A method for constructing a camera extrinsic parameter calibration field and a calibration field

Abstract

This invention discloses a method for constructing a camera extrinsic parameter calibration field and the calibration field itself. The method includes driving different types of unmanned vehicles to move to a corresponding target area in the calibration field; moving a calibration frame with calibration plates in the calibration field so that each camera on each unmanned vehicle captures images of at least a predetermined number of calibration plates, and the captured images of the calibration plates do not obstruct each other; and marking the target area with identification lines. The technical solution of this invention makes the calibration of camera extrinsic parameters more accurate, adaptable to various vehicle models, and saves time and effort.

Description

Technical Field

[0001] The embodiments of the present invention relate to camera parameter calibration technology, and more particularly to a method for constructing a camera extrinsic parameter calibration field and a calibration field. Background Technology

[0002] As autonomous vehicles become more widely used, the number of autonomous vehicles and the accuracy of their calibration are also gradually increasing. In order to meet the requirements of rapid, assembly-line calibration, a calibration site is needed to adapt to the calibration of the extrinsic parameters of autonomous vehicle cameras.

[0003] Currently, in the calibration field, camera extrinsic parameter calibration involves using a calibration frame to calibrate the extrinsic parameters of a single camera. The calibration frame is placed in front of the camera, and a measuring tape is used to measure the distance between the QR code sticker on the calibration frame and the center of the vehicle. A laser level is used to position the QR code sticker on the calibration frame so that it is horizontal and vertical. Calibration is carried out in this calibration field, which is difficult to operate, time-consuming and labor-intensive, and has low calibration accuracy. Summary of the Invention

[0004] This invention provides a method for constructing a camera extrinsic parameter calibration field and a calibration field that can make the calibration of camera extrinsic parameters more accurate, adapt to various vehicle models, and save time and effort.

[0005] In a first aspect, embodiments of the present invention provide a method for constructing camera extrinsic calibration field, the method comprising:

[0006] Drive different types of unmanned vehicles to the corresponding target areas in the calibration field;

[0007] The calibration frame with calibration plates in the calibration field is moved so that each camera on each unmanned vehicle can capture images of at least a set number of calibration plates, and the captured images of the calibration plates do not obstruct each other.

[0008] The target area is marked with a marker line.

[0009] Furthermore, there are multiple calibration frames, including a first calibration frame, a second calibration frame, and a third calibration frame, wherein the height of the first calibration frame is greater than the height of the second calibration frame, and the height of the second calibration frame is greater than the height of the third calibration frame.

[0010] Furthermore, the calibration frame with the calibration plate disposed in the calibration field is moved, comprising:

[0011] The calibration field is equipped with a calibration plate and a calibration frame that matches the height of the unmanned vehicle; wherein unmanned vehicles of different heights are matched with calibration frames of different heights.

[0012] Furthermore, the orientation angles of the calibration plates set on the multiple calibration frames are between 0° and 270°.

[0013] Furthermore, the method for constructing the camera extrinsic calibration field also includes:

[0014] The origin of the calibration field is taken as the starting position of the longitudinal center line of the calibration field, and the origin of the calibration plate is taken as the upper left corner of the calibration plate. The coordinates of the origin of the calibration plate relative to the origin of the calibration field are determined.

[0015] Furthermore, the longitudinal centerline of the unmanned vehicle coincides with the longitudinal centerline of the calibration field.

[0016] Furthermore, a QR code sticker is provided on the calibration plate, and the calibration frame is perpendicular to the ground.

[0017] Furthermore, the method for constructing the camera extrinsic calibration field also includes:

[0018] The captured images are processed using toolchain software to obtain the camera's extrinsic parameter calibration results.

[0019] Furthermore, the method for constructing the camera extrinsic calibration field also includes:

[0020] A reference camera is set on a target calibration frame, and the target calibration frame is abutted against the rear wheel of the unmanned vehicle;

[0021] Accordingly, the captured images are processed using toolchain software to obtain the camera extrinsic calibration results, including:

[0022] Determine the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field;

[0023] The extrinsic parameters of the reference camera relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field.

[0024] The extrinsic parameters of the unmanned vehicle relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration field and the extrinsic parameters of the reference camera relative to the unmanned vehicle.

