A multi-sensor alignment extrinsic calibration method, device, system and medium

Abstract

The present application relates to the field of sensor calibration, and more particularly to a multi-sensor alignment extrinsic calibration method, device, equipment and medium. According to the odometer sequence collected by different sensors, the target odometer sequence and the to-be-aligned odometer sequence are determined, the to-be-aligned odometer sequence and the target odometer sequence are aligned through the difference value, the aligned odometer sequence is obtained, and the extrinsic calibration is performed according to the target odometer sequence and the aligned odometer sequence. In the present application, different frequency odometer data corresponding to different sensors are converted to the same frequency, in the same frequency, the odometer parameters corresponding to each sensor are determined, the rotation matrix and the translation vector between the corresponding sensors are calculated according to the odometer parameters corresponding to each sensor, and the rotation matrix and the translation vector are used as the rotation matrix and the translation vector in the extrinsic calibration, thereby improving the efficiency of the sensor extrinsic calibration.

Description

Technical Field

[0001] This invention relates to the field of sensor calibration, and in particular to a method, apparatus, system, and medium for calibrating extrinsic parameters of multiple sensors. Background Technology

[0002] Currently, most SLAM (Simultaneous Localization and Mapping) methods are based on multi-sensor approaches, such as LiDAR (Light Detection and Ranging) + IMU (Inertial Measurement Unit) methods, vision + IMU methods, and LiDAR + vision + IMU methods. Multi-sensor fusion can significantly improve the accuracy of mapping and localization, but it first requires calibrating the extrinsic parameters between sensors, i.e., determining the rotational and translational relationships from one sensor to another. In existing technologies, the calibration of rotational matrices representing the rotational relationship and translational matrices representing the translational relationship between sensors is generally achieved through methods such as triangulation and calibration plates. However, these methods require specific experimental environments and equipment, involve complex calibration processes, and consume significant time and effort, resulting in low efficiency in sensor extrinsic parameter calibration. Therefore, improving the efficiency of sensor extrinsic parameter calibration is a pressing issue that needs to be addressed. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a method, apparatus, system and storage medium for extrinsic parameter calibration of multi-sensor alignment, in order to solve the problem of low calibration efficiency when calibrating sensor extrinsic parameters.

[0004] In a first aspect, embodiments of the present invention provide a method for extrinsic parameter calibration of multi-sensor alignment, the method comprising:

[0005] The system acquires a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period. It compares the frequency of the first odometer sequence with the frequency of the second odometer sequence, determines the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned. Each odometer sequence consists of odometer data sorted according to the sampling time, and the odometer data includes angular velocity and linear acceleration.

[0006] Each target sampling time in the target odometer sequence is obtained. Odometer data corresponding to adjacent sampling times of each target sampling time is determined from the odometer sequence to be aligned. For any target sampling time, the odometer interpolation of the target sampling time is calculated based on the odometer data corresponding to adjacent sampling times of the target sampling time. All target sampling times are traversed, and the odometer interpolations corresponding to all target sampling times are arranged to obtain the aligned odometer sequence corresponding to the odometer sequence to be aligned.

[0007] Based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor, as well as the angular acceleration at each target sampling moment in the target odometer sequence and the aligned odometer sequence, are calculated.

[0008] Based on the rotation matrix, the angular and linear accelerations in the target odometer sequence, and the angular and linear accelerations in the aligned odometer sequence, a translation vector characterizing the translational relationship between the first sensor and the second sensor is calculated.

[0009] Secondly, embodiments of the present invention provide an extrinsic parameter calibration device for multi-sensor alignment, the extrinsic parameter calibration device comprising:

[0010] The acquisition module is used to acquire a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period, compare the frequency of the first odometer sequence and the frequency of the second odometer sequence, determine the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned, wherein each odometer sequence is composed of odometer data sorted according to the sampling time, and the odometer data includes angular velocity and linear acceleration;

[0011] An interpolation module is used to acquire each target sampling time in the target odometer sequence, determine the odometer data corresponding to the adjacent sampling times of each target sampling time from the odometer sequence to be aligned, calculate the odometer interpolation value of the target sampling time based on the odometer data corresponding to the adjacent sampling times of the target sampling time for any target sampling time, traverse all target sampling times, and arrange the odometer interpolation values ​​corresponding to all target sampling times to obtain the aligned odometer sequence corresponding to the odometer sequence to be aligned;

[0012] The first calculation module is used to calculate, based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor, and the angular acceleration at each target sampling time in the target odometer sequence and the aligned odometer sequence;

[0013] The second calculation module is used to calculate a translation vector characterizing the translation relationship between the first sensor and the second sensor based on the rotation matrix, the angular acceleration and linear acceleration in the target odometer sequence, and the angular acceleration and linear acceleration in the aligned odometer sequence.

[0014] Thirdly, embodiments of the present invention provide a multi-sensor aligned extrinsic parameter calibration system, including a computer device, a first sensor, and a second sensor. The computer device is connected to the first sensor and the second sensor and receives the odometer sequence sent by the first sensor and the second sensor. The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the multi-sensor aligned extrinsic parameter calibration method as described in the first aspect.

[0015] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the extrinsic parameter calibration method for multi-sensor alignment as described in the first aspect.

