Vehicle calibration method and device based on multi-source data and electronic equipment

Abstract

The invention relates to a vehicle calibration method and device based on multi-source data and electronic equipment. The method comprises the steps of obtaining a driving state of a vehicle; when the driving state is determined to be a static state, determining any road side end which has a common sensing range with the vehicle at present as a first target road side end; acquiring a first sensing comparison result of the vehicle and the first target roadside end; and when the first sensing comparison result does not reach the first preset threshold value, performing first parameter calibration on the sensing equipment of the vehicle according to a first correction strategy. According to the scheme provided by the invention, the dynamic calibration of the vehicle sensing equipment in a real road environment can be realized, and the sensing data deviation of the vehicle side equipment and the road side equipment is reduced, so that the decision accuracy of an automatic driving system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving vehicle calibration, and in particular to a vehicle calibration method and device based on multi-source data and an electronic device. BACKGROUND

[0002] With the development of automatic driving and intelligent transportation systems, roadside devices (such as cameras, lidar, etc.) and vehicle-side devices (such as vehicle-mounted sensors) together constitute a cooperative perception network to provide data support for vehicle automatic driving decision-making.

[0003] In related technologies, vehicle perception calibration usually relies on device-level calibration in a laboratory static environment. However, vehicles are easily affected by factors such as jolt impact, temperature deformation, or installation looseness during long-term driving, which causes the vehicle-side devices on the vehicle to easily have pose shifts. The above shifts cause structural deviations between the existing calibration parameters on the vehicle and the actual perception requirements, which in turn causes the recognition data of the same perceived object by the vehicle-mounted devices and the roadside devices to be inconsistent, directly causing the accumulation of perception errors, and ultimately affecting the decision-making accuracy of the automatic driving system. SUMMARY

[0004] To solve or partially solve the problems in related technologies, the present application provides a vehicle calibration method and device based on multi-source data and an electronic device, which can realize dynamic calibration of vehicle sensing devices in a real road environment, reduce the perception data deviation between vehicle-side devices and roadside devices, and thus improve the decision-making accuracy of the automatic driving system.

[0005] The first aspect of the present application provides a vehicle calibration method based on multi-source data, comprising: obtaining a driving state of a vehicle; when the driving state is determined to be a static state, determining any roadside end currently having a common sensing range with the vehicle as a first target roadside end; obtaining a first sensing comparison result of the vehicle and the first target roadside end; when the first sensing comparison result is determined not to reach a first preset threshold, performing first parameter calibration on the sensing device of the vehicle according to a first correction strategy.

[0006] In some embodiments, the obtaining of the first sensing comparison result of the vehicle and the first target roadside end comprises: obtaining first perception data of the vehicle and second perception data of the first target roadside end; converting the first perception data and the second perception data into the same preset coordinate system for feature matching of at least one same perceived object, to obtain first feature data and second feature data corresponding to at least one same perceived object; The first feature data is compared with the second feature data to obtain a first feature comparison result.

[0007] In some embodiments, when the first sensing comparison result determines that a first preset threshold is not reached, the sensing device of the vehicle is calibrated according to a first calibration strategy, including: When the coincidence degree of at least one feature of the first feature data and the second feature data in the first feature comparison result is less than a first preset threshold, a first calibration parameter is calculated based on a preset optimization algorithm according to the first feature data and the second feature data, and the internal parameter of the sensing device of the vehicle is corrected by the first calibration parameter.

[0008] In some embodiments, the method further includes: Obtaining a second sensing comparison result of the vehicle and the first target road side end; When the second sensing comparison result determines that a second preset threshold is not reached, the first calibration strategy is re-executed.

[0009] In some embodiments, the obtaining of the second sensing comparison result of the vehicle and the first target road side end includes: Obtaining third perception data of the vehicle after the parameter correction is completed and fourth perception data of the first target road side end; Identifying at least one same perception object from the third perception data and the fourth perception data, obtaining third feature data and fourth feature data corresponding to the at least one same perception object; Comparing the third feature data with the fourth feature data to obtain a second sensing comparison result; The re-execution of the first calibration strategy when the second sensing comparison result determines that the second preset threshold is not reached, including: When the coincidence degree of at least one feature of the third feature data and the fourth feature data in the second sensing comparison result is less than a second preset threshold, the first calibration strategy is re-executed.

[0010] In some embodiments, the method further includes: When the second sensing comparison result determines that a second preset threshold is reached, a second parameter calibration is performed on the sensing device of the vehicle according to a second calibration strategy.

