CN111190153A

CN111190153A - External parameter calibration method and device, intelligent robot and computer readable storage medium

Info

Publication number: CN111190153A
Application number: CN202010272202.7A
Authority: CN
Inventors: 吴波; 宋乐; 秦宝星; 程昊天
Original assignee: Shanghai Gaussian Automation Technology Development Co Ltd
Current assignee: Shanghai Gaussian Automation Technology Development Co Ltd
Priority date: 2020-04-09
Filing date: 2020-04-09
Publication date: 2020-05-22
Anticipated expiration: 2040-04-09
Also published as: CN111190153B

Abstract

The application discloses an external reference calibration method, which is used for external reference calibration for a radar, and comprises the following steps: acquiring point clouds which are on a reference plane, a first plane and a second plane and are not parallel to each other and are related to external parameters of a radar to be calibrated; acquiring fitting errors of the reference surface, the first plane and the second plane according to the point cloud; acquiring a geometric relation error between the reference plane, the first plane and the second plane according to the point cloud; obtaining external parameters to be selected of the radar to be calibrated through the optimization fitting error and the geometric relation error; judging whether the difference between the height of the point cloud of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not; and if so, determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated. The application also discloses an external reference calibration device, an intelligent robot and a computer readable storage medium. According to the method and the device, a high-precision external reference result can be obtained, and a foundation is provided for accurate navigation and obstacle avoidance of the intelligent robot.

Description

External parameter calibration method and device, intelligent robot and computer readable storage medium

Technical Field

The present disclosure relates to the field of radar parameter calibration technologies, and in particular, to an external reference calibration method and apparatus, an intelligent robot, and a computer-readable storage medium.

Background

The laser radar sensor is widely applied to the aspects of map building, positioning, navigation, obstacle avoidance and the like of the mobile robot. The mobile robot can identify the obstacle on the premise of avoiding the obstacle, and for the obstacle on the ground, if the obstacle needs to be accurately identified, the laser radar sensor is the first choice. However, the actual installation position and design of the laser radar are different due to installation errors, mechanical structure abrasion and the like, so that the manual calibration is inefficient and has low precision, and therefore, the automatic calibration of the external parameters of the laser radar is very urgently needed.

Disclosure of Invention

In view of the above, the present invention is directed to solving, at least to some extent, one of the problems in the related art. Therefore, the embodiment of the application provides an external reference calibration method and device, an intelligent robot and a computer readable storage medium.

The external reference calibration method is used for external reference calibration of the radar, and comprises the following steps: acquiring point clouds which are on a reference plane, a first plane and a second plane and are not parallel to each other and are related to external parameters of a radar to be calibrated; acquiring fitting errors of the reference surface, the first plane and the second plane according to the point cloud; acquiring a geometric relation error between the reference plane, the first plane and the second plane according to the point cloud; optimizing the fitting error and the geometric relation error to obtain a candidate external parameter of the radar to be calibrated; judging whether the difference between the height of the point cloud of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not; and if so, determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated.

In the external reference calibration method of the embodiment of the application, firstly, point clouds of external references of a radar to be calibrated on a reference plane, a first plane and a second plane which are not parallel to each other are obtained, fitting errors of the reference plane, the first plane and the second plane are obtained according to the point clouds, geometric relation errors between the reference plane, the first plane and the second plane are obtained according to the point clouds, then, optimization processing is carried out on the fitting errors and the geometric relation errors to obtain external references to be selected, whether the difference value between the height of the point clouds of the reference plane under the external references to be selected and the reference height of the reference plane is in a preset range is judged, if yes, the external references to be selected are determined to be the external references calibrated by the radar to be calibrated, the point clouds on the three planes related to the external references of the radar to be calibrated are obtained, the fitting errors and the geometric relation errors are both related to the external references of the radar to be calibrated, and further, the external references to be selected can be obtained by optimizing the fitting errors and the geometric, therefore, the external reference calibration method can automatically calibrate a plurality of external references of the radar, can obtain a high-precision external reference result, and provides a foundation for accurate navigation and obstacle avoidance of the intelligent robot.

In some embodiments, the acquiring point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to external parameters of the radar to be calibrated includes: acquiring a first point cloud on the reference surface, a second point cloud on the first plane and a third point cloud on the second plane, which are detected by the radar to be calibrated, wherein the first point cloud, the second point cloud and the third point cloud are all related to external parameters of the radar to be calibrated; and acquiring a fourth point cloud on the first plane and a fifth point cloud on the second plane, which are detected by a reference radar, wherein the fourth point cloud and the fifth point cloud are related to the external parameters of the reference radar.

In the embodiment, the point clouds respectively detected by the radar to be calibrated and the reference radar are obtained, the point clouds on the reference surface, the first plane and the second plane are respectively detected by the radar to be calibrated, and the point clouds on the first plane and the second plane are respectively obtained by the reference radar.

In some embodiments, the obtaining fitting errors of the reference plane, the first plane, and the second plane from the point cloud comprises: converting the first point cloud, the second point cloud and the third point cloud from the radar coordinate system to be calibrated to the reference radar coordinate system; acquiring a reference surface expression of the reference surface according to the first point cloud; acquiring a first plane expression of the first plane according to the second point cloud and the fourth point cloud; acquiring a second plane expression of the second plane according to the third point cloud and the fifth point cloud; and acquiring fitting errors of the reference surface, the first plane and the second plane respectively according to the reference surface expression, the first plane expression and the second plane expression.

In this embodiment, the first point cloud, the second point cloud, and the third point cloud are converted from the radar coordinate system to be calibrated to the reference radar coordinate system, such that the first point cloud, the second point cloud, and the third point cloud include the external reference relationship of the radar coordinate system to be calibrated, the reference surface expression of the reference surface is further obtained according to the first point cloud, the first plane expression of the first plane is obtained through the second point cloud and the fourth point cloud, the second plane expression is obtained through the third point cloud and the fifth point cloud, and then the fitting errors of the reference surface, the first plane, and the second plane are respectively obtained according to the reference surface expression, the first plane expression, and the second plane expression, so that the obtained fitting error is related to the external reference of the radar to be calibrated, and the external reference to be selected obtained in the following steps is more accurate by obtaining the fitting error.

In some embodiments, the obtaining a geometric relationship error between the reference plane, the first plane, and the second plane from the point cloud includes: acquiring a geometric relation error between the reference surface and the first plane according to the geometric relation between the reference surface and the first plane; acquiring a geometric relation error of the reference surface and the second plane according to the geometric relation between the reference surface and the second plane; and acquiring the geometric relation error of the first plane and the second plane according to the geometric relation between the first plane and the second plane.

In this embodiment, since the reference plane, the first plane, and the second plane are not parallel to each other, that is, the reference plane, the first plane, and the second plane intersect each other, and according to the geometric relationship between the planes, the geometric relationship error between the reference plane and the first plane, the geometric relationship error between the reference plane and the second plane, and the geometric relationship error between the first plane and the second plane are obtained, so that the external parameter to be selected of the radar to be calibrated can be better obtained by obtaining the geometric relationship error between the planes, and correlating the geometric relationship error with the external parameter of the radar to be calibrated, and the external parameter calibrated of the radar to be calibrated is more accurate.

In some embodiments, the angle between the first plane and the second plane is [10 °,170 °, and/or the angle between the reference plane and the first plane is [10 °,170 °, and/or the angle between the reference plane and the second plane is [10 °,170 °.

In this embodiment, an included angle between the first plane and the second plane is [10 °,170 ° ], an included angle between the reference plane and the first plane is [10 °,170 ° ], and/or an included angle between the reference plane and the second plane is [10 °,170 ° ], which can avoid that when an included angle between the reference plane, the first plane, and the second plane is too small or too large, the two planes are close to coincide with each other, and the constraint between the two planes is weak, so that the two planes may be fitted into one plane in the fitting process, resulting in inaccurate obtained external parameters.

In some embodiments, the optimizing the fitting error and the geometric relationship error to obtain a candidate external parameter of the radar to be calibrated includes: establishing an optimization equation according to the fitting error and the geometric relation error; along the gradient descending direction of the optimization equation, generating a group of test external parameters from the initial value of the external parameters by increasing a preset step length every time, and obtaining a plurality of groups of test external parameters; acquiring gradients of an optimization equation under a plurality of groups of test external parameters, and taking the current test external parameter as the external parameter to be selected when a difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is smaller than a preset difference value; or when the number of the test external parameters is equal to or greater than the preset number, taking the last group of test external parameters as the external parameters to be selected.