[0025] Determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board based on the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and the extrinsic parameters of the calibration board in the calibration field;

[0026] The extrinsic parameters of the camera on the unmanned vehicle relative to the unmanned vehicle are determined based on the extrinsic parameters of the unmanned vehicle relative to the calibration field and the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field.

[0027] Secondly, embodiments of the present invention also provide a camera extrinsic calibration field, the calibration field including: a target area and a calibration frame, wherein a calibration plate is provided on the calibration frame;

[0028] The target area is designated for parking different types of driverless vehicles.

[0029] Specifically, the calibration frame in the calibration field is moved so that each camera on each unmanned vehicle can capture images of at least a set number of calibration boards, and the captured images of the calibration boards do not obstruct each other. After the calibration frame has been moved, the target area is marked with a marker line.

[0030] This invention addresses the problems of high operational difficulty, time-consuming and labor-intensive operation, and low calibration accuracy in existing technologies by driving different types of unmanned vehicles to move to the corresponding target area in the calibration field; moving the calibration frame with calibration plates in the calibration field so that each camera on each unmanned vehicle can capture images of at least a set number of calibration plates, and the captured images of the calibration plates do not obstruct each other; and marking the target area with marking lines. This makes the calibration of camera extrinsic parameters more accurate, adaptable to various vehicle models, and saves time and effort. Attached Figure Description

[0031] Figure 1A This is a flowchart of the method for constructing camera extrinsic calibration field according to Embodiment 1 of the present invention;

[0032] Figure 1B This is a schematic diagram of a scene for camera extrinsic calibration provided in Embodiment 1 of the present invention;

[0033] Figure 1C This is a schematic diagram of setting a QR code sticker on a calibration plate according to Embodiment 1 of the present invention;

[0034] Figure 2A This is a schematic diagram of a toolchain software login interface provided in Embodiment 1 of the present invention;

[0035] Figure 2B This is a schematic diagram of a business selection interface for toolchain software provided in Embodiment 1 of the present invention;

[0036] Figure 2C This is a schematic diagram of the calibration module interface of a toolchain software provided in Embodiment 1 of the present invention;

[0037] Figure 2D This is a schematic diagram illustrating the viewing, saving, and uploading of calibration results for toolchain software provided in Embodiment 1 of the present invention;

[0038] Figure 2 This is a schematic diagram of camera external parameter calibration provided in Embodiment 2 of the present invention. Detailed Implementation

[0039] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0040] Example 1

[0041] Figure 1A This is a flowchart of a method for constructing a camera extrinsic calibration field according to Embodiment 1 of the present invention. This embodiment is applicable to calibrating the extrinsic parameters of cameras for various vehicle models, and specifically includes the following steps:

[0042] S110 drives different types of unmanned vehicles to the corresponding target area in the calibration field.

[0043] Different types of autonomous vehicles can include those at different heights. The target area can be the central region of the calibration field where the autonomous vehicle is positioned. This central region can encompass the longitudinal center of the calibration field; alternatively, the central region of the calibration field can be the central region of the vehicle's parking calibration area. In other words, the longitudinal centerline of the calibration field coincides with the longitudinal centerline of the vehicle's parking calibration area. The vehicle's parking calibration area, within the calibration field, represents the area of ​​activity for calibrating the extrinsic parameters of the autonomous vehicle's cameras. The reference... Figure 1B Area 01 can be the vehicle parking calibration area of ​​the calibration field, while the middle area 02 (i.e. the middle area of ​​the calibration field) of the vehicle parking calibration area can be the target area, that is, the unmanned vehicle can be driven to move to area 02.

[0044] Specifically, the autonomous vehicle can be pre-driven to the target area in the calibration field to calibrate the extrinsic parameters of the cameras of different types of autonomous vehicles, such as the position and attitude of the cameras. The autonomous vehicle can be driven manually to the target area in the calibration field, or it can be driven through intelligent control. Optionally, the longitudinal centerline of the autonomous vehicle coincides with the longitudinal centerline of the calibration field, which can prevent the autonomous vehicle from tilting in the target area and make the camera extrinsic parameter calibration more accurate.

[0045] S120. Move the calibration frame with calibration plates in the calibration field so that each camera on each unmanned vehicle can capture images of at least a set number of calibration plates, and the captured images of the calibration plates do not obstruct each other.