[0016] The advantages of this invention compared to the prior art are:

[0017] The process involves acquiring a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period. The frequencies of the first and second odometer sequences are compared, and the odometer data with the lower frequency is identified as the target odometer sequence, while the odometer data with the higher frequency is identified as the odometer sequence to be aligned. Each odometer sequence consists of odometer data ordered by sampling time, including angular velocity and linear acceleration. For each target sampling time in the target odometer sequence, the odometer data corresponding to adjacent sampling times of each target sampling time is determined from the odometer sequence to be aligned. For any given target sampling time, the odometer data corresponding to adjacent sampling times is calculated. The process involves calculating the odometer interpolation at the target sampling time from the odometer data. It iterates through all target sampling times and arranges the corresponding odometer interpolations to obtain an aligned odometer sequence. Based on the angular velocities in the target and aligned odometer sequences, a rotation matrix representing the rotational relationship between the first and second sensors is calculated, along with the angular acceleration at each target sampling time in both sequences. Finally, a translation vector representing the translational relationship between the first and second sensors is calculated based on the rotation matrix, the angular and linear accelerations in the target and aligned odometer sequences. In this application, by converting odometer data from different sensors at different frequencies to the same frequency, determining the odometer parameters for each sensor within that frequency, and calculating the rotation matrix and translation vector between the corresponding sensors based on these parameters, the efficiency of sensor extrinsic parameter calibration is improved. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of an application environment for a multi-sensor alignment extrinsic calibration method provided in an embodiment of the present invention;

[0020] Figure 2 This is a flowchart illustrating an extrinsic parameter calibration method for multi-sensor alignment provided in an embodiment of the present invention;

[0021] Figure 3 This is a schematic diagram of the structure of a multi-sensor alignment extrinsic calibration device provided in an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the structure of a multi-sensor alignment extrinsic calibration system provided in an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0025] It should be understood that, when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0026] It should also be understood that the term “and / or” as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0027] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if [described condition or event] is detected" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once [described condition or event] is detected," or "in response to detection of [described condition or event]."

[0028] Furthermore, in the description of this invention and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0029] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0030] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0031] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0032] This invention provides an extrinsic parameter calibration method for multi-sensor alignment, which can be applied to applications such as... Figure 1 In this application environment, the client communicates with the server. The client includes, but is not limited to, handheld computers, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, cloud terminal devices, and personal digital assistants (PDAs). The server can be a standalone server or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0033] See Figure 2 This is a flowchart illustrating a multi-sensor alignment extrinsic calibration method according to an embodiment of the present invention. The aforementioned multi-sensor alignment extrinsic calibration method can be applied to... Figure 1 The server side, such as Figure 2 As shown, the extrinsic parameter calibration method for multi-sensor alignment may include the following steps.

[0034] S201: Obtain the first odometer sequence collected by the first sensor and the second odometer sequence collected by the second sensor within a preset time period, compare the frequency of the first odometer sequence and the frequency of the second odometer sequence, determine the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned.

[0035] In step S201, odometer data for corresponding unit time from the first and second sensors are acquired. Each odometer sequence consists of odometer data sorted according to the sampling time. The odometer data includes angular velocity and linear acceleration. Since the frequencies of the first and second sensors may be different, the number of odometer data in the first and second odometer sequences is different. The number of odometer data with lower frequency is less, and the number of odometer data with higher frequency is more. The frequencies of the first and second odometer sequences are compared, and the odometer data with lower frequency is determined as the target odometer sequence, and the odometer data with higher frequency is the odometer sequence to be aligned.

[0036] In this embodiment, two sensors with different data acquisition frequencies, including a first sensor and a second sensor, are used to acquire corresponding odometer data. The first and second sensors acquire corresponding odometer data at their respective acquisition frequencies, obtaining odometer data for each sampling moment. The odometer data includes angular velocity and linear acceleration. Within a preset time period, multiple odometer data corresponding to multiple sampling moments are obtained. The odometer data are sorted according to the sampling moment to obtain a first odometer sequence acquired by the first sensor and a second odometer sequence acquired by the second sensor. The sensors can be any two of the following: lidar sensors, inertial navigation sensors, or vision sensors. The sensor devices can be positioning devices installed on mobile products such as vehicles, aircraft, and mobile robots, or other devices with the same or similar functions; this embodiment does not impose any limitations on this.

[0037] It should be noted that odometer data can also include the distance traveled relative to the initial position. Odometer is a method that uses data obtained from motion sensors to estimate the change of an object's position over time. This method is highly sensitive to errors introduced when estimating the position by integrating velocity over time. Fast and accurate data acquisition, device calibration, and processing are essential for the efficient use of this method. Specifically, odometer data is calculated from the data obtained from the sensors. The relative positions of the odometer between each sampling time are then obtained from the odometer data. The set of relative positions of the odometer at each sampling time is called the odometer relative position set. The relative positions can be used for subsequent detection of extrinsic parameter accuracy.

[0038] The frequencies of the first odometer sequence and the second odometer sequence are compared. The odometer data with the lower frequency is identified as the target odometer sequence, and the odometer data with the higher frequency is identified as the odometer sequence to be aligned. The odometer sequence to be aligned is then frequency-aligned with the target odometer sequence by converting the frequencies of the odometer data in the odometer sequence to match the frequencies of the odometer data in the target odometer sequence. For example, if the first odometer sequence corresponding to the first sensor includes odometer data from 5 sampling times, and the second odometer sequence includes odometer data from 4 sampling times, and the first odometer sequence has a higher frequency than the second odometer sequence, then the first odometer sequence is identified as the odometer sequence to be aligned, and the second odometer sequence is identified as the target odometer sequence.