In the embodiment, an optimization equation is established according to the fitting error and the geometric relation error, then a group of test external parameters is generated by increasing a preset step length from an initial value of the external parameters along the gradient descending direction of the optimization equation, a plurality of groups of test external parameters are obtained, finally the gradient of the optimization equation under the plurality of groups of test external parameters is obtained, when the difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is less than a preset difference value, the current test external parameter is taken as the external parameter to be selected, or when the number of the test external parameters is more than a preset number, the last group of test external parameters is taken as the external parameter to be selected, because the gradient of the optimization equation is descending along with the increase of the test external parameters, when the difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is less than the preset difference value, the gradient of the optimization equation is, and at the moment, the corresponding test external parameter is the proper external parameter, and the current test external parameter is acquired as the external parameter to be selected, so that the external parameter to be selected is more accurate. Meanwhile, because the gradient of the optimization equation is always reduced along with the increase of the test external parameters, when the number of the test external parameters is enough, the gradient corresponding to the optimization equation tends to a critical value, and the last group of test external parameters are obtained as the external parameters to be selected, so that the external parameters to be selected are more accurate.

In some embodiments, when the radar to be calibrated is capable of rotating around a rotation axis relative to the mounting carrier of the radar to be calibrated, the external reference calibration method further includes: setting an angle interval according to the angle range of the radar to be calibrated rotating around the rotating shaft to form at least one calibration area; and performing external reference calibration for the radar to be calibrated in each calibration area at least once.

In the embodiment, the radar to be calibrated can rotate around the rotating shaft relative to the installation carrier of the radar to be calibrated, therefore, the installation position of the radar to be calibrated on the installation carrier is not fixed, firstly, an angle interval is set according to the angle range of the radar to be calibrated rotating around the rotating shaft, at least one calibration area is formed, and meanwhile, the radar to be calibrated is calibrated at least once in each calibration area.

In some embodiments, when the mounting carrier of the radar to be calibrated is an intelligent robot and the reference surface is a driving surface of the intelligent robot, the external reference calibration method further includes: and controlling the intelligent robot to move to an area with a smooth running surface, wherein no obstacle exists in a preset range.

In the embodiment, the radar to be calibrated is installed on the intelligent robot, the reference surface is the running surface of the intelligent robot, and the intelligent robot is controlled to move to the region with no obstacle in the preset range and a smooth running surface, so that the influence of point clouds on the obstacle on calculated fitting errors and geometric relation errors can be avoided, the inaccuracy of the height of the reference surface finally calculated due to the unevenness of the running surface can be avoided, the inaccuracy of the external parameter finally calibrated can be avoided, and the radar to be calibrated can obtain the accurate external parameter.

The external reference calibration device comprises a first acquisition module, a second acquisition module, a third acquisition module, a calculation module, a judgment module and a determination module, wherein the first acquisition module is used for acquiring point clouds which are not parallel to each other and are related to external references of a radar to be calibrated; the second acquisition module is used for acquiring the fitting errors of the reference surface, the first plane and the second plane by the point cloud; the third acquisition module is used for acquiring a geometric relationship error between the reference plane, the first plane and the second plane according to the point cloud; the calculation module is used for optimizing the fitting error and the geometric relation error so as to obtain a candidate external parameter of the radar to be calibrated; the judging module is used for judging whether the difference value between the point cloud height of the reference surface under the external reference to be selected and the reference height of the reference surface is within a preset range; and the determining module is used for determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated when the judgment result of the judging module is yes.

In the external reference calibration device of the embodiment of the application, firstly, point clouds of external references of a radar to be calibrated on a reference plane, a first plane and a second plane which are not parallel to each other are obtained, fitting errors of the reference plane, the first plane and the second plane are obtained according to the point clouds, geometric relation errors between the reference plane, the first plane and the second plane are obtained according to the point clouds, then, optimization processing is carried out on the fitting errors and the geometric relation errors to obtain external references to be selected, whether the difference value between the height of the point clouds of the reference plane under the external references to be selected and the reference height of the reference plane is in a preset range is judged, if yes, the external references to be selected are determined to be the external references calibrated by the radar to be calibrated, the point clouds on the three planes related to the external references of the radar to be calibrated are obtained, the fitting errors and the geometric relation errors are both related to the external references of the radar to be calibrated, and further, the external references to be selected can be obtained by optimizing the fitting errors and the geometric, therefore, the external reference calibration method can automatically calibrate a plurality of external references of the radar, can obtain a high-precision external reference result, and provides a foundation for accurate navigation and obstacle avoidance of the intelligent robot.

In some embodiments, the first obtaining module is further configured to: acquiring a first point cloud on the reference surface, a second point cloud on the first plane and a third point cloud on the second plane, which are detected by the radar to be calibrated, wherein the first point cloud, the second point cloud and the third point cloud are all related to external parameters of the radar to be calibrated; and acquiring a fourth point cloud on the first plane and a fifth point cloud on the second plane, which are detected by a reference radar, wherein the fourth point cloud and the fifth point cloud are related to the external parameters of the reference radar.

In some embodiments, the second obtaining module is further configured to: converting the first point cloud, the second point cloud and the third point cloud from the radar coordinate system to be calibrated to the reference radar coordinate system; acquiring a reference surface expression of the reference surface according to the first point cloud; acquiring a first plane expression of the first plane according to the second point cloud and the fourth point cloud; acquiring a second plane expression of the second plane according to the third point cloud and the fifth point cloud; and acquiring fitting errors of the reference surface, the first plane and the second plane respectively according to the reference surface expression, the first plane expression and the second plane expression.

In this embodiment, the first point cloud, the second point cloud, and the third point cloud are converted from the radar coordinate system to be calibrated to the reference radar coordinate system, so that the first point cloud, the second point cloud, and the third point cloud include the external reference relationship of the radar coordinate system to be calibrated, the reference surface expression of the reference surface is further obtained according to the first point cloud, the first plane expression of the first plane is obtained through the second point cloud and the fourth point cloud, the second plane expression is obtained through the third point cloud and the fifth point cloud, and then the sum errors of the reference surface, the first plane, and the second plane are respectively obtained according to the reference surface expression, the first plane expression, and the second plane expression, so that the obtained fitting error is related to the external reference of the radar to be calibrated, and the external reference to be selected obtained in the following step is more accurate by obtaining the fitting error.

In some embodiments, the third obtaining module is further configured to: acquiring a geometric relation error between the reference surface and the first plane according to the geometric relation between the reference surface and the first plane; acquiring a geometric relation error of the reference surface and the second plane according to the geometric relation between the reference surface and the second plane; and acquiring the geometric relation error of the first plane and the second plane according to the geometric relation between the first plane and the second plane.

In this embodiment, since the reference plane, the first plane, and the second plane are not parallel to each other, that is, the reference plane, the first plane, and the second plane intersect each other, a geometric relationship exists between the reference plane, the first plane, and the second plane, and a geometric relationship error between the reference plane and the first plane, a geometric relationship error between the reference plane and the second plane, and a geometric relationship error between the first plane and the second plane are obtained according to the geometric relationship between the planes, so that the external parameters to be selected of the radar to be calibrated can be better obtained by obtaining the geometric relationship error between the planes, and correlating the geometric relationship error with the external parameters of the radar to be calibrated, so that the external parameters to be calibrated of the radar to be calibrated are more accurate.

In some embodiments, the computing module is further configured to: establishing an optimization equation according to the fitting error and the geometric relation error; along the gradient descending direction of the optimization equation, generating a group of test external parameters from the initial value of the external parameters by increasing a preset step length every time, and obtaining a plurality of groups of test external parameters; acquiring gradients of an optimization equation under a plurality of groups of test external parameters, and taking the current test external parameter as the external parameter to be selected when a difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is smaller than a preset difference value; or when the number of the test external parameters is equal to or greater than the preset number, taking the last group of test external parameters as the external parameters to be selected.

In the embodiment, an optimization equation is established according to the fitting error and the geometric relation error, then a group of test external parameters is generated by increasing a preset step length from an initial value of the external parameters along the gradient descending direction of the optimization equation, a plurality of groups of test external parameters are obtained, finally the gradient of the optimization equation under the plurality of groups of test external parameters is obtained, when the difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is less than a preset difference value, the current test external parameter is taken as the external parameter to be selected, or when the number of the test external parameters is more than a preset number, the last group of test external parameters is taken as the external parameter to be selected, because the gradient of the optimization equation is descending along with the increase of the test external parameters, when the difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is less than the preset difference value, the gradient of the optimization, and at the moment, the corresponding test external parameter is the proper external parameter, and the current test external parameter is acquired as the external parameter to be selected, so that the external parameter to be selected is more accurate. Meanwhile, because the gradient of the optimization equation is always reduced along with the increase of the test external parameters, when the number of the test external parameters is enough, the gradient corresponding to the optimization equation tends to a critical value, and the last group of test external parameters are obtained as the external parameters to be selected, so that the external parameters to be selected are more accurate.