[0011] In some embodiments, the second parameter calibration of the sensing device of the vehicle according to the second calibration strategy when the second sensing comparison result determines that the second preset threshold is reached, including: when the second feature coincidence degree in the second feature comparison result is greater than or equal to a second preset threshold, selecting a next roadside end having a common sensing range with the vehicle as a second target roadside end according to a preset selection rule; obtaining a third sensing comparison result of the vehicle and the second target roadside end; when the third sensing comparison result determines that a third preset threshold is not reached, performing second parameter calibration on the sensing device of the vehicle according to a second calibration strategy.

[0012] The second aspect of the present application provides a vehicle calibration device based on multi-source data, comprising: a state determination module configured to obtain a driving state of a vehicle; a target determination module configured to determine any roadside end currently having a common sensing range with the vehicle as a first target roadside end when the driving state is determined as a static state; a sensing data comparison module configured to obtain a first sensing comparison result of the vehicle and the first target roadside end; a first calibration strategy execution module configured to perform first parameter calibration on the sensing device of the vehicle according to a first calibration strategy when the first sensing comparison result determines that a first preset threshold is not reached.

[0013] The third aspect of the present application provides an electronic device, comprising: a processor; and a memory having executable code stored thereon, when the executable code is executed by the processor, the processor executes the method as described above.

[0014] The fourth aspect of the present application provides a computer-readable storage medium having executable code stored thereon, when the executable code is executed by the processor of the electronic device, the processor executes the method as described above.

[0015] The technical solution provided by the present application can include the following beneficial effects: The technical solution of the present application realizes the dynamic calibration process of the sensing device of the vehicle on the real road by comparing the sensing data in the common sensing range between the vehicle and the roadside end in the static state of the vehicle, dynamically triggering the parameter calibration of the sensing device of the vehicle according to the sensing data comparison result, effectively solving the pose offset problem of the sensing device caused by bumps, temperature deformation or installation looseness in long-term driving, eliminating the accumulation of perception errors caused by inconsistent road and vehicle data, and effectively improving the decision accuracy of the automatic driving system.

[0016] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0017] The above and other objects, features and advantages of the present application will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings in which like reference characters refer to like parts throughout the figures, and in which:

[0018] Figure 1 is a flow diagram of a vehicle calibration method based on multi-source data according to an embodiment of the present application; Figure 2 is another flow diagram of a vehicle calibration method based on multi-source data according to an embodiment of the present application; Figure 3 is another flow diagram of a vehicle calibration method based on multi-source data according to an embodiment of the present application; Figure 4 is a structural diagram of a vehicle calibration apparatus based on multi-source data according to an embodiment of the present application; Figure 5 is another structural diagram of a vehicle calibration apparatus based on multi-source data according to an embodiment of the present application; Figure 6 is a structural diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] Embodiments of the present application will be described more fully hereinafter with reference to the accompanying drawings, in which embodiments of the application are shown. This application may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the application to those skilled in the art.

[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0021] It should be understood that, although the terms "first", "second", "third", etc. can be used in this application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information without departing from the scope of the present application. Therefore, the features defined as "first", "second" can explicitly or implicitly include one or more features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0022] In related technologies, vehicle perception calibration usually relies on device-level calibration in a laboratory static environment; however, vehicles are easily affected by factors such as jolt impact, temperature deformation or installation loosening during long-term driving, which causes the pose of the vehicle-side device on the vehicle to easily deviate, and the above-mentioned change of external parameters easily causes structural deviation between the existing calibration parameters on the vehicle and the actual perception demand, thereby causing the recognition data of the same perception object by the vehicle measurement device and the roadside device to be inconsistent. The above inconsistency is easy to gradually accumulate perception errors after multi-level sensor data interaction and transmission, and finally affects the decision accuracy and safety of the automatic driving system.

[0023] To solve the above problems, the embodiment of the present application provides a vehicle calibration method based on multi-source data, which can improve the dynamic calibration efficiency of the vehicle sensing device in the real road environment, reduce the perception data deviation between the vehicle-side device and the roadside device, and thus improve the decision accuracy of the automatic driving system.

[0024] The technical solutions of the embodiments of the present application are described in detail below with reference to the drawings.

[0025] Figure 1 is a flowchart of the vehicle calibration method based on multi-source data shown in the embodiments of the present application.

[0026] Referring to Figure 1 , the vehicle calibration method based on multi-source data of the present application comprises: S110, acquiring the driving state of the vehicle.

[0027] In this step, the current driving state of the vehicle is acquired.

[0028] The driving state can refer to the motion state of the vehicle in motion or at rest. The driving state of the vehicle can be determined by collecting speed and acceleration data by using the vehicle-mounted sensor or by using the state identifier output by the vehicle-mounted system. The vehicles in the static state are selected to exclude sensor noise interference in dynamic driving.