[0039] For example, the sensors can be vision sensors and inertial sensors. The camera's image output frequency is 30Hz, and the inertial sensor's odometer measurement frequency is 100-150Hz. Using the camera's image frame sampling time as an alignment marker, information from multiple sensors is synchronized. The camera's odometer data is used as the target odometer sequence, and the inertial sensor's odometer data is used as the odometer sequence to be aligned.

[0040] It should be noted that inertial navigation technology is the most important part of external parameter calibration when realizing intelligence, and inertial sensors are also the core sensors for multi-sensor external parameter calibration. The reasons are as follows: First, inertial sensor measurements are less affected by the environment. Visual sensors experience a decrease in accuracy or even positioning failure in poor lighting conditions, and lidar is subject to additional measurement interference in rainy or snowy weather. In comparison, inertial sensor measurements are relatively stable and reliable in most operating conditions. Second, inertial sensor measurements do not depend on external signals and are not attacked by external interference signals.

[0041] S202: Obtain each target sampling time in the target odometer sequence, determine the odometer data corresponding to the adjacent sampling time of each target sampling time from the odometer sequence to be aligned, calculate the odometer interpolation of the target sampling time based on the odometer data corresponding to the adjacent sampling time of the target sampling time for any target sampling time, traverse all target sampling times, and arrange the odometer interpolation corresponding to all target sampling times to obtain the aligned odometer sequence of the corresponding odometer sequence to be aligned.

[0042] In step S02, the target sampling time in the target odometer sequence is used as the sampling time of the standard frequency. Based on the odometer data corresponding to the adjacent sampling times of each target sampling time in the odometer sequence to be aligned, the interpolation of the odometer data of the corresponding frequency in the odometer sequence to be aligned with the corresponding frequency in the target odometer sequence is calculated. The frequencies of the odometer data of the corresponding frequency in the odometer sequence to be aligned are aligned with the frequencies of the odometer data of the corresponding frequency in the target odometer sequence at the same frequency, so as to facilitate the calculation of the relationship between the odometer data in the odometer sequence to be aligned and the odometer data in the target odometer sequence at the same sampling time.

[0043] In this embodiment, each target sampling time in the target odometer sequence is between adjacent sampling times in the odometer sequence to be aligned. For any target sampling time, the odometer interpolation for the target sampling time is calculated based on the odometer data corresponding to the adjacent sampling times. The interpolation is calculated when the odometer data in the odometer sequence to be aligned and the odometer data in the target odometer sequence have the same frequency. This means that the odometer data of the sensor with the higher frequency at that target sampling time can be aligned with the odometer data of the sensor with the higher frequency. The relationship between different odometer data can be directly calculated to obtain the rotation and translation relationships between different sensors, which is convenient for extrinsic parameter calibration of different sensors, thereby improving the calibration efficiency of the sensors.

[0044] It should be noted that when calculating the odometer interpolation at the target sampling time, the odometer interpolation at the target sampling time can be calculated using a linear interpolation algorithm based on the odometer data corresponding to the adjacent sampling times of the target sampling time. Alternatively, a nonlinear interpolation algorithm can be used for interpolation, such as the spline interpolation algorithm. For any target sampling time, the odometer interpolation at the target sampling time can be calculated based on the odometer data corresponding to the two adjacent sampling times before and after the target sampling time. Spline interpolation algorithms include, but are not limited to, the basis spline (B-spline) interpolation algorithm, the Bezier interpolation algorithm, etc. This embodiment does not limit the implementation method of the interpolation algorithm.

[0045] By iterating through all target sampling times and arranging the odometer interpolations corresponding to all target sampling times, an aligned odometer sequence is obtained for the corresponding odometer sequence to be aligned. By iterating through all target sampling times, the odometer data of the corresponding sensor with the larger frequency can be calculated at each target sampling time. The number of odometer data in the obtained aligned odometer sequence is equal to the number of odometer data in the target odometer sequence. The extrinsic parameter relationship between the odometer data in the target odometer sequence and the odometer data in the odometer sequence to be aligned can be calculated at each target sampling time. Multiple extrinsic parameter relationships are obtained. Based on multiple extrinsic parameter relationships, the optimal solution can be calculated to obtain the optimal extrinsic parameter relationship, thereby improving the accuracy of extrinsic parameter calibration.

[0046] It should be noted that before selecting any target sampling time, it is necessary to filter out odometer data in the odometer sequence to be aligned that have valid time confidence. Time confidence is used to determine whether the target sampling time can participate in linear interpolation; odometer data with valid time confidence can participate in linear interpolation. The target sampling time and the sampling times of the odometer data in the odometer sequence to be aligned can be compared, and the corresponding time confidence status can be determined based on the comparison result. Specifically, if the difference between the timestamps of the sampling time of a certain odometer data in the odometer sequence to be aligned and the target sampling time is greater than a time threshold, then the time confidence of the odometer data in the odometer sequence to be aligned is considered valid. For example, the time threshold can be set to 20ms to 1000ms.

[0047] Optionally, based on the odometer data corresponding to adjacent sampling times of the target sampling time, the odometer interpolation for the target sampling time is calculated, including:

[0048] Based on the odometer data corresponding to adjacent sampling times of the target sampling time, determine the first odometer data and the second odometer data corresponding to adjacent sampling times;

[0049] Calculate the difference between the first odometer data and the second odometer data. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference, and the first odometer data, calculate the odometer interpolation at the target sampling time.