In some embodiments, when the radar to be calibrated is capable of rotating around a rotation axis relative to the mounting carrier of the radar to be calibrated, the external reference calibration apparatus is further configured to: setting an angle interval according to the angle range of the radar to be calibrated rotating around the rotating shaft to form at least one calibration area; and performing external reference calibration for the radar to be calibrated in each calibration area at least once.

In some embodiments, when the mounting carrier of the radar to be calibrated is an intelligent robot and the reference surface is a driving surface of the intelligent robot, the external reference calibration method is further configured to: and controlling the intelligent robot to move to an area with a smooth running surface, wherein no obstacle exists in a preset range.

The intelligent robot of the embodiment of the application comprises one or more processors and a memory; and one or more programs, wherein the one or more programs are stored in the memory and executed by the one or more processors, the programs comprising instructions for performing the external reference calibration method of any of the embodiments described above.

In the intelligent robot of the embodiment of the application, the point clouds of external references of the radar to be calibrated on the non-parallel reference plane, the first plane and the second plane are firstly obtained, the fitting errors of the reference plane, the first plane and the second plane are obtained according to the point clouds, the geometric relation error between the reference plane, the first plane and the second plane is obtained according to the point clouds, then the fitting error and the geometric relation error are optimized to obtain the external references to be selected, whether the difference value between the height of the point clouds of the reference plane under the external references to be selected and the reference height of the reference plane is in a preset range is judged, if yes, the external references to be selected are determined to be the external references calibrated by the radar to be calibrated, the point clouds on the three planes related to the external references of the radar to be calibrated are obtained, the fitting errors and the geometric relation errors are both related to the external references of the radar to be calibrated, and the fitting errors and the geometric relation errors are further optimized to obtain the external references to be selected, therefore, the external reference calibration method can automatically calibrate a plurality of external references of the radar, can obtain a high-precision external reference result, and provides a foundation for accurate navigation and obstacle avoidance of the intelligent robot.

The computer-readable storage media of the embodiments of the present application, when executed by one or more processors, cause the processors to perform the external reference calibration method of any of the above embodiments.

In the computer-readable storage medium of the embodiment of the application, the point clouds of external references of the radar to be calibrated on a reference plane, a first plane and a second plane which are not parallel to each other are firstly obtained, the fitting errors of the reference plane, the first plane and the second plane are obtained according to the point clouds, then the geometric relation errors between the reference plane, the first plane and the second plane are obtained according to the point clouds, then the fitting errors and the geometric relation errors are optimized to obtain external references to be selected, whether the difference value between the height of the point clouds of the reference plane under the external references to be selected and the reference height of the reference plane is in a preset range is judged, if yes, the external references to be selected are determined to be the external references calibrated by the radar to be calibrated, the point clouds on the three planes and related to the external references of the radar to be calibrated are obtained, the fitting errors and the geometric relation errors are both related to the external references of the radar to be calibrated, and further the optimization errors and the geometric relation errors can obtain the external references to be selected, therefore, the external reference calibration method can calibrate a plurality of external references of the radar, can obtain a high-precision external reference result, and provides a foundation for navigation and obstacle avoidance of the robot.

Additional aspects and advantages of embodiments of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of embodiments of the present application.

Drawings

The above and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a schematic flow diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 2 is a block schematic diagram of an intelligent robot according to some embodiments of the present application;

FIG. 3 is a block schematic diagram of an external reference calibration apparatus according to some embodiments of the present application;

FIG. 4 is a schematic structural diagram of an intelligent robot according to some embodiments of the present application;

FIG. 5 is a schematic structural diagram of an intelligent robot according to some embodiments of the present application;

FIG. 6 is a schematic structural diagram of an intelligent robot according to some embodiments of the present application;

FIG. 7 is a schematic view of a scenario of an external reference calibration method according to some embodiments of the present application;

FIG. 8 is a schematic view of a scenario of an external reference calibration method according to some embodiments of the present application;

FIG. 9 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 10 is a schematic view of a scenario of an external reference calibration method according to some embodiments of the present application;

FIG. 11 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 12 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 13 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 14 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 15 is a schematic view of a scenario of an external reference calibration method according to some embodiments of the present application;

FIG. 16 is a schematic view of a scenario of an external reference calibration method according to some embodiments of the present application;

FIG. 17 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 18 is a schematic flow chart diagram of an external reference calibration method according to some embodiments of the present application;

FIG. 19 is a schematic diagram of a connection between a computer-readable storage medium and a processor according to some embodiments of the present application.

Detailed Description

Embodiments of the present application will be further described below with reference to the accompanying drawings. The same or similar reference numbers in the drawings identify the same or similar elements or elements having the same or similar functionality throughout.

In addition, the embodiments of the present application described below in conjunction with the accompanying drawings are exemplary and are only for the purpose of explaining the embodiments of the present application, and are not to be construed as limiting the present application.

In this application, unless expressly stated or limited otherwise, the first feature "on" or "under" the second feature may be directly contacting the first and second features or indirectly contacting the first and second features through intervening media. Also, a first feature "on," "over," and "above" a second feature may be directly or diagonally above the second feature, or may simply indicate that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature may be directly under or obliquely under the first feature, or may simply mean that the first feature is at a lesser elevation than the second feature.

Description of the main element symbols:

the intelligent robot comprises an intelligent robot 100, a processor 10, a memory 20, a communication interface 30, a radar to be calibrated 40, a reference radar 50, an external reference calibration device 200, a first acquisition module 210, a second acquisition module 220, a third acquisition module 230, a calculation module 240, a judgment module 250, a determination module 260, a computer-readable storage medium 300, computer-executable instructions 301 and a processor 400.

Referring to fig. 1 and fig. 2, an external reference calibration method according to an embodiment of the present application is used for performing external reference calibration for a radar, and the external reference calibration method includes the steps of:

s010: acquiring point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to external parameters of the radar 40 to be calibrated;

s020: acquiring fitting errors of the reference surface, the first plane and the second plane according to the point cloud;

s030: acquiring a geometric relation error between the reference plane, the first plane and the second plane according to the point cloud;

s040: optimizing the fitting error and the geometric relation error to obtain external parameters to be selected of the radar 40 to be calibrated;

s050: judging whether the difference between the height of the point cloud of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not; and

s060: if yes, determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated 40.

The intelligent robot 100 of the present embodiment includes one or more processors 10, a memory 20, and one or more programs, where the one or more programs are stored in the memory 20 and executed by the one or more processors 10, the programs including instructions for performing the extrinsic calibration method of the present embodiment. When the processor 10 executes the program, the processor 10 may be configured to perform step S010, step S020, step S030, step S040, step S050, and step S060, that is, the processor 10 may be configured to: acquiring point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to external parameters of the radar 40 to be calibrated; acquiring fitting errors of the reference surface, the first plane and the second plane according to the point cloud; acquiring a geometric relation error between the reference plane, the first plane and the second plane according to the point cloud; optimizing the fitting error and the geometric relation error to obtain external parameters to be selected of the radar 40 to be calibrated; judging whether the difference between the height of the point cloud of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not; and if so, determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated 40.

Referring to fig. 3, the external parameter calibration apparatus 200 according to the embodiment of the present disclosure includes a first obtaining module 210, a second obtaining module 220, a third obtaining module 230, a calculating module 240, a determining module 250, and a determining module 260, where the first obtaining module 210, the second obtaining module 220, the third obtaining module 230, the calculating module 240, the determining module 250, and the determining module 260 may respectively perform steps S010, S020, S030, S040, S050, and S060, that is, the first obtaining module 210 may be configured to obtain point clouds on a reference plane, a first plane, and a second plane that are not parallel to each other and are related to external parameters of the radar 40 to be calibrated; the second obtaining module 220 may be configured to obtain a fitting error of the reference plane, the first plane, and the second plane according to the point cloud; the third obtaining module 230 may be configured to obtain a geometric relationship error between the reference plane, the first plane, and the second plane according to the point cloud; the calculation module 240 may be configured to optimize the fitting error and the geometric relationship error to obtain a candidate external parameter of the radar 40 to be calibrated; the judging module 250 may be configured to judge whether a difference between a point cloud height of a reference surface under the candidate outlier and a reference height of the reference surface is within a preset range; and the determining module 260 may be configured to determine the external parameter to be selected as the external parameter calibrated by the radar 40 to be calibrated when the determination result of the determining module 250 is yes.

