Information processing device

Abstract

To provide an information processing device which enables an absolute position for point cloud information to be corrected efficiently.SOLUTION: In a server device (10), a point cloud integrating section (11) acquires point cloud data generated based on sensor data obtained by a sensor installed in a vehicle V and based on absolute position data of the vehicle V at the time of acquiring the sensor data, and a deviation amount calculating section (14) acquires aerial photograph data with a bird's view on a ground surface from sky. Then, the deviation amount calculating section (14) calculates deviation amounts between absolute positions on the aerial photograph data and absolute positions for point clouds included in the point cloud data, wherein the absolute positions on the aerial photograph data and for point clouds correspond to predetermined reference points P and Q. A trajectory correcting section (15) corrects a traveling trajectory extracted from the point cloud data by a trajectory extracting section (12), based on the deviation amounts calculated in the deviation amount calculating section (14), wherein the absolute positions for point clouds in the point cloud data are corrected based on the correction result of the trajectory correcting section (15).SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an information processing device that performs predetermined processing on point cloud information generated based on sensor information obtained by a sensor mounted on a moving body and information regarding the absolute position of the moving body when the sensor information was acquired. [Background technology]

[0002] In recent years, vehicles and other moving objects have been developed that are equipped with sensors such as laser scanners and cameras to collect information about their surroundings. Point cloud information acquired by these types of sensors can be used, for example, as material for high-precision maps used in autonomous driving of vehicles.

[0003] However, the absolute position of the point cloud acquired by the above-mentioned sensor is often not very accurate, so manual corrections such as adjusting the position of the point cloud are required, which is time-consuming, costly, and inefficient.

[0004] To address this problem, for example, Patent Document 1 discloses a method for calculating coordinate position data of a three-dimensional point cloud with high accuracy by driving the same road multiple times, using fixed objects whose positions do not change in the point cloud results for each drive as reference points, and then performing post-processing to stretch or shrink the point cloud so that the reference points are superimposed. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2013-40886 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, while the method described in Patent Document 1 can obtain a highly accurate point cloud, it mainly uses GPS (Global Positioning System), and therefore there are cases where the accuracy of GPS is reduced, for example, when there is a lot of multipath due to the influence of surrounding buildings, and there is room for improvement.

[0007] One example of a problem to be solved by the present invention is how to efficiently correct the absolute position of point cloud information. [Means for solving the problem]

[0008] In order to solve the above problem, the invention described in claim 1 is characterized by comprising a point cloud information acquisition unit that acquires point cloud information generated based on sensor information obtained by a sensor mounted on a moving body and information regarding the absolute position of the moving body when the sensor information was acquired, a bird's-eye view information acquisition unit that acquires bird's-eye view information of a view of the ground surface from the sky, a calculation unit that calculates the amount of deviation between the absolute position of the bird's-eye view information corresponding to a set reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point, and a correction unit that corrects the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit.

[0009] Claim 2 The invention described in the above relates to an information processing method executed by an information processing device that performs a predetermined process. in The method includes a point cloud information acquisition process for acquiring point cloud information generated based on sensor information obtained by a sensor mounted on a moving body and information regarding the absolute position of the moving body when the sensor information was acquired; a bird's-eye view information acquisition process for acquiring bird's-eye view information of a view of the ground surface from the sky; a calculation process for calculating an amount of deviation between an absolute position of the bird's-eye view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; and a correction process for correcting the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated in the calculation process.

[0012] Claim 3The invention described in is characterized by comprising a sensor information acquisition unit that acquires sensor information obtained by a sensor mounted on a moving body; an absolute position information acquisition unit that acquires information regarding the absolute position of the moving body when the sensor information is acquired; a point cloud information generation unit that generates point cloud information based on the sensor information and information regarding the absolute position; an overhead information acquisition unit that acquires overhead information of a view down on the ground surface from the sky; a calculation unit that calculates the amount of deviation between the absolute position of the overhead information corresponding to a set reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point; and a correction unit that corrects the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit.

[0013] Claim 4 The invention described in the above relates to an information processing method executed by an information processing device that performs a predetermined process. in the method includes a sensor information acquisition process for acquiring sensor information obtained by a sensor mounted on a moving body; an absolute position information acquisition process for acquiring information regarding the absolute position of the moving body when the sensor information was acquired; a point cloud information generation process for generating point cloud information based on the sensor information and information regarding the absolute position; an overhead information acquisition process for acquiring overhead information overlooking the ground surface from the sky; a calculation process for calculating an amount of deviation between an absolute position of the overhead information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; and a correction process for correcting the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated in the calculation process.

[0014] Claim 5 The invention described in claim 2 or 4 The present invention is characterized in that the information processing method described above is executed by a computer.