[0050] In this embodiment, when calculating the odometer interpolation at the target sampling time, it is necessary to calculate based on the odometer data corresponding to the adjacent sampling times of the target sampling time in the odometer sequence to be aligned. The odometer data of the moment before the adjacent sampling time of the target sampling time in the odometer sequence to be aligned is taken as the first odometer data, and the odometer data of the moment after the adjacent sampling time of the target sampling time in the odometer sequence to be aligned is taken as the second odometer data. The change value between the odometer data of the moment before and the moment after is considered to be a uniform change. The difference between the first odometer data and the second odometer data is calculated. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference, and the first odometer data, the odometer interpolation at the target sampling time is calculated.

[0051] Optionally, based on the target sampling time, adjacent sampling times, the difference, and the first odometer data, the odometer interpolation for the target sampling time is calculated, including:

[0052] Calculate the difference between the angular velocity in the first odometer data and the angular velocity in the second odometer data. Based on the target sampling time, the adjacent sampling times of the target sampling time, the angular velocity difference, and the angular velocity in the first odometer data, calculate the interpolated angular velocity at the target sampling time.

[0053] The difference between the linear acceleration in the first odometer data and the linear acceleration in the second odometer data is calculated. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference in linear acceleration, and the linear acceleration in the first odometer data, the interpolated linear acceleration at the target sampling time is calculated.

[0054] In this embodiment, the difference between the angular velocity in the first odometer data and the angular velocity in the second odometer data is calculated. Based on the target sampling time, the adjacent sampling times of the target sampling time, the angular velocity difference, and the angular velocity in the first odometer data, the interpolated angular velocity at the target sampling time is calculated. The calculation formula is as follows:

[0055]

[0056] in, For the target sampling time, The sampling time corresponding to the first odometer data. This refers to the sampling time corresponding to the second odometer data. The angular velocity in the first odometer reading. The angular velocity in the second odometer. The interpolated angular velocity at the target sampling time.

[0057] The difference between the linear acceleration in the first odometer data and the linear acceleration in the second odometer data is calculated. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference in linear acceleration, and the linear acceleration in the first odometer data, the interpolated linear acceleration at the target sampling time is calculated using the following formula:

[0058]

[0059] in, For the target sampling time, The sampling time corresponding to the first odometer data. This refers to the sampling time corresponding to the second odometer data. The linear acceleration in the first odometer. The linear acceleration in the second odometer. The interpolation line acceleration is the value at the target sampling time.

[0060] S203: Based on the angular velocities in the target odometer sequence and the aligned odometer sequence, calculate the rotation matrix characterizing the rotational relationship between the first and second sensors, as well as the angular acceleration at each target sampling moment in the target odometer sequence and the aligned odometer sequence.

[0061] In step S203, the frequencies of the angular velocities in the alignment odometer sequence and the angular velocities in the target odometer sequence are equal. In the target odometer sequence, the angular velocity at each target sampling time has a corresponding interpolated angular velocity in the alignment odometer sequence. Based on the angular velocities in the target odometer sequence and the alignment odometer sequence, the rotation matrix representing the rotational relationship between the first sensor and the second sensor is directly calculated, which can improve the efficiency of calculating the rotational relationship between the first sensor and the second sensor.

[0062] In this embodiment, based on the angular velocities in the target odometer sequence and the angular velocities in the aligned odometer sequence, the relationship between the angular velocities of different sensors at the corresponding sampling time at the same frequency is first determined. One sensor can be rotated to align with the angular velocity of another sensor. Therefore, the corresponding rotation matrix can be directly calculated based on the angular velocities of one sensor and the other sensor.

[0063] Based on the angular velocity at each target sampling time, the increase in angular velocity at adjacent target sampling times can be calculated, thus obtaining the angular acceleration at each target sampling time. Similarly, based on the interpolated angular velocity at each target sampling time in the aligned odometry sequence, the increase in interpolated angular velocity at adjacent target sampling times can be calculated, thus obtaining the angular acceleration of the interpolated angular velocity at each target sampling time.

[0064] It should be noted that in the process of extrinsic parameter calibration, the existing technology generally uses the PnP (pespective-n-point) algorithm for calculation. However, the PnP algorithm requires calibration using a calibration board, which has a large calibration time cost. In this application, the rotation matrix representing the rotation relationship between the first sensor and the second sensor is calculated based on the angular velocity in the target odometer sequence and the angular velocity in the aligned odometer sequence, thereby improving the efficiency of calculating the rotation relationship between the first sensor and the second sensor.

[0065] It should be noted that when calculating the rotation matrix representing the rotation relationship between the first and second sensors based on the angular velocities in the target odometer sequence and the aligned odometer sequence, all target sampling times can be traversed to obtain the rotation matrix between the first and second sensors at each target sampling time, resulting in a rotation matrix sequence. Based on the rotation matrix sequence, the optimal solution of the rotation matrix is ​​calculated using the least squares error algorithm to obtain the optimal rotation matrix.

[0066] It should be noted that the optimal solution of the rotation matrix can also be calculated using the gradient descent method. This embodiment does not limit the implementation method of the optimal solution of the rotation matrix.

[0067] Optionally, based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first and second sensors is calculated, including:

[0068] For any target sampling time, calculate the rotation matrix of the rotation relationship between the first sensor and the second sensor corresponding to the target sampling time based on the angular velocity in the alignment sequence of the target sampling time and the angular velocity in the target odometer sequence of the target sampling time;

[0069] By iterating through all target sampling times, the rotation matrix between the first and second sensors at each target sampling time is obtained, resulting in a rotation matrix sequence. Based on the rotation matrix sequence, the optimal solution of the rotation matrix is ​​calculated using the least squares error algorithm, and the optimal rotation matrix is ​​obtained. The optimal rotation matrix is ​​then used as the rotation matrix characterizing the rotation relationship between the first and second sensors.

