An extrinsic parameter determination method, device, medium and equipment based on semi-automatic calibration

Abstract

The application discloses a kind of based on semi-automatic calibration external parameter determination method, device medium and equipment, belong to high-precision map technical field.The method mainly includes: using the high-precision inertial navigation of vehicle, calibration sensor, respectively collect self-vehicle driving trajectory, calibration sensor sensing information;Calibration sensor sensing information is used to determine the position of reference target in calibration sensor coordinate system;With the position of reference target in world coordinate system and the position of reference target in calibration sensor coordinate system, the intermediate value of calibration sensor's external parameter is determined;The external parameter of calibration sensor determined is assigned as the initial value of calibration sensor's external parameter, the above process is iterated no less than 3 times, and the final value of calibration sensor's external parameter is determined.A semi-automatic external parameter calibration method is provided, which does not depend on a specific calibration site and can complete the calibration task in a natural scene, while ensuring calibration accuracy, expanding the application range of the calibration method.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of high-precision maps, and in particular to a semi-automatic calibration-based extrinsic parameter determination method and device, medium and equipment. BACKGROUND

[0002] In the prior art, the calibration method selects the output data of the sensor, wherein the INSD outputs the pose information of the vehicle body of the autonomous vehicle, the radar outputs the 3D point cloud, and the camera outputs the RGB pixel; and is divided into two calibration methods according to whether a calibration target is needed, namely, targetless (no calibration target) and target (calibration target needed). For the targetless calibration method, the method can be performed in a natural scene, the constraint adjustment is less, and the application range is wide, but the calibration accuracy is limited when the method is used for calibration. For the calibration method requiring a calibration target, the calibration accuracy is high when the coordinates of the calibration target in the world coordinate system are known, and the calibration target and the site have high requirements. Hand-eye calibration is a commonly used extrinsic parameter calibration method, which needs to obtain the extrinsic parameter through the data collected by the radar and / or camera sensor and the pose data of the target (INSD) sensor, but requires that the radar and / or camera sensor and the target (INSD) sensor have good trajectory data, and the trajectory accuracy is high, that is, the method has strong constraints and high constraints on the site. The trajectory of the radar and the camera needs to be matched with the odometer or the map and the calibration target, and the trajectory accuracy depends on the odometer or the map matching and the calibration site. SUMMARY

[0003] In view of the problem of high constraints on the site in the prior art, the present application mainly provides a semi-automatic calibration-based extrinsic parameter determination method, device, medium and equipment.

[0004] To achieve the above-mentioned purpose, one technical scheme adopted by the present application is to provide a semi-automatic calibration-based extrinsic parameter determination method, which comprises: during driving of an autonomous vehicle on a predetermined road, using a high-precision inertial navigation system and a calibration sensor carried by the vehicle to collect a self-vehicle driving trajectory and calibration sensor perception information, respectively, wherein a plurality of columnar objects with a predetermined longitudinal spacing are arranged on both sides of the predetermined road along the road direction, and a plurality of reference targets with a predetermined height interval are arranged on the columnar objects; determining the positions of the reference targets in a calibration sensor coordinate system by using the calibration sensor perception information; determining an intermediate value of the extrinsic parameter of the calibration sensor by using the self-vehicle driving trajectory and the positions of the reference targets in the world coordinate system and the positions of the reference targets in the calibration sensor coordinate system; and assigning the determined extrinsic parameter of the calibration sensor as an initial value of the extrinsic parameter of the calibration sensor, and iterating the above process not less than 3 times to determine a final value of the extrinsic parameter of the calibration sensor.

