Alignment method and program

Abstract

To provide an alignment method and program capable of improving alignment accuracy.SOLUTION: An alignment method is executed by an alignment system 1 that performs a process of aligning a reference point cloud 3A, which contains position information for roads and surrounding structures in a specified road section with a maintenance point cloud 3B acquired in the road section at a different time than the time the reference point cloud 3A was acquired. The maintenance point cloud 3B includes position information PPn, indicating the position and orientation of a measurement vehicle 2B in the road section, and information representing the relative positional relationship between the measurement vehicle 2B and the road and surrounding structures. In a division step, the maintenance point cloud 3B is divided into at least three divided road sections Tn. In an alignment step, a process is performed to align a portion of the maintenance point cloud 3B with the reference point cloud 3A. The alignment process is performed for each divided road section Tn.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a registration method and a program. [Background technology]

[0002] In order to generate map data, various sensors are mounted on a vehicle, and while the vehicle is moving, the various sensors are used to generate point cloud data by measuring roads and their surroundings in three dimensions. Patent Document 1 discloses a technology for updating (extracting change points) map data generated based on point cloud data measured with equipment and methods capable of high accuracy measurement, using low-accuracy measurement data, and a technology for aligning the point clouds in this process. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2020-144226 Summary of the Invention [Problem to be solved by the invention]

[0004] The point cloud data is calculated based on position information indicating the position and orientation detected by sensors mounted on a moving vehicle. Therefore, the point cloud data contains errors contained in the detection results of the sensors, and these errors accumulate as the vehicle moves. Such accumulated errors cause a decrease in alignment accuracy.

[0005] The present invention has been made in light of the above circumstances, and an object of the present invention is to provide a registration method and a program that can improve registration accuracy. [Means for solving the problem]

[0006] In order to achieve the above object, a positioning method according to a first aspect of the present invention comprises: 1. A registration method in which an information processing device executes a process of aligning a reference point cloud including position information of a road and surrounding structures in a predetermined road section with a maintenance point cloud acquired in the road section at a time different from a time when the reference point cloud was acquired, the method comprising: the maintenance point cloud includes position information indicating the position and attitude of a vehicle in the road section, and information indicating a relative positional relationship between the vehicle and the road and surrounding structures; a dividing step of dividing the maintenance point cloud into at least three divided road sections; a registration step of performing a process for registering a part of the maintenance point cloud with the reference point cloud, The alignment process is performed for each divided road section.

[0007] A program according to a second aspect of the present invention comprises: a computer that aligns a reference point cloud containing position information of a road and surrounding structures in a specified road section with a maintenance point cloud that is acquired in the specified road section at a time different from the time when the reference point cloud was acquired, the maintenance point cloud containing position information indicating the position and attitude of a vehicle in the road section and information indicating the relative positional relationship between the vehicle and the road and surrounding structures; a dividing means for dividing the maintenance point cloud into at least three divided road sections; an alignment means for executing a process for aligning a part of the maintenance point cloud with the reference point cloud; It functions as The alignment process is performed for each divided road section. [Effects of the Invention]

[0008] According to the present invention, it is possible to improve the alignment accuracy. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing a functional configuration of an alignment system according to a first embodiment of the present invention. [Figure 2] 1A is a schematic diagram showing the flow of aligning the maintenance point cloud with the reference point cloud in each divided road section, and FIG. 1B is a schematic diagram showing how the position information of the measurement vehicle changes for the maintenance point cloud whose point cloud information has been aligned. [Figure 3] 10(A) and 10(B) are first schematic diagrams showing how the maintenance point cloud and the reference point cloud in each road section are aligned and the position information is corrected. [Figure 4] 10(A), 10(B), and 10(C) are second schematic diagrams showing how the maintenance point cloud and the reference point cloud in each road section are aligned and the position information is corrected. [Figure 5] FIG. 2 is a diagram showing the hardware configuration of the alignment system of FIG. 1. [Figure 6] 10 is a flowchart of the alignment process of the alignment system of FIG. [Figure 7] (A) is a schematic diagram showing the process of determining whether alignment has failed due to an outlier in the position information. (B) is a schematic diagram showing the process of determining whether alignment has failed due to roll and pitch. (C) is a schematic diagram showing the process of determining whether longitudinal deviation has occurred due to the cumulative travel distance. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each drawing, the same or equivalent parts are denoted by the same reference numerals. In the following embodiments, the terms "have," "include," or "contain" also mean "consist of" or "consist of."