[0015] Claim 6 The invention described in claim 5 The information processing program according to the present invention is characterized in that it stores the information processing program described above. [Brief description of the drawings]

[0016] [Figure 1] 1 is a schematic configuration diagram of a system including an information processing device according to a first embodiment of the present invention. [Diagram 2] FIG. 2 is a functional configuration diagram of a server device shown in FIG. [Diagram 3] 3 is an explanatory diagram of a method for correcting point cloud data in the server device shown in FIG. 2. [Figure 4] 3 is an explanatory diagram of a method of correcting a travel locus in the locus correction unit shown in FIG. 2. [Diagram 5] 3 is a flowchart of an information processing method in the server device shown in FIG. 2. [Figure 6] FIG. 11 is a functional configuration diagram of a server device according to a second embodiment of the present invention. [Figure 7] 7 is an explanatory diagram of the operation of a deviation amount calculation unit shown in FIG. 6. [Figure 8] FIG. 11 is a functional configuration diagram of a server device according to a third embodiment of the present invention. [Figure 9] FIG. 13 is an explanatory diagram of a new reference point. [Figure 10] FIG. 13 is a functional configuration diagram of a server device according to a fourth embodiment of the present invention. [Figure 11] FIG. 11 is an explanatory diagram of a boundary reference point. [Figure 12] 11 is a flowchart of an information processing method in the server device shown in FIG. [Figure 13] FIG. 13 is a functional configuration diagram of a server device according to a fifth embodiment of the present invention. [Figure 14] 14 is a flowchart of an information processing method in the server device shown in FIG. 13. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] An information processing device according to an embodiment of the present invention will be described below. In the information processing device according to an embodiment of the present invention, a point cloud information acquisition unit acquires point cloud information generated based on sensor information acquired by a sensor mounted on a moving body and information on the absolute position of the moving body when the sensor information was acquired, and a bird's-eye view information acquisition unit acquires bird's-eye view information of the ground surface from the sky. Then, a calculation unit calculates the amount of deviation between the absolute position of the bird's-eye view information corresponding to a set reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point, and a correction unit corrects the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit. In this way, the amount of deviation of the absolute position at the reference point can be corrected based on the bird's-eye view information. Since existing aerial photographs or the like can be used as the bird's-eye view information, there is no need to acquire high-precision data again, and the absolute position of the point cloud information can be efficiently corrected.

[0018] In addition, the positional accuracy of the overhead view information corresponding to at least the reference point may be higher than the absolute position indicated by the information on the absolute position acquired when generating the point cloud information. In this way, the absolute position of the point cloud information can be corrected based on the overhead view information.

[0019] The point cloud information acquisition unit may acquire a movement trajectory of the moving object when the sensor acquires the sensor information as information on the absolute position, and the correction unit may correct the movement trajectory based on the amount of deviation calculated by the calculation unit. By correcting the movement trajectory, it is possible to correct the absolute position of the point cloud calculated based on the movement trajectory.

[0020] The calculation unit may calculate the amount of deviation between the first and second reference points, which are the two reference points set, the absolute position of the overhead view information corresponding to the first reference point, the absolute position of the overhead view information corresponding to the second reference point, and the absolute positions of the point clouds included in the point cloud information corresponding to the first and second reference points, respectively, and the correction unit may correct the position of the point cloud information in the range between the first and second reference points based on the amount of deviation calculated by the calculation unit. In this way, the absolute position of the point cloud between the two reference points can be corrected based on the amount of deviation between the two reference points.

[0021] The apparatus may further include a determination unit that determines the accuracy of the absolute position of the point cloud information corrected by the correction unit, and if the accuracy of the absolute position determined by the determination unit is equal to or lower than a predetermined value, the calculation unit adds a third reference point, which is a new reference point, between the first reference point and the second reference point, and calculates the amount of deviation between the absolute position of the overhead view information corresponding to the third reference point and the absolute position of the point cloud included in the point cloud information corresponding to the third reference point, and the correction unit may correct the absolute position of the point cloud based on the amount of deviation between the absolute position of the overhead view information corresponding to the third reference point calculated by the calculation unit and the absolute position of the point cloud included in the point cloud information corresponding to the third reference point. By doing so, for example, even if the amount of deviation at the first reference point or the second reference point is large and sufficient accuracy cannot be ensured even after correction, re-correction can be performed to improve accuracy.

[0022] The system may further include a conversion unit that converts the point cloud information acquired by the point cloud information acquisition unit into image information, the overhead information being image information, the overhead information acquisition unit acquiring image information of a predetermined range including a reference point in the overhead information, and the calculation unit performing image matching processing between the image information of the predetermined range and the image information converted from the point cloud information to calculate the amount of deviation between the absolute position of the overhead information corresponding to the reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point. In this way, it is possible to calculate not only the deviation in the x direction, which is the horizontal axis, and the y direction, which is the vertical axis (or latitude and longitude), at the reference point, but also the deviation in the rotation angle by a known image matching processing for the range including the reference point.

