Calibration method and device for road end infrastructure sensor
By setting distance-related weight coefficients in the iterative nearest point IICP algorithm of the road-end infrastructure sensor, the external parameters of the sensor are calculated, and the problem of poor perceived performance of the sensor at long distance targets is solved, and the balanced improvement of perceived performance is achieved.
Patent Information
- Application Number
- CN202311691870.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-08
- Publication Date
- 2025-06-10
AI Technical Summary
Road-end infrastructure sensors (such as road-end cameras) have poor perceived performance at long-distance targets, making it difficult to meet the project requirements in vehicle-road collaborative development.
The sensor's external parameters such as rotation matrix and translation matrix are calculated based on the distance between the target and the road-end infrastructure sensor, thereby improving perceptual performance.
The perceived performance at both long-distance and close-distance targets is achieved to ensure that the detection target information output by the sensor meets the project requirements.
Smart Images
Figure CN120122064A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of calibration of roadside infrastructure sensors, and more specifically, to a method and device for calibrating roadside infrastructure sensors, a computer-readable storage medium, a computer program product, and a roadside infrastructure system. Background Art
[0002] The dynamic calibration of the external parameters of a large-scale roadside infrastructure sensor involves a large amount of work. At the same time, how to make the detection target information output by the sensor with calibrated external parameters meet the project requirements to the greatest extent is a thorny problem encountered in the process of vehicle-road collaboration development. Summary of the Invention
[0003] The inventors of this application noticed that: currently, roadside infrastructure sensors (such as roadside cameras) have worse perception performance for "distant" targets than for "close" targets. In view of this phenomenon, one or more embodiments of this application propose to determine a weight coefficient based on the distance between the target and the roadside infrastructure sensor, and finally obtain the external parameters (such as rotation matrix and translation matrix) of the roadside infrastructure sensor, so that the perception performance of both "distant" and "close" targets can meet the requirements.
[0004] According to one aspect of this application, a method for calibrating a roadside infrastructure sensor is provided. The method includes: receiving first geographical location information related to a target, where the first geographical location information is true value information; receiving second geographical location information related to the target, where the second geographical location information is calculated based on the perception result of the roadside infrastructure sensor; and after synchronizing the first geographical location information and the second geographical location information in time, calculating the external parameters of the roadside infrastructure sensor based on an improved iterative closest point (IICP) algorithm, where a weight coefficient is set in the improved iterative closest point (IICP) algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor.
[0005] As a supplement or replacement to the above solution, in the above method, the target is a vehicle with a positioning system, and the first geographical location information and the second geographical location information include position and longitude and latitude.
[0006] As a supplement or replacement to the above solution, in the above method, receiving the first geographical location information related to the target includes: receiving the first position and first longitude and latitude of the vehicle itself from the vehicle with a positioning system via one or more routers.
[0007] As a supplement or replacement for the above solution, in the above method, receiving the second geographical location information related to the target includes: receiving the second location and the second longitude and latitude related to the vehicle from the roadside infrastructure computing unit ICU, where the roadside infrastructure computing unit ICU receives the sensing result from the roadside infrastructure sensor, and calculates the second location and the second longitude and latitude after performing coordinate system transformation on the sensing result.
[0008] As a supplement or replacement for the above solution, in the above method, after synchronizing the time of the first geographical location information and the second geographical location information, calculating the external parameters of the roadside infrastructure sensor based on the improved iterative closest point (IICP) algorithm includes: Step A: Matching the first location and the first longitude and latitude with the second location and the second longitude and latitude respectively based on the time synchronization; Step B: Constructing the external parameters of the roadside infrastructure sensor, where the external parameters are a rotation matrix R and a translation matrix T; Step C: Obtaining an objective function, where the objective function includes the weight coefficient and is related to the distance between the vehicle and the roadside infrastructure sensor; Step D: Judging whether the objective function is less than a preset threshold and whether the iteration times of the algorithm reach the upper limit; and Step E: When the objective function is less than the preset threshold or the iteration times of the algorithm have reached the upper limit, obtaining the rotation matrix R and the translation matrix T, otherwise returning to Step B.
[0009] As a supplement or replacement for the above solution, in the above method, the roadside infrastructure sensor includes a camera, a millimeter-wave radar, and a lidar.
