving scenario, the system retrieves reference driving data collected in the target driving scenario from the driving dataset in response to the request. This reference data is then optimized according to a data optimization method that matches the testing requirements. This optimization increases the matching degree between the testing requirements and the optimized target driving data. The optimized target driving data is then used to re-feed back the assisted driving system, thereby better testing the assisted driving performance of the system in the target driving scenario. This technical solution solves the problem of low testing efficiency for the assisted driving performance of assisted driving systems in related technologies, achieving the technical effect of improving the testing efficiency of assisted driving performance.

[0105] Optionally, the optimization module includes at least one of the following: an adjustment unit, configured to adjust the frame rate of the reference driving image in the reference driving data according to the frame rate requirement of the driving assistance system, to obtain a first driving image matching the frame rate requirement as the target driving data, wherein the test requirement includes the frame rate requirement; and an optimization unit, configured to perform image optimization on the image region where the target object is located in the reference driving image according to the target object involved in the driving assistance system, to obtain a second driving image as the target driving data, wherein the test requirement includes the target object, and the target object is a road object that affects the driving assistance process of the driving assistance system.

[0106] Optionally, the adjustment unit is configured to: determine the timestamp of the image frame to be inserted in the reference driving image based on the frame rate requirement and the reference frame rate of the reference driving image; determine the target position of the target object in the image at the timestamp from the local motion model, wherein the local motion model is used to characterize the motion trajectory of the moving object in the reference driving image; add the target object to the target position of the reference image frame to obtain a candidate image frame, wherein the image size of the reference image frame is the same as the image size of the reference driving image; fill the background of the candidate image frame to obtain a target image frame; and insert the target image frame into the position corresponding to the timestamp in the reference driving image to obtain the first driving image.

[0107] Optionally, the apparatus further includes: an extraction module, configured to extract feature descriptors corresponding to each image frame of the reference driving image before determining the target position of the target object in the image at the timestamp from the local motion model, wherein the feature descriptors are used to describe the features of corner points in the image frames; a calculation module, configured to calculate affine model transformation parameters corresponding to two adjacent image frames based on the feature descriptors of adjacent image frames in the reference driving image, wherein the affine model transformation parameters are used to characterize the motion trajectory of objects in two adjacent image frames; and a construction module, configured to construct the local motion model using the affine model transformation parameters.

[0108] Optionally, the optimization unit is configured to: use a target image recognition model to recognize the reference driving image to obtain a target image region output by the target image recognition model, wherein the target image region is the region where the target object is located, and the target image recognition model is obtained by training an initial image recognition model using image samples labeled with the image region corresponding to the reference object; and enlarge the target image region in the reference driving image to obtain the second driving image.

[0109] Optionally, the optimization unit is configured to: segment the target image region from the reference driving image; and perform bicubic interpolation to enlarge the image region to obtain the second driving image.

[0110] Optionally, the acquisition module includes: a collection unit, used to acquire collected driving data from multiple collection angles in the target driving scenario from the driving data set of the assisted driving system; and a processing unit, used to perform timestamp alignment on the collected driving data from the multiple collection angles to obtain the reference driving data.

[0111] Embodiments of this application also provide a storage medium including a stored program, wherein the program executes the test data re-feeding method described above when it runs.

[0112] Optionally, in this embodiment, the storage medium may be configured to store program code for performing the following steps: receiving a data feedback request, wherein the data feedback request is used to request the feedback of test data of the driver assistance system to be tested in a target driving scenario, the test data being used to test the driver assistance performance of the driver assistance system in the target driving scenario; responding to the data feedback request, obtaining reference driving data collected in the target driving scenario from the driving data set of the driver assistance system; optimizing the reference driving data according to a data optimization method matching the test requirements of the driver assistance system to obtain target driving data with increased matching degree with the test requirements; and feeding the target driving data back to the driver assistance system.

[0113] Embodiments of this application also provide an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above-described test data re-feedback method embodiments.

[0114] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0115] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program: receiving a data feedback request, wherein the data feedback request is used to request the feedback of test data of the driver assistance system to be tested in a target driving scenario, the test data being used to test the driver assistance performance of the driver assistance system in the target driving scenario; responding to the data feedback request, obtaining reference driving data collected in the target driving scenario from the driving data set of the driver assistance system; optimizing the reference driving data according to a data optimization method matching the test requirements of the driver assistance system to obtain target driving data with increased matching degree with the test requirements; and feeding the target driving data back to the driver assistance system.