[0023] In the information processing method according to one embodiment of the present invention, in a point cloud information acquisition step, point cloud information generated based on sensor information obtained by a sensor mounted on a moving body and information on the absolute position of the moving body when the sensor information was acquired is acquired, and in a bird's-eye view information acquisition step, bird's-eye view information of the ground surface viewed from the sky is acquired. Then, in a calculation step, a deviation amount between the absolute position of the bird's-eye view information corresponding to a set reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point is calculated, and in a correction step, the absolute position of the point cloud included in the point cloud information is corrected based on the deviation amount calculated in the calculation step. In this way, the deviation amount of the absolute position at the reference point can be corrected based on the bird's-eye view information. Since existing aerial photographs or the like can be used as the bird's-eye view information, there is no need to acquire high-precision data again, and the absolute position of the point cloud information can be efficiently corrected.

[0024] Moreover, an information processing program according to an embodiment of the present invention causes a computer to execute the above-mentioned information processing method. In this way, it is possible to correct the amount of deviation of the absolute position at the reference point using the overhead information as a reference, using a computer. Since existing aerial photographs, etc. can be used as the overhead information, there is no need to obtain high-precision data again, and the absolute position of the point cloud information can be corrected efficiently.

[0025] The information processing program may be stored in a computer-readable storage medium, which allows the program to be distributed as a standalone program rather than being incorporated into a device, and allows for easy version upgrades.

[0026] In addition, in the information processing device according to another embodiment of the present invention, the sensor information acquisition unit acquires sensor information obtained by a sensor mounted on a moving body, the absolute position information acquisition unit acquires information on the absolute position of the moving body when the sensor information is acquired, the point cloud information generation unit generates point cloud information based on the sensor information and the information on the absolute position, and the bird's-eye view information acquisition unit acquires bird's-eye view information of the ground surface viewed from the sky. Then, the calculation unit calculates the amount of deviation between the absolute position of the bird's-eye view information corresponding to the set reference point and the absolute position of the point cloud included in the point cloud information corresponding to the reference point, and the correction unit corrects the absolute position of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit. In this way, the amount of deviation of the absolute position at the reference point can be corrected based on the bird's-eye view information. Since existing aerial photographs, etc. can be used as the bird's-eye view information, there is no need to acquire high-precision data again, and the absolute position of the point cloud information can be efficiently corrected. In addition, the configuration for generating the point cloud information does not need to be integrated, and may be separated into, for example, a moving body such as a vehicle and a server device, etc.

[0027] In addition, in the information processing method according to another embodiment of the present invention, in a sensor information acquisition step, sensor information obtained by a sensor mounted on a moving body is acquired, in an absolute position information acquisition step, information on the absolute position of the moving body when the sensor information is acquired is acquired, in a point cloud information generation step, point cloud information is generated based on the sensor information and information on the absolute position, and in a bird's-eye view information acquisition step, bird's-eye view information of a ground surface viewed from the sky is acquired. Then, in a calculation step, a deviation amount between an absolute position of the bird's-eye view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point is calculated, and in a correction step, the absolute position of the point cloud included in the point cloud information is corrected based on the deviation amount calculated in the calculation step. In this way, the deviation amount of the absolute position at the reference point can be corrected based on the bird's-eye view information. Since an existing aerial photograph or the like can be used as the bird's-eye view information, there is no need to acquire high-precision data again, and the absolute position of the point cloud information can be efficiently corrected. In addition, the configuration for generating the point cloud information does not need to be integrated, and may be separated into, for example, a moving body such as a vehicle and a server device.

[0028] Moreover, the above-mentioned information processing method is executed by a computer. In this way, the computer can be used to correct the amount of deviation of the absolute position at the reference point based on the overhead information. Since existing aerial photographs or the like can be used as the overhead information, there is no need to obtain high-precision data anew, and the absolute position of the point cloud information can be corrected efficiently. Furthermore, the configuration for generating the point cloud information does not have to be integrated, and may be separated into, for example, a moving body such as a vehicle and a server device.

[0029] The information processing program may be stored in a computer-readable storage medium, which allows the program to be distributed as a standalone program rather than being incorporated into a device, and allows for easy version upgrades. EXAMPLES

[0030] An information processing device according to a first embodiment of the present invention will be described with reference to Fig. 1 to Fig. 8. In Fig. 1, the information processing device according to this embodiment is configured as a server device 10. The server device 10 acquires data (information) obtained by a sensor 1 mounted on a vehicle V as a moving body via a network N such as the Internet, and corrects position information (absolute positions) of a group of points included in the data based on an aerial photograph described below.

[0031] The sensor 1 includes a LiDAR (Light Detection and Ranging), a GPS receiver, etc. The sensor 1 may include an on-board camera that captures an image of the surroundings, such as the front of the vehicle.

[0032] A lidar measures the distance to an object in the outside world discretely by emitting a laser beam as an electromagnetic wave. A pulsed laser is output in a predetermined detection area while changing the output direction, and the reflected wave of the laser is received to generate sensor information. The lidar outputs multiple pulses of a laser within the detection area, and generates a point cloud based on the reflected wave of the multiple pulses of the laser. Each point that constitutes the point cloud at this point contains information indicating the output direction of the laser, the distance to the object that reflected the laser, and the intensity of the reflected wave (reflection intensity) as sensor data.