[0010] According to another aspect of the present application, there is provided a calibration device for a roadside infrastructure sensor, the device including: a first receiving device for receiving the first geographical location information related to a target, where the first geographical location information is true value information; a second receiving device for receiving the second geographical location information related to the target, where the second geographical location information is calculated based on the sensing result of the roadside infrastructure sensor; and a calculation device for calculating the external parameters of the roadside infrastructure sensor based on the improved iterative closest point (IICP) algorithm after synchronizing the time of the first geographical location information and the second geographical location information, where a weight coefficient is set in the improved iterative closest point (IICP) algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor.
[0011] As a supplement or replacement for the above solution, in the above device, the target is a vehicle with a positioning system, and the first geographical location information and the second geographical location information include a location and longitude and latitude.
[0012] As a supplement or replacement to the above solution, in the above device, the first receiving device is configured to: receive the first position and the first longitude and latitude of the vehicle itself from the vehicle with a positioning system via one or more routers.
[0013] As a supplement or replacement to the above solution, in the above device, the second receiving device is configured to: receive the second position and the second longitude and latitude related to the vehicle from the roadside infrastructure computing unit ICU, where the roadside infrastructure computing unit ICU receives the sensing result from the roadside infrastructure sensor and calculates the second position and the second longitude and latitude after performing coordinate system transformation on the sensing result.
[0014] As a supplement or replacement to the above solution, in the above device, the computing device is configured to perform the following steps: Step A: match the first position and the first longitude and latitude with the second position and the second longitude and latitude respectively based on the time synchronization; Step B: construct the external parameters of the roadside infrastructure sensor, where the external parameters are the rotation matrix R and the translation matrix T; Step C: obtain the objective function, where the objective function includes the weight coefficient and is related to the distance between the vehicle and the roadside infrastructure sensor; Step D: determine whether the objective function is less than a preset threshold and whether the iteration times of the algorithm reach the upper limit; and Step E: when the objective function is less than the preset threshold or the iteration times of the algorithm have reached the upper limit, obtain the rotation matrix R and the translation matrix T, otherwise return to Step B.
[0015] As a supplement or replacement to the above solution, in the above device, the roadside infrastructure sensor includes a camera, a millimeter-wave radar, and a lidar.
[0016] According to another aspect of the present application, there is provided a computer-readable storage medium, where the medium includes instructions that, when running, execute the method as described above.
[0017] According to another aspect of the present application, there is provided a computer program product, including a computer program that, when executed by a processor, implements the method as described above.
[0018] According to another aspect of the present application, there is provided a roadside infrastructure system, where the roadside infrastructure system includes the device as described above.
[0019] One or more embodiments of the present application provide a calibration scheme for the external parameters of roadside infrastructure sensors (such as cameras, millimeter-wave radars, lidar, etc.) under vehicle-road collaborative technology. After synchronizing the received first geographical location information and second geographical location information in time, the calibration scheme of the roadside infrastructure sensors in the embodiments of the present application calculates the external parameters of the roadside infrastructure sensors based on an improved Iterative Closest Point (IICP) algorithm, so as to obtain more accurate external parameters (such as the optimal rotation matrix R and translation matrix T). In one or more embodiments, a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor (for example, as the distance between the two becomes farther, the weight coefficient becomes smaller). In this way, the external parameters can meet the requirements within the detection area of interest, avoiding the adverse impact on the external parameter calibration of the sensor caused by the poor performance in the detection area of no interest. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] The above and other objects and advantages of the present application will become more fully apparent from the following detailed description in conjunction with the accompanying drawings, in which like or similar elements are denoted by like reference numerals.
[0021] Figure 1 FIG. shows a schematic flow chart of a calibration method for roadside infrastructure sensors according to an embodiment of the present application;
[0022] Figure 2 FIG. shows a schematic structural diagram of a calibration device for roadside infrastructure sensors according to an embodiment of the present application;
[0023] Figure 3 FIG. shows a schematic flow chart of an improved Iterative Closest Point (IICP) algorithm according to an embodiment of the present application; and
[0024] Figure 4 FIG. shows a schematic diagram of a vehicle-road collaborative calibration scenario according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0025] Hereinafter, a calibration scheme for roadside infrastructure sensors according to various exemplary embodiments of the present application will be described in detail with reference to the accompanying drawings.
[0026] Figure 1 FIG. shows a schematic flow chart of a calibration method 1000 for roadside infrastructure sensors according to an embodiment of the present application. As Figure 1 shown, the calibration method 1000 for roadside infrastructure sensors includes:
[0027] In step S110, first geographical location information related to a target is received, and the first geographical location information is true value information;
[0028] In step S120, receive second geographical location information related to the target, where the second geographical location information is calculated based on the sensing result of the roadside infrastructure sensor; and
[0029] In step S130, after time synchronization of the first geographical location information and the second geographical location information, calculate the external parameters of the roadside infrastructure sensor based on an improved Iterative Closest Point (IICP) algorithm, where a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor.