In the external reference calibration method, the external reference calibration device 200 and the intelligent robot 100 according to the embodiment of the present application, the point clouds of the external references of the radar 40 to be calibrated on the reference plane, the first plane and the second plane which are not parallel to each other are first obtained, the fitting errors of the reference plane, the first plane and the second plane are obtained according to the point clouds, the geometric relationship error between the reference plane, the first plane and the second plane is obtained according to the point clouds, then the fitting error and the geometric relationship error are optimized to obtain the external reference to be selected, the height of the point cloud of the reference plane under the external reference to be selected is judged whether the difference value between the height of the point cloud of the reference plane and the height of the reference plane is within a preset range, if yes, the external reference to be selected is determined to be the external reference calibrated of the radar 40 to be calibrated, the point clouds related to the radar 40 to be calibrated on the three planes are obtained, and the fitting error and the geometric relationship error are both related to the external reference of the radar, further, the external reference to be selected can be obtained by optimizing the fitting error and the geometric relation error, and therefore, the external reference calibration method can automatically calibrate a plurality of external references of the radar, can obtain a high-precision external reference result, and provides a basis for precise navigation and obstacle avoidance of the intelligent robot 100.

The intelligent robot 100 may be an industrial robot, an agricultural robot, a home robot, a service robot, a cleaning robot, etc., which is not limited herein. Further, the cleaning robot may be an intelligent robot 100 such as a sweeper, a scrubber, a vacuum cleaner, etc. The intelligent robot 100 may also include elements such as a communication interface 30, a cleaning implement, and the like. The intelligent robot 100 may be used to clean surfaces such as floors, floor tiles, pavements, or cement grounds. The radar may be a laser radar, a microwave radar, a millimeter wave radar, etc., which is not limited herein.

Further, in the embodiment of the present application, the mounting carrier of the radar 40 to be calibrated is taken as an example for the intelligent robot 100 to perform explanation, and it is understood that the mounting carrier of the radar 40 to be calibrated may be other, and is not limited herein. Meanwhile, the radar 40 to be calibrated is functionally illustrated as a laser radar, and the radar 40 to be calibrated may be other types of radars, which is not limited herein. The radar 40 to be calibrated is installed on the intelligent robot 100, and can be used for the aspects of map building, positioning, navigation, obstacle avoidance and the like of the intelligent robot 100.

Specifically, referring to fig. 1 again, in step S010, point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to external references of the radar 40 to be calibrated are obtained, wherein the reference plane, the first plane and the second plane are not parallel to each other, and the reference plane, the first plane and the second plane are crossed in pairs, namely the reference plane is intersected with the first plane, the reference plane is intersected with the second plane, the first plane is intersected with the second plane, wherein, the point cloud related to the external parameters of the radar 40 to be calibrated can be the point cloud acquired by the radar 40 to be calibrated under the corresponding external parameters, namely, the point cloud related to the external reference of the radar 40 to be calibrated on the reference plane, the point cloud related to the external reference of the radar 40 to be calibrated on the first plane, the point cloud related to the radar 40 to be calibrated on the second plane, the point cloud obtained in this way can provide support for subsequently obtaining the optimal external reference of the radar 40 to be calibrated.

Further, in step S020, fitting errors of the reference plane, the first plane and the second plane are obtained according to the point cloud, the point clouds on the reference plane, the first plane and the second plane of the point cloud obtained in step S010 are respectively fitted, and the reference plane, the first plane and the second plane are fitted through the obtained point cloud, so that errors exist between the fitted reference plane, the first plane and the second plane and an actual reference plane, the actual first plane and the actual second plane respectively, and therefore, the fitting errors of the reference plane, the first plane and the second plane can be obtained through calculation and the like according to the point cloud, and support is further provided for obtaining the optimal external reference of the radar 40 to be calibrated subsequently.

Further, in step S030, a geometric relationship error between the reference plane, the first plane, and the second plane is obtained according to the point cloud, and the reference plane, the first plane, and the second plane are not parallel to each other, and the reference plane, the first plane, and the second plane intersect with each other, so that a geometric relationship exists between the reference plane and the first plane, a geometric relationship exists between the first plane and the second plane, and a geometric relationship exists between the reference plane and the second plane, so that a geometric relationship error exists between the reference plane, the first plane, and the second plane, and since the obtained point cloud is on the reference plane, the first plane, and the second plane, the point cloud can be obtained by the point cloud: the geometric relationship error between the reference plane and the first plane, the geometric relationship error between the first plane and the second plane, and the geometric relationship error between the reference plane and the second plane further provide support for subsequently obtaining the optimal external reference of the radar 40 to be calibrated.

In step S040, the fitting error and the geometric relation error are optimized to obtain a candidate external reference of the radar 40 to be calibrated, in steps S020 and S030, the fitting error and the geometric relation error are respectively obtained, and are both obtained through point clouds, and the point clouds are related to the external reference of the radar 40 to be calibrated, so that the fitting error and the geometric relation error can be solved through optimizing the fitting error and the geometric relation error to obtain the candidate external reference of the radar 40 to be calibrated, and the fitting error and the geometric relation error corresponding to the candidate external reference are minimum, so that the radar 40 to be calibrated has a small error when using the candidate external reference, and can accurately identify obstacles, build images and other functions.

In step S050, judging whether the difference value between the point cloud height of the reference surface under the candidate external reference and the reference height of the reference surface is in a preset range or not, in step S040, the candidate outlier is obtained, the radar 40 to be calibrated detects the point cloud of the reference surface under the candidate outlier, by judging whether the difference between the detected point cloud height and the reference height of the reference surface is within a preset range, so as to determine whether the error of the candidate external reference is within the error acceptable by the user and other groups, if the difference between the point cloud height of the reference surface under the candidate external reference and the reference height of the reference surface is within the preset range, namely, step S060 is executed to determine that the external parameter to be selected is the external parameter calibrated by the radar 40 to be calibrated, therefore, the external reference of the radar 40 to be calibrated is more accurate, and the radar 40 to be calibrated can more accurately execute the functions of identifying obstacles, building images and the like.

Referring to fig. 4 to 6, in some embodiments, the intelligent robot 100 is installed with a radar 40 to be calibrated and a reference radar 50, where the radar 40 to be calibrated and the reference radar 50 are both in a reference coordinate system, the reference coordinate system includes X, Y, Z three directions, and an external reference of the reference radar 50 in the reference coordinate system is fixed, where the radar 40 to be calibrated includes a plurality of external references, specifically, the external reference of the radar 40 to be calibrated includes an installation position z, a pitch angle pitch, a roll angle roll, a heading angle yaw, an offset X of the radar 40 to be calibrated and the reference radar 50 in an X-axis direction, an offset Y of the radar 40 to be calibrated and the reference radar 50 in a Y-axis direction, and by obtaining the offset X of the radar 40 to be calibrated and the reference radar 50 in the X-axis direction, the offset Y of the radar 40 to be calibrated and the reference radar 50 in the Y-axis direction can obtain installation positions of the radar 40 to be calibrated and the Y-axis, the embodiment is used for calibrating the installation position (x, y, z) of the radar 40 to be calibrated and the pitch angle pitch, roll angle roll and course angle yaw of the radar 40 to be calibrated, and finally obtaining the optimal installation position x, y, z, pitch angle pitch, roll angle roll and course angle yaw, so that the radar 40 to be calibrated can more accurately realize the functions of identifying obstacles, building images and the like. Of course, in addition to the mounting positions x, y, z and the pitch, roll and yaw angles pitch, yaw may also be calibrated.

In some embodiments, the external reference calibration apparatus 200 may be used to calibrate the mounting position x, y, z and the pitch angle pitch, roll angle roll, and heading angle yaw of the radar 40 to be calibrated, and the processor 10 may be used to calibrate the mounting position x, y, z and the pitch angle pitch, roll angle roll, and heading angle yaw of the radar 40 to be calibrated.

Referring to fig. 7, in some embodiments, the intelligent robot 100 moves to obtain a point cloud on a reference surface, when the intelligent robot 100 obtains the point cloud on the reference surface, no obstacle should be kept on the reference surface within a preset range centered on the intelligent robot 100, and the reference surface should be flat, wherein the reference surface may be a ground surface, a wall surface, or a plate surface, and may also be a plane constructed by a user, which is not limited herein. After the point cloud on the reference plane is acquired, the intelligent robot 100 may drive to a preset distance where two planes intersect, for example, 1 meter, 0.5 meter, 1.5 meter, 1.6 meter, etc. away from the intersection line of the two planes, without limitation, and the radar 40 to be calibrated starts to acquire the point cloud on the two intersecting planes.