[0033] As is well known, the GPS receiver periodically receives radio waves transmitted from a plurality of GPS satellites to detect current position information (latitude and longitude) and time. That is, this current position information becomes the absolute position of the vehicle V. Note that the detection of the absolute position is not limited to the GPS receiver, and other sensors such as an IMU (inertial measurement unit) may also be used. Then, this absolute position data and the above-mentioned sensor data are associated with each other and transmitted to the server device 10.

[0034] 2 shows the functional configuration of the server device 10. The server device 10 includes a point cloud integration unit 11, a trajectory extraction unit 12, a reference point acquisition unit 13, a deviation amount calculation unit 14, a trajectory correction unit 15, and a point cloud correction unit 16. These functions are realized by a CPU of the server device 10 executing a program (information processing program) stored in a storage device such as a hard disk of the server device 10.

[0035] The server device 10 also stores sensor data 21, absolute position data 22, aerial photograph data 23, and corrected point cloud data 24. These data are stored in a storage device such as a hard disk of the server device 10. The sensor data 21 and absolute position data 22 are data transmitted from the sensor 1. That is, the sensor data 21 is sensor information, and the absolute position data is information related to the absolute position. The sensor data 21 and the absolute position data 22 are associated with absolute position data at the absolute position where the sensor data was obtained. Note that these data are not limited to being directly acquired from the sensor 1, and may be acquired from another server device outside the server device 10.

[0036] The aerial photograph data 23 is photograph data taken by a camera from a flying object such as an aircraft while looking down on the ground surface. It is not limited to aerial photographs, but may be satellite photographs or point cloud data acquired by an aerial lidar (a lidar is mounted on a flying object such as an aircraft and irradiates laser light toward the ground surface). In other words, the aerial photograph data 23 is bird's-eye view information.

[0037] The corrected point cloud data 24 is point cloud data (data in which the positions of the point cloud are absolute positions) whose absolute positions have been corrected by a point cloud correction unit 16, which will be described later. That is, the point cloud data becomes point cloud information.

[0038] The point cloud integration unit 11 first obtains the absolute position of the point cloud obtained by the lidar based on the sensor data 21 and the absolute position data 22, and generates and acquires point cloud data consisting of the absolute positions of the point cloud. That is, the point cloud integration unit 11 functions as a point cloud information acquisition unit that acquires point cloud data. The point cloud integration unit 11 also integrates multiple pieces of point cloud data obtained by the vehicle V traveling the same route multiple times. The integration method may be, for example, selecting a representative one from the multiple pieces of trajectory and point cloud for the traveling trajectory and point cloud of the vehicle V obtained from the absolute position, or averaging the multiple pieces of trajectory and point cloud. When integrating multiple pieces, a method such as ICP (Iterative Closest Point) is used at a position that is commonly present between the point clouds to be integrated, so that there is no inconsistency as a point cloud. In this embodiment, the processing described below may be performed based on point cloud data obtained in one travel without traveling multiple times. In the case of point cloud data obtained in one travel, the above-mentioned integration process can be omitted.

[0039] The trajectory extraction unit 12 extracts (generates) a traveling trajectory (movement trajectory) of the vehicle V based on the absolute position data acquired by the point cloud integration unit 11. That is, the trajectory extraction unit 12 functions as a part of a point cloud information acquisition unit that extracts (acquires) the traveling trajectory as information related to the absolute position.

[0040] The reference point acquisition unit 13 acquires data indicating a reference point, which will be described later. The data indicating the reference point may be acquired from an external source, or may be acquired by generating it by the reference point acquisition unit 13 itself based on the aerial photograph data 23. Alternatively, the operator or the like may set it manually by referring to the aerial photograph or the like. The reference point, which will be described in detail later, is a position that is used as a reference when the deviation amount calculation unit 14 calculates the deviation amount between the point cloud data and the aerial photograph data 23, and can be set at a characteristic point on the map, such as the center of an intersection or a corner of an intersection (a point where lines indicating the widthwise ends of intersecting roads intersect with each other).

[0041] The deviation amount calculation unit 14 calculates the amount of deviation between the aerial photograph data 23 at the reference point and the point cloud data based on the point cloud data integrated by the point cloud integration unit 11, the data indicating the reference point acquired by the reference point acquisition unit 13, and the aerial photograph data 23. The calculation of the amount of deviation will be described in detail later. The deviation amount calculation unit 14 functions as an overhead information acquisition unit that acquires overhead information.

[0042] The trajectory correction unit 15 corrects the travel trajectory extracted by the trajectory extraction unit 12 based on the deviation amount calculated by the deviation amount calculation unit 14 .

[0043] The point cloud correction unit 16 corrects the absolute positions of the point cloud data generated by the point cloud integration unit 11 based on the traveling trajectory corrected by the trajectory correction unit 15 to generate corrected point cloud data 24. The corrected point cloud data 24 may include the traveling trajectory corrected by the trajectory correction unit 15.