[0046] In this embodiment of the invention, the set quantity can refer to the lower limit of the number of calibration plates captured by the cameras on each unmanned vehicle. Optionally, the set quantity can be 4 or other quantities can be set as needed. Optionally, there can be multiple calibration frames, and each calibration frame can be provided with at least one calibration plate. The calibration frames can include a first calibration frame, a second calibration frame, and a third calibration frame, wherein the height of the first calibration frame is greater than the height of the second calibration frame, and the length of the second calibration frame is greater than the height of the third calibration frame. For example, the calibration field contains 16 calibration frames, and a total of 28 calibration plates are provided on the calibration frames, including 6 long calibration frames (i.e., the first calibration frames), 6 medium-length calibration frames (i.e., the second calibration frames), and 4 short calibration frames (i.e., the third calibration frames). The number of calibration frames of various types in the calibration field can be set as needed.

[0047] In one embodiment of the present invention, optionally, moving the calibration frame with the calibration plate in the calibration field may include: moving a calibration frame with the calibration plate in the calibration field that matches the height of the unmanned vehicle; wherein, unmanned vehicles of different heights are matched with calibration frames of different heights. Since there are height differences between different types of unmanned vehicles, it is necessary to match calibration frames of different heights for unmanned vehicles of different heights when calibrating the unmanned vehicle camera. Specifically, based on the height of the currently processed unmanned vehicle, a calibration frame matching the height of the current unmanned vehicle can be selected from all calibration frames with calibration plates in the calibration field, and this type of calibration frame is moved to calibrate the currently processed unmanned vehicle camera. Optionally, when the unmanned vehicle is high, a higher-height calibration frame can be moved; when the unmanned vehicle is low, a lower-height calibration frame can be moved to complete the camera calibration, thus making it applicable to different vehicle models. Therefore, by moving the calibration frame that matches the height of the unmanned vehicle, it is possible to adapt to the calibration of camera extrinsic parameters on different vehicle models.

[0048] In this embodiment of the invention, after the unmanned vehicle moves to the target area in the calibration field, the camera on the unmanned vehicle can be turned on, and the calibration frame with calibration plates can be moved. The images captured by each camera on the unmanned vehicle at the current parking position of the unmanned vehicle and the current position of the calibration frame can be viewed. It can be checked whether the number of calibration plates in the captured images reaches the set number and whether there is any occlusion between the calibration plates in the captured images. If the number of calibration plates in the captured images does not reach the set number or there is occlusion between the calibration plates in the captured images, the calibration frame with calibration plates can be moved manually or intelligently controlled to ensure that each camera on each unmanned vehicle can capture at least the set number of calibration plates, for example, at least the calibration plates on 4 calibration frames, and the calibration plates captured by the cameras on each unmanned vehicle cannot be occluded.

[0049] In this embodiment of the invention, optionally, the orientation angle of the calibration plates disposed on the plurality of calibration frames is between 0° and 270°. Specifically, the calibration field includes calibration frames with multiple orientations, for example... Figure 1B The calibration plate on the calibration frame 04 indicated in the diagram has multiple orientations, where the orientation angle can be the angle between the front of the calibration plate and the longitudinal center line of the calibration field. Optionally, if the orientation angle of the calibration plate is 90°, the calibration plate is perpendicular to the longitudinal center line of the calibration field. By setting multiple orientation angles for the calibration plate, cameras on the autonomous vehicle from different directions can capture images of at least a set number of calibration plates, making the calibration of camera extrinsic parameters more accurate.

[0050] In this embodiment of the invention, optionally, a constant light source can be set in the calibration field so that the light intensity in the calibration field can be controlled between 150-200 lux, and external light source interference can be avoided in the calibration field; the ground in the calibration field can be completely horizontal (for example, the ground drop of 10 square meters is less than or equal to 3 millimeters) and the ground is non-reflective, so that the calibration board image captured by the camera on the unmanned vehicle can be better and the calibration from the camera's external parameters can be more accurate.

[0051] In an embodiment of the invention, optionally, a QR code sticker is provided on the calibration plate, and the calibration frame is perpendicular to the ground. Providing the QR code sticker on the calibration plate simplifies the algorithm during image processing for camera extrinsic parameter calibration. Specifically, refer to... Figure 1C The QR code sticker on the calibration plate can be affixed to the calibration plate. The QR code sticker is 0.7 meters long and wide, and can be fixed to the calibration frame with one side facing up. The QR code sticker number can be marked on the back of the calibration plate in advance. Furthermore, the calibration plate can be fixed, for example, by using 4 nuts to avoid deformation of the calibration plate due to too many nuts. The crossbar of the mounting bracket should not protrude to prevent the calibration plate edge from warping.