[0011] As shown in Fig. 1, a positioning system 1 according to this embodiment is incorporated into a map data maintenance system that maintains map data. This positioning system 1 is an information processing device that performs positioning between two point groups in three-dimensional space, a reference point group 3A as a first point group for reference, and a maintenance point group 3B as a second point group. In other words, the positioning system 1 performs a positioning method.

[0012] First, a measurement vehicle 2A equipped with a group of sensors capable of high-precision measurements is first driven along a road, and position information indicating the position and orientation of the measurement vehicle 2A detected at a predetermined sampling period and point cloud information representing the relative positional relationship between the measurement vehicle 2A and surrounding roads or structures such as buildings are detected. From this position information and point cloud information, a reference point cloud 3A, which is a point cloud for reference representing the outlines of structures around the moving measurement vehicle 2A, is generated. The reference point cloud 3A is generated by converting relative point cloud information of the position reference of the measurement vehicle 2A, which represents the outlines of structures around the measurement vehicle 2A, into point cloud information in an absolute coordinate system, using the position information of the measurement vehicle 2A as a reference. Initial map data is generated using this reference point cloud 3A. In this embodiment, the reference point cloud 3A is point cloud information including position information indicating the position and orientation of the measurement vehicle 2A in a predetermined road section and information representing the relative positional relationship between the measurement vehicle 2A and surrounding structures. Although the reference point cloud 3A uses information obtained by vehicle measurement as a data source, the reference point cloud 3A itself is position information of roads and structures (position information defined by absolute coordinates of latitude, longitude, and height) that integrates information obtained from data sources (vehicle measurements), and the reference point cloud 3A is not limited to points measured by the measurement vehicle 2A. In other words, the reference point cloud 3A is point cloud information that includes position information of roads and surrounding structures in a specified road section.

[0013] Then, for example, several years later, a measurement vehicle 2B equipped with a group of general-purpose sensors is driven along a road, and position information indicating the position and attitude of the measurement vehicle 2B detected at a predetermined sampling period and point cloud information representing the relative positional relationship between the measurement vehicle 2B and surrounding roads or structures such as buildings are detected. From this position information and point cloud information, a maintenance point cloud 3B, which is a maintenance point cloud, is generated. In other words, the maintenance point cloud 3B is point cloud information including position information indicating the position and attitude of the measurement vehicle 2B in a predetermined road section and information representing the relative positional relationship between the measurement vehicle 2B and the road and surrounding structures, and is point cloud information acquired in the predetermined road section at a time different from the time when the reference point cloud 3A was acquired.

[0014] The registration system 1 executes a process of aligning the reference point group 3A with the maintenance point group 3B. The aligned maintenance point group 3B is used to update the map data generated from the reference point group 3A.

[0015] The measurement vehicle 2A is equipped with a sensor group consisting of a high-precision Global Navigation Satellite System (GNSS) sensor, an inertial measurement unit (IMU), a rangefinder, and a light detection and ranging (LiDAR). The high-precision GNSS detects the position information of the measurement vehicle 2A with high accuracy based on satellite information. The inertial measurement unit is a sensor that integrates a gyro sensor (angular velocity sensor) and an acceleration sensor, and the position and attitude of the measurement vehicle 2A can be calculated by integrating the detected values. The rangefinder measures the cumulative travel distance of the measurement vehicle 2A. The high-precision GNSS sensor and inertial measurement unit can detect position information indicating the position and attitude of the measurement vehicle 2A at a predetermined sampling period. The LiDAR is a sensor that measures distance by emitting a pulse of laser light instantaneously, reflecting off an object, and returning.

[0016] The LiDAR is a scanning lidar that uses a movable mirror to change the direction of the emitted laser light for scanning. The LiDAR obtains a point cloud representing the outline of structures around the measurement vehicle 2A, and can detect point cloud information indicating the relative position of structures, such as roads or buildings, that exist in the scanned direction. Based on the position information of the measurement vehicle 2A detected by the high-precision GNSS sensor and the inertial measurement unit and the point cloud information detected by the LiDAR, a reference point cloud 3A representing the outline of three-dimensional structures around the route traveled by the measurement vehicle 2A is generated.

[0017] The measurement vehicle 2B is equipped with a sensor group including a GNSS sensor, an inertial measurement unit (IMU), a rangefinder, and a LiDAR. The GNSS sensor can detect position information indicating the position and attitude of the measurement vehicle 2B at a predetermined sampling period. The position information of the measurement vehicle 2B is detected based on the position detected by the GNSS sensor, the position and attitude detected by the IMU, and the traveled distance measured by the rangefinder.