[0044] Next, a method of correcting point cloud data in the server device 10 having the above-mentioned configuration will be described with reference to Fig. 3. The left side of Fig. 3 shows sensor data 21 (point cloud) acquired on a certain route traveled by the vehicle V, and the travel trajectory of the vehicle V (dashed line in Fig. 3) based on absolute position data 22. The right side of Fig. 3 shows an aerial photograph corresponding to the range on the left side of Fig. 3.

[0045] Here, the absolute position of the point cloud data may have a reduced position measurement accuracy due to the effects of multipath in GPS, etc. On the other hand, the position accuracy of some aerial photographs corresponding to reference points is higher than the measurement accuracy of GPS, and in this embodiment, an aerial photograph having a higher position accuracy than the absolute position of such point cloud data is used. An example of such aerial photograph data is the Geospatial Information Authority of Japan map (aerial photograph). Note that the position accuracy of the aerial photograph needs to be higher at least at the reference point than the position accuracy of the absolute position acquired by the sensor 1.

[0046] Furthermore, P and Q in the figure are reference points. The absolute position (latitude and longitude) of this reference point PQ can be found on the left side of Fig. 3 from the absolute position of the point cloud corresponding to the reference point included in the point cloud data. The right side of Fig. 3 (aerial photograph) is an image with location information, such as GeoTIFF, that includes the latitude and longitude of the four corners of the photograph. In such an image with location information, if the accuracy (resolution) is known in advance, the absolute positions (latitude and longitude) of the reference points P and Q can be found from this information.

[0047] Then, the deviation amount calculation unit 14 calculates the difference (deviation amount) between the absolute positions at the calculated reference points P and Q. That is, the deviation amount calculation unit 14 functions as a calculation unit that calculates the deviation amount between the two set reference points, that is, the reference point P (first reference point) and the reference point Q (second reference point), the absolute position of the aerial photograph data 23 (bird's-eye view information) corresponding to the reference point P (first reference point), the absolute position of the aerial photograph data 23 (bird's-eye view information) corresponding to the reference point Q (second reference point), and the absolute positions of the point clouds included in the point cloud data (point cloud information) corresponding to the reference point P (first reference point) and the reference point Q (second reference point), respectively.

[0048] Then, the trajectory correction unit 15 corrects the travel trajectory between the reference points PQ based on the deviation amount, and the point cloud correction unit 16 corrects the absolute position of the sensor information (point cloud) based on the correction result of the travel trajectory. That is, the trajectory correction unit 15 and the point cloud correction unit 16 function as a correction unit that corrects the position of the point cloud information in the range sandwiched between the reference point P (first reference point) and the reference point Q (second reference point) based on the deviation amount calculated by the deviation amount calculation unit 14 (calculation unit).

[0049] The correction of the travel trajectory based on the reference point will be described with reference to Fig. 4. Regarding the absolute position of the reference point PQ acquired from the sensor data 21 and the absolute position data 22 (Fig. 4(a)), the deviation amount from the absolute position of the reference point P'Q' calculated from the aerial photograph data 23 is calculated (Fig. 4(b)). Then, the travel trajectory between the reference points PQ is corrected by, for example, linear or nonlinear interpolation, assuming that the reference point PQ is shifted to the reference point P'Q' (Fig. 4(c)). In Fig. 4(c), the solid line indicates the travel trajectory after correction, and the dashed line indicates the travel trajectory before correction.

[0050] Next, the operation (information processing method) of the server device 10 described above will be described with reference to the flowchart in Fig. 5. First, the point cloud integration unit 11 integrates point cloud data for multiple runs from the sensor data 21 and the absolute position data 22 (step S11), then the deviation amount calculation unit 14 calculates the deviation amount of the absolute position at the reference point (step S12), then the trajectory correction unit 15 corrects the running trajectory between the reference points (step S13), and finally the point cloud correction unit 16 corrects the point cloud (step S14).

[0051] That is, step S11 functions as a point cloud information acquisition step, step S12 functions as a calculation step, and steps S13 and S14 function as correction steps.

[0052] From the viewpoint of constraint conditions for correction, it is preferable to set two reference points as described above, but one may be used. In the case of one reference point, the travel trajectory may be corrected using the deviation amount at the one reference point as the correction amount. That is, the deviation amount calculation unit 14 may calculate the deviation amount between the absolute position of the aerial photograph data 23 (bird's-eye view information) corresponding to the set reference point and the absolute position of the point cloud included in the point cloud data (point cloud information) corresponding to the reference point. Then, the trajectory correction The shift amount calculation unit 15 and the point cloud correction unit 16 may correct the absolute positions of the point cloud included in the point cloud data (point cloud information) based on the amount of deviation calculated by the deviation amount calculation unit 14 (calculation unit).