[0005] Another technical scheme adopted by the present application is to provide a semi-automatic calibration-based external parameter determination device, which comprises: a module for collecting self-vehicle driving track and calibration sensor sensing information by using high-precision inertial navigation and calibration sensors carried by an autonomous vehicle during driving on a predetermined road, wherein a plurality of columnar objects with a predetermined longitudinal interval are arranged on both sides of the predetermined road along the road direction, and a plurality of reference targets with a predetermined height interval are arranged on the columnar objects; a module for determining the position of the reference target in the calibration sensor coordinate system by using the calibration sensor sensing information; a module for determining the intermediate value of the external parameter of the calibration sensor by using the self-vehicle driving track, the position of the reference target in the world coordinate system, and the position of the reference target in the calibration sensor coordinate system; and a module for assigning the determined external parameter of the calibration sensor as the initial value of the external parameter of the calibration sensor, and iterating the above process no less than 3 times to determine the final value of the external parameter of the calibration sensor.

[0006] Another technical scheme adopted by the present application is to provide a computer readable storage medium storing computer instructions, which are operated to execute the semi-automatic calibration-based external parameter determination method in scheme one.

[0007] Another technical scheme adopted by the present application is to provide a computer device, which comprises: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores computer instructions executable by the at least one processor, and the at least one processor operates the computer instructions to execute the semi-automatic calibration-based external parameter determination method in scheme one.

[0008] The beneficial effects that can be achieved by the technical scheme of the present application are: the present application designs a semi-automatic calibration-based external parameter determination method, device, medium and equipment. By providing a semi-automatic external parameter calibration method, the calibration task can be completed in a natural scene without relying on a specific calibration site, and the application range of the calibration method is expanded while ensuring the calibration accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical schemes in the embodiments of the present application or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0010] Figure 1 is a schematic diagram of an optional embodiment of a semi-automatic calibration-based external parameter determination method of the present application;

[0011] Figure 2is a schematic diagram of an optional embodiment of a device for determining an external parameter based on semi-automatic calibration.

[0012] The specific embodiments of the present application have been shown through the above-described drawings, and will be described in more detail hereinafter. These drawings and written descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0013] The preferred embodiments of the present application will be described in detail hereinafter with reference to the accompanying drawings, so that the advantages and features of the present application can be more easily understood by those skilled in the art, and the scope of protection of the present application can be more clearly defined.

[0014] It should be noted that, in this document, relational terms such as first and second, and the like, are used solely to distinguish one entity or action from another entity or action, without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises... a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the listed element.

[0015] A plurality of sensors, and / or a plurality of types of sensors, are usually installed on an autonomous vehicle, and the coordinate relationship between the sensors needs to be determined before multi-sensor fusion. Sensor calibration is divided into internal parameter calibration and external parameter calibration. The internal parameter determines the mapping relationship inside the sensor, such as the focal length, eccentricity, and pixel aspect ratio (+ distortion coefficient) of a camera, while the external parameter determines the conversion relationship between the sensor and an external coordinate system, such as the attitude parameter (rotation and translation degrees of freedom). The external parameter calibration of the sensor is to solve the transformation relationship between the coordinate systems by correlating the measurement data under different sensors.

[0016] In the prior art, the output data of the selected sensor in the calibration method is selected, wherein the INSD outputs the pose information of the vehicle body of the autonomous vehicle, the radar outputs the 3D point cloud, and the camera outputs the pixels of the RGB; and according to whether a calibration target is needed, the calibration method is divided into two kinds of calibration methods, namely targetless (without calibration target) and target (with calibration target). For the calibration method without calibration target, the method can be performed in a natural scene, the constraint adjustment is less, and the application range is wide, but when the calibration method is used, the accuracy of the calibration is limited. For the calibration method with calibration target, in the case that the coordinates of the calibration target in the world coordinate system are known, the calibration accuracy is high, and the requirements for the calibration target and the site are high. Hand-eye calibration is a commonly used external parameter calibration method, which needs to collect the data of the radar and / or camera sensors and the pose data of the target (INSD) sensor, and calculate the external parameters by solving the equation AX=XB, but the radar and / or camera sensors and the target (INSD) sensor need to have good trajectory data, and the trajectory accuracy is required to be high, that is, the method has strong constraints, and the site constraints are high. The trajectory of the radar and the camera needs to be matched with the odometer or the map and the calibration target, and the trajectory accuracy depends on the odometer or the map matching and the calibration site.