[0018] The LiDAR is a scanning lidar, and with each scan, point cloud information representing the shapes of structures around the measurement vehicle 2B is obtained. Based on the position information of the measurement vehicle 2B detected by a GNSS sensor or the like and the point cloud information detected by the LiDAR, a maintenance point cloud 3B representing the outlines of structures around the route traveled by the measurement vehicle 2B is generated. The maintenance point cloud 3B is stored in a storage device, and after being aligned with the reference point cloud 3A in the alignment system 1, is used to update the map data.

[0019] To align the reference point group 3A and the maintenance point group 3B, the alignment system 1 includes a first alignment unit 10, a second alignment unit 20, and a correction unit 30. The first alignment unit 10 roughly aligns the reference point group 3A and the maintenance point group 3B. The second alignment unit 20 more precisely aligns the reference point group 3A and the maintenance point group 3B, which have been aligned by the first alignment unit 10.

[0020] The position information of the measurement vehicles 2A and 2B includes the position coordinates (X, Y, Z) in a three-dimensional Cartesian coordinate system and the pitch, roll, and yaw rotation angles (θx, θy, θz) of the measurement vehicles 2A and 2B (see Figure 7(B)).

[0021] The maintenance point cloud 3B is a point cloud calculated from the position information or point cloud information of the measurement vehicle 2B detected by a sensor group with lower accuracy than the sensor group used to detect the reference point cloud 3A.

[0022] (First alignment part) The first alignment unit 10 aligns the reference point group 3A with the maintenance point group 3B by pattern matching some of the feature points of the reference point group 3A with some of the feature points of the maintenance point group 3B. For example, the first alignment unit 10 sets a point group near one end of the reference point group 3A and the maintenance point group 3B as a whole (for example, near the measurement start point of the measurement vehicles 2A and 2B) and a point group near the other end opposite the one end (for example, near the measurement end point of the measurement vehicles 2A and 2B) as feature point groups, and aligns the reference point group 3A with the maintenance point group 3B by pattern matching these feature point groups. Alternatively, the first alignment unit 10 may extract the overall general outlines of the reference point group 3A and the maintenance point group 3B, and align the reference point group 3A with the maintenance point group 3B by pattern matching the general outlines.

[0023] (Second alignment part) The second alignment unit 20 aligns the reference point group 3A aligned by the first alignment unit 10 through pattern matching with the maintenance point group 3B.

[0024] First, the second alignment unit 20 divides the maintenance point cloud 3B into at least three divided road sections according to the sections for which the position information and point cloud information were detected. For example, as shown in FIG. 1, the first point in a predetermined road section measured by the measurement vehicle 2B is designated as the first point t1, and a point after the first point t1 is designated as the second point t2. The first point t1 may be the measurement start point, and the second point t2 may be the measurement end point, or the first point t1 and the second point t2 may be points during the measurement. In this case, once processing between the set first point t1 and second point t2 is completed, the first point t1 and the second point t2 are changed, and processing is performed on the changed road section.

[0025] The road section between the first point t1 and the second point t2 is divided into at least three divided road sections T1 to T N (N is a natural number of 3 or more). The second alignment unit 20 divides the maintenance point cloud 3B into at least three divided road sections T n The road section T is divided into the point cloud. n The length T of the measurement vehicle 2B is set to the section where one scan is performed, for example, by LiDAR. During one scan, the measurement vehicle 2B moves, and the alignment system 1 calculates the amount of movement from the detection values ​​of the GNSS sensor, the inertial measurement unit, and the rangefinder, and generates and aligns the maintenance point cloud 3B.

[0026] The second alignment unit 20 aligns the individual divided road sections T n Regarding the maintenance point cloud 3B, some of the point clouds PG n Specifically, the second alignment unit 20 performs a process of aligning each divided road section T n A part of the maintenance point cloud 3B corresponding to PG n and a part of the maintenance point group 3B roughly aligned by the first alignment unit 10, PG n In other words, the process for aligning is performed by aligning the reference points 3A of the area overlapping the divided road section T nThis is done every time.

[0027] The reference point group 3A and the maintenance point group 3B are aligned by ICP (Iterative Closest Point). The second alignment unit 20 performs coordinate transformation of the maintenance point group 3B so that the residual error between the points included in the reference point group 3A and the points included in the maintenance point group 3B is minimized when the points are aligned one-to-one.