[0053] Moreover, it is preferable that the reference point is used only for correcting the point cloud involved in the calculation of the deviation amount at the reference point. For example, in the case of FIG. 3, if the point cloud used for calculating the deviation amount at the reference point P and the point cloud used for calculating the deviation amount at the reference point Q are acquired by different vehicles, correction (linear interpolation, etc.) between the reference points PQ is not performed. This is because when the sensor information is acquired by different vehicles, there is a possibility that an error occurs between the data due to differences in the performance of the sensor itself, such as a lidar, and the mounting conditions of the sensor. In this case, correction is performed based on the one reference point described above or other reference points included in the point cloud data acquired by the same vehicle.

[0054] According to this embodiment, the server device 10 has a point cloud integration unit 11 that acquires point cloud data generated based on sensor data acquired by a sensor 1 mounted on a vehicle V and absolute position data of the vehicle V when the sensor data was acquired, and a deviation calculation unit 14 acquires aerial photograph data overlooking the ground surface from the sky. Then, the deviation calculation unit 14 calculates the deviation between the absolute position on the aerial photograph data corresponding to the set reference point PQ and the absolute position of the point cloud included in the point cloud data corresponding to the reference point PQ, and the trajectory correction unit 15 corrects the running trajectory extracted from the point cloud data by the trajectory extraction unit 12 based on the deviation calculated by the deviation calculation unit 14, and corrects the absolute position of the point cloud of the point cloud data based on the correction result of the trajectory correction unit 15. In this way, the deviation amount of the absolute position at the reference point PQ can be corrected using the aerial photograph as a reference. Since existing aerial photographs, etc. can be used in this way, there is no need to acquire high-precision data again, and the absolute position of the point cloud information can be efficiently corrected.

[0055] Furthermore, since the trajectory correction unit 15 corrects the travel trajectory based on the deviation amount calculated by the deviation amount calculation unit 14, it is possible to correct the absolute position of the point cloud calculated based on the travel trajectory.

[0056] In addition, the deviation amount calculation unit 14 calculates the amount of deviation between the absolute positions on the aerial photograph corresponding to each of the two reference points PQ and the absolute positions of the point clouds included in the point cloud data corresponding to each of the two reference points PQ. The trajectory correction unit 15 corrects the driving trajectory extracted from the point cloud data by the trajectory extraction unit 12 based on the amount of deviation calculated by the deviation amount calculation unit 14, and corrects the absolute positions of the point clouds in the point cloud data based on the correction result of the trajectory correction unit 15. Therefore, the absolute positions of the point clouds between the two reference points can be corrected based on the amount of deviation between the two reference points.

[0057] In addition, the positional accuracy of the aerial photograph corresponding to at least the reference point is higher than the absolute position acquired when generating the point cloud data, so that the absolute position of the point cloud information can be corrected based on the aerial photograph. EXAMPLES

[0058] Next, an information processing device according to a second embodiment of the present invention will be described with reference to Figures 6 and 7. Note that the same parts as those in the first embodiment described above are given the same reference numerals and the description thereof will be omitted.

[0059] Fig. 6 shows a functional configuration of a server device 10A as an information processing device according to this embodiment. The server device 10A shown in Fig. 6 includes a point cloud processing unit 17 added thereto, and the deviation amount calculation unit 14 is changed to a deviation amount calculation unit 14A.

[0060] The point cloud processing unit 17 converts the point cloud data generated by the point cloud integrating unit 11 into two-dimensional image data viewed from above. That is, the point cloud processing unit 17 functions as a conversion unit that converts the point cloud data (point cloud information) into image information.

[0061] The deviation amount calculation unit 14A performs image adjustments such as edge extraction and blurring on the aerial photograph data 23 and the two-dimensional image data converted by the point cloud processing unit 17, and then performs a well-known image template matching process. For example, the center position of the range matched by the image template matching is set as a reference point, and the deviation amount between the aerial photograph data 23 and the point cloud data (two-dimensional image data) at the reference point is calculated. The process of the deviation amount calculation unit 14A will be described with reference to FIG. 7.

[0062] The leftmost image in Fig. 7 is an aerial photograph. As with the first embodiment, the absolute positions of the four corners and the resolution of this aerial photograph are known in advance. In other words, it is image information of a specified range including a reference point in the overhead view information. The second image from the left in Fig. 7 is an image obtained by performing edge extraction and blurring processing on the aerial photograph.

[0063] The rightmost image in Fig. 7 shows point cloud data converted into two-dimensional image data. In this example, a point cloud categorized on the ground surface is imaged, but other methods may be used, such as extracting and imaging a point cloud up to a certain height from the road surface. The second image from the right in Fig. 7 shows an image in which edge extraction and blurring have been performed on the two-dimensional image data on the rightmost image.

[0064] Then, the geodetic system is unified between the image obtained by edge extraction and blurring of the aerial photograph and the image obtained by edge extraction and blurring of the two-dimensional image data, the scale is adjusted, and a comparison is made by template matching using the aerial photograph as a template image, for example.The amount of deviation (vertical and horizontal deviation, rotation angle, etc.) at a reference point defined within the range of the matching images is then calculated.