[0070] In this embodiment, when calculating the rotation matrix between the first and second sensors at each target sampling time, the relationship between the angular velocities of the different sensors is as follows:

[0071]

[0072] in, The angular velocity in the target odometry sequence at the target sampling time. To align the interpolated angular velocities at the target sampling time in the sequence, Let be the rotation matrix of the first and second sensors at the target sampling time.

[0073] By iterating through all target sampling times, the rotation matrix between the first and second sensors at each target sampling time is obtained, resulting in a rotation matrix sequence. Based on this sequence, the optimal solution for the rotation matrix is ​​calculated using a least-squares error algorithm. This optimal rotation matrix is ​​then used as the rotation matrix characterizing the rotational relationship between the first and second sensors. The calculation formula is as follows:

[0074]

[0075] Where R is the optimal rotation matrix. The angular velocity in the target odometry sequence at the target sampling time. To align the interpolated angular velocities of the odometer interpolation in the sequence, Let be the rotation matrix of the first and second sensors at the target sampling time.

[0076] Optionally, the angular acceleration at each target sampling time in the target odometer sequence and the aligned odometer sequence is calculated, including:

[0077] For any target sampling time, calculate the first derivative of the angular velocity of the target odometry sequence to obtain the angular acceleration of the target sampling time in the target odometry sequence. Iterate through all target sampling times to obtain the angular acceleration of each target sampling time in the target odometry sequence.

[0078] For any target sampling time, calculate the first derivative of the angular velocity in the aligned odometer sequence to obtain the angular acceleration at the target sampling time in the aligned odometer sequence. Iterate through all target sampling times to obtain the angular acceleration at each target sampling time in the aligned odometer sequence.

[0079] In this embodiment, the angular acceleration is calculated based on the angular velocity at each target sampling time. When calculating the angular acceleration, the first derivative of the angular velocity at the target sampling time is calculated by differentiation. The change in angular velocity corresponding to the first derivative per unit time is taken as the angular acceleration at each target sampling time. By traversing all target sampling times, the angular acceleration at each target sampling time in the target odometer sequence and the angular acceleration at each target sampling time in the aligned odometer sequence are obtained.

[0080] S204: Based on the rotation matrix, the angular and linear accelerations in the target odometer sequence, and the angular and linear accelerations in the aligned odometer sequence, calculate the translation vector characterizing the translational relationship between the first and second sensors.

[0081] In step S204, the translation vector represents the translation relationship between the first sensor and the second sensor. Based on the rotation matrix, the angular acceleration and linear acceleration in the target odometer sequence, and the angular acceleration and linear acceleration in the aligned odometer sequence, the translation vector representing the translation relationship between the first sensor and the second sensor is calculated. The translation matrix is ​​directly calculated using the rotation matrix, which improves the calculation efficiency of the translation vector and thus improves the sensor calibration efficiency.

[0082] In this embodiment, the data in the odometer sequence to be aligned is rotated and then translated to align with the odometer data in the target odometer sequence. This allows different sensors to use the same sensor to process the acquired data in the same coordinate system. This embodiment requires no external auxiliary devices and can simultaneously calibrate the sensor's extrinsic parameters online. For the odometer information from multiple sensors with different timestamps and frequencies used in this paper, the linear interpolation alignment measurement method helps improve the efficiency of data reading, synchronization, and processing.

[0083] It should be noted that in the prior art, the transformation matrix is ​​generally calculated by using the PnP (pespective-n-point) algorithm, and the translation vector is determined based on the transformation matrix and the rotation matrix. In this application, the translation vector representing the translation relationship between the first sensor and the second sensor is calculated based on the rotation matrix, the angular acceleration and linear acceleration in the target odometer sequence, and the angular acceleration and linear acceleration in the aligned odometer sequence, without the need for external auxiliary devices.

[0084] It should be noted that in this application, when calculating the translation vector, all target sampling times can be traversed to obtain the translation vector between the first sensor and the second sensor at each target sampling time, resulting in a translation vector sequence. Based on the translation vector sequence, the optimal solution of the translation vector is calculated using the least squares error algorithm to obtain the optimal translation vector. Alternatively, the gradient descent method can be used for calculation. This embodiment does not limit the implementation method of the optimal solution of the rotation matrix.

[0085] Optionally, for any target sampling time, a translation vector representing the translation relationship between the first sensor and the second sensor is calculated based on the rotation matrix, the angular acceleration and linear acceleration at the target sampling time, and the angular acceleration and linear acceleration in the aligned odometer sequence.

[0086] By iterating through all target sampling times, the translation vector between the first and second sensors at each target sampling time is obtained, resulting in a translation vector sequence. Based on the translation vector sequence, the optimal solution of the translation vector is calculated using the least squares error algorithm, and the optimal translation vector is obtained. The optimal translation vector is then used as the translation vector characterizing the translation relationship between the first and second sensors.