[0052] In this embodiment of the invention, a laser level can be used to adjust the horizontal position of the calibration frame so that the upper and lower QR code stickers are on the same horizontal line, and the calibration frame is perpendicular to the horizontal ground (with an allowable error within 1 mm). A laser level can also be used to adjust the horizontal relationship between the calibration plate and the calibration field. Specifically, the laser level projects three lines: one line coincides with the y-axis; the other two lines are horizontal to the QR code stickers on the calibration plate, with the light source precisely positioned to avoid obstructing the surface of the QR code stickers. The purpose of this setup is to ensure that the calibration plate, facing directly forward, is perpendicular to the longitudinal center line of the calibration field.

[0053] S130. The target area is marked with a marker line.

[0054] Specifically, after the calibration frame has been moved, a rectangular area can be marked with marker lines to identify the target area. (Reference) Figure 1B The target area can be identified using the boundary line shown in area 02.

[0055] In the image processing, the starting position of the longitudinal center line of the calibration field is taken as the origin of the calibration field, and the upper left corner of the calibration board is taken as the origin of the calibration board. The coordinates of the origin of the calibration board relative to the origin of the calibration field are determined, thereby determining the coordinates of each pixel of the QR code sticker on the calibration board.

[0056] The longitudinal centerline of the calibration field can be as follows: Figure 1B The lines indicated by 03 are shown in Table 1. As shown in Table 1, the coordinates x, y, and z values ​​of multiple calibration plate origins (i.e., target positions) relative to the calibration field origin can be determined. Among them, the Z-axis difference can represent the distance between the origins of two calibration plates on the calibration frame, and the YAW orientation can represent the orientation angle of the calibration plate.

[0057] Table 1

[0058]

[0059]

[0060] In this embodiment of the invention, the calibration of the intrinsic parameters of the autonomous vehicle's camera can be performed using toolchain software. Specifically, the toolchain software can achieve semi-automatic calibration of the camera's intrinsic parameters, as well as local saving and cloud uploading of the calibrated intrinsic parameters. Semi-automatic calibration means that the software needs to be used with a calibration turntable; the camera only needs to be manually fixed on the turntable and the calibration plate placed, and then the software can automatically capture and calibrate images from multiple angles. The basic functions of the toolchain software mainly include: semi-automatic calibration of camera intrinsic parameters, local saving of intrinsic parameters, cloud uploading of intrinsic parameters, and support for manual calibration of camera intrinsic parameters; the toolchain software runs on an Ubuntu 16.04 system under x86 and ARM platforms.

[0061] The operation process of the toolchain software is as follows:

[0062] Step 1: Log in and enter the calibration interface: (Reference) Figure 2A Enter the toolchain program, input your username and password, select the project, and click the login button below. After successful login, click the business module icon to enter the business selection interface; (Reference) Figure 2B Click the "Calibration" module, then click the camera icon (see reference). Figure 2C (Region 01), then click the Internal Reference tab (see reference). Figure 2C Enter the internal parameter calibration interface in area 02.

[0063] Step 2, Connect the calibration turntable: Install the camera onto the turntable and connect the turntable and camera to the PC;

[0064] Step 3, Set up the camera: Refer to Figure 2C In area 03, select the supported camera device node for "Camera"; select pinhole camera or fisheye camera for "Camera Type"; select grayscale or color image for "Color"; select the manufacturer code for "Byte"; and select PNG or JPG for "Format".

[0065] Step 4, Calibration Settings: Refer to Figure 2C In area 04, select the image save path under "Path"; select the calibration board type under "Board Type", supporting Chess Board and ChAruco Board; select the connected calibration turntable under "USB Device"; and set the single rotation angle of the calibration turntable in the X-axis and Y-axis directions, respectively, under "X" and "Y".

[0066] Step 5, Automatic Calibration: Reference Figure 2C In area 05, click the "Automatic Calibration" button to start automatic calibration.