[0017] The application concept of the present application is: by providing a semi-automatic external parameter calibration method, device, medium and equipment, that is, during the driving of the autonomous vehicle on the predetermined road, the high-precision inertial navigation system and the calibration sensor on the vehicle are used to collect the self-vehicle driving trajectory and the calibration sensor perception information respectively, wherein a plurality of columnar objects with a longitudinal interval not less than a predetermined longitudinal interval are arranged on both sides of the predetermined road along the road direction, and a plurality of reference targets with an upper and lower interval not less than a predetermined height interval are arranged on the columnar objects respectively; the position of the reference target in the calibration sensor coordinate system is determined by using the calibration sensor perception information; the intermediate value of the external parameter of the calibration sensor is determined by using the self-vehicle driving trajectory and the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system; and the determined external parameter of the calibration sensor is assigned as the initial value of the external parameter of the calibration sensor, and the above process is iterated not less than 3 times to determine the final value of the external parameter of the calibration sensor. The present application does not depend on a specific calibration site, and can complete the calibration task in a natural scene, thereby expanding the application range of the calibration method while ensuring the calibration accuracy.

[0018] The technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be described in detail in the specific embodiments below. The specific embodiments below can be combined with each other, and the same or similar concepts or processes can not be described in detail in some embodiments. The embodiments of the present application will be described below with reference to the drawings.

[0019] Figure 1An optional embodiment of the external parameter determination method based on semi-automatic calibration is shown.

[0020] In Figure 1 In the embodiment shown, the external parameter determination method based on semi-automatic calibration mainly includes step S101, during the driving of the autonomous vehicle on the predetermined road, the vehicle-mounted high-precision inertial navigation and calibration sensor are used to collect the self-vehicle driving trajectory and the calibration sensor perception information, respectively, wherein a plurality of reference targets with an upper and lower interval not less than a predetermined height interval are arranged on a plurality of columnar objects with an interval not less than a predetermined longitudinal interval along the road direction on both sides of the predetermined road.

[0021] In the embodiment, when the calibration sensor is a radar, taking a laser radar as an example, the laser radar collects 3D point cloud data of the predetermined road when the autonomous vehicle drives through the predetermined road; when the calibration sensor is a camera, the camera takes 2D images of the predetermined road when the autonomous vehicle drives through the predetermined road. The calibration sensor perception information includes but is not limited to the data of the reference target corresponding to the columnar object, the ground markings on the road, and the data of the objects such as the road along the road.

[0022] In the embodiment, the columnar object includes but is not limited to trees and poles on both sides of the predetermined road.

[0023] Optionally, the laser reflection sheet and / or point light source are used as the reference target, and the laser reflection sheet and / or point light source are arranged on the columnar object of the predetermined road. Using the laser reflection sheet and / or point light source as the reference target can improve the reflectivity of the reference target and improve the calibration accuracy, wherein the point light source includes but is not limited to a light bulb, and the shape of the reflective strip includes but is not limited to a circle and a square.

[0024] In Figure 1 In the optional embodiment shown, the external parameter determination method based on semi-automatic calibration further includes step S102, determining the position of the reference target in the calibration sensor coordinate system by using the calibration sensor perception information.

[0025] In the embodiment, the field personnel measure and plot the position of the reference target in the calibration sensor coordinate system by using the calibration sensor perception information, obtain the latitude, longitude, and height coordinate data of the position, and use the data as a known condition for subsequent calculation and acquisition of the external parameter.