[0028] In Figure 2(B), a part of the reference point cloud 3A, PG n and the point cloud PG of maintenance point cloud 3B n The outline of a part of the maintenance point cloud 3B is shown by a square, and the ICP is used to n 2B shows how the outline of the measurement vehicle 2B is aligned with the outline of the reference point group 3A. In FIG. 2B, the position information PP n As shown in FIG. 2(B), when the maintenance point cloud 3B is aligned with the reference point cloud 3A by the ICP, the position information PP n The position of is also adjusted.

[0029] (correction section) Returning to FIG. 1, the correction unit 30 calculates the maintenance point group PG n Each time the process of aligning the reference point group 3A with the reference point group 3B is performed, the aligned divided road section T n The correction unit 30 corrects the position information in the corrected position information PP n Based on this, the remaining divided road sections T that have not yet been aligned are n Location information detected by PP n For example, the point group PG nAfter the alignment is performed, the correction unit 30 calculates the maintenance point cloud PG 3B, which is a part of the maintenance point cloud 3B corresponding to the divided road section T1 for which the alignment is performed, as shown in FIG. 2(B). n Based on the alignment result, the position information PP n Correct the following.

[0030] Furthermore, the correction unit 30 n The corrected position information PP n Based on this, the remaining divided road sections T that have not yet been aligned are n Location information detected by PP n That is, correct any one of the divided road sections T n Location information PP n When the correction is made, the correction result is applied to other divided road sections T n Location information PP n This also corrects the distance between the divided road sections T n The alignment result in other divided road sections T n For example, the aligned divided road section T n Location information PP n The correction vector of the other divided road section T is corrected by the same amount as the correction vector of n Location information PP n Correction is performed.

[0031] The second alignment unit 20 performs the alignment process on at least three divided road sections T n The second alignment unit 20 and the correction unit 30 perform the alignment from both ends of the first divided road section T1 and the last divided road section T N Regarding the maintenance point cloud 3B, point clouds PG1 and PG N and the reference point group 3A, and the associated alignment of each divided road section T n Location information PP n After making the correction, the remaining divided road section T n The point cloud PG of the reference point cloud 3A corresponding to n and the maintenance point group 3B, and the associated position information PP2 to PP N-1Specifically, as shown in FIG. 3B, the correction unit 30 corrects the divided road sections T1 and T2. N Location information PP1, PP N By interpolation using N-1 Location information PP2~PP N-1 In FIG. 3(A), N=5.

[0032] Furthermore, the second alignment unit 20 aligns the remaining divided road sections T 1 except for the first and last ones as shown in FIG. n Regarding the maintenance point cloud 3B, some of the point clouds PG n and the reference point group 3A, and the associated alignment of each divided road section T n Location information PP n For example, in the case shown in Figure 4(A), when the position adjustment of the divided road section T2 is performed, the correction unit 30 corrects the position information of the position information PP2 of the divided road section T2, and also corrects the position information of the divided road sections T3 and T4 that have not yet been adjusted.

[0033] Thereafter, for example, as shown in Figure 4(B), when the divided road section T4 has been aligned, the correction unit 30 corrects the position information P4 for the divided road section T4, and then corrects the position information P3 for the divided road section T3 that has not yet been aligned. Finally, as shown in Figure 4(C), when the divided road section T3 has been aligned, the position information P3 for the divided road section T3 is corrected.

[0034] (Hardware configuration) The alignment system 1 shown in Fig. 1 is realized, for example, by a computer 50 having the hardware configuration shown in Fig. 5 executing a software program. Specifically, the computer 50 includes a CPU (Central Processing Unit) 51, which is a processor that controls the entire device, a main memory 52 such as a RAM (Random Access Memory), an external memory 53 configured from a non-volatile memory such as a flash memory or a hard disk, an operation unit 54 configured from devices such as a keyboard and a mouse, a display 55 configured from a display device such as a CRT (Cathode Ray Tube) or a liquid crystal monitor, a communication interface 56 that communicates data with an external computer, and an internal bus 58 that connects these.

[0035] The program 59 is loaded from the external memory 53 into the main memory 52 and executed by the CPU 51. This realizes the functions of the computer 50. When executing the program 59, the CPU 51 performs data communication with an external computer via the communication interface 56 as necessary. In this embodiment, the program 59 executed by the CPU 51 includes a program for the alignment system 1.

[0036] The functions of the alignment system 1 can be implemented in a computer system consisting of one or more computers including one or more processors and one or more storage devices including non-transitory storage media. The multiple computers realize the functions of the alignment system 1 while communicating with each other via an interconnected communication network. For example, some of the functions of the alignment system 1 may be implemented in one computer, and other parts may be implemented in other computers. The functions of the alignment system 1 may also be realized by a cloud computer.