[0029] S120, when the driving state is determined to be a static state, determining any roadside end currently having a common sensing range with the vehicle as a first target roadside end.

[0030] In this step, when the driving state of the vehicle is determined to be a static state, one of all roadside ends currently having a common sensing range with the vehicle is determined as a first target roadside end.

[0031] Among them, the common sensing range can refer to the perception area in which the sensing device of the vehicle and the roadside end have an overlap in the spatial coverage area. Among them, the sensing range of the roadside end can be greater than the sensing range of the vehicle. It should be understood that between the vehicle and the first target roadside end having a common sensing range, the same perception object can be sensed from the common sensing range.

[0032] Among them, the first target roadside end can be the roadside end closest to the vehicle and having a common sensing range, so as to further reduce the influence of the difference in environmental conditions caused by too large distance on the sensing accuracy between the two.

[0033] S130, obtaining a first sensing comparison result of the vehicle and the first target roadside end.

[0034] In this step, the first sensing comparison result obtained by comparing and analyzing the sensing data between the sensing device of the vehicle itself and the first target roadside end is obtained.

[0035] Among them, the first sensing comparison result can be used to represent the matching degree of the sensing data of the sensing device of the vehicle itself and the roadside end to the same perception object. Among them, the first sensing comparison result can be obtained by extracting the feature data of the same perception object and calculating the similarity.

[0036] It should be understood that through the first sensing comparison result, it can be judged whether there is a perception deviation caused by the external parameter offset of the sensing device of the vehicle.

[0037] S140, when the first sensing comparison result is determined to not reach a first preset threshold, performing first parameter calibration on the sensing device of the vehicle according to a first correction strategy.

[0038] In this step, when it is determined that the first sensing comparison result does not reach the first preset threshold, it is judged that there is a deviation in the sensing data between the sensing device of the vehicle and the first target roadside end at this time, and the first parameter calibration is performed on the sensing device of the vehicle according to the first correction strategy.

[0039] Among them, the first parameter calibration process can be to correct and calibrate the external parameters of the sensing device of the vehicle, so as to compensate for the pose offset of the sensing device of the vehicle.

[0040] The first correction strategy can be a parameter adjustment method based on the difference between the sensing device of the vehicle and the perception data of the roadside terminal. For example, an optimization algorithm can be used to calculate the calibration parameter difference or a preset compensation model to perform parameter iterative correction, so as to dynamically correct the external parameters of the sensing device of the vehicle to realize alignment with the perception data of the roadside terminal.

[0041] In this embodiment, the vehicle calibration method based on multi-source data provided by the application compares the sensing data in the co-sensing range between the vehicle and the roadside terminal in the stationary state of the vehicle, dynamically triggers the parameter calibration of the sensing device of the vehicle according to the comparison result of the perception data, realizes the dynamic calibration process of the sensing device of the vehicle on the real road, effectively solves the problem of pose offset of the sensing device caused by bumps, temperature deformation or installation looseness during long-term driving, eliminates the accumulation of perception errors caused by inconsistent road and vehicle data, and effectively improves the decision accuracy of the automatic driving system.

[0042] Figure 2 is another flowchart of the vehicle calibration method based on multi-source data provided by the embodiment of the application. Based on the embodiment shown in Figure 1 , the technical solutions of the application are further described in detail.

[0043] Referring to Figure 2 , the vehicle calibration method based on multi-source data provided by the application comprises: S210, obtaining the driving state of the vehicle.

[0044] S220, when the driving state is determined as a stationary state, determining any roadside terminal having a co-sensing range with the vehicle as a first target roadside terminal.

[0045] The steps S210 and S220 are similar to the steps S110 and S120, and the specific process is described above with reference to the steps S110 and S120, which will not be repeated here.

[0046] S230, obtaining the first perception data of the vehicle and the second perception data of the first target roadside terminal.

[0047] In this step, the first perception data corresponding to the sensing device on the vehicle and the second perception data of the second target roadside terminal are obtained.

[0048] It should be understood that the first perception data and the second perception data correspond to the same time. The sensing device for obtaining perception data on the vehicle and the sensing device of all roadside terminals are pre-calibrated with data time stamp, so as to ensure that the obtained first perception data and second perception data are consistent in time sequence.

[0049] The first perception data can be obtained by multiple types of sensing devices on the vehicle. The second perception data can be obtained by a single type of sensing device at the roadside end. That is, the acquisition dimensions of the first perception data and the second perception data can be the same or different.