[0026] In an optional embodiment of the present application, when the calibration sensor is a laser radar, the reference target is a laser reflector, the calibration sensor sensing information is laser radar information, and the calibration sensor coordinate system is a laser radar coordinate system. In an optional example of the present application, when calibrating the radar, a road section with good GPS signal, few vehicles, and a length and width of not less than 20 meters is selected as the predetermined road of the present application; a plurality of columnar reflectors are decorated on both sides of the predetermined road along the road direction at a predetermined longitudinal interval, the laser reflector is used as the reference target, the predetermined longitudinal interval is set to 1 meter, and the distance between the reference target farthest from the vertical distance of the road center and the road center is not less than 10 meters; the data collection vehicle loaded with the radar is started, and after the starting time is recorded, the data collection vehicle starts to circle back and forth on the predetermined road at a slow speed to ensure that the data collection is complete. The vehicle turns around when it is 10-20 meters away, circles back and forth for 3-5 times, returns to the starting position, records the current time as the ending time, and then lets the data collection vehicle stand for a period of time, closes the collection program, and completes the data collection. In the subsequent steps, only the data collected from the starting time to the ending time is operated to reduce the amount of data, so that the amount of calculation of the system in the subsequent steps is reduced and the calibration speed is accelerated. Before the starting time is recorded, the state of the equipment in the data collection vehicle is first tested, that is, the data collection vehicle is parked on the predetermined road for a period of time, and then circles on the predetermined road. The circle time is consistent with the parking time, and whether the equipment collects data is checked to complete the test.

[0027] Preferably, in the optional example described above, the predetermined longitudinal interval can be set to 3 meters.

[0028] In an optional embodiment of the present application, when the calibration sensor is a laser radar, the position of the reference target in the calibration sensor coordinate system is determined by using the calibration sensor sensing information, which includes: selecting the point cloud corresponding to the laser reflector in the laser radar coordinate system according to the laser radar information; selecting at least one point cloud point with a laser echo intensity greater than a predetermined laser echo intensity threshold from the point cloud; and determining the point cloud point with the smallest distance between the center point of the laser reflector in the world coordinate system as the laser reflector point cloud point.

[0029] In the present embodiment, when calibrating the laser radar, the data collected from the starting time to the ending time, i.e., the point cloud data, is selected, and the intensity threshold can be 200. The point cloud point with a laser echo intensity greater than 200 is selected from the point cloud, and according to the point cloud point with a laser echo intensity greater than 200, the point cloud point with the smallest distance between the center point of the laser reflector is selected as the laser reflector point cloud point. Before the distance comparison is performed, the point cloud point with a laser echo intensity greater than 200 needs to be converted to the world coordinate system, and then subsequent operations are performed.

[0030] In an optional example of the present application, the point cloud points with laser echo intensity greater than 200 are respectively converted to the world coordinate system by formula (1) to obtain their coordinate data in the world coordinate system, and formula (1) is:

[0031] P w =T wi *(T i1 *P l ) (1)

[0032] Wherein, P l is the coordinate data of the point cloud points with laser echo intensity greater than 200 in the radar point cloud coordinate system; T i1 is the extrinsic matrix between the IMU coordinate system and the radar point cloud coordinate system; T wi is the pose of the IMU coordinate system in the world coordinate system at the current time, that is, the trajectory data output by the INSD; P w is the coordinate data of the point cloud points with laser echo intensity greater than 200 in the world coordinate system.

[0033] In an optional embodiment of the present application, when the calibration sensor is a camera, the reference target is a point light source, and the calibration sensor sensing information is a 2D picture, and the calibration sensor coordinate system is the camera coordinate system.

[0034] In an optional example of the present application, when calibrating the camera, a road section with good GPS signal, less vehicles and can drive back and forth is selected as the predetermined road of the present application, wherein the length and width of the predetermined road are not less than 20 meters; point light sources are installed on multiple columns, roadsides and wall surfaces on both sides of the predetermined road along the road direction at a predetermined longitudinal interval, the predetermined longitudinal interval is set to 3-5 meters, the predetermined height interval is set to 4 meters, and the distance between the reference target farthest from the vertical distance of the road center and the road center is not less than 10 meters; start the data collection vehicle loaded with the camera, record the starting time, then start to drive back and forth on the predetermined road, and drive slowly to ensure complete data collection; turn around when the vehicle drives out for 10-20 meters, drive back and forth for 3-5 circles, return to the starting position, record the current time as the ending time, and then let the data collection vehicle stand for a period of time, close the collection program, and complete the data collection. In the subsequent steps, only the data collected from the starting time to the ending time is operated to reduce the data volume, so that the system operation amount in the subsequent steps can be reduced and the calibration speed can be accelerated. Before recording the starting time, first test the state of the equipment in the data collection vehicle, that is, let the data collection vehicle stand on the predetermined road for a period of time, then drive around the predetermined road, the driving time is consistent with the standing time, and check whether the equipment collects data to complete the test.