Referring again to fig. 7 to 10, in some embodiments, step S010 includes:

step S011: acquiring a first point cloud D1 on a reference plane M, a second point cloud D2 on a first plane M1 and a third point cloud D3 on a second plane M2, which are detected by a radar 40 to be calibrated, wherein the first point cloud D1, the second point cloud D2 and the third point cloud D3 are all related to external parameters of the radar 40 to be calibrated; and

step S012: the fourth point cloud D4 on the first plane M1, the fifth point cloud D5 on the second plane M2, the fourth point cloud D4 and the fifth point cloud D5 detected by the reference radar 50 are acquired to be related to the external parameters of the reference radar 50.

Further, the order of execution of step S011 and step S012 is not limited herein, and step S011 and step S012 may be performed simultaneously, step S011 may be performed first, step S012 is performed later, step S012 may be performed first, and step S011 is performed later, which is not limited herein.

Specifically, the intelligent robot 100 moves relative to the reference plane M, the radar 40 to be calibrated detects a first point cloud D1 on the reference plane M, after the radar 40 to be calibrated moves to a preset distance at the intersection position of the first plane M1 and the second plane M2 when the intelligent robot 100 moves, the radar 40 to be calibrated detects a second point cloud D2 on the first plane M1 through the current external reference, and detects a third point cloud D3 on the second plane M2, wherein the number of the first point cloud D1 is at least three, the number of the second point cloud D2 and the number of the third point cloud D3 are at least two, and the first point cloud D1, the second point cloud D2 and the third point cloud D3 are all related to the external reference of the radar 40 to be calibrated. When the intelligent robot 100 moves to the position where the first plane M1 and the second plane M2 intersect, the reference radar 50 detects a fourth point cloud D4 on the first plane M1 and a fifth point cloud D5 on the second plane M2 according to the external parameters of the reference radar 50, wherein the number of the fourth point cloud D4 and the fifth point cloud D5 is at least 2, and the fourth point cloud D4 and the fifth point cloud D5 are both related to the external parameters of the reference radar 50.

In this embodiment, the radar 40 to be calibrated respectively detects point clouds on the reference plane M, the first plane M1, and the second plane M2 according to the current external parameters, the reference radar 50 respectively detects point clouds on the first plane M1 and the second plane M2 according to the current external parameters, and the point clouds respectively detected by the radar 40 to be calibrated and the reference radar 50 are obtained, so that the point clouds detected by the reference radar 50 can provide references for the radar 40 to be calibrated, and the accuracy of the external parameters to be selected of the subsequent radar 40 to be calibrated is enhanced. Meanwhile, the radar 40 to be calibrated and the reference radar 50 respectively acquire a plurality of point clouds on the first plane M1 and the second plane M2, and the point clouds are related to the external parameters of the radar 40 to be calibrated, so that data can be provided for subsequently acquiring the external parameters of the radar 40 to be calibrated, and the subsequent calculation can obtain accurate external parameters.

In one embodiment, the reference plane M is a ground plane, the first plane M1 is a wall plane, and the second plane M2 is another wall plane, where the two wall planes intersect to form a corner, and both the two wall planes intersect with the ground plane, it can be understood that the intelligent robot 100 first drives on the ground plane, detects the point cloud on the ground plane, then moves to a preset distance from the corner, the radar 40 to be calibrated detects the point cloud of the two wall planes, and the reference radar 50 detects the point cloud of the two wall planes. The intelligent robot 100 may be located inside the corner (as shown in fig. 8) or outside the corner (as shown in fig. 10), which is not limited herein.

In some embodiments, the first obtaining module 210 can be further configured to perform steps S011 and S012, and the processor 10 can be further configured to perform steps S011 and S012.

Referring to fig. 11, in some embodiments, step S020 includes the steps of:

s021: converting the first point cloud D1, the second point cloud D2 and the third point cloud D3 from the radar coordinate system to be calibrated to a reference radar coordinate system;

s022: acquiring a reference surface M expression of a reference surface M according to the first point cloud D1;

s023: acquiring a first plane M1 expression of a first plane M1 according to the second point cloud D2 and the fourth point cloud D4;

s024: acquiring a second plane M2 expression of a second plane M2 according to the third point cloud D3 and the fifth point cloud D5; and

s025: and respectively acquiring the fitting errors of the reference plane M, the first plane M1 and the second plane M2 according to the expression of the reference plane M, the expression of the first plane M1 and the expression of the second plane M2.

Specifically, because the first point cloud D1, the second point cloud D2, and the third point cloud D3 are detected by the radar 40 to be calibrated, and the first point cloud D1, the second point cloud D2, and the third point cloud D3 are all generated in the radar coordinate system to be calibrated, in order to unify data in the calculation process, the first point cloud D1, the second point cloud D2, and the third point cloud D3 are converted from the radar coordinate system to be calibrated to the reference radar coordinate system, so that data differences caused by different coordinate systems can be avoided, and thus obtained external parameters are inaccurate.

In one embodiment, the external parameters of the radar 40 to be calibrated include the installation position (x, y, z), the pitch angle pitch, the roll angle, and the heading angle yaw of the radar 40 to be calibrated, and the coordinates of the point cloud in the coordinate system of the radar to be calibrated are

Wherein the coordinate transformation formula is as follows:

wherein

For the coordinates of the point cloud under the radar 40 to be calibrated under the reference radar coordinate system, further:

，

，

wherein:

and

are all expressed in vector form; t represents vector transposition; r is a change matrix formed by external parameters under a reference radar coordinate system and consists of raw, pitch and roll in the external parameters of the radar 40 to be calibrated; t is a translation matrix formed by external parameters under a reference radar coordinate system and consists of mounting positions x, y and z in the external parameters of the radar 40 to be calibrated.

The first point cloud D1, the second point cloud D2, and the third point cloud D3 can be converted from the radar coordinate system to be calibrated to the reference radar coordinate system by a coordinate conversion formula, and further, the coordinates of the first point cloud D1, the second point cloud D2, and the third point cloud D3 in the reference radar coordinate system include the relevant external parameters of the radar 40 to be calibrated.

Further, the first point cloud D1 is detected by the radar 40 to be calibrated when the intelligent robot 100 moves relative to the reference plane M, so that the reference plane M expression of the reference plane M can be obtained by fitting the first point cloud D1, specifically, by detecting a plurality of first point clouds D1 on the reference plane M, by fitting some of the first point clouds D1 into one line segment, by fitting another part of the point clouds into another line segment, and by fitting two line segments, the reference plane M expression of the reference plane M is obtained. Further, a second point cloud D2 detected on the first plane M1 by the radar 40 to be calibrated is fitted into a first line segment, a fourth point cloud D4 detected on the first plane M1 by the reference radar 50 is fitted into a second line segment, and a first plane M1 expression of the first plane M1 is obtained through the fitting of the first line segment and the second line segment. Further, a third point cloud D3 on a second plane M2 detected by the radar 40 to be calibrated is fitted to form a third line segment, a fifth point cloud D5 of a second plane M2 detected by the reference radar 50 is fitted to form a fourth line segment, and a second plane M2 expression of the second plane M2 is obtained through the third line segment and the fourth line segment.

Certainly, the coordinate of at least three first point clouds D1 on the reference surface M may be directly solved to obtain an expression of the reference surface M, the coordinates of the second point cloud D2 and the fourth point cloud D4 are combined to obtain a first expression of the reference surface M1, the coordinates of the third point cloud D3 and the fourth point cloud D4 are used to obtain an expression of the second plane M2, the minimum sum of the second point cloud D2 and the fourth point cloud D4 is three, the minimum sum of the third point cloud D3 and the fifth point cloud D5 is three, and the more the detected first point cloud D1, the second point cloud D2, the third point cloud D3, the detected fourth point cloud D4 and the detected fifth point cloud D5 are, the obtained plane expression is more accurate.

In one embodiment, the reference plane M obtained by fitting the first point cloud D1 is expressed as:

the expression of the first plane M1 obtained by fitting the second point cloud D2 and the fourth point cloud D4 is:

the expression of the second plane M2 obtained by fitting the third point cloud D3 and the fifth point cloud D5 is:

。

because the expression of the reference plane M is obtained by fitting the first point cloud D1, the expression of the first plane M1 is obtained by fitting the second point cloud D2 and the fourth point cloud D4, and the expression of the second plane M2 is obtained by fitting the third point cloud D3 and the fifth point cloud D5, a certain fitting error exists, and thus the fitting errors of the reference plane M, the first plane M1 and the second plane M2 can be obtained by calculation according to the expression of the reference plane M, the expression of the first plane M1 and the expression of the second plane M2. Wherein, the calculation formula of the fitting error is as follows:

wherein the content of the first and second substances,

: referring to coordinate values of point cloud data under a laser radar coordinate system; a, b, c, d are coefficients of a plane, which is generally an expression. Thereby substituting the coefficients respectively corresponding to the expression of the reference plane M, the expression of the first plane M1, and the expression of the second plane M2 acquired in the steps S022, S023, and S024, and the coordinates of the corresponding point cloud into the fitting error

The fitting errors of the reference plane M, the first plane M1 and the second plane M2 can be obtained by the calculation formula (2). Wherein

，

，

Are correlated with external parameters of the radar 40 to be calibrated. Error of fit

Refers to the distance of the corresponding point cloud to the corresponding plane, wherein the fitting error

The closer to zero, the more accurate the fitted planar expression is shown.