[0116] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0117] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0118] Obviously, those skilled in the art should understand that the modules or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0119] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method of backfilling test data, characterized by, The method comprises: receiving a data backfill request, wherein the data backfill request is used to request backfilling test data of an auxiliary driving system to be tested in a target driving scene, and the test data is used to test the auxiliary driving performance of the auxiliary driving system in the target driving scene; obtaining reference driving data collected in the target driving scene from a driving data set of the auxiliary driving system in response to the data backfill request; performing data optimization on the reference driving data in a data optimization manner matched with a test requirement of the auxiliary driving system to obtain target driving data with increased matching degree with the test requirement; backfilling the target driving data to the auxiliary driving system; The data optimization manner matched with the test requirement of the auxiliary driving system comprises at least one of the following: adjusting a frame rate of reference driving images in the reference driving data according to a frame rate requirement of the auxiliary driving system on the driving images to obtain first driving images matched with the frame rate requirement as the target driving data, wherein the test requirement comprises the frame rate requirement; and performing image optimization on an image region where a target object involved in the auxiliary driving system is located in the reference driving images according to the target object to obtain second driving images as the target driving data, wherein the test requirement comprises the target object, and the target object is a road object affecting the auxiliary driving process of the auxiliary driving system; The adjusting of the frame rate of the reference driving images in the reference driving data according to the frame rate requirement of the auxiliary driving system on the driving images to obtain the first driving images matched with the frame rate requirement as the target driving data comprises: determining a timestamp of an image frame to be inserted in the reference driving images according to the frame rate requirement and a reference frame rate of the reference driving images; determining a target position of the target object in the image at the timestamp from a local motion model, wherein the local motion model is used to represent a motion trajectory of a moving object in the reference driving images; adding the target object to the target position of a reference image frame to obtain a candidate image frame, wherein an image size of the reference image frame is the same as an image size of the reference driving images; performing background filling on the candidate image frame to obtain a target image frame; and inserting the target image frame into a position corresponding to the timestamp in the reference driving images to obtain the first driving images. The image region where the target object is located in the reference driving image is optimized according to the target object involved in the auxiliary driving system, and a second driving image is obtained as the target driving data, including: using a target image recognition model to recognize the reference driving image, and obtaining a target image region output by the target image recognition model, wherein the target image region is a region where the target object is located, and the target image recognition model is obtained by training an initial image recognition model using image samples labeled with image regions corresponding to reference objects; and performing region enlargement on the target image region in the reference driving image to obtain the second driving image; The target image region in the reference driving image is enlarged to obtain the second driving image, including: segmenting the target image region from the reference driving image; and performing bicubic interpolation enlargement on the image region to obtain the second driving image.

2. The method of claim 1, wherein, Before determining the target position of the target object in the image at the timestamp from the local motion model, the method further comprises: extracting feature descriptors corresponding to each image frame of the reference driving image, wherein the feature descriptors are used to describe the features of the corner points in the image frame; calculating affine model transformation parameters corresponding to adjacent two image frames according to the feature descriptors of the adjacent image frames in the reference driving image, wherein the affine model transformation parameters are used to represent the motion trajectories of objects in adjacent two image frames; constructing the local motion model using the affine model transformation parameters.

3. The method of claim 1, wherein, The reference driving data collected in the target driving scene is obtained from the driving data set of the auxiliary driving system, including: obtaining collected driving data in multiple collection angles under the target driving scene from the driving data set of the auxiliary driving system; aligning the timestamps of the collected driving data in multiple collection angles to obtain the reference driving data.