[0035] In an optional embodiment of the present application, when the calibration sensor is a camera, the position of the reference target in the calibration sensor coordinate system is determined by using the calibration sensor sensing information, including: projecting the point light source from the world coordinate system into the 2D image of the calibration sensor coordinate system to obtain the corresponding point light source projection point; and searching for the pixel point with the maximum pixel intensity near the point light source projection point in the 2D image, and determining the pixel point as the candidate point.

[0036] In the embodiment, when calibrating the camera, the data collected from the start time to the end time, i.e., the point light source, is selected. The point light source is projected into the 2D image of the calibration sensor coordinate system to obtain the point light source projection point corresponding to the electric light source; the pixel point corresponding to the point light source projection point is taken as the center to search for the pixel point with the maximum intensity near the center, and the pixel point is taken as the candidate point.

[0037] In an optional example of the present application, the point light source is projected into the camera coordinate system by formula (2) to obtain the coordinate data of the point light source projection point in the camera coordinate system, and formula (2) is:

[0038] P c =T cw *P w (2)

[0039] wherein P w is the coordinate data of the point light source in the world coordinate system, P c is the coordinate data of the point light source projection point in the camera coordinate system, and T cw is the pose of the camera coordinate system in the world coordinate system at the current time.

[0040] Different camera types have different distortion coefficients, and the distortion coefficient of the camera may also be used when obtaining the coordinate data of the point light source projection point in the camera coordinate system. Commonly used cameras are pinhole cameras and fisheye cameras; wherein the distortion model of the pinhole camera is:

[0041] r 2 =x 2 +y 2

[0042]

[0043]

[0044] u=f x x'+c x

[0045] v=f y y'+c y

[0046] wherein r is the distance from the point light source projection point to the image center; x, y are the coordinate data of the point light source projection point in the pinhole camera coordinate system; x', y' are the actual coordinate data of the point light source projection point; f x is the coordinate in the x direction, f y is the coordinate in the y direction; (c x , c y ) is the principal point of the pinhole camera, that is, the center of the image; k1, k2, k3, k4, k5, k6, p1, p2 are all distortion coefficients of the pinhole camera.

[0047] The distortion model of the fisheye camera is:

[0048] a = x / z and b = y / z

[0049] r 2 = a 2 +b 2

[0050] θ = atan(r)

[0051] θ d = θ (1 + k1θ 2 + k2θ 4 + k3θ 6 + k4θ 8 )

[0052] x' = (θ d / r) a

[0053] y' = (θ d / r) b

[0054] wherein r is the distance from the point light source projection point to the image center; x, y, z are the coordinate data of the point light source projection point in the camera coordinate system; a, b are the coordinate data of the point light source projection point in the fisheye camera imaging coordinate system; x', y' are the actual coordinate data of the point light source projection point; θ is the incident angle of the fisheye camera, θ d is the exit angle of the fisheye camera; k1, k2, k3, k4 are all distortion coefficients of the fisheye camera.

[0055] In an optional embodiment of the present application, the position of the reference target in the calibration sensor coordinate system is determined by using the calibration sensor sensing information, and the method further comprises: in the 2D image, selecting a plurality of pixel points around the candidate point whose pixel intensity is greater than a pixel intensity threshold, and determining the average value of the coordinates of the plurality of pixel points in the camera coordinate system as the position of the point light source in the camera coordinate system.