[0050] For example, the vehicle collects point cloud data of the surrounding environment through the vehicle-mounted laser radar and image data of the surrounding environment through the vehicle-mounted camera as the first perception data. The first target roadside end collects image data of the surrounding environment of the vehicle through the roadside camera as the second perception data.

[0051] S240, converting the first perception data and the second perception data into the same preset coordinate system to perform feature matching of at least one same perception object, and obtaining first feature data and second feature data corresponding to the at least one same perception object.

[0052] In this step, the obtained first perception data and second perception data are converted into the same coordinate system, and feature matching of at least one same perception object is performed, so as to obtain first feature data and second feature data corresponding to two data sources and corresponding to at least one same perception object.

[0053] It should be understood that the first feature data and the second feature data are two data sets, and the number of same perception objects corresponding to the first feature data and the second feature data is one or more than one.

[0054] The feature parameters matched in the feature matching process can include but are not limited to position, size, shape.

[0055] The preset coordinate system can be a vehicle coordinate system. For example, the second perception data collected by the roadside end is converted to the origin of the vehicle coordinate system through a coordinate transformation matrix. The preset coordinate system can also be a pre-constructed global coordinate system. For example, the UTM coordinate system is used as the preset coordinate system, and when performing feature matching, the perception data of the vehicle and the roadside end are respectively converted to the UTM coordinate system through spatial mapping of geographic positioning information.

[0056] The feature matching process can use the coincidence degree of three-dimensional point cloud data to realize feature matching. For example, the first perception data and the second perception data converted to the same coordinate system are calculated through an iterative closest point algorithm, and the optimal matching relationship between the spatial point sets corresponding to the first perception data and the second perception data is used. When the point cloud coincidence degree reaches 85%, it is determined that it is the same perception object.

[0057] S250, performing feature comparison on the first feature data and the second feature data to obtain a first feature comparison result.

[0058] In this step, the obtained first feature data and the second feature data are compared for each parameter, and a first feature comparison result is obtained.

[0059] In the feature comparison process, the coincidence degree can be compared for the position coordinates, size, and shape information of the perceived object.

[0060] S260, when the at least one feature coincidence degree of the first feature data and the second feature data in the first feature comparison result is less than the first preset threshold, a first calibration parameter is calculated based on a preset optimization algorithm according to the first feature data and the second feature data, and the parameter of the sensing device of the vehicle is corrected by the first calibration parameter.

[0061] In this step, when the at least one feature coincidence degree of the first feature data and the second feature data in the first feature comparison result is less than the first preset threshold, it is determined that the sensing device of the vehicle has a pose offset, and a first calibration parameter is calculated based on a preset optimization algorithm according to the first feature data and the second feature data, and the parameter of the sensing device of the vehicle is corrected by the first calibration parameter.

[0062] The preset optimization algorithm can be at least one of least square method, genetic algorithm, or gradient descent method. Through the preset optimization algorithm, an iterative calculation process can be realized, and a global optimal solution is calculated according to the first feature data and the second feature data, and the optimal solution is used as the first calibration parameter, so as to ensure the calibration and correction effect of the sensing device of the vehicle.

[0063] The calculation process of the first calibration parameter can combine the feature data of multiple perceived objects, and different perceived objects are configured with different weights according to the distance from the vehicle. For example, the weight of the perceived object data within 10 meters from the vehicle is set to 0.6, and the weight decreases by 0.1 for every 5 meters of distance increase.

[0064] The parameter correction process can be realized by modifying the parameter configuration file. The parameter configuration file can include the self-positioning parameter of the vehicle, the radar parameter, and the camera parameter. For example, when the self-positioning error of the vehicle exceeds the threshold, the self-positioning parameter in the parameter configuration file is adjusted; when the shape error of the perceived object exceeds the threshold, the camera distortion parameter in the parameter configuration file is adjusted.

[0065] In this embodiment, the vehicle calibration method based on multi-source data of the application can effectively eliminate the spatial reference deviation caused by the differences in device installation position, sensor type and other factors after converting the perception data of the vehicle and the roadside end to the same preset coordinate system and then performing feature matching, so that the perception data of different sources can be matched in the same spatial dimension, effectively ensuring the accuracy of feature matching, and when the feature coincidence degree does not meet the standard, using a preset optimization algorithm to accurately calculate the calibration parameters for calibration, so that the optimal calibration parameters can be calculated for different perception scenes, effectively improving the accuracy of the calibration process.

[0066] Figure 3 is another flowchart of the vehicle calibration method based on multi-source data according to an embodiment of the application. Based on the embodiment shown in Figure 1 , the technical solutions of the application are further elaborated.

[0067] Referring to Figure 3 , the vehicle calibration method based on multi-source data of the application comprises: S310, obtaining the driving state of the vehicle.