[0030] In one or more embodiments, the term "roadside infrastructure sensor" includes cameras, millimeter-wave radars, lidars, etc. In one or more embodiments, the term "target" is a vehicle equipped with a positioning system (such as GNSS). Since this vehicle can obtain accurate geographical location information (i.e., ground truth information) of the vehicle itself using the positioning system installed on it, this vehicle is also referred to as a "Ground Truth Vehicle" in the context of this application.
[0031] The first geographical location information and the second geographical location information may include, for example, position (such as coordinate position in the world coordinate system), longitude and latitude, and heading angle, etc. In one embodiment, step S110 includes: receiving the first position and the first longitude and latitude of the vehicle itself from the vehicle equipped with the positioning system (i.e., the ground truth vehicle) via one or more routers. In one embodiment, step S120 includes: receiving the second position and the second longitude and latitude related to the vehicle from the Roadside Infrastructure Computing Unit (ICU), where the Roadside Infrastructure Computing Unit ICU receives the sensing result from the roadside infrastructure sensor and calculates the second position and the second longitude and latitude after coordinate transformation of the sensing result.
[0032] In step S130, after time synchronization of the first geographical location information and the second geographical location information, calculate the external parameters of the roadside infrastructure sensor based on an improved Iterative Closest Point (IICP) algorithm, where a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor. For example, refer to Figure 3 , which shows a schematic flow diagram of an improved Iterative Closest Point (IICP) algorithm according to an embodiment of the present application. As Figure 3As shown, in step S310, the position and longitude and latitude information sent by the ground truth vehicle and the roadside infrastructure computing unit ICU are collected; then, in step S320, the collected data (including the position and longitude and latitude information sent by the ground truth vehicle and the roadside infrastructure computing unit ICU) is preprocessed (such as time synchronization, etc.) and matched; in step S330, the transformation matrices R (i.e., rotation matrix) and T (i.e., translation matrix) of the roadside infrastructure sensor are constructed - the external parameters of the sensor; then, in step S340, an objective function related to the distance between the ground truth vehicle and the sensor is formed; in step S350, it is determined whether the threshold of the objective function is satisfied or whether the number of iterations of the algorithm reaches the upper limit; if so, step S360 is executed to obtain the transformation matrices R and T of the sensor - the external parameters of the sensor; otherwise, return to step S330 to construct / adjust the transformation matrices R (i.e., rotation matrix) and T (i.e., translation matrix) of the roadside infrastructure sensor.
[0033] In one embodiment, step S330 may include: (step A) matching the first position and the first longitude and latitude with the second position and the second longitude and latitude respectively based on the time synchronization; (step B) constructing the external parameters of the roadside infrastructure sensor, where the external parameters are the rotation matrix R and the translation matrix T; (step C) obtaining an objective function, where the objective function includes the weight coefficient such that the objective function is related to the distance between the vehicle and the roadside infrastructure sensor; (step D) determining whether the objective function is less than a preset threshold and whether the number of iterations of the algorithm reaches the upper limit; and (step E) when the objective function is less than the preset threshold or the number of iterations of the algorithm has reached the upper limit, obtaining the rotation matrix R and the translation matrix T, otherwise returning to step B.
[0034] In one embodiment, the objective function F may be constructed in the following form:
[0035]
[0036] Wherein, K represents a weight coefficient, which depends on the distance d between the target (such as a vehicle) and the infrastructure sensor (such as a roadside camera). For example, K decreases as the distance d increases, for example, in inverse proportion. In addition, the perceived target position coordinates are pi', and the ground truth target position coordinates are pi, and the number thereof is n. By subtracting the perceived target position coordinates after coordinate transformation from the ground truth target position coordinates and multiplying this difference (i.e., the residual) by the weight coefficient, the final objective function value is obtained. It can be understood that the smaller the objective function value, the better the matching between the perceived target position coordinates and the ground truth target position coordinates. That is to say, as long as F obtains 0, or is close to 0, it proves that the matching effect is very good.