Further, the coefficients of the planar expression and the external parameters of the radar 40 to be calibrated may be optimized through the optimization library, that is, a, b, c, d and the external parameters (x, y, z, pitch, roll, yaw) of the radar 40 to be calibrated may be put into the optimization library for optimization, so that the corresponding planar expression may obtain more accurate coefficients, and specifically, the coefficients in the expression of the reference plane M may be optimized

And placing external parameters (x, y, z, pitch, roll and yaw) of the radar 40 to be calibrated into an optimization library for optimization to obtain more accurate parameters

The expression of the first plane M1 can be used

In the expression of the second plane M2

。

Further, the optimization library may be an optimization library such as *** ceres, Nlopt, Snopt, IPOPT, and the like, which is not limited herein. The coefficients a, b, c and d of the expression of the optimal plane can be obtained through optimization by putting data of a plurality of point clouds and the like into an optimization library, so that the obtained fitting error is more accurate, and the finally obtained external parameters to be selected are more accurate. The optimization library is mainly used for solving a plurality of point clouds in a plane in a least square method and other modes to obtain optimal coefficients a, b, c and d of the plane.

In some embodiments, the second obtaining module 220 is further configured to perform steps S021, S022, S023, S024 and S025, and the processor 10 is further configured to perform steps S021, S022, S023, S024 and S025.

Referring to fig. 12, in some embodiments, step S030 includes the steps of:

s031: acquiring a geometric relation error between the reference plane M and the first plane M1 according to the geometric relation between the reference plane M and the first plane M1;

s032: acquiring a geometric relation error between the reference plane M and the second plane M2 according to the geometric relation between the reference plane M and the second plane M2; and

s033: according to the geometric relationship between the first plane M1 and the second plane M2, the geometric relationship error of the first plane M1 and the second plane M2 is obtained.

Specifically, referring to fig. 7, since the reference plane M intersects the first plane M1, an included angle θ 1 exists between the reference plane M and the first plane M1, meanwhile, the expression of the reference plane M, the expression of the first plane M1 and the expression of the second plane M2 are acquired in step S020, therefore, the geometric relation error of the reference plane M and the first plane M1 can be calculated according to the included angle theta 1 between the reference plane M and the first plane M1, and similarly, the included angle theta 2 exists between the reference plane M and the second plane M2, therefore, the geometric relationship error of the reference plane M and the second plane M2 can be calculated according to the included angle theta 2 between the reference plane M and the first plane M1, theta 3 exists between the first plane M1 and the second plane M2, therefore, the geometric relationship error of the first plane M1 and the second plane M2 can be calculated according to the included angle theta 3 between the first plane M1 and the second plane M2. Wherein the corresponding calculation formula can be obtained through the geometrical relationship:

further, the geometric relationship error between the two planes is obtained:

where θ is the angle between the two planes, A₁、B₁、C₁Coefficient of a planar expression, A₂、B₂、C₂Is a coefficient of another plane expression, i.e.

Through the included angle θ 1 between the reference plane M and the first plane M1, the geometric relationship error between the reference plane M and the first plane M1 can be obtained:

through the included angle θ 2 between the reference plane M and the second plane M2, the geometric relationship error between the reference plane M and the second plane M2 can be obtained:

through the included angle θ 3 between the first plane M1 and the second plane M2, the geometric relationship error between the first plane M1 and the second plane M2 can be obtained:

wherein when

The closer the value of (A) is to zero, the smaller the error of the geometric relationship between the two planes is, and the closer the fitted geometric relationship between the two planes is to the actual geometric relationship in which the two planes intersect in practice.

In one embodiment, the reference plane M, the first plane M1 and the second plane M2 are perpendicular to each other, so that the geometric relationship error between the reference plane M and the first plane M1 when θ 1, θ 2 and θ 3 are all 90 °, i.e. cos θ 1=0, cos θ 2=0 and cos θ 3=0, i.e. the reference plane M, the first plane M1 and the second plane M2 are perpendicular to each other can be obtained:

geometric relationship error of the reference plane M and the second plane M2:

geometric relationship error of the first plane M1 and the second plane M2:

。

in some embodiments, the third acquisition module 230 can also be used to perform step S031, step S032 and step S033, and the processor 10 can also be used to perform step S031, step S032 and step S033.

Further, referring again to FIG. 7, in some embodiments, the angle θ 1 between the reference plane M and the first plane M1 is [10 °,170 ° ], and/or the included angle theta 2 between the reference plane M and the second plane M2 is [10 degrees, 170 degrees ], and/or the angle θ 3 between the first plane M1 and the second plane M2 is [10 °,170 ° ], specifically, the angle θ 1 between the reference plane M and the first plane M1 may be 10 °, 30 °, 50 °, 60 °, 90 °, 110 °, 120 °, 150 °,170 °, and the like, the angle θ 2 between the reference plane M and the second plane M2 may be 10 °, 30 °, 50 °, 60 °, 90 °, 110 °, 120 °, 150 °,170 °, and the angle θ 3 between the first plane M1 and the second plane M2 may be 10 °, 30 °, 50 °, 60 °, 90 °, 110 °, 120 °, 150 °,170 °, without limitation.

In this embodiment, an included angle θ 1 between the reference plane M and the first plane M1 is [10 °,170 ° ], and/or an included angle θ 2 between the reference plane M and the second plane M2 is [10 °,170 ° ], and/or an included angle θ 3 between the first plane M1 and the second plane M2 is [10 °,170 ° ], which can avoid that when an included angle θ between each two of the reference plane M, the first plane M1, and the second plane M2 is too small or too large, the two planes are close to coincide with each other, so that the constraint between the two planes is weak, and the two planes may be fitted into one plane in the fitting process, so that the obtained external reference is inaccurate.

Referring to fig. 13 and 14, in some embodiments, step S040 includes the steps of:

s041: establishing an optimal equation according to the fitting error and the geometric relation error;

s042: along the gradient descending direction of the optimization equation, generating a group of test external parameters from the initial value of the external parameters by increasing a preset step length every time to obtain a plurality of groups of test external parameters;

s043: acquiring gradients of an optimization equation under a plurality of groups of test external parameters, and taking the current test external parameter as a candidate external parameter when a difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is smaller than a preset difference value; or

S044: and when the number of the test external parameters is equal to or greater than the preset number, taking the last group of test external parameters as the external parameters to be selected.

Specifically, step S041, step S042 and step S043 may be executed, as shown in fig. 13, or step S041, step S042 and step S044 may be executed, as shown in fig. 14, which is not limited herein. In step S020, fitting errors of the reference plane M, the first plane M1 and the second plane M2 are obtained, in step S030, a geometric relationship error between each two of the reference plane M, the first plane M1 and the second plane M2 is obtained, and both the obtained fitting error and the obtained geometric relationship error are related to external parameters of the radar 40 to be calibrated, so that an optimization equation can be established according to the fitting error and the geometric relationship error, so that the optimization equation can be solved to obtain the external parameters to be selected.

Wherein, in one embodiment, the optimization equation is established as:

(x,y,z,yaw,pitch,roll）=argmin∑（|ei|+|ej|）

wherein, x, y, z, yaw, pitch, roll: referring to the current external reference under a laser radar coordinate system;

is the error of the fit, and is,

in order to be a geometric relationship error,

represents the sum of fitting errors of the reference plane M, the first plane M1 and the second plane M2,

is the sum of the geometric relationship errors between the reference plane M, the first plane M1 and the second plane M2, wherein argmin is the calculation

Of the radar 40 to be calibrated, it is understood that there is one for each external parameter of the radar 40 to be calibrated

When is coming into contact with

When the minimum value is obtained, the fitting error and the geometric relation error obtained at the moment are added to be minimum, and when the corresponding external parameter is used for the radar to be calibrated 40, the error of the radar to be calibrated 40 is minimum.