[0087] In this embodiment, the translation vector of the odometer data in the aligned odometer sequence to the odometer data in the target odometer sequence at each target sampling time is first calculated. At any target sampling time, based on the rotation matrix, the angular and linear accelerations at the target sampling time, and the angular and linear accelerations in the aligned odometer sequence, the translation vector representing the translation relationship between the first and second sensors at the target sampling time is calculated. The calculation formula is as follows:

[0088]

[0089] in, Let a be the rotation matrix representing the rotation relationship between the first and second sensors at the target sampling time. A For the interpolation line acceleration in the alignment sequence at the target sampling time, ω B Let a be the angular velocity at the target sampling time. B Let Ω be the linear acceleration at the target sampling time. B The angular acceleration of the interpolated angular velocity at the target sampling time. This is the concatenated translation vector representing the translation relationship between the first and second sensors at the target sampling time. This refers to the antisymmetric matrix corresponding to *.

[0090] By iterating through all target sampling times, the translation vector between the first and second sensors at each target sampling time is obtained, resulting in a translation vector sequence. Based on this sequence, the optimal solution for the translation vectors is calculated using a least squares error algorithm, yielding the optimal translation vector. This optimal translation vector is then used as a table. The rotation matrix, The interpolation line acceleration in the aligned sequence at the k-th sampling time is... Let be the angular velocity at the k-th sampling time. Let x be the linear acceleration at the k-th sampling time. Let be the angular acceleration of the interpolated angular velocity at the k-th sampling time. This is the concatenated translation vector representing the translation relationship between the first and second sensors at the k-th sampling time. This refers to the antisymmetric matrix corresponding to *.

[0091] The process involves acquiring a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period. The frequencies of the first and second odometer sequences are compared, and the odometer data with the lower frequency is identified as the target odometer sequence, while the odometer data with the higher frequency is identified as the odometer sequence to be aligned. Each odometer sequence consists of odometer data ordered by sampling time, including angular velocity and linear acceleration. For each target sampling time in the target odometer sequence, the odometer data corresponding to adjacent sampling times of each target sampling time is determined from the odometer sequence to be aligned. For any given target sampling time, the odometer data corresponding to adjacent sampling times is calculated. The process involves calculating the odometer interpolation at the target sampling time from the odometer data. It iterates through all target sampling times and arranges the corresponding odometer interpolations to obtain an aligned odometer sequence. Based on the angular velocities in the target and aligned odometer sequences, a rotation matrix representing the rotational relationship between the first and second sensors is calculated, along with the angular acceleration at each target sampling time in both sequences. Finally, a translation vector representing the translational relationship between the first and second sensors is calculated based on the rotation matrix, the angular and linear accelerations in the target and aligned odometer sequences. In this application, by converting odometer data from different sensors at different frequencies to the same frequency, determining the odometer parameters for each sensor within that frequency, and calculating the rotation matrix and translation vector between the corresponding sensors based on these parameters, the efficiency of sensor extrinsic parameter calibration is improved.

[0092] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a multi-sensor alignment extrinsic calibration device provided in an embodiment of the present invention. The units included in this embodiment are used to perform... Figure 2 The steps in the corresponding embodiments. Please refer to the details. Figure 2 The relevant descriptions in the corresponding embodiments are shown below. For ease of explanation, only the parts relevant to this embodiment are shown. See also... Figure 3 The external parameter calibration device 30 includes an acquisition module 31, an interpolation module 32, a first calculation module 33, and a second calculation module 34.

[0093] The acquisition module 31 is used to acquire a first odometer sequence collected by the first sensor and a second odometer sequence collected by the second sensor within a preset time period, compare the frequency of the first odometer sequence and the frequency of the second odometer sequence, determine the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned, wherein each odometer sequence is composed of odometer data sorted according to the sampling time, and the odometer data includes angular velocity and linear acceleration.

[0094] Interpolation module 32 is used to obtain each target sampling time in the target odometer sequence, determine the odometer data corresponding to the adjacent sampling time of each target sampling time from the odometer sequence to be aligned, calculate the odometer interpolation of the target sampling time based on the odometer data corresponding to the adjacent sampling time of any target sampling time for any target sampling time, traverse all target sampling times, and arrange the odometer interpolation corresponding to all target sampling times to obtain the aligned odometer sequence of the corresponding odometer sequence to be aligned.

[0095] The first calculation module 33 is used to calculate, based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor, as well as the angular acceleration at each target sampling moment in the target odometer sequence and the aligned odometer sequence.

[0096] The second calculation module 34 is used to calculate a translation vector characterizing the translation relationship between the first sensor and the second sensor based on the rotation matrix, the angular acceleration and linear acceleration in the target odometer sequence, and the angular acceleration and linear acceleration in the aligned odometer sequence.

[0097] Optionally, the interpolation module 32 includes:

[0098] The determining unit is used to determine the first odometer data and the second odometer data corresponding to adjacent sampling times based on the odometer data corresponding to adjacent sampling times of the target sampling time.

[0099] The calculation unit is used to calculate the difference between the first odometer data and the second odometer data, and to calculate the odometer interpolation at the target sampling time based on the target sampling time, the adjacent sampling times of the target sampling time, the difference and the first odometer data.

[0100] Optionally, the above-mentioned computing unit includes:

[0101] The interpolation angular velocity calculation subunit is used to calculate the difference between the angular velocity in the first odometer data and the angular velocity in the second odometer data. Based on the target sampling time, the adjacent sampling times of the target sampling time, the angular velocity difference, and the angular velocity in the first odometer data, the interpolation angular velocity at the target sampling time is calculated.

[0102] The interpolation linear acceleration calculation subunit is used to calculate the difference between the linear acceleration in the first odometer data and the linear acceleration in the second odometer data. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference in linear acceleration, and the linear acceleration in the first odometer data, the interpolation linear acceleration at the target sampling time is calculated.