[0037] (Alignment method) Next, the alignment process, i.e., the alignment method, executed by the alignment system 1 will be described. As shown in FIG. 6, first, in the alignment system 1, the first alignment unit 10 reads the reference point group 3A and the maintenance point group 3B (step S1). Next, the first alignment unit 10 performs coarse alignment (step S2; pattern matching step). Here, the first alignment unit 10 aligns the reference point group 3A with the maintenance point group 3B by pattern matching some of the feature points of the reference point group 3A with some of the feature points of the maintenance point group 3B. In the subsequent alignment steps, the first alignment unit 10 performs coarse alignment between the reference point group 3A and the maintenance point group 3B by pattern matching some of the feature points of the reference point group 3A with some of the feature points of the maintenance point group 3B. n The maintenance point group 3B corresponding to the area is aligned with the reference point group 3A in the area that overlaps with the maintenance point group 3B. As a result, the reference point group 3A and the maintenance point group 3B are aligned by template matching, and it becomes possible to align the reference point group 3A and the maintenance point group 3B in detail.

[0038] Next, the second alignment unit 20 aligns the maintenance point group 3B with at least three divided road sections T n The maintenance point cloud is divided into the divided road sections T (step S3; division step). n Point cloud PG n Next, the second alignment unit 20 aligns the first divided road section T1 (step S4; alignment step). Here, a process is executed to align the point group PG1 corresponding to the first divided road section T1 with the reference point group 3A. Furthermore, the second alignment unit 20 corrects the position information PP1 of the divided road section T1 in accordance with the alignment in step S4 (step S5; correction step). Furthermore, the second alignment unit 20 corrects the position information PP1 of the divided road section T1 in accordance with the alignment in step S4 (step S6; correction step). n Location information PP n In this way, the second alignment unit 20 corrects the point group PGn Each time the process of aligning the reference point group 3A with the reference point group 3B is performed, the aligned divided road section T n Location information PP n Correct the following.

[0039] Next, the second alignment unit 20 aligns the final divided road section T N (Step S7: Alignment step), and N Location information PP N (Step S8: correction step). Furthermore, the second alignment unit 20 corrects the divided road section T n Location information PP n Then, the above correction is performed (step S9; correction step).

[0040] Next, the second alignment unit 20 initializes the variable i to 2, the variable j to N-1, and the variable k to 1 (step S10). Next, the second alignment unit 20 determines whether the variable k is an odd number (step S11). If the variable k is an odd number (step S11; Yes), the second alignment unit 20 assigns the value of the variable i to the variable n (step S12), and if the variable k is not an odd number (step S11; No), the second alignment unit 20 assigns the value of the variable j to the variable n (step S13).

[0041] Next, the second alignment unit 20 aligns the divided road section T n (step S14; alignment step), and the correction unit 30 n Location information PP n (Step S15: Correction step). Furthermore, the correction unit 30 corrects the divided road section T n Location information PP n (Step S16: Correction step) n If alignment has been performed in step S16, nothing is done in step S16.

[0042] Next, the second alignment unit 20 increments the variable i by 1, decrements the variable j by 1, and increments the variable k by 1 (step S17). n It is determined whether there is an unaligned divided road section T (step S18). n If there is (step S18; Yes), the second alignment unit 20 returns to step S11.

[0043] Unaligned divided road section T n While it is determined that there is a divided road section T (step S18; Yes), the second alignment unit 20 and the correction unit 30 repeat steps S11 to S18. n (Step S14: Alignment step), n Location information PP n (Step S15; correction step), n Location information PP n In this way, in the alignment step, in the process for alignment, the point group PG n Each time the process of aligning the reference point group 3A with the reference point group 3B is performed, the aligned divided road section T n Location information PP n is corrected.

[0044] Unaligned divided road section T n If there is no such position (step S18; No), the registration system 1 ends the registration process.

[0045] According to this alignment process, for example, the maintenance point group 3B is divided into five divided road sections T n When the point group PG1 is divided into two, the point groups PG1 and PG5 at both ends are aligned, and then the second and fourth point groups PG2 and PG4 are aligned, and so on. Then, alignment is performed from both ends in order, and the position information PP nThis makes it possible to reduce misalignment due to accumulated errors.

[0046] In this alignment process, the divided road section T n Part of the maintenance point cloud 3B point cloud PG n The reference point group 3A is aligned with the reference point group 3A, and the divided road section T n Location information PP n Each time the correction is performed, it may be determined whether the alignment has failed.