[0056] In an optional example of the present application, when calibrating the camera, a first pixel point in a first preset range is obtained with the pixel point corresponding to the projected point of the post-projection point light source as the center, wherein the first preset range is a circle with a radius of 1.5 meters; in the first pixel point, a plurality of pixel points with pixel intensity greater than a pixel intensity threshold are selected; and then the average coordinates of the plurality of pixel points are calculated, and the average coordinates are taken as the position of the point light source in the camera coordinate system.

[0057] In Figure 1 In the optional embodiment shown, the method for determining the extrinsic parameter of the calibration sensor based on the semi-automatic calibration further includes a step S103 of determining an intermediate value of the extrinsic parameter of the calibration sensor by using the driving trajectory of the ego vehicle and the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system.

[0058] In an optional embodiment of the present application, the intermediate value of the extrinsic parameter of the calibration sensor is determined by using the driving trajectory of the ego vehicle and the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system, including: determining the laser radar extrinsic parameter that minimizes the distance between the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system as the intermediate value of the extrinsic parameter of the laser radar.

[0059] In an optional example of the present application, the distance between the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system is minimized by using formula (3) to obtain the intermediate value of the extrinsic parameter of the laser radar, and formula (3) is:

[0060]

[0061] wherein P l is the coordinate data of the reference target in the calibration sensor coordinate system, T il is the extrinsic parameter matrix between the imu coordinate system and the calibration sensor coordinate system, T wi is the pose of the imu coordinate system in the world coordinate system at the current time, Q w is the coordinate data of the reference target in the world coordinate system.

[0062] In an optional embodiment of the present application, the intermediate value of the extrinsic parameter of the calibration sensor is determined by using the driving trajectory of the ego vehicle and the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system, including: determining the laser radar extrinsic parameter that minimizes the distance between the laser reflection sheet point cloud point and the laser reflection sheet center point as the intermediate value of the extrinsic parameter of the laser radar.

[0063] In an optional embodiment of the present application, after the coordinates (u, v) of the point light source projection point in the camera coordinate system are calculated, the intrinsic parameters of the camera and the extrinsic parameters of the camera and the INSD can be optimized by minimizing the re-projection error. The projection error between the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system is minimized by using formula (4) to obtain the extrinsic parameters, formula (4) is:

[0064]

[0065] Wherein, u, v are the position of the reference target in the calibration sensor coordinate system, u', v' are the position of the reference target in the world coordinate system.

[0066] In Figure 1 In the optional embodiment shown in the figure, based on the semi-automatic calibration extrinsic parameter determination method, the determined calibration sensor extrinsic parameter is assigned as the calibration sensor extrinsic parameter initial value, and the above process is iterated not less than 3 times to determine the calibration sensor extrinsic parameter final value.

[0067] In the present embodiment, the intermediate value of the extrinsic parameter obtained by the foregoing calculation is taken as the extrinsic parameter initial value of the calibration sensor; and the foregoing S101-S103 process is iterated according to the extrinsic parameter initial value to determine the extrinsic parameter final value of the calibration sensor, thereby improving the accuracy of the extrinsic parameter.

[0068] In an optional embodiment of the present application, the determined calibration sensor extrinsic parameter is assigned as the calibration sensor extrinsic parameter initial value, and the above process is iterated not less than 3 times to determine the calibration sensor extrinsic parameter final value, including: after iteration for 3 times, it is judged whether the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system converges to 0, if yes, the iteration is stopped, if not, the iteration is continued.

[0069] In the present embodiment, according to the convergence domain of the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system, the final iteration number is determined, when the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system converges to 0 after iteration for 3 times, the extrinsic parameter obtained by the third iteration is determined as the extrinsic parameter final value; when the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system does not converge to 0 after iteration for 3 times, the iteration is continued until the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system converges to 0.

[0070] Figure 2 An optional embodiment of an extrinsic parameter determination device based on semi-automatic calibration of the present application is shown.