[0065] That is, the deviation calculation unit 14A (calculation unit) performs template matching (image matching processing) between the aerial photograph data 23 (image information of a specified range) and image information converted from the point cloud data (point cloud information) to calculate the deviation amount between the absolute position of the aerial photograph data 23 (bird's-eye view information) corresponding to the reference point and the absolute position of the point cloud included in the point cloud data (point cloud information) corresponding to the reference point.

[0066] Such image template matching processing is performed at two locations. In other words, if the above-mentioned image template matching processing corresponds to, for example, the calculation of the amount of deviation at the reference point Q in Fig. 3, then by performing the image template matching processing on the aerial photograph corresponding to the reference point P and an image obtained by converting the point cloud data into two-dimensional image data, correction can be performed in the same manner as the correction of the traveling trajectory between the reference points P and Q described in the first embodiment.

[0067] According to this embodiment, the point cloud processing unit 17 converts the point cloud data generated by the point cloud integration unit 11 into a two-dimensional image. The deviation amount calculation unit 14A acquires an image of a predetermined range including a reference point in the aerial photograph data 23, and performs image matching processing between the image of the predetermined range and a two-dimensional image converted from the point cloud data to calculate the deviation amount between the absolute position of the aerial photograph corresponding to the reference point and the absolute position of the point cloud included in the point cloud data corresponding to the reference point. In this way, not only the deviation in the vertical and horizontal directions (latitude and longitude) at the reference point but also the deviation in the rotation angle can be calculated by a well-known image matching processing for the range including the reference point. EXAMPLES

[0068] Next, an information processing device according to a third embodiment of the present invention will be described with reference to Figures 8 and 9. Note that the same parts as those in the first and second embodiments described above are given the same reference numerals and the description thereof will be omitted.

[0069] Fig. 8 shows a functional configuration of a server device 10B as an information processing device according to this embodiment. The server device 10B shown in Fig. 8 is obtained by adding an error evaluation unit 18 to the server device 10B shown in Fig. 6, and changing the reference point acquisition unit 13 to a reference point acquisition unit 13A. The configuration in Fig. 8 is based on the configuration in Fig. 6, but may also be based on the configuration in Fig. 2.

[0070] The error evaluation unit 18 evaluates (judges) the error (accuracy) of the driving trajectory corrected by the trajectory correction unit 15. The contents of this evaluation include, for example, a case where the absolute position is incorrect, such as when the corrected driving trajectory is on a building, or a case where the deviation between the driving trajectory before and after correction is equal to or greater than a predetermined threshold. In this embodiment, these contents are defined as a state where the accuracy of the absolute position is equal to or less than a predetermined value. That is, the error evaluation unit 18 functions as a judgment unit that judges the accuracy of the absolute position of the point cloud data (point cloud information) corrected by the trajectory correction unit 15 and the point cloud correction unit 16 (correction unit).

[0071] Then, if the evaluation result indicates that there is an abnormality in the absolute position or that the error is greater than a predetermined threshold, that is, the accuracy of the absolute position is equal to or less than a predetermined value, the reference point acquisition unit 13A is instructed to acquire a new reference point, which will be described later. That is, a third reference point is added. If the error is smaller than the threshold, a new reference point is not added, and the point cloud correction unit 16 corrects the absolute positions of the point cloud based on the corrected traveling trajectory, as in the first and second embodiments.

[0072] In addition to acquiring the reference points described in the first embodiment, the reference point acquiring unit 13A acquires new reference points in response to an instruction from the error evaluating unit 18. The new reference points will be described with reference to FIG.

[0073] The upper part of Fig. 9 shows reference point P, reference point Q, and a travel trajectory connecting these reference points PQ. In this case, if the travel trajectory corrected by the trajectory correction unit 15 is evaluated by the error evaluation unit 18 to have an error greater than a predetermined threshold, for example, a new reference point R is set on the travel trajectory between the reference points PQ. This reference point R becomes a third reference point. Then, the deviation amount calculation unit 14A calculates the amount of deviation between the absolute position of the aerial photograph corresponding to the reference point R and the absolute position of the point cloud included in the point cloud data corresponding to the reference point R.

[0074] Then, based on the calculated deviation amount at reference point R and the already calculated deviation amount at reference point P, the trajectory correction unit 15 corrects the travel trajectory between reference points PR in a manner similar to that of the first embodiment, and the trajectory correction unit 15 corrects the travel trajectory between reference points RQ in a manner similar to that of the first embodiment. Then, the error of the corrected travel trajectory is evaluated again, and a new reference point is recursively set until it becomes equal to or less than the threshold value (for example, a new reference point is set between PR).

[0075] The new reference point is set on the driving trajectory before correction, not on the driving trajectory after correction. The setting position of the new reference point may be the midpoint of two reference points, a curve or corner (topographical inflection point), or a position that reflects a weight according to the reliability of the reference point.