[0068] S320, when the driving state is determined as a static state, determining any roadside end currently having a common sensing range with the vehicle as a first target roadside end.

[0069] S330, obtaining a first sensing comparison result of the vehicle and the first target roadside end.

[0070] S340, when the first sensing comparison result is determined not to meet a first preset threshold, performing first parameter calibration on the sensing device of the vehicle according to a first correction strategy.

[0071] Among them, steps S310 and S340 are similar to steps S110 and S140, and the specific process is referred to the related content of steps S110 and S120 above, which will not be repeated here.

[0072] S350, obtaining a second sensing comparison result of the vehicle and the first target roadside end.

[0073] In this step, the second sensing comparison result is obtained by comparing and analyzing the sensing data reacquired between the vehicle and the first target roadside end after the first parameter calibration correction.

[0074] Among them, the second sensing comparison result can be generated by feature comparison between the third perception data reacquired by the vehicle after completing the parameter correction and the fourth perception data reacquired by the first target roadside end.

[0075] Among them, the second sensing comparison result can be obtained by the following way: S351, acquire third perception data of the vehicle after parameter correction and fourth perception data of the first target road side.

[0076] S352, identify at least one same perception object from the third perception data and the fourth perception data, and acquire third feature data and fourth feature data corresponding to the at least one same perception object.

[0077] S353, perform feature comparison between the third feature data and the fourth feature data to obtain a second sensing comparison result.

[0078] The steps S351 to S353 are similar to the steps S230 to S250, and will not be described here.

[0079] It should be understood that in the method of the present application, after the initial parameter calibration (first parameter calibration) of the sensing device of the vehicle, there may be a situation that the calibrated parameters do not completely eliminate the device bias or the dynamic change of the environment causes the calibration to fail, resulting in inconsistency between the perception data of the vehicle and the road side. The present application obtains the second sensing comparison result between the third perception data of the vehicle after correction and the fourth perception data of the first target road side, thereby verifying the correction result.

[0080] Among them, the third perception data and the fourth perception data have continuity in the time dimension. For example, the third perception data and the fourth perception data are synchronously collected within 5-10 seconds after the first parameter calibration.

[0081] S360, when the second sensing comparison result determines that the second preset threshold is not reached, re-execute the first correction strategy.

[0082] In this step, when it is judged that the second feature comparison result does not reach the preset second preset threshold, it is judged that there is still deviation or calibration failure after the first parameter calibration of the vehicle, and the first correction strategy is re-executed. The calibration parameters are recalculated based on the latest perception data by the preset optimization algorithm and calibrated.

[0083] Among them, the steps S350 and S360 in the present application are iterative optimization processes that are continuously repeated until the feature coincidence degree reaches the preset standard, effectively eliminating the residual deviation, thereby ensuring the effectiveness and accuracy of the parameter calibration process.

[0084] Among them, the second preset threshold can be less than or equal to the first preset threshold, which is not limited here.

[0085] S370, when the second sensing comparison result determines that the second preset threshold is reached, perform second parameter calibration on the sensing device of the vehicle according to the second correction strategy.

[0086] In this step, according to the second sensing comparison result, when the second preset threshold is reached, it is judged that the vehicle and the current road side end complete the calibration, and the second parameter calibration is performed on the sensing device of the vehicle according to the second correction strategy and the next road side end.

[0087] It should be understood that the vehicle sensing device after completing the preliminary parameter correction (first parameter calibration) may still have unmet calibration needs, such as the vehicle sensing device after preliminary parameter correction still has sensing errors due to different environmental conditions. By introducing other road side ends in one step for further collaborative verification, the universality and stability of the calibration result under different road side end environments can be further ensured. That is, in the method of the present application, after the vehicle completes the preliminary parameter correction, it can enter the second parameter correction stage, and the second parameter calibration is performed on the sensing device of the vehicle according to the second correction strategy.

[0088] Among them, the second correction strategy can correspond to single or multiple second parameter calibration processes. That is, after completing the preliminary parameter correction, the sensing device of the vehicle can be calibrated through the next road side end or another preset number of road side ends for single or multiple second parameter calibration.

[0089] Among them, the second parameter correction stage can include the following steps: S371, when the third feature data and the fourth feature data in the second feature comparison result match at least one feature coincidence degree greater than or equal to the second preset threshold, the next road side end with a common sensing range with the vehicle is selected as the second target road side end according to the preset selection rule.