[0071] In Figure 2 In the optional embodiment shown, the external parameter determination device based on semi-automatic calibration mainly comprises: a module 201 for collecting the self-vehicle driving trajectory and the calibration sensor perception information by using the high-precision inertial navigation and the calibration sensor carried by the autonomous vehicle during driving on the predetermined road, wherein a plurality of reference targets with an upper and lower interval not less than a predetermined height interval are arranged on a plurality of columnar objects with an interval not less than a predetermined longitudinal interval along the road direction on both sides of the predetermined road; a module 202 for determining the position of the reference target in the calibration sensor coordinate system by using the calibration sensor perception information; a module 203 for determining the intermediate value of the external parameter of the calibration sensor by using the self-vehicle driving trajectory, and the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system; and a module 204 for assigning the determined external parameter of the calibration sensor as the initial value of the external parameter of the calibration sensor, and iterating the above process not less than 3 times to determine the final value of the external parameter of the calibration sensor.

[0072] The external parameter determination device based on semi-automatic calibration provided in the application can be used to execute the external parameter determination method based on semi-automatic calibration described in any of the above embodiments, and has similar implementation principles and technical effects, which will not be described here.

[0073] In another optional embodiment of the application, a computer readable storage medium stores computer instructions, which are operated to execute the external parameter determination method based on semi-automatic calibration described in the above embodiments.

[0074] In one optional embodiment of the application, each functional module in the external parameter determination method based on semi-automatic calibration can be directly in hardware, in a software module executed by a processor, or in a combination of both.

[0075] The software module can reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. The exemplary storage medium is coupled to the processor, so that the processor can read information from the storage medium and write information to the storage medium.

[0076] The processor can be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. The general-purpose processor can be a microprocessor, but in the alternative, the processor can be any conventional processor, controller, microcontroller, or state machine. The processor can also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. The storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0077] In an optional implementation of the present disclosure, a computer device includes at least one processor, and a memory connected to the at least one processor in communication; wherein the memory stores computer instructions executable by the at least one processor, and the at least one processor operates the computer instructions to perform the method for determining the extrinsic parameter based on semi-automatic calibration described in the above embodiments.

[0078] In the several embodiments provided in the present disclosure, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. There can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0079] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0080] The above merely describes the embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structural transformation or direct or indirect application in other related technical fields based on the content of the present application specification and drawings is also included in the patent protection scope of the present application.

Claims

1. A method for determining extrinsic parameters based on semi-automatic calibration, characterized in that, include: During the process of an autonomous vehicle driving on a predetermined road, the vehicle's driving trajectory and calibration sensor information are collected by the vehicle's high-precision inertial navigation and calibration sensors. In this process, multiple reference targets are pre-arranged on multiple columnar objects on both sides of the predetermined road at intervals of not less than a predetermined longitudinal distance along the road direction, with vertical intervals of not less than a predetermined height interval. The position of the reference target in the calibration sensor coordinate system is determined using the information sensed by the calibration sensor. Using the vehicle's driving trajectory and the positions of the reference target in the world coordinate system and the calibration sensor's coordinate system, intermediate values ​​of the extrinsic parameters of the calibration sensor are determined. This determination includes: The lidar extrinsic parameter that minimizes the distance between the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system is determined as the intermediate value of the lidar extrinsic parameter; and The determined extrinsic parameters of the calibration sensor are assigned as initial values ​​for the extrinsic parameters of the calibration sensor. The above process is iterated at least 3 times to determine the final values ​​of the extrinsic parameters of the calibration sensor.

2. The method for determining extrinsic parameters based on semi-automatic calibration according to claim 1, characterized in that, When the calibration sensor is a lidar, the reference target is a laser reflector, the information sensed by the calibration sensor is lidar information, and the coordinate system of the calibration sensor is the lidar coordinate system, wherein determining the position of the reference target in the calibration sensor coordinate system using the information sensed by the calibration sensor includes: Based on the lidar information, select the point cloud in the lidar coordinate system corresponding to the lidar reflector; Select at least one point cloud point from the point cloud whose laser echo intensity is greater than a predetermined laser echo intensity threshold; and The point cloud point with the smallest distance to the center point of the laser reflector in the world coordinate system among the at least one point cloud points is determined as the laser reflector point cloud point.