4. A system for re-injection of test data, characterized in that including a host computer and a data processing module, wherein the host computer is connected with the auxiliary driving system, the host computer is also connected with the data processing module, and the data processing module is connected with the auxiliary driving system; The host computer is configured to generate a data backfill request, wherein the data backfill request is used to request backfilling of test data of the auxiliary driving system in a target driving scene, and the test data is used to test the auxiliary driving performance of the auxiliary driving system in the target driving scene. The data processing module is configured to receive the data backfill request, obtain reference driving data collected in the target driving scene from the driving data set of the auxiliary driving system in response to the data backfill request, perform data optimization on the reference driving data according to a data optimization mode matching the test requirements of the auxiliary driving system, obtain target driving data with increased matching degree with the test requirements, and backfill the target driving data to the auxiliary driving system. The host computer is further configured to monitor the data backfill result of the auxiliary driving system. The reference driving data is data-optimized in a manner matching the test requirement of the auxiliary driving system, to obtain target driving data with increased matching degree with the test requirement, including at least one of the following: adjusting a frame rate of reference driving images in the reference driving data according to a frame rate requirement of the auxiliary driving system on driving images, to obtain first driving images matching the frame rate requirement as the target driving data, wherein the test requirement includes the frame rate requirement; performing image optimization on an image region in which a target object is located in the reference driving images according to the target object involved in the auxiliary driving system, to obtain second driving images as the target driving data, wherein the test requirement includes the target object, and the target object is a road object affecting an auxiliary driving process of the auxiliary driving system; The adjusting of the frame rate of the reference driving images in the reference driving data according to the frame rate requirement of the auxiliary driving system on driving images, to obtain the first driving images matching the frame rate requirement as the target driving data, includes: determining a timestamp of an image frame to be inserted in the reference driving images according to the frame rate requirement and a reference frame rate of the reference driving images; determining a target position of the target object in the image at the timestamp from a local motion model, wherein the local motion model is used to represent a motion trajectory of a moving object in the reference driving images; adding the target object to the target position of a reference image frame to obtain a candidate image frame, wherein an image size of the reference image frame is the same as an image size of the reference driving images; performing background filling on the candidate image frame to obtain a target image frame; and inserting the target image frame into a position corresponding to the timestamp in the reference driving images to obtain the first driving images; The performing of the image optimization on the image region in which the target object is located in the reference driving images according to the target object involved in the auxiliary driving system, to obtain the second driving images as the target driving data, includes: identifying the reference driving images using a target image recognition model to obtain a target image region output by the target image recognition model, wherein the target image region is a region in which the target object is located, and the target image recognition model is obtained by training an initial image recognition model using image samples in which image regions corresponding to reference objects are labeled; and performing region enlargement on the target image region in the reference driving images to obtain the second driving images; The performing of the region enlargement on the target image region in the reference driving images to obtain the second driving images includes: segmenting the target image region from the reference driving images; and performing bicubic interpolation enlargement on the image region to obtain the second driving images.

5. A device for backfilling test data, characterized in that including: The receiving module is configured to receive a data backfill request, wherein the data backfill request is used to request backfilling of test data of an auxiliary driving system to be tested in a target driving scene, and the test data is used to test an auxiliary driving performance of the auxiliary driving system in the target driving scene. The obtaining module is configured to obtain reference driving data collected in the target driving scene from a driving data set of the auxiliary driving system in response to the data backfill request. The optimization module is configured to perform data optimization on the reference driving data in a data optimization manner matched with a test requirement of the auxiliary driving system, to obtain target driving data with increased matching degree with the test requirement. The backfilling module is configured to backfill the target driving data to the auxiliary driving system. The optimization module includes at least one of the following: an adjusting unit configured to adjust a frame rate of reference driving images in the reference driving data according to a frame rate requirement of the auxiliary driving system on the driving images, to obtain first driving images matched with the frame rate requirement as the target driving data, wherein the test requirement includes the frame rate requirement; and an optimization unit configured to perform image optimization on an image region in which a target object involved in the auxiliary driving system is located in the reference driving images according to the target object, to obtain second driving images as the target driving data, wherein the test requirement includes the target object, and the target object is a road object that affects an auxiliary driving process of the auxiliary driving system. The adjusting unit is configured to: determine a timestamp of an image frame to be inserted in the reference driving images according to the frame rate requirement and a reference frame rate of the reference driving images; determine a target position of the target object in the image at the timestamp from a local motion model, wherein the local motion model is used to represent a motion trajectory of a moving object in the reference driving images; add the target object to the target position of a reference image frame to obtain a candidate image frame, wherein an image size of the reference image frame is the same as an image size of the reference driving images; perform background filling on the candidate image frame to obtain a target image frame; and insert the target image frame into a position corresponding to the timestamp in the reference driving images to obtain the first driving images. The optimization unit is configured to: identify the reference driving images using a target image recognition model to obtain a target image region output by the target image recognition model, wherein the target image region is a region in which the target object is located, and the target image recognition model is obtained by training an initial image recognition model using image samples in which image regions corresponding to reference objects are labeled; and perform region enlargement on the target image region in the reference driving images to obtain the second driving images. The optimization unit is configured to: segment the target image region from the reference driving images; and perform bicubic interpolation enlargement on the image region to obtain the second driving images.

6. A computer readable storage medium, characterized in that, The computer readable storage medium comprises a stored program, wherein the program performs the method of any one of claims 1 to 3 when executed. 7.An electronic device comprising a memory and a processor, the electronic device characterized by, The memory stores a computer program, and the processor is configured to execute the method of any one of claims 1 to 3 by using the computer program.

Images