[0047] For example, as shown in FIG. 7(A), the correction unit 30 calculates the road division section T n Part of the maintenance point cloud 3B point cloud PG n and the reference point cloud 3A, and the divided road section T n Location information PP n The divided road section T n Location information PP n However, other divided road sections T n-1 , T n+1 Location information PP in etc. n-1 , PP n+1 When it is determined that the value is an outlier that deviates by more than the first threshold E1 from the value of the divided road section T n The divided road section T before and after n-1 , T n+1 Location information PP n-1 , PP n+1 Based on the above, the divided road section T n Location information PP n For example, the position information PP n Location information PP n-1 , PP n+1 If the deviation is equal to or greater than the first threshold E1 from the movement trajectory of the measurement vehicle 2B formed by connecting the above, the position information PP n , location information PP n-1 , PP n+1 This determination is made by updating the location information to the location information interpolated by the divided road section T n Alternatively, the alignment and correction may be performed after the completion of alignment and correction in all divided road sections T nAlternatively, the correction may be performed after the alignment and correction in step 1 is completed.

[0048] As shown in FIG. 7B, the position information PP n contains position coordinates (X, Y, Z) in an XYZ Cartesian coordinate system, and the +Y direction is the traveling direction of the measurement vehicle 2B. In this case, the position information PP n The corrected position information PP n When the pitch θx or roll θy of the measurement vehicle 2B included in the divided road section T exceeds the second threshold value E2, n The divided road section T before and after n-1 , T n+1 Location information PP in etc. n-1 , PP n+1 Based on this, the divided road section T n Location information PP n For example, if the measurement vehicle 2B is unnaturally tilted with respect to the direction of gravity after the correction, it is considered that the alignment has failed, and the front and rear position information PP n-1 , PP n+1 By interpolating with n This determination is made based on the divided road section T n Alternatively, the alignment and correction may be performed after the completion of alignment and correction in all divided road sections T n Alternatively, the correction may be performed after the alignment and correction in step 1 is completed.

[0049] As described above, the sensors mounted on the measurement vehicle 2B include a distance meter that measures the cumulative travel distance of the measurement vehicle 2B. The accuracy of the cumulative travel distance measured by the distance meter is guaranteed in the traveling direction of the measurement vehicle 2B. Therefore, the correction unit 30 corrects the position information PP n The position of the measurement vehicle 2B in the traveling direction based on the cumulative travel distance measured by the distance meter is separated by a third threshold E3 or more from the position of the measurement vehicle 2B in the traveling direction based on the cumulative travel distance measured by the distance meter. n If there is a divided road section T nThe divided road section T before and after n-1 , T n+1 Location information PP n-1 , PP n+1 Based on this, the divided road section T n Location information PP n In this case, for example, as shown in FIG. 7C, the position information PP n If there is a deviation in the intervals of the position information PP n This bias can be corrected by using the position information PP n This determination can be made by checking whether the ratio of the intervals between the divided road sections T1 and T2 (T1 / T2 in FIG. 7C) exceeds a fourth threshold value E4. n Alternatively, the alignment and correction may be performed after the completion of alignment and correction in all divided road sections T n Alternatively, the correction may be performed after the alignment and correction in step 1 is completed.

[0050] (1) As explained in detail above, the alignment method according to this embodiment is an alignment method in which the alignment system 1 executes a process of aligning a reference point group 3A including position information of a road and surrounding structures in a predetermined road section with a maintenance point group 3B acquired in the road section at a time different from the time when the reference point group 3A was acquired. The maintenance point group 3B includes position information indicating the position and attitude of the measurement vehicle 2B in the road section, and information indicating the relative positional relationship between the measurement vehicle 2B and the road and surrounding structures. In the alignment method, the division step divides the maintenance point group 3B into at least three divided road sections T n In the alignment step, the maintenance point cloud 3B is divided into the point cloud PG n The process for aligning the divided road section T n In this way, the accumulated error contained in the alignment result can be reduced, and therefore the alignment accuracy can be improved.

[0051] (2) According to the registration method of this embodiment, the registration process is performed for at least three divided road sections T n By performing the alignment in this order, it is possible to perform the alignment with the least accumulated error.