Further, in step S042, a group of test external parameters is generated every time a preset step length is added from an initial value of the external parameter along a direction in which a gradient of the optimization equation decreases, so as to obtain a plurality of groups of test external parameters. The smaller the value of the optimization equation is, the more accurate the corresponding external parameter is, so that along the gradient descending direction of the optimization equation, the corresponding external parameter approaches to the optimal value, and therefore, a group of test external parameters are generated every time a preset step length is added along the initial value of the external parameter, and finally, a plurality of groups of test external parameters are obtained, wherein the preset step length can be calculated through a step length optimizer.

Further, since the gradient is gradually decreased, since

Is zero, so that as the gradient approaches zero, it indicates that the corresponding external parameter of the gradient is more accurate. And calculating the gradient of the optimization equation at the test external parameter every time a group of test external parameters are obtained, wherein the test external parameter is generated along the descending direction of the gradient of the optimization equation, so that a difference value exists between the gradient values of the optimization equation between every two groups of adjacent test external parameters, and when the difference value of the corresponding gradient of the two groups of adjacent test external parameters in the optimization equation is smaller, the gradient of the optimization equation gradually tends to a stable value, which indicates that the change of the gradient of the optimization equation approaches to zero and the corresponding test external parameter is more accurate. In step S043, when the gradient of the optimization equation corresponding to the current test external parameter and the gradient of the optimization equation corresponding to the previous group of test external parameters are smaller than the predetermined difference, the current test external parameter is taken as the external parameter to be selected, and thus, the external parameter to be selected is the more accurate external parameter of the radar 40 to be calibrated.

Specifically, in step S044, the test external parameters are generated in the direction in which the gradient of the optimization equation decreases, it can be understood that the greater the number of the test external parameters, the smaller the final gradient of the optimization equation, and when the number of the test external parameters is sufficiently large, the final gradient of the optimization equation will approach zero infinitely, so that by setting the predetermined number of the test external parameters, when the number of the test external parameters is less than the predetermined number, it is indicated that the final corresponding external parameters are not the most accurate external parameters, and when the number of the test external parameters is equal to or greater than the predetermined number, it is indicated that the final corresponding external parameters are more accurate, and therefore, the last group of the test external parameters is taken as the external parameters to be selected, so that the calibrated external parameters of the radar to be calibrated 40 are more accurate.

In some embodiments, the computing module 240 may be further configured to perform steps S041, S042 and S043, and the processor 10 may be further configured to perform steps S041, S042 and S043; the calculation module 240 may be further configured to perform steps S041, S042 and S044, and the processor 10 may be further configured to perform steps S041, S042 and S044.

Of course, in step S040, the manner of determining the candidate outlier through the optimization equation further includes other manners, for example, when the gradient of the optimization equation is equal to the gradient threshold, the test outlier corresponding to the gradient threshold is taken as the candidate outlier, and when the gradient of the optimization equation is the gradient threshold, it indicates that the corresponding test outlier at this time is more accurate. The gradient threshold may be set by the user, or may be an empirical value obtained through multiple calculation processes, for example, the gradient threshold may be 0, 0.01, 0.011, 0.02, and the like, which is not limited herein. The computing module 240 may also be used to execute the present embodiment, and the processor may also be used in the present embodiment.

Referring to fig. 1 and fig. 7 again, the extrinsic parameters of the radar 40 to be calibrated are six extrinsic parameters, i.e., an installation position x, y, z, a pitch angle pitch, a roll angle roll, and a heading angle yaw of the radar 40 to be calibrated, and the extrinsic parameter to be selected of the radar 40 to be calibrated is obtained in step S040, further, in step S050, the height of the point cloud of the reference plane M under the extrinsic parameter to be selected is determined, and the difference between the height of the point cloud of the reference plane M and the height of the reference plane M is within a preset range, where the point cloud on the reference plane M is the point cloud obtained by the radar 40 to be calibrated according to the detection reference plane M of the extrinsic parameter to be selected, and the calculation formula for calculating the height of the point cloud is:

h_i=z-q_xisin(pitch)+q_yicos(pitch)sin(roll)

，

to calibrate the coordinate values of the ground point in the lidar coordinate system,

z, pitch, roll is a to-be-selected external reference which is calibrated under a reference laser radar coordinate system;

by substituting the coordinates of a plurality of point clouds on the reference plane M detected by the radar 40 to be calibrated

In the calculation formula (2), a plurality of point clouds are searched

And determining whether a difference between the maximum value and a reference height of a reference plane M is within a preset range, wherein the reference height is an objective height of the reference plane M, for example, when the reference plane M is a ground, the reference height is zero, and if the candidate is accurate, the difference should be close to zero. The preset range is an acceptable error range, and may be 0.01 m, 0.02 m, 0.03 m, and the like, and specific numerical values are not limited herein. The smaller the difference value is, the more accurate the obtained external parameters to be selected are.

The external reference calibration apparatus 200 may also be used to implement the above embodiments, and the processor 10 may also be used to implement the above embodiments.

In some embodiments, if the determination result in step S050 is negative, it is determined that the difference between the height of the point cloud of the reference plane M under the candidate outlier and the reference height of the reference plane M is not within the preset range, and step S010, step S020, step S030, step S040, step S050, and step S060 are executed again until the calibrated outlier of the radar 40 to be calibrated is determined. The radar 40 to be calibrated acquires the point clouds on the reference plane M, the first plane M1 and the second plane M2 which are not parallel to each other and are related to the external parameters of the radar 40 to be calibrated again, and the acquired point clouds should be different from the previously acquired point clouds. The external reference calibration apparatus 200 may also be used to implement the above embodiments, and the processor 10 may also be used to implement the above embodiments.

Referring to fig. 15 to 17, in some embodiments, when the radar 40 to be calibrated can rotate around the rotation axis H relative to the mounting carrier of the radar 40 to be calibrated, the external reference calibration method further includes the steps of:

s001 setting an angular interval α to form at least one calibration area Y depending on an angular range β of rotation of the radar 40 to be calibrated about the rotation axis H, and

s002: and performing external reference calibration at least once in each calibration area Y for the radar 40 to be calibrated.

Specifically, taking an installation carrier of the radar 40 to be calibrated as the intelligent robot 100 for example to perform functional explanation, the radar 40 to be calibrated is installed on the intelligent robot 100, and the installation position of the radar 40 to be calibrated on the intelligent robot 100 is not fixed, and the external parameters of the radar 40 to be calibrated are not the same at different installation positions, so calibration is required for the external parameters at different installation positions.

Further, please refer to fig. 15 and 16, an angle range in which the radar 40 to be calibrated can rotate relative to the intelligent robot 100 is β, where the angle range β may be 60 °, 90 °, 120 °, 140 °, 180 °, 210 °, 260 °, 300 °, 360 °, and the like, and is not limited herein, according to the set angle interval α, the angle range β is divided into a plurality of calibration regions Y, and finally, the radar 40 to be calibrated performs at least one external reference calibration in each calibration region Y, so that the radar 40 to be calibrated can constantly maintain appropriate external reference data during the rotation process, which is beneficial for avoiding obstacles and building images and the like for the radar 40 to be calibrated, where the angle interval α may be set by a user according to a mechanical structure condition of a mounting carrier of the radar 40 to be calibrated, the angle interval α may also be determined according to the angle range β, where the angle interval α is restricted by the driving motor, and when the angle interval α is smaller, the requirement for the driving motor is higher, where the angle interval α may be 5 °, 10 °, 30 °, 120 °, 180 °, 360 °, and the like.

In one embodiment, the radar 40 to be calibrated can rotate 140 ° around the rotation axis H, wherein the angle interval α is [5 °, 140 ° ], and when the angle interval α is 5 °, the number of times N =140 °/5 ° =28 that the radar 40 to be calibrated needs to perform at least calibration is N =140 °/140 ° =28, when the angle interval α is 140 °, the number of times N =140 °/140 ° =1 that the radar 40 to be calibrated needs to perform at least calibration is not large, wherein when the angle interval α is smaller than 5 °, the external reference change when the radar 40 to be calibrated rotates by one angle interval α is not large, and the smaller angle interval α is, the larger number of times is calculated when the test height is calculated, the longer time is needed, the working efficiency is reduced, and therefore, the angle gap between [5 °, 140 ° ] can achieve better effect, and the calculation amount is smaller.

In some embodiments, the external reference calibration apparatus 200 may be further configured to perform steps S001 and S002, and the processor 10 may be further configured to perform steps S001 and S002.

Referring to fig. 18, in some embodiments, when the installation carrier of the radar 40 to be calibrated is the intelligent robot 100 and the reference plane M is the driving surface of the intelligent robot 100, the external reference calibration method further includes the steps of:

s003: and controlling the intelligent robot 100 to move to an area with a smooth surface and without obstacles within a preset range.