[0090] In this step, when the feature coincidence degree of the third feature data and the fourth feature data in the second feature comparison result is greater than or equal to the second preset threshold, it is judged that the sensing device of the vehicle and the perception data of the first target road side end have reached preliminary consistency according to the first correction strategy, and the next road side end associated with the vehicle driving route is selected as the second target road side end according to the preset selection rule.

[0091] Among them, the preset selection rule can include at least one of the following selection rules: selecting the next target road side end in different directions or angles of the vehicle, such as the road side end on the left or right side of the vehicle; selecting the road side end on different roads to cover the perception scene of the intersection or parallel road section; selecting the road side end in different environmental conditions, wherein the environmental conditions can include but are not limited to: weather conditions, lighting conditions, temperature conditions, such as areas with a difference in light intensity of more than 50% or a temperature change of more than 10 degrees Celsius; selecting a road side end with a distance from the previous target road side end exceeding a preset distance, such as selecting a road side end more than 20 meters away from the previous target road side end to verify the long-distance space consistency.

[0092] S372, acquire a third sensing comparison result of the vehicle and the second target roadside end.

[0093] In this step, after determining the second target roadside end, a third sensing comparison result between the vehicle and the second target roadside end at this time is acquired through sensing data comparison.

[0094] In the step 372, the third sensing comparison result is acquired, please refer to the part of acquiring the first sensing comparison result in the steps S230 to S250 in the foregoing description, which will not be repeated here.

[0095] S373, when the third sensing comparison result is determined not to reach the third preset threshold, performing second parameter calibration on the sensing device of the vehicle according to the second correction strategy.

[0096] In this step, when the third sensing comparison result is determined not to reach the third preset threshold, it is judged that there is deviation in the sensing data between the vehicle and the second target roadside end at this time, and the second parameter calibration is performed on the sensing device of the vehicle according to the second correction strategy.

[0097] In the step 373, the third preset threshold is the same as or different from the first preset threshold.

[0098] In the step S373, the second correction strategy and the second parameter calibration are the same as those in the step S260 in the foregoing description, which will not be repeated here. It should be understood that in the second correction strategy, the calibration parameter is calculated based on the sensing data of the vehicle at this time and the sensing data of the second target roadside end, and then the parameter calibration is performed on the sensing device of the vehicle based on the calculated calibration parameter.

[0099] In this embodiment, the vehicle calibration method based on multi-source data provided by the application can verify the actual effect of the calibration parameter by acquiring the new sensing data of the corrected vehicle and the same roadside end after completing the preliminary parameter calibration, so as to construct a closed-loop verification mechanism, thereby ensuring the effectiveness of the parameter correction of the vehicle sensing device. The method provided by the application also sets the second correction strategy, compares the sensing data with different roadside ends, supplements the potential deviation not covered by the first correction strategy, and constitutes a cooperative verification mechanism of multi-view roadside equipment, thereby effectively overcoming the limitation of a single verification source in the spatial coverage range, and effectively improving the parameter calibration robustness of the vehicle sensing device in complex road scenes.

[0100] Corresponding to the foregoing application function implementation method embodiment, the application also provides a vehicle calibration device based on multi-source data, an electronic device and corresponding embodiments.

[0101] Figure 4 is a structural schematic diagram of the vehicle calibration device based on multi-source data provided by the embodiment of the application.

[0102] Referring to Figure 4 The vehicle calibration device based on multi-source data 400 of the present application comprises a state determination module 410, a target determination module 420, a sensing data comparison module 430, and a first calibration strategy execution module 440.

[0103] The state determination module 410 is configured to obtain the driving state of the vehicle.

[0104] The target determination module 420 is configured to determine the first target roadside end when the driving state is determined as the static state.

[0105] The sensing data comparison module 430 is configured to obtain the first sensing comparison result of the vehicle and the first target roadside end.

[0106] In some embodiments, the sensing data comparison module 430 can obtain the first perception data of the vehicle and the second perception data of the first target roadside end; convert the first perception data and the second perception data into the same preset coordinate system for feature matching of at least one same perception object, obtain the first feature data and the second feature data corresponding to the at least one same perception object; and perform feature comparison on the first feature data and the second feature data to obtain the first feature comparison result.

[0107] The first calibration strategy execution module 440 is configured to perform first parameter calibration on the sensing device of the vehicle according to the first correction strategy when the first sensing comparison result does not reach the first preset threshold.

[0108] In some embodiments, when the coincidence degree of at least one feature of the first feature data and the second feature data in the first feature comparison result is less than the first preset threshold, the first calibration strategy execution module 440 can calculate the first calibration parameter based on the preset optimization algorithm according to the first feature data and the second feature data, and perform parameter correction on the intrinsic parameter of the sensing device of the vehicle through the first calibration parameter.