[0037] In one embodiment, when the above IICP algorithm matches points, considering that the position sending frequency of the ground truth vehicle is higher than the sending frequency of the target sensed by the roadside infrastructure sensor (such as a roadside camera). Therefore, in actual matching, the ground truth (GT) point that is closest to the target sensed by the roadside infrastructure sensor within a certain time difference can be selected. Additionally, since the performance of the roadside infrastructure sensor for target sensing varies with the distance between the target and the sensor, for example, the position deviation is larger at a long distance and smaller at a short distance. Therefore, in one or more embodiments of the present application, the IICP algorithm comprehensively considers the factor of the distance between the target and the roadside infrastructure sensor, so as to obtain better performance within the target sensing range of the roadside infrastructure sensor.
[0038] In one embodiment, the coordinate information (x, y, z) in the world coordinate system and the coordinate information (u, v) in the pixel coordinate system satisfy the following conversion relationship:
[0039]
[0040] In the above formula, Zc is a scale factor (not zero), representing the effective focal length (the distance from the optical center to the image plane), which is the homogeneous coordinate of the image point in the image coordinate system. f x and f y are respectively called the normalized focal lengths on the x-axis and y-axis, f x = f / dx, f y = f / dy, where f is the focal length of the camera, with the unit of mm, dx and dy are the pixel sizes, u 0 and v 0 are the image centers. R is the rotation matrix, and T is the translation matrix.
[0041] Additionally, represents the internal parameters of the camera, while represents the external parameters of the camera. It should be noted that those skilled in the art can understand that the external parameters can also be represented in other forms besides the translation matrix T and the rotation matrix R, for example, represented by (x, y, z, Φ, θ, ψ), where Φ represents the roll angle, θ represents the pitch angle, and ψ represents the yaw angle.
[0042] In addition, those skilled in the art can easily understand that the calibration method 1000 of the roadside infrastructure sensor provided by one or more of the above embodiments of the present application can be implemented by a computer program. For example, the computer program is included in a computer program product, and when the computer program is executed by a processor, it implements the calibration method 1000 of the roadside infrastructure sensor of one or more embodiments of the present application. Another example is that when a computer-readable storage medium (such as a USB flash drive) storing the computer program is connected to a computer, running the computer program can execute the calibration method 1000 of the roadside infrastructure sensor of one or more embodiments of the present application.
[0043] Reference Figure 2 , which shows a schematic structural diagram of a calibration device 2000 for a roadside infrastructure sensor according to an embodiment of the present application. As Figure 2 shown, the calibration device 2000 for the roadside infrastructure sensor includes: a first receiving device 210, configured to receive first geographical location information related to a target, where the first geographical location information is true value information; a second receiving device 220, configured to receive second geographical location information related to the target, where the second geographical location information is calculated based on the sensing result of the roadside infrastructure sensor; and a calculation device 230, configured to calculate external parameters of the roadside infrastructure sensor based on an improved iterative closest point (IICP) algorithm after synchronizing the first geographical location information and the second geographical location information in time, where a weight coefficient is set in the improved iterative closest point (IICP) algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor.
[0044] In one or more embodiments, the term "roadside infrastructure sensor" includes cameras, millimeter-wave radars, lidars, etc. In one or more embodiments, the term "target" is a vehicle equipped with a positioning system (such as GNSS). Since this vehicle can obtain accurate geographical location information (i.e., true value information) of the vehicle itself by using the positioning system installed on it, this vehicle is also referred to as a "Ground Truth Vehicle" in the context of the present application.
[0045] The first geographical location information and the second geographical location information may include, for example, a location (such as a coordinate position in a world coordinate system), longitude and latitude, and a heading angle, etc. In one embodiment, the first receiving device 210 is configured to: receive the vehicle's own first position and first longitude and latitude from the vehicle with a positioning system (i.e., the ground truth vehicle) via one or more routers. In one embodiment, the second receiving device 220 is configured to: receive the second position and second longitude and latitude related to the vehicle from the roadside infrastructure computing unit ICU, where the roadside infrastructure computing unit ICU receives the sensing result from the roadside infrastructure sensor and calculates the second position and the second longitude and latitude after performing a coordinate system transformation on the sensing result.