Specifically, the reference plane M is a driving surface of the intelligent robot 100, that is, the intelligent robot 100 drives on the reference plane M, and the theoretical reference height of the reference plane M should be zero, and if the reference plane M is the driving surface, if the running surface is uneven, the heights of the first point clouds D1 detected by the radar 40 to be calibrated are greatly different, and the reference plane M expression obtained by fitting is greatly different, therefore, the finally obtained external reference has a large error, which affects the radar 40 to be calibrated to identify the obstacle, further, if the obstacle exists in the preset range of the intelligent robot 100, the first point cloud D1 possibly detected by the radar 40 to be calibrated when detecting the first point cloud D1 of the reference plane M is the point cloud of the obstacle, thereby causing inaccurate external reference results and incapability of correctly identifying the obstacle in subsequent work of the radar 40 to be calibrated. The intelligent robot 100 is moved to a region with a smooth running surface and no obstacle within a preset range, so that the error of calculating the final external parameter can be reduced, and the radar 40 to be calibrated can be favorably used for identifying the obstacle. The preset range may be a range set by a user, or may be a range that can be detected by the radar 40 to be calibrated, which is not limited herein.

In some embodiments, the external reference calibration apparatus 200 may be further configured to perform step S003, and the processor 10 may be further configured to perform step S003.

Referring again to fig. 2, in some embodiments, the memory 20 is used for storing a computer program that can be executed on the processor 10, and the processor 10 executes the computer program to implement the external reference calibration method in any of the above embodiments.

The memory 20 may comprise high-speed RAM memory and may also include non-volatile memory (non-volatile memory), such as at least one disk memory. Further, the intelligent robot 100 may further include a communication interface 30, and the communication interface 30 is used for communication between the memory 20 and the processor 10.

If the memory 20, the processor 10 and the communication interface 30 are implemented independently, the communication interface 30, the memory 20 and the processor 10 may be connected to each other through a bus and perform communication with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (enhanced Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 10, but this is not intended to represent only one bus or type of bus.

Optionally, in a specific implementation, if the memory 20, the processor 10, and the communication interface 30 are integrated on a chip, the memory 20, the processor 10, and the communication interface 30 may complete communication with each other through an internal interface.

The processor 10 may be a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits configured to implement embodiments of the present Application.

Referring to fig. 19, a non-transitory computer-readable storage medium 300 of an embodiment of the present application includes computer-executable instructions 301 that, when executed by one or more processors 400, cause the processors 400 to perform a referencing method of any embodiment of the present application.

For example, when the computer-executable instructions are executed by the processor 400, the processor 400 is configured to perform the steps of:

s010: acquiring point clouds on a reference plane M, a first plane M1 and a second plane M2 which are not parallel to each other and are related to external parameters of the radar 40 to be calibrated;

s020: acquiring fitting errors of the reference plane M, the first plane M1 and the second plane M2 according to the point cloud;

s030: acquiring a geometric relationship error between the reference plane M, the first plane M1 and the second plane M2 according to the point cloud;

s050: judging whether the difference between the point cloud height of the reference surface M under the candidate external reference and the reference height of the reference surface M is within a preset range or not; and

On which a computer program is stored which, when executed by the processor 400, implements the external reference calibration method as described above.

Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.

The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.

It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.

It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.

In addition, functional units in the embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium. The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.

In the description herein, reference to the description of the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the feature. In the description of the present application, "a plurality" means at least two, e.g., two, three, unless specifically limited otherwise.

Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations of the above embodiments may be made by those of ordinary skill in the art within the scope of the present application, which is defined by the claims and their equivalents.

Claims

1. An external reference calibration method of a radar is characterized by comprising the following steps:

acquiring point clouds which are on a reference plane, a first plane and a second plane and are not parallel to each other and are related to external parameters of a radar to be calibrated;

acquiring fitting errors of the reference surface, the first plane and the second plane according to the point cloud;

acquiring a geometric relation error between the reference plane, the first plane and the second plane according to the point cloud;

optimizing the fitting error and the geometric relation error to obtain a candidate external parameter of the radar to be calibrated;

judging whether the difference between the height of the point cloud of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not; and

and if so, determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated.

2. The method for calibrating external parameters according to claim 1, wherein the acquiring point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to the external parameters of the radar to be calibrated comprises:

acquiring a first point cloud on the reference surface, a second point cloud on the first plane and a third point cloud on the second plane, which are detected by the radar to be calibrated, wherein the first point cloud, the second point cloud and the third point cloud are all related to external parameters of the radar to be calibrated; and

and acquiring a fourth point cloud on the first plane and a fifth point cloud on the second plane, which are detected by a reference radar, wherein the fourth point cloud and the fifth point cloud are related to the external parameters of the reference radar.

3. The method according to claim 2, wherein the obtaining of the fitting errors of the reference plane, the first plane and the second plane according to the point cloud comprises:

converting the first point cloud, the second point cloud and the third point cloud from the radar coordinate system to be calibrated to the reference radar coordinate system;

acquiring a reference surface expression of the reference surface according to the first point cloud;

acquiring a first plane expression of the first plane according to the second point cloud and the fourth point cloud;

acquiring a second plane expression of the second plane according to the third point cloud and the fifth point cloud; and

and respectively acquiring the fitting errors of the reference surface, the first plane and the second plane according to the reference surface expression, the first plane expression and the second plane expression.

4. The external reference calibration method according to claim 1, wherein the obtaining of the geometric relationship error between the reference plane, the first plane and the second plane according to the point cloud comprises:

acquiring a geometric relation error between the reference surface and the first plane according to the geometric relation between the reference surface and the first plane;

acquiring a geometric relation error of the reference surface and the second plane according to the geometric relation between the reference surface and the second plane; and

and acquiring the geometric relation error of the first plane and the second plane according to the geometric relation between the first plane and the second plane.

5. The external reference calibration method according to claim 4, wherein the included angle between the first plane and the second plane is [10 degrees, 170 degrees ], and/or the included angle between the reference plane and the first plane is [10 degrees, 170 degrees ], and/or the included angle between the reference plane and the second plane is [10 degrees, 170 degrees ].

6. The method for calibrating external parameters according to claim 1, wherein the optimizing the fitting error and the geometric relationship error to obtain the external parameters to be selected of the radar to be calibrated comprises:

establishing an optimization equation according to the fitting error and the geometric relation error;

along the gradient descending direction of the optimization equation, generating a group of test external parameters from the initial value of the external parameters by increasing a preset step length every time, and obtaining a plurality of groups of test external parameters;

acquiring gradients of an optimization equation under a plurality of groups of test external parameters, and taking the current test external parameter as the external parameter to be selected when a difference value between the gradient corresponding to the current test external parameter and the gradient corresponding to the previous group of test external parameters is smaller than a preset difference value; or

And when the number of the test external parameters is equal to or greater than the preset number, taking the last group of test external parameters as the external parameters to be selected.

7. The external reference calibration method according to claim 1, wherein when the radar to be calibrated is rotatable around a rotation axis relative to the mounting carrier of the radar to be calibrated, the external reference calibration method further comprises:

setting an angle interval according to the angle range of the radar to be calibrated rotating around the rotating shaft to form at least one calibration area; and

and performing external reference calibration for the radar to be calibrated in each calibration area at least once.

8. The external reference calibration method according to claim 1, wherein when the mounting carrier of the radar to be calibrated is an intelligent robot and the reference surface is a driving surface of the intelligent robot, the external reference calibration method further comprises:

and controlling the intelligent robot to move to an area with a smooth running surface, wherein no obstacle exists in a preset range.

9. An external reference calibration device for a radar, comprising:

the first acquisition module is used for acquiring point clouds on a reference plane, a first plane and a second plane which are not parallel to each other and are related to external parameters of the radar to be calibrated;

a second obtaining module, configured to obtain, by the point cloud, fitting errors of the reference plane, the first plane, and the second plane;

a third obtaining module, configured to obtain, according to the point cloud, a geometric relationship error between each two of the reference plane, the first plane, and the second plane;

the calculation module is used for optimizing the fitting error and the geometric relation error so as to obtain a candidate external parameter of the radar to be calibrated;

the judging module is used for judging whether the difference value between the point cloud height of the reference surface under the candidate external reference and the reference height of the reference surface is within a preset range or not;

and the determining module is used for determining the external parameter to be selected as the external parameter calibrated by the radar to be calibrated when the judgment result of the judging module is yes.

10. An intelligent robot, comprising:

one or more processors, memory; and

one or more programs, wherein the one or more programs are stored in the memory and executed by the one or more processors, the programs comprising instructions for performing the external reference calibration method of any of claims 1 to 8.

11. A non-transitory computer-readable storage medium containing computer-executable instructions that, when executed by one or more processors, cause the processors to perform the extrinsic calibration method of any one of claims 1 to 8.