[0067] Step 6: View, save, and upload calibration results: (Refer to...) Figure 2D ,according to Figure 2D After the image calibration of region 01 is completed, in Figure 2D The calibration results of the internal parameters can be viewed in area 02, which is the "Calibration Results" column on the right side of the interface; click... Figure 2D The "Data Upload" icon on the right side of the module bar in area 03 can be used to open... Figure 2D In the "Upload" tab shown in area 04, after selecting the vehicle, clicking the "Save" button will save the result locally, and clicking the "Upload" button will upload the result to the cloud.

[0068] In this embodiment of the invention, after the intrinsic parameters of the camera on the unmanned vehicle are determined, the captured images can be processed by toolchain software to obtain the extrinsic parameter calibration results of the camera.

[0069] In an embodiment of the present invention, optionally, the method provided by the embodiment of the present invention may further include: setting a reference camera on a target calibration frame and abutting the target calibration frame against the rear wheel of the unmanned vehicle; correspondingly, processing the captured images using toolchain software to obtain camera extrinsic parameter calibration results, including: determining the extrinsic parameters of the reference camera relative to the calibration plate and the extrinsic parameters of the calibration plate in the calibration field; determining the extrinsic parameters of the reference camera relative to the calibration field based on the extrinsic parameters of the reference camera relative to the calibration plate and the extrinsic parameters of the calibration plate in the calibration field; determining the extrinsic parameters of the unmanned vehicle relative to the calibration field based on the extrinsic parameters of the reference camera relative to the calibration field and the extrinsic parameters of the reference camera relative to the unmanned vehicle; determining the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration plate, and determining the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field based on the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field and the extrinsic parameters of the calibration plate in the calibration field; determining the extrinsic parameters of the camera on the unmanned vehicle relative to the unmanned vehicle based on the extrinsic parameters of the unmanned vehicle relative to the calibration field and the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field.

[0070] The reference camera can be a camera mounted separately on a target calibration frame. When calibrating the extrinsic parameters of the cameras on the autonomous vehicle, the calibration frame is placed against the rear wheel of the autonomous vehicle to determine the positional relationship of the separately mounted reference camera relative to the autonomous vehicle (i.e., the extrinsic parameters are determined). This camera serves as the reference camera. Optionally, the target calibration frame can be a target tripod, on which the reference camera is mounted.

[0071] Specifically, a reference camera can be pre-set on a target calibration frame, which is then abutted against the rear wheel of the autonomous vehicle. Images captured by the reference camera are processed, and the reference camera is calibrated using OpenCV interfaces to obtain its extrinsic parameters relative to the calibration plate. The extrinsic parameters of the calibration plate within the calibration field can be determined by the relationship between the calibration plate and the origin of the calibration field. Furthermore, based on the extrinsic parameters of the reference camera relative to the calibration plate and the extrinsic parameters of the calibration plate within the calibration field, the extrinsic parameters of the reference camera relative to the calibration field can be obtained through relative position transformation. Thus, based on the extrinsic parameters of the reference camera relative to the calibration field and the pre-determined extrinsic parameters of the reference camera relative to the autonomous vehicle, the extrinsic parameters of the autonomous vehicle relative to the calibration field (e.g., the autonomous vehicle's position and attitude within the calibration field) can be determined through relative position transformation.

[0072] In this embodiment of the invention, images captured by cameras on an autonomous vehicle can be processed, and the cameras on the autonomous vehicle can be calibrated through OpenCV related interfaces to obtain the extrinsic parameters of the cameras on the autonomous vehicle relative to the calibration board. The extrinsic parameters of the cameras on the autonomous vehicle relative to the calibration board are combined with the extrinsic parameters of the calibration board in the calibration field, and the extrinsic parameters of the cameras on the autonomous vehicle relative to the calibration field are obtained through relative position relationship transformation. Thus, based on the extrinsic parameters of the autonomous vehicle relative to the calibration field and the extrinsic parameters of the cameras on the autonomous vehicle relative to the calibration field, the extrinsic parameters of the cameras on the autonomous vehicle relative to the autonomous vehicle are obtained through relative position relationship transformation.

[0073] Therefore, by setting up a reference camera and using it to calibrate the extrinsic parameters of the cameras on the autonomous vehicle, accurate calibration can be achieved, making calibration faster and more convenient. It should be noted that the parking position of the autonomous vehicle in the calibration field can be adaptively adjusted. For example, if the cameras on the autonomous vehicle cannot consistently capture at least the required number of images within the target area, the vehicle's position in the calibration field can be moved to ensure that it captures the required images.