[0052] (3) According to the alignment method of this embodiment, in the alignment step, in the process for alignment, the point group PG n Each time the process of aligning the reference point group 3A with the reference point group 3B is performed, the aligned divided road section T n Location information PP n In this way, the part of the maintenance point group 3B, PG n Each time the position of the reference point group 3A is aligned with the position information PP n It is possible to reduce the cumulative error contained in

[0053] (4) According to the alignment method of this embodiment, in the correction step, the corrected position information PP n However, other divided road sections PP n-1 , PP n+1 When the position information PP n The divided road section T is determined to be an outlier. n The divided road section T before and after n-1 , T n+1 Location information PP n=1 , PP n+1 Based on this, the position information PP n In this way, the position information PP n can be excluded from the registration process.

[0054] (5) According to the alignment method of this embodiment, in the correction step, the corrected position information PP nWhen the pitch or roll of the measurement vehicle 2B included in the road section T exceeds the second threshold E2, the measurement vehicle 2B is n The divided road section T before and after n Location information PP n Based on this, the divided road section T n Location information PP n In this way, the abnormal value of the position information PP n can be excluded from the registration process.

[0055] (6) According to the alignment method of this embodiment, in the correction step, the position information PP n The position of the measurement vehicle 2B in the traveling direction based on the distance meter that measures the traveling distance of the measurement vehicle 2B is separated by a third threshold value E3 or more from the position of the measurement vehicle 2B in the traveling direction based on the cumulative traveling distance measured by the distance meter that measures the traveling distance of the measurement vehicle 2B. n If there is a divided road section T n The divided road section T before and after n-1 , T n+1 Location information PP n Based on this, the divided road section T n Location information PP n In this way, the position information PP n can be excluded from the registration process.

[0056] (7) According to the alignment method of this embodiment, the alignment step further includes a pattern matching step of aligning the reference point group 3A with the maintenance point group 3B by pattern matching some of the feature points of the reference point group 3A with some of the feature points of the maintenance point group 3B. n The maintenance point group 3B corresponding to the area is aligned with the reference point group 3A in the area that overlaps with the maintenance point group 3B. By performing alignment in two stages in this way, it is possible to prevent the point groups from being significantly misaligned with each other.

[0057] (8) According to the alignment method of this embodiment, the maintenance point cloud 3B is a point cloud calculated from the position information or point cloud information of the measurement vehicle 2B detected by a sensor group with lower accuracy than the sensor group used to detect the reference point cloud 3A. In this way, high-accuracy alignment is possible using a point cloud measured by a sensor group with lower accuracy.

[0058] (9) The program according to this embodiment includes a reference point group 3A including position information of roads and surrounding structures in a predetermined road section, and position information PP which is a point group acquired in the predetermined road section at a time different from the time when the reference point group 3A was acquired and indicates the position and attitude of the measurement vehicle 2B in the road section. n and a maintenance point group 3B including information indicating the relative positional relationship between the measurement vehicle 2B and the road and surrounding structures. n a division means for dividing the maintenance point cloud 3B into a part of the point cloud PG n , the reference point cloud PG n The process for alignment is performed by the divided road section T n In this way, the position information PP n Since the accumulated error included in can be reduced, the alignment accuracy can be improved.

[0059] In the above embodiment, the divided road section T n The length of the divided road section T n The length of the period can be any length, for example, a length corresponding to half a period or multiple periods of the LiDAR scan.

[0060] In the above embodiment, the reference point cloud 3A for reference is a high-precision point cloud measured by a high-precision measurement vehicle 2A. However, this is not limiting, and the second point cloud for reference may be a point cloud measured by a measurement vehicle 2B, or information other than the position information measured by a measurement vehicle may be used.

[0061] The hardware and software configurations of the alignment system 1 are merely examples and can be changed and modified as desired.

[0062] The core processing portion of the computer 50, which is composed of a CPU 51, a main memory 52, an external memory 53, an operation unit 54, a display 55, a communication interface 56, an internal bus 58, etc., can be realized by using an ordinary computer system rather than a dedicated system. For example, a computer program for executing the above operations may be stored and distributed on a computer-readable recording medium (such as a flexible disk, CD-ROM, or DVD-ROM), and the computer program may be installed on a computer to configure the alignment system 1 that executes the above processing. Alternatively, the computer program may be stored in a storage device of a server device on a communication network such as the Internet, and the alignment system 1 may be configured by downloading the computer program into an ordinary computer system.

[0063] When the functions of the alignment system 1 are realized by sharing the functions between an OS (operating system) and an application program, or by cooperation between the OS and the application program, only the application program portion may be stored in a recording medium or storage device.