[0103] Optionally, the first computing module 33 includes:

[0104] The rotation matrix calculation unit is used to calculate the rotation matrix of the rotation relationship between the first sensor and the second sensor corresponding to any target sampling time, based on the angular velocity in the alignment sequence of the target sampling time and the angular velocity in the target odometer sequence of the target sampling time.

[0105] The first optimization unit is used to traverse all target sampling times, obtain the rotation matrix between the first sensor and the second sensor at each target sampling time, obtain a rotation matrix sequence, and calculate the optimal solution of the rotation matrix using the least squares error algorithm based on the rotation matrix sequence to obtain the optimal rotation matrix. The optimal rotation matrix is ​​used as the rotation matrix characterizing the rotation relationship between the first sensor and the second sensor.

[0106] Optionally, the first computing module 33 includes:

[0107] The first angular acceleration calculation unit is used to calculate the first derivative of the angular velocity of the target odometer sequence for any target sampling time, so as to obtain the angular acceleration of the target sampling time in the target odometer sequence. It then iterates through all target sampling times to obtain the angular acceleration of each target sampling time in the target odometer sequence.

[0108] The second angular acceleration calculation unit is used to calculate the first derivative of the angular velocity in the aligned odometer sequence for any target sampling time, so as to obtain the angular acceleration of the target sampling time in the aligned odometer sequence. It can then traverse all target sampling times to obtain the angular acceleration of each target sampling time in the aligned odometer sequence.

[0109] Optionally, the second computing module 34 includes:

[0110] The translation vector calculation unit is used to calculate, for any target sampling time, the translation vector representing the translation relationship between the first sensor and the second sensor, based on the rotation matrix, the angular acceleration and linear acceleration at the target sampling time, and the angular acceleration and linear acceleration in the aligned odometer sequence.

[0111] The second optimization unit is used to traverse all target sampling times, obtain the translation vector between the first sensor and the second sensor at each target sampling time, obtain the translation vector sequence, calculate the optimal solution of the translation vector using the least squares error algorithm based on the translation vector sequence, obtain the optimal translation vector, and use the optimal translation vector as the translation vector characterizing the translation relationship between the first sensor and the second sensor.

[0112] It should be noted that the information interaction and execution process between the above modules, units, and sub-units are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0113] Figure 4 This is a schematic diagram of the structure of a multi-sensor aligned extrinsic parameter calibration system 40 provided by the present invention. It includes a computer device, a first sensor, and a second sensor. The computer device is connected to the first sensor and the second sensor, and receives the odometer sequence sent by the first sensor and the second sensor, such as... Figure 4 As shown, the computer device of this embodiment includes: at least one processor ( Figure 4 Only one is shown in the diagram), a memory, and a computer program stored in the memory that can run on at least one processor, which, when executing the computer program, implements the steps in any of the above-described embodiments of the extrinsic parameter method for multi-sensor alignment.

[0114] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 4 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown, or combinations of certain components, or different components.

[0115] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0116] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of a computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal storage units and external storage devices of a computer device. Memory is used to store the operating system, applications, bootloader, data, and other programs, such as program code for computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0117] Those skilled in the art will understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the functions described above can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this invention. The specific working process of the units and modules in the above device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the processes in the methods of the above embodiments by instructing related hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable medium can include at least: any entity or device capable of carrying computer program code, a recording medium, a computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0118] The present invention can implement all or part of the processes in the methods of the above embodiments, or it can be accomplished by a computer program product. When the computer program product is run on a computer device, the computer device executes the steps in the above method embodiments.

[0119] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0120] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0121] In the embodiments provided by this invention, it should be understood that the disclosed apparatus / computer devices and methods can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0122] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0123] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for calibrating extrinsic parameters of multi-sensor alignment, characterized in that, The external parameter calibration method includes: The system acquires a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period. It compares the frequency of the first odometer sequence with the frequency of the second odometer sequence, determines the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned. Each odometer sequence consists of odometer data sorted according to the sampling time, and the odometer data includes angular velocity and linear acceleration. Each target sampling time in the target odometer sequence is obtained. Odometer data corresponding to adjacent sampling times of each target sampling time is determined from the odometer sequence to be aligned. For any target sampling time, the odometer interpolation of the target sampling time is calculated based on the odometer data corresponding to adjacent sampling times of the target sampling time. All target sampling times are traversed, and the odometer interpolations corresponding to all target sampling times are arranged to obtain the aligned odometer sequence corresponding to the odometer sequence to be aligned. Based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor, as well as the angular acceleration at each target sampling moment in the target odometer sequence and the aligned odometer sequence, are calculated. Based on the rotation matrix, the angular and linear accelerations in the target odometer sequence, and the angular and linear accelerations in the aligned odometer sequence, a translation vector characterizing the translational relationship between the first sensor and the second sensor is calculated.

2. The external parameter calibration method as described in claim 1, characterized in that, The step of calculating the odometer interpolation for the target sampling time based on the odometer data corresponding to adjacent sampling times of the target sampling time includes: Based on the odometer data corresponding to the adjacent sampling times of the target sampling time, determine the first odometer data and the second odometer data corresponding to the adjacent sampling times; Calculate the difference between the first odometer data and the second odometer data, and calculate the odometer interpolation at the target sampling time based on the target sampling time, the adjacent sampling times of the target sampling time, the difference, and the first odometer data.