[0074] The technical solution of this embodiment solves the problems of high operational difficulty, time-consuming and labor-intensive operation, and low calibration accuracy of the prior art by driving different types of unmanned vehicles to move to the corresponding target area in the calibration field; moving the calibration frame with calibration plates in the calibration field so that each camera on each unmanned vehicle can capture at least a set number of calibration plates, and the captured calibration plates do not obstruct each other; and marking the target area with marking lines. This makes the calibration of camera extrinsic parameters more accurate, adaptable to various vehicle models, and saves time and effort.

[0075] Example 2

[0076] Figure 2 This is a schematic diagram of a camera external parameter calibration field provided in Embodiment 2 of the present invention. The calibration field may include: a target area 210 and a calibration frame 220, wherein a calibration plate is disposed on the calibration frame;

[0077] Target area 210 is used to park different types of driverless vehicles;

[0078] The calibration frame 220 in the calibration field is moved so that each camera on each unmanned vehicle can capture images of at least a set number of calibration boards, and the captured images of the calibration boards do not obstruct each other. After the calibration frame 220 has been moved, the target area 210 is marked with marking lines.

[0079] The technical solution of this embodiment solves the problems of high operational difficulty, time-consuming and labor-intensive operation, and low calibration accuracy caused by parking different types of unmanned vehicles in the target area, moving the calibration frame so that each camera on each unmanned vehicle can capture at least a set number of calibration plates, and the captured calibration plates do not obstruct each other. After the calibration frame is moved, the target area is marked with marking lines. This method solves the problems of high operational difficulty, time-consuming and labor-intensive operation, and low calibration accuracy caused by placing the calibration frame in front of the camera, measuring the distance between the QR code sticker on the calibration frame and the center of the vehicle with a tape measure, and using a laser level to place the QR code sticker on the calibration frame so that it is horizontal and vertical. This method makes the calibration of camera extrinsic parameters more accurate, adaptable to multiple vehicle models, and saves time and labor.

[0080] In the above calibration field, optionally, there are multiple calibration frames 220. Each calibration frame 220 may include a first calibration frame, a second calibration frame, and a third calibration frame, wherein the height of the first calibration frame is greater than the height of the second calibration frame, and the length of the second calibration frame is greater than the height of the third calibration frame.

[0081] Optionally, in the above-mentioned calibration field, the step of moving the calibration frame in the calibration field, which is equipped with the calibration plate, may include:

[0082] The calibration field is equipped with a calibration plate and a calibration frame that matches the height of the unmanned vehicle; wherein unmanned vehicles of different heights are matched with calibration frames of different heights.

[0083] In the aforementioned calibration field, optionally, the orientation angle of the calibration plates set on multiple calibration frames is between 0° and 270°.

[0084] Optionally, the calibration field mentioned above may also include:

[0085] The origin of the calibration field is taken as the starting position of the longitudinal center line of the calibration field, and the origin of the calibration plate is taken as the upper left corner of the calibration plate. The coordinates of the origin of the calibration plate relative to the origin of the calibration field are determined.

[0086] Optionally, in the aforementioned calibration field, the longitudinal centerline of the unmanned vehicle coincides with the longitudinal centerline of the calibration field.

[0087] Optionally, in the above calibration field, a QR code sticker is provided on the calibration plate, and the calibration frame 220 is perpendicular to the ground.

[0088] Optionally, the calibration field mentioned above may also include:

[0089] The captured images are processed using toolchain software to obtain the camera's extrinsic parameter calibration results.

[0090] Optionally, the calibration field mentioned above may also include:

[0091] A reference camera is set on a target calibration frame, and the target calibration frame is abutted against the rear wheel of the unmanned vehicle;

[0092] Accordingly, the captured images are processed using toolchain software to obtain the camera extrinsic calibration results, including:

[0093] Determine the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field;

[0094] The extrinsic parameters of the reference camera relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field.

[0095] The extrinsic parameters of the unmanned vehicle relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration field and the extrinsic parameters of the reference camera relative to the unmanned vehicle.

[0096] Determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board based on the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and the extrinsic parameters of the calibration board in the calibration field;

[0097] The extrinsic parameters of the camera on the unmanned vehicle relative to the unmanned vehicle are determined based on the extrinsic parameters of the unmanned vehicle relative to the calibration field and the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field.