[0064] It is also possible to superimpose a computer program on a carrier wave and distribute it over a communications network. For example, the computer program may be posted on a bulletin board system (BBS) on the communications network and distributed over the network. The computer program may then be started and executed under the control of an operating system in the same way as any other application program, thereby enabling the above-mentioned processing to be performed.

[0065] This invention allows various embodiments and modifications without departing from the broad spirit and scope of this invention. Furthermore, the above-described embodiments are intended to explain this invention and do not limit the scope of this invention. That is, the scope of this invention is defined by the claims, not the embodiments. Various modifications made within the scope of the claims and the meaning of the invention equivalent thereto are considered to be within the scope of this invention. [Industrial Applicability]

[0066] The present invention can be applied to calculating the outer shape of a structure based on vehicle position information indicating the position and attitude of the vehicle detected at a predetermined sampling period by a group of sensors mounted on the vehicle while the vehicle is moving, and point cloud information indicating the outer shape of the structure based on the vehicle's position and attitude. [Explanation of symbols]

[0067] 1 Alignment system, 2A, 2B Measurement vehicle (vehicle), 3A Reference point cloud (first point cloud), 3B Maintenance point cloud (second point cloud), 10 First alignment unit, 20 Second alignment unit, 30 Correction unit, 50 Computer, 51 CPU, 52 Main memory, 53 External memory, 54 Operation unit, 55 Display, 56 Communication interface, 58 Internal bus, 59 Program, PP n Location information, PG n Point cloud, T1~T N Divided road section

Claims

1. 1. A registration method in which an information processing device executes a process of aligning a reference point cloud including position information of a road and surrounding structures in a predetermined road section with a maintenance point cloud acquired in the road section at a time different from a time when the reference point cloud was acquired, the method comprising: the maintenance point cloud includes position information indicating the position and attitude of a vehicle in the road section, and information indicating a relative positional relationship between the vehicle and the road and surrounding structures; a dividing step of dividing the maintenance point cloud into at least three divided road sections; a registration step of performing a process for registering a part of the maintenance point cloud with the reference point cloud, the alignment process is performed for each divided road section; Alignment method.

2. In the alignment step, performing the process for alignment from both ends of the at least three divided road sections divided in the dividing step; The alignment method according to claim 1 .

3. the alignment step includes a correction step of correcting the position information in the divided road section where the alignment has been performed each time a process of aligning a part of the maintenance point cloud with the reference point cloud is performed in the process for alignment, The alignment method according to claim 2 .

4. In the correction step, if the corrected position information is determined to be an outlier that deviates by a first threshold or more from the position information in another divided road section, correcting the position information determined to be an outlier based on the position information in the divided road sections before and after the divided road section whose position information is determined to be an outlier. The alignment method according to claim 3 .

5. In the correction step, when the pitch or roll of the vehicle included in the corrected position information exceeds a second threshold, correcting the position information in the divided road section based on the position information in the divided road sections before and after the divided road section in which the pitch or roll of the vehicle exceeded the second threshold. The alignment method according to claim 3 or 4.

6. In the correction step, If there is a divided road section in which the position in the traveling direction of the vehicle based on the position information is away by a third threshold or more from the position in the traveling direction of the vehicle based on an accumulated traveling distance measured by a distance meter that measures the traveling distance of the vehicle, correct the position information of the divided road section in question based on the position information of the divided road sections before and after the divided road section in question. The alignment method according to claim 3 or 4.

7. The alignment step further comprises: a pattern matching step of aligning the reference point cloud with the maintenance point cloud by pattern matching some feature points of the reference point cloud with some feature points of the maintenance point cloud; performing registration between the maintenance point clouds corresponding to the individual divided road sections registered in the pattern matching step and a reference point cloud in an area overlapping with the maintenance point clouds; The alignment method according to claim 1 or 2.

8. The maintenance point cloud is a point cloud calculated from the position information of the vehicle detected by a sensor group having lower accuracy than the sensor group used to detect the reference point cloud or the point cloud information. The alignment method according to claim 1 or 2.

9. a computer that aligns a reference point cloud containing position information of a road and surrounding structures in a specified road section with a maintenance point cloud that is acquired in the specified road section at a time different from the time when the reference point cloud was acquired, the maintenance point cloud containing position information indicating the position and attitude of a vehicle in the road section and information indicating the relative positional relationship between the vehicle and the road and surrounding structures; a dividing means for dividing the maintenance point cloud into at least three divided road sections; an alignment means for executing a process for aligning a part of the maintenance point cloud with the reference point cloud; It functions as the alignment process is performed for each divided road section; program.

Images