3. The external parameter calibration method as described in claim 2, characterized in that, The step of calculating the odometer interpolation for the target sampling time based on the target sampling time, the adjacent sampling times of the target sampling time, the difference, and the first odometer data includes: Calculate the difference between the angular velocity in the first odometer data and the angular velocity in the second odometer data. Based on the target sampling time, the adjacent sampling times of the target sampling time, the angular velocity difference, and the angular velocity in the first odometer data, calculate the interpolated angular velocity at the target sampling time. The difference between the linear acceleration in the first odometer data and the linear acceleration in the second odometer data is calculated. Based on the target sampling time, the adjacent sampling times of the target sampling time, the difference in linear acceleration, and the linear acceleration in the first odometer data, the interpolated linear acceleration at the target sampling time is calculated.

4. The external parameter calibration method according to claim 1, characterized in that, The step of calculating a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor based on the angular velocities in the target odometer sequence and the aligned odometer sequence includes: For any target sampling time, a rotation matrix is ​​calculated based on the angular velocity in the aligned odometer sequence at the target sampling time and the angular velocity in the target odometer sequence at the target sampling time to determine the rotation relationship between the first sensor and the second sensor corresponding to the target sampling time. By iterating through all target sampling times, the rotation matrix between the first sensor and the second sensor at each target sampling time is obtained, resulting in a rotation matrix sequence. Based on the rotation matrix sequence, the optimal solution of the rotation matrix is ​​calculated using the least squares error algorithm to obtain the optimal rotation matrix. The optimal rotation matrix is ​​then used as the rotation matrix characterizing the rotation relationship between the first sensor and the second sensor.

5. The external parameter calibration method according to claim 3, characterized in that, The calculation of the angular acceleration at each target sampling time in the target odometer sequence and the aligned odometer sequence includes: For any target sampling time, calculate the first derivative of the angular velocity of the target odometer sequence to obtain the angular acceleration of the target sampling time in the target odometer sequence. Iterate through all target sampling times to obtain the angular acceleration of each target sampling time in the target odometer sequence. For any target sampling time, calculate the first derivative of the angular velocity in the aligned odometer sequence to obtain the angular acceleration at the target sampling time in the aligned odometer sequence. Iterate through all target sampling times to obtain the angular acceleration at each target sampling time in the aligned odometer sequence.

6. The external parameter calibration method as described in claim 3, characterized in that, The step of calculating a translation vector characterizing the translation relationship between the first sensor and the second sensor based on the rotation matrix, the angular and linear accelerations in the target odometer sequence, and the angular and linear accelerations in the aligned odometer sequence includes: For any target sampling time, based on the rotation matrix, the angular acceleration and linear acceleration of the target sampling time, and the angular acceleration and linear acceleration in the aligned odometer sequence, a translation vector representing the translation relationship between the first sensor and the second sensor corresponding to the target sampling time is calculated. By iterating through all target sampling times, the translation vector between the first sensor and the second sensor at each target sampling time is obtained, resulting in a translation vector sequence. Based on the translation vector sequence, the optimal solution of the translation vector is calculated using the least squares error algorithm to obtain the optimal translation vector. The optimal translation vector is then used as the translation vector characterizing the translation relationship between the first sensor and the second sensor.

7. A extrinsic parameter calibration device for multi-sensor alignment, characterized in that, The external parameter calibration device includes: The acquisition module is used to acquire a first odometer sequence collected by a first sensor and a second odometer sequence collected by a second sensor within a preset time period, compare the frequency of the first odometer sequence and the frequency of the second odometer sequence, determine the odometer data with the lower frequency as the target odometer sequence, and the odometer data with the higher frequency as the odometer sequence to be aligned, wherein each odometer sequence is composed of odometer data sorted according to the sampling time, and the odometer data includes angular velocity and linear acceleration; An interpolation module is used to acquire each target sampling time in the target odometer sequence, determine the odometer data corresponding to the adjacent sampling times of each target sampling time from the odometer sequence to be aligned, calculate the odometer interpolation value of the target sampling time based on the odometer data corresponding to the adjacent sampling times of the target sampling time for any target sampling time, traverse all target sampling times, and arrange the odometer interpolation values ​​corresponding to all target sampling times to obtain the aligned odometer sequence corresponding to the odometer sequence to be aligned; The first calculation module is used to calculate, based on the angular velocities in the target odometer sequence and the aligned odometer sequence, a rotation matrix characterizing the rotational relationship between the first sensor and the second sensor, and the angular acceleration at each target sampling time in the target odometer sequence and the aligned odometer sequence; The second calculation module is used to calculate a translation vector characterizing the translation relationship between the first sensor and the second sensor based on the rotation matrix, the angular acceleration and linear acceleration in the target odometer sequence, and the angular acceleration and linear acceleration in the aligned odometer sequence.

8. The external parameter calibration device as described in claim 7, characterized in that, The interpolation module includes: The determining unit is used to determine the first odometer data and the second odometer data corresponding to the adjacent sampling times based on the odometer data corresponding to the adjacent sampling times of the target sampling time; The calculation unit is used to calculate the difference between the first odometer data and the second odometer data, and to calculate the odometer interpolation at the target sampling time based on the target sampling time, the adjacent sampling times of the target sampling time, the difference and the first odometer data.

9. A extrinsic parameter calibration system for multi-sensor alignment, characterized in that, The device includes a computer, a first sensor, and a second sensor. The computer is connected to the first sensor and the second sensor and receives the odometer sequence sent by the first sensor and the second sensor. The computer includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the external parameter calibration method as described in any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the external parameter calibration method as described in any one of claims 1 to 6.

Images