[0046] The computing device 230 is configured to: after synchronizing the first geographical location information and the second geographical location information in terms of time, calculate the external parameters of the roadside infrastructure sensor based on an improved Iterative Closest Point (IICP) algorithm, where a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor. In one embodiment, the computing device 230 may be configured to perform the following steps: (Step A) match the first position and the first longitude and latitude with the second position and the second longitude and latitude respectively based on the time synchronization; (Step B) construct the external parameters of the roadside infrastructure sensor, where the external parameters are a rotation matrix R and a translation matrix T; (Step C) obtain an objective function, where the objective function includes the weight coefficient and is related to the distance between the vehicle and the roadside infrastructure sensor; (Step D) determine whether the objective function is less than a preset threshold and whether the iteration count of the algorithm has reached an upper limit; and (Step E) when the objective function is less than the preset threshold or the iteration count of the algorithm has reached the upper limit, obtain the rotation matrix R and the translation matrix T, otherwise return to Step B.
[0047] In one embodiment, the objective function F may be constructed in the following form:
[0048]
[0049] Among them, K represents a weight coefficient, which depends on the distance d between the target (such as a vehicle) and the roadside infrastructure sensor (such as a roadside camera). For example, K decreases as the distance d increases, for example, in inverse proportion. In addition, the perceived target position coordinates are pi', and the true target position coordinates are pi, and the number thereof is n. The final objective function value is obtained by subtracting the perceived target position coordinates after coordinate transformation from the true target position coordinates and multiplying this difference (i.e., the residual) by the weight coefficient. It can be understood that the smaller the objective function value, the better the matching between the perceived target position coordinates and the true target position coordinates. That is to say, as long as F obtains 0, or is close to 0, it proves that the matching effect is very good.
[0050] In one or more embodiments, the calibration device 2000 of the above-mentioned roadside infrastructure sensor can be integrated into various types of roadside infrastructure systems. In addition to the calibration device 2000, in one or more embodiments, the roadside infrastructure system may further include other devices, such as a wireless router, etc.
[0051] Figure 4 A schematic diagram of a vehicle-road collaborative calibration scenario according to an embodiment of the present application is shown. As Figure 4 shown, the vehicle 410 is equipped with a GNSS system (which can also be called a true value vehicle or a GT vehicle) and is driving rightward on the lane 440. In one embodiment, the vehicle 410 can broadcast information such as its own location, longitude / latitude, heading angle, etc. through socket communication based on the UDP protocol; at the same time, to prevent the scenario where the vehicle is too far away from the ICU 420 and cannot receive the information sent by the vehicle, the wireless router 415 installed on the vehicle 410 can communicate with the roadside wireless router 425, and the roadside wireless router can further communicate with the wireless router 435 on the roadside infrastructure computing unit ICU 420, thus realizing long-distance bridging / communication. The roadside infrastructure computing unit ICU 420 can also receive the outputs of roadside infrastructure sensors (such as traffic cameras 432 and 434), and calculate information such as the position, longitude / latitude, etc. of the vehicle 415 after coordinate transformation. In Figure 4 the embodiment, the roadside infrastructure computing unit ICU 420 can calculate the optimal external parameters of the sensor by using the improved iterative closest point (IICP) algorithm based on the received information and the calculated information.
[0052] In summary, one or more embodiments of the present application provide a calibration scheme for the external parameters of roadside infrastructure sensors (such as cameras, millimeter-wave radars, lidars, etc.) under the vehicle-road collaborative technology. After synchronizing the time of the received first geographical location information and the second geographical location information, the calibration scheme of the roadside infrastructure sensors in the embodiments of the present application calculates the external parameters of the roadside infrastructure sensors based on the improved Iterative Closest Point (IICP) algorithm, so as to obtain more accurate external parameters (such as the optimal rotation matrix R and translation matrix T). In one or more embodiments, a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor (for example, as the distance between the two increases, the weight coefficient decreases). In this way, the external parameters can meet the requirements within the detection area of interest, avoiding the adverse impact of poor performance in the detection area of no interest on the external parameter calibration of the sensor.
[0053] The above examples mainly illustrate the calibration scheme of the roadside infrastructure sensors in the embodiments of the present application. Although only some embodiments of the present application are described, those of ordinary skill in the art should understand that the present application can be implemented in many other forms without departing from its gist and scope. Therefore, the examples and embodiments shown are regarded as illustrative rather than restrictive, and the present application may cover various modifications and substitutions without departing from the spirit and scope of the present application as defined by the various claims.
Claims
1. A calibration method for roadside infrastructure sensors, characterized in that, the method includes: Receiving first geographical location information related to a target, where the first geographical location information is ground truth information; Receiving second geographical location information related to the target, where the second geographical location information is calculated based on the sensing result of the roadside infrastructure sensor; and After synchronizing the first geographical location information and the second geographical location information in time, calculating the external parameters of the infrastructure sensor based on an improved Iterative Closest Point (IICP) algorithm, where a weight coefficient is set in the improved IICP algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensor.