3. The method for determining extrinsic parameters based on semi-automatic calibration according to claim 2, characterized in that, The step of determining the intermediate values ​​of the extrinsic parameters of the calibration sensor using the vehicle's driving trajectory and the positions of the reference target in the world coordinate system and the calibration sensor's coordinate system includes: The extrinsic parameter of the lidar that minimizes the distance between the point cloud of the laser reflector and the center point of the laser reflector is determined as the intermediate value of the extrinsic parameter of the lidar.

4. The method for determining extrinsic parameters based on semi-automatic calibration according to claim 1, characterized in that, When the calibration sensor is a camera, the reference target is a point light source, the information sensed by the calibration sensor is a 2D image, and the coordinate system of the calibration sensor is the camera coordinate system, wherein determining the position of the reference target in the calibration sensor coordinate system using the information sensed by the calibration sensor includes: Projecting the point light source from the world coordinate system onto the 2D image of the calibration sensor coordinate system yields the corresponding point light source projection point; and Search for the pixel with the highest pixel intensity near the projection point of the point light source in the 2D image, and determine the pixel as a candidate point.

5. The method for determining extrinsic parameters based on semi-automatic calibration according to claim 4, characterized in that, The step of determining the position of the reference target in the calibration sensor coordinate system using the information sensed by the calibration sensor further includes: In the 2D image, multiple pixels with pixel intensity greater than a pixel intensity threshold around the candidate point are selected, and the average value of the coordinates of the multiple pixels in the camera coordinate system is determined as the position of the point light source in the camera coordinate system.

6. The method for determining extrinsic parameters based on semi-automatic calibration according to claim 1, characterized in that, The step of assigning the determined extrinsic parameters of the calibration sensor as initial extrinsic parameters of the calibration sensor, and iterating the above process at least 3 times to determine the final extrinsic parameters of the calibration sensor includes: After the iteration is performed 3 times, it is determined whether the distance between the position of the reference target in the calibration sensor coordinate system and the position of the reference target in the world coordinate system converges to 0. If they converge, the iteration stops; if they do not converge, the iteration continues.

7. A device for determining extrinsic parameters based on semi-automatic calibration, characterized in that, include: This module is used to collect the vehicle's driving trajectory and calibration sensor perception information by using onboard high-precision inertial navigation and calibration sensors during the driving of an autonomous vehicle on a predetermined road. The predetermined road has multiple reference targets arranged in advance on multiple columnar objects with a vertical spacing of not less than a predetermined height spacing along the road direction on both sides. A module for determining the position of the reference target in the calibration sensor coordinate system using the information sensed by the calibration sensor; A module for determining intermediate values ​​of extrinsic parameters of a calibration sensor using the vehicle's trajectory, the position of the reference target in the world coordinate system, and the position of the reference target in the calibration sensor coordinate system, wherein determining the intermediate values ​​of extrinsic parameters of the calibration sensor using the vehicle's trajectory, the position of the reference target in the world coordinate system, and the position of the reference target in the calibration sensor coordinate system includes: The lidar extrinsic parameter that minimizes the distance between the position of the reference target in the world coordinate system and the position of the reference target in the calibration sensor coordinate system is determined as the intermediate value of the lidar extrinsic parameter; and This module is used to assign the determined extrinsic parameters of the calibration sensor as the initial extrinsic parameters of the calibration sensor, iterate the above process at least 3 times, and determine the final extrinsic parameters of the calibration sensor.

8. A computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are operated to perform the external parameter determination method based on semi-automatic calibration as described in any one of claims 1-6.

9. A computer device, characterized in that, include: At least one processor; as well as A memory that is communicatively connected to the at least one processor; The memory stores computer instructions executable by the at least one processor, which operates the computer instructions to perform the extrinsic parameter determination method based on semi-automatic calibration as described in any one of claims 1-6.

Images