[0109] Figure 5 FIG. 4 is another structural schematic diagram of the vehicle calibration device based on multi-source data according to an embodiment of the present application.

[0110] Referring to Figure 5 The vehicle calibration device based on multi-source data 400 of the present application comprises a state determination module 410, a target determination module 420, a sensing data comparison module 430, a first calibration strategy execution module 440, an effect verification module 450, and a second calibration strategy execution module 460.

[0111] The effect verification module 450 is configured to obtain a second sensing comparison result of the vehicle and the first target roadside end; and control the first calibration strategy execution module 440 to re-execute the first correction strategy when the second sensing comparison result determines that a second preset threshold is not reached.

[0112] In some embodiments, the effect verification module 450 can obtain third perception data of the vehicle after the parameter correction is completed and fourth perception data of the first target roadside end; identify at least one same perception object from the third perception data and the fourth perception data, obtain third feature data and fourth feature data corresponding to the at least one same perception object; perform feature comparison on the third feature data and the fourth feature data to obtain a second sensing comparison result; and control the first calibration strategy execution module 440 to re-execute the first correction strategy when at least one feature coincidence degree of the third feature data and the fourth feature data in the second sensing comparison result is less than a second preset threshold.

[0113] The second calibration strategy execution module 460 is configured to perform second parameter calibration on the sensing device of the vehicle according to a second correction strategy when the second sensing comparison result determines that a second preset threshold is reached.

[0114] In some embodiments, the second calibration strategy execution module 460 can select a next roadside end having a common sensing range with the vehicle as a second target roadside end according to a preset selection rule when at least one feature coincidence degree of the third feature data and the fourth feature data in the second feature comparison result is greater than or equal to the second preset threshold; obtain a third sensing comparison result of the vehicle and the second target roadside end; and perform second parameter calibration on the sensing device of the vehicle according to the second correction strategy when the third sensing comparison result determines that a third preset threshold is not reached.

[0115] In this embodiment, the vehicle calibration device based on multi-source data provided by the present application can realize the dynamic calibration process of the sensing device of the vehicle on the real road by performing sensing data comparison of the common sensing range between the vehicle and the roadside end in the stationary state of the vehicle, dynamically triggering parameter calibration of the sensing device of the vehicle according to the perception data comparison result, effectively solving the problem of pose offset of the sensing device caused by bumps, temperature deformation or installation looseness in long-term driving, eliminating the accumulation of perception errors caused by inconsistent road and vehicle data, and effectively improving the decision accuracy of the automatic driving system.

[0116] As to the device in the above-mentioned embodiments, the specific manners in which various modules perform operations have been described in detail in the embodiments of the method, and will not be described in detail here.

[0117] Figure 6 FIG. 1 is a structural schematic diagram of an electronic device according to an embodiment of the present application.

[0118] Referring toFigure 6 The electronic device 1000 includes a memory 1010 and a processor 1020.

[0119] The processor 1020 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gates or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can be any conventional processor.

[0120] The memory 1010 can include various types of storage units, such as a system memory, a read-only memory (ROM), and a permanent storage device. Among them, the ROM can store static data or instructions required by the processor 1020 or other modules of the computer. The permanent storage device can be a read and write storage device. The permanent storage device can be a non-volatile storage device that does not lose stored instructions and data even after the computer is powered off. In some embodiments, the permanent storage device uses a mass storage device (such as a magnetic or optical disk, flash memory) as a permanent storage device. In some other embodiments, the permanent storage device can be a removable storage device (such as a floppy disk, an optical drive). The system memory can be a read and write storage device or a volatile read and write storage device, such as a dynamic random access memory. The system memory can store some or all of the instructions and data required by the processor during runtime. In addition, the memory 1010 can include a combination of any computer readable storage media, including various types of semiconductor storage chips (such as DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), magnetic disks and / or optical disks. In some embodiments, the memory 1010 can include a read and / or write removable storage device, such as a compact disc (CD), a read-only digital versatile disc (such as DVD-ROM, double-layer DVD-ROM), a read-only Blu-ray disc, an ultra-density optical disc, a flash memory card (such as an SD card, a min SD card, a Micro-SD card, etc.), a magnetic floppy disk, etc. The computer readable storage medium does not include a carrier wave and a transient electronic signal transmitted through a wireless or wired transmission.

[0121] The memory 1010 stores executable code, which, when processed by the processor 1020, can cause the processor 1020 to perform part or all of the above-mentioned methods.

[0122] Furthermore, the method according to the present application can also be implemented as a computer program or a computer program product, which comprises computer program code instructions for executing some or all of the steps of the above-mentioned method according to the present application.