2. The method according to claim 1, wherein, the target is a vehicle equipped with a positioning system, and the first geographical location information and the second geographical location information include position, longitude, and latitude.
3. The method according to claim 2, wherein, Receiving first geographical location information related to a target includes: Receiving the first position and the first longitude and latitude of the vehicle itself from the vehicle equipped with a positioning system via one or more routers.
4. The method according to claim 3, wherein, Receiving second geographical location information related to the target includes: Receiving the second position and the second longitude and latitude related to the vehicle from a roadside infrastructure computing unit (ICU), where the roadside infrastructure computing unit (ICU) receives the sensing result from the roadside infrastructure sensor, and calculates the second position and the second longitude and latitude after coordinate system transformation of the sensing result.
5. The method according to claim 4, wherein, After synchronizing the first geographical location information and the second geographical location information in time, calculating the external parameters of the roadside infrastructure sensor based on an improved IICP algorithm includes: Step A: Matching the first position and the first longitude and latitude with the second position and the second longitude and latitude respectively based on the time synchronization; Step B: Constructing the external parameters of the roadside infrastructure sensor, where the external parameters are a rotation matrix R and a translation matrix T; Step C: Obtaining an objective function, where the objective function contains the weight coefficient and is related to the distance between the vehicle and the roadside infrastructure sensor; Step D: Judging whether the objective function is less than a preset threshold and whether the iteration times of the algorithm reach the upper limit; and Step E: When the objective function is less than the preset threshold or the iteration times of the algorithm have reached the upper limit, obtaining the rotation matrix R and the translation matrix T, otherwise returning to Step B.
6. The method according to any one of claims 1 to 5, wherein, the roadside infrastructure sensor includes a camera, a millimeter-wave radar, and a lidar.
7. A calibration device for roadside infrastructure sensors, characterized in that, the device includes: A first receiving device for receiving first geographical location information related to a target, where the first geographical location information is ground truth information; A second receiving device for receiving second geographical location information related to the target, the second geographical location information being calculated based on the sensing results of the roadside infrastructure sensors; and A calculating device for calculating the external parameters of the roadside infrastructure sensors based on an improved Iterative Closest Point (IICP) algorithm after synchronizing the first geographical location information and the second geographical location information in time, wherein a weight coefficient is set in the improved Iterative Closest Point (IICP) algorithm, and the weight coefficient depends on the distance between the target and the roadside infrastructure sensors.
8. The device according to claim 7, wherein, the target is a vehicle with a positioning system, and the first geographical location information and the second geographical location information include position, longitude and latitude.
9. The device according to claim 8, wherein, the first receiving device is configured to: receive the first position and the first longitude and latitude of the vehicle itself from the vehicle with the positioning system via one or more routers.
10. The device according to claim 9, wherein, the second receiving device is configured to: receive a second position and a second longitude and latitude related to the vehicle from a Roadside Infrastructure Computing Unit (ICU), wherein the Roadside Infrastructure Computing Unit (ICU) receives the sensing results from the roadside infrastructure sensors, and calculates the second position and the second longitude and latitude after performing coordinate system transformation on the sensing results.
11. The device according to claim 10, wherein, the calculating device is configured to perform the following steps: Step A: Matching the first position and the first longitude and latitude with the second position and the second longitude and latitude respectively based on the time synchronization; Step B: Constructing the external parameters of the roadside infrastructure sensors, wherein the external parameters are a rotation matrix R and a translation matrix T; Step C: Obtaining an objective function, wherein the objective function includes the weight coefficient and is related to the distance between the vehicle and the roadside infrastructure sensors; Step D: Judging whether the objective function is less than a preset threshold and whether the number of iterations of the algorithm reaches an upper limit; and Step E: When the objective function is less than the preset threshold or the number of iterations of the algorithm has reached the upper limit, obtaining the rotation matrix R and the translation matrix T, otherwise returning to Step B.
12. The device according to any one of claims 7 to 11, wherein, the roadside infrastructure sensors include cameras, millimeter-wave radars and lidars.
13. A computer-readable storage medium, characterized in that the medium includes instructions that, when running, execute the method according to any one of claims 1 to 6.
14. A computer program product comprising a computer program, characterized in that when the computer program is executed by a processor, it implements the method according to any one of claims 1 to 6.
15. A roadside infrastructure system, characterized in that the roadside infrastructure system includes the device according to any one of claims 7 to 12.