[0098] The above description can be referred to the description of the above embodiments, and will not be repeated here.

[0099] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A method for constructing a camera extrinsic calibration field, characterized in that, include: Drive different types of unmanned vehicles to the corresponding target areas in the calibration field; The calibration frame, which is equipped with calibration plates in the calibration field and is matched to the height of the unmanned vehicle, is moved so that each camera on each unmanned vehicle can capture images of at least a predetermined number of calibration plates, and the captured images of the calibration plates do not obstruct each other. There are multiple calibration frames, including a first calibration frame, a second calibration frame, and a third calibration frame. The height of the first calibration frame is greater than the height of the second calibration frame, and the height of the second calibration frame is greater than the height of the third calibration frame. Unmanned vehicles of different heights are matched with calibration frames of different heights. The orientation angle of the calibration plates on the multiple calibration frames is [missing information]. The orientation angle is the angle between the front of the calibration plate and the longitudinal center line of the calibration field; the longitudinal center line of the unmanned vehicle coincides with the longitudinal center line of the calibration field; a QR code sticker is provided on the calibration plate, and the calibration frame is perpendicular to the ground; the QR code sticker is fixed on the calibration frame with one side facing upwards, and the number of the QR code sticker is marked on the back of the calibration plate; the starting position of the longitudinal center line of the calibration field is taken as the origin of the calibration field, the upper left corner of the calibration plate is taken as the origin of the calibration plate, and the coordinates of the origin of the calibration plate relative to the origin of the calibration field are determined, and the coordinates of each pixel of the QR code sticker on the calibration plate are determined; The target area is marked with a marker line.

2. The method according to claim 1, characterized in that, The method further includes: The captured images are processed using toolchain software to obtain the camera's extrinsic parameter calibration results.

3. The method according to claim 2, characterized in that, Also includes: A reference camera is set on a target calibration frame, and the target calibration frame is abutted against the rear wheel of the unmanned vehicle; Accordingly, the captured images are processed using toolchain software to obtain the camera extrinsic calibration results, including: Determine the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field; The extrinsic parameters of the reference camera relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration board and the extrinsic parameters of the calibration board in the calibration field. The extrinsic parameters of the unmanned vehicle relative to the calibration field are determined based on the extrinsic parameters of the reference camera relative to the calibration field and the extrinsic parameters of the reference camera relative to the unmanned vehicle. Determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and determine the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board based on the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration board, and the extrinsic parameters of the calibration board in the calibration field; The extrinsic parameters of the camera on the unmanned vehicle relative to the unmanned vehicle are determined based on the extrinsic parameters of the unmanned vehicle relative to the calibration field and the extrinsic parameters of the camera on the unmanned vehicle relative to the calibration field.

4. A camera extrinsic calibration field, characterized in that, include: The target area and the calibration frame, wherein a calibration plate is provided on the calibration frame; The target area is designated for parking different types of driverless vehicles. Specifically, the calibration frame in the calibration field is moved so that each camera on each unmanned vehicle can capture images of at least a set number of calibration boards, and the captured images of the calibration boards do not obstruct each other. After the calibration frame has been moved, the target area is marked with marking lines. The calibration frames are multiple, including a first calibration frame, a second calibration frame, and a third calibration frame. The height of the first calibration frame is greater than the height of the second calibration frame, and the height of the second calibration frame is greater than the height of the third calibration frame. Autonomous vehicles of different heights are matched with calibration frames of different heights. The orientation angle of the calibration plates mounted on the multiple calibration frames is... The orientation angle is the angle between the front of the calibration plate and the longitudinal center line of the calibration field; the longitudinal center line of the unmanned vehicle coincides with the longitudinal center line of the calibration field; a QR code sticker is provided on the calibration plate, and the calibration frame is perpendicular to the ground; the QR code sticker is fixed on the calibration frame with one side facing upwards, and the number of the QR code sticker is marked on the back of the calibration plate; the starting position of the longitudinal center line of the calibration field is taken as the origin of the calibration field, the upper left corner of the calibration plate is taken as the origin of the calibration plate, and the coordinates of the origin of the calibration plate relative to the origin of the calibration field are determined, and the coordinates of each pixel of the QR code sticker on the calibration plate are determined.

Images