[0123] Alternatively, the present application can also be implemented as a computer readable storage medium (or a non-transitory machine readable storage medium or a machine readable storage medium) having stored thereon executable codes (or computer programs or computer instruction codes) which, when executed by a processor of an electronic device (or a server, etc.), cause the processor to perform some or all of the steps of the above-mentioned method according to the present application.

[0124] The above has described the embodiments of the present application, the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes are obvious to those skilled in the art without departing from the scope and spirit of the described embodiments. The selection of the terms used herein is intended to best explain the principles of the embodiments, practical application, or improvement to the technology in the market, or to enable other ordinary skilled in the art to understand the embodiments disclosed herein.

Claims

1. A vehicle calibration method based on multi-source data, characterized in that, include: Obtain the vehicle's driving status; When the driving state is determined to be a stationary state, any roadside end that currently has a common sensing range with the vehicle is determined as the first target roadside end; Obtain the first sensing comparison result between the vehicle and the first target roadside end; When the first sensing comparison result determines that the first preset threshold has not been reached, the first parameter calibration of the vehicle's sensing device is performed according to the first correction strategy.

2. The method according to claim 1, characterized in that, The step of obtaining the first sensing comparison result between the vehicle and the first target roadside end includes: Acquire the first perception data of the vehicle and the second perception data of the first target roadside end; The first and second sensing data are converted to the same preset coordinate system to perform feature matching of at least one identical sensing object, and the first and second feature data corresponding to at least one of the identical sensing objects are obtained. The first feature data and the second feature data are compared to obtain the first feature comparison result.

3. The method according to claim 2, characterized in that, When the first sensing comparison result determines that the first preset threshold has not been reached, the first parameter calibration of the vehicle's sensing device is performed according to the first correction strategy, including: When the overlap of at least one feature matching the first feature data and the second feature data in the first feature comparison result is less than a first preset threshold, a first calibration parameter is calculated based on the first feature data and the second feature data according to a preset optimization algorithm, and the intrinsic parameter execution parameter of the vehicle's sensing device is corrected through the first calibration parameter.

4. The method according to any one of claims 1 to 3, characterized in that, The method also includes: Obtain the second sensing comparison result between the vehicle and the first target roadside end; When the second sensing comparison result determines that the second preset threshold has not been reached, the first correction strategy is re-executed.

5. The method according to claim 4, characterized in that, The step of obtaining the second sensing comparison result between the vehicle and the first target roadside end includes: Acquire the third perception data of the vehicle after parameter correction and the fourth perception data of the first target roadside end; At least one identical sensory object is identified from the third sensory data and the fourth sensory data, and third feature data and fourth feature data corresponding to at least one of the identical sensory objects are obtained; The third feature data is compared with the fourth feature data to obtain the second sensing comparison result; When the second sensing comparison result determines that the second preset threshold has not been reached, the first correction strategy is re-executed, including: When the overlap of at least one feature between the third feature data and the fourth feature data in the second sensing comparison result is less than a second preset threshold, the first correction strategy is re-executed.

6. The method according to claim 5, characterized in that, The method also includes: When the second sensing comparison result is determined to reach the second preset threshold, the second parameter calibration of the vehicle's sensing device is performed according to the second correction strategy.

7. The method according to claim 6, characterized in that, When the second sensing comparison result is determined to reach the second preset threshold, the second parameter calibration of the vehicle's sensing device is performed according to the second correction strategy, including: When the overlap of at least one feature in the second feature comparison result where the third feature data and the fourth feature data match is greater than or equal to the second preset threshold, the next roadside end with the same sensing range as the vehicle is selected as the second target roadside end according to the preset selection rule. Obtain the third sensing comparison result between the vehicle and the second target roadside end; When the third sensing comparison result determines that the third preset threshold has not been reached, the second parameter calibration of the vehicle's sensing device is performed according to the second correction strategy.

8. A vehicle calibration device based on multi-source data, characterized in that, include: The status determination module is used to obtain the vehicle's driving status; The target determination module is used to determine any roadside end that currently has a common sensing range with the vehicle as the first target roadside end when the driving state is determined to be a stationary state. The sensing data comparison module is used to obtain the first sensing comparison result between the vehicle and the first target roadside end; The first calibration strategy execution module is used to perform first parameter calibration on the vehicle's sensing device according to the first correction strategy when the first sensing comparison result determines that the first preset threshold has not been reached.

9. An electronic device, characterized in that, include: processor; as well as A memory having executable code stored thereon, which, when executed by the processor, causes the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium having executable code stored thereon, characterized in that: When the executable code is executed by the processor of the electronic device, the processor performs the method as described in any one of claims 1-7.

Images