[0076] According to this embodiment, the trajectory correction unit 15 further includes an error evaluation unit 18 that evaluates the error of the travel trajectory corrected by the trajectory correction unit 15. When the error of the travel trajectory determined by the error evaluation unit is greater than a threshold value, the reference point acquisition unit 13A adds, for example, a reference point R, which is a new reference point, between the reference points PQ, and the deviation amount calculation unit 14A calculates the deviation amount between the absolute position of the aerial photograph corresponding to the reference point R and the absolute position of the point cloud included in the point cloud data corresponding to the reference point R. Then, the trajectory correction unit 15 corrects the trajectory between the reference points PR based on the deviation amount at the reference point R and the deviation amount at the reference point P calculated by the deviation amount calculation unit 14A. By doing so, even if the deviation amount at the reference point PQ is large and sufficient accuracy cannot be ensured even after correction, re-correction can be performed to improve accuracy.

[0077] In the first to third embodiments described above, the point cloud information acquisition unit is configured to acquire sensor information and information related to absolute position to generate point cloud information, but it may be divided into three blocks: a block that acquires point cloud information, a block that acquires information related to absolute position, and a block that generates point cloud information. EXAMPLES

[0078] Next, an information processing device according to a fourth embodiment of the present invention will be described with reference to Figures 10 to 12. Note that the same parts as those in the first to third embodiments described above are given the same reference numerals and the description thereof will be omitted.

[0079] This embodiment is an embodiment in which updates are performed when the accuracy of the point cloud corrected by the methods of the first to third embodiments deteriorates due to changes in road surface conditions, for example due to road construction, or when periodic updates are performed depending on the season.

[0080] 10 shows the functional configuration of a server device 10C as an information processing device according to this embodiment. The server device 10C includes a data acquisition unit 31, an error deterioration determination unit 32, a boundary reference point generation unit 33, a point cloud matching unit 34, a trajectory correction unit 35, and a point cloud correction unit 36. As in the first to third embodiments, these functions are realized by a CPU of the server device 10C executing a program (information processing program) stored in a storage device such as a hard disk of the server device 10C.

[0081] The server device 10C also stores the corrected point cloud data 24, an error DB 41, and updated point cloud data 42. The corrected point cloud data 24 as the first point cloud information is point cloud data corrected by the method described in the first embodiment (see FIG. 2). The corrected point cloud data 24 also has information about the reference point used when generating the data. The information about the reference point is not limited to the absolute position of the reference point, but may be information indicating a characteristic point on the map indicating the reference point (such as a corner of an intersection).

[0082] The error DB 41 is a database in which the current accuracy of the corrected point cloud data 24 is stored for each specified area. If correction is performed using aerial photograph data with the same accuracy for all areas, the data may simply be accuracy information rather than a database. The updated point cloud data 42 is point cloud data whose absolute positions have been corrected (updated) by the point cloud correction unit 36 ​​described below.

[0083] The data acquisition unit 31 acquires new sensor data and absolute position data, for example, from the vehicle V, another vehicle, another server device, etc. Then, similar to the point cloud integrating unit 11, new point cloud data is generated as second point cloud information based on the sensor data and the absolute position data. That is, the data acquisition unit 31 functions as a second acquisition unit that acquires second point cloud information that is newer than the first point cloud information.

[0084] In addition, the sensor data etc. acquired by the data acquisition unit 31 need only be acquired within a specified range, and therefore may not be limited to data acquired from a vehicle but may be data acquired from, for example, a lidar installed on a roadside.

[0085] The error deterioration determination unit 32 determines whether or not the error (accuracy) of the corrected point cloud data 24 has deteriorated based on the corrected point cloud data 24, the point cloud data acquired by the data acquisition unit 31, and the error DB 41. That is, the error deterioration determination unit 32 functions as a first acquisition unit that acquires first point cloud information.

[0086] Furthermore, if the error has worsened, the error deterioration determination unit 32 instructs the boundary reference point generation unit 33 to generate a boundary reference point, which will be described later. As a method of determining the error deterioration, for example, if the absolute position of a specific point indicated by the new point cloud data is outside the range in which the absolute position of the same point in the corrected point cloud data 24 is included in the error DB 41 and the error of the position is considered, it is determined that the error has worsened. Specifically, for example, if the absolute position of a certain point in the corrected point cloud data 24 is (X, Y) and the error of the position in the error DB 41 is 20 cm, it is determined that the error has worsened if the absolute position of the same point in the new point cloud data exceeds (X±20 cm, Y±20 cm). Furthermore, the error deterioration may be determined in a system other than the metric system (such as latitude and longitude difference). That is, the error deterioration determination unit 32 functions as a part of an update unit that performs an update based on the comparison result between the first point cloud information and the second point cloud information.

[0087] When the error deterioration determination unit 32 determines that the error has deteriorated, the boundary reference point generation unit 33 generates a boundary reference point at the boundary between the area indicated by the new point cloud data (the area to be updated) and the existing point cloud data (the corrected point cloud data 24). The boundary reference point will be described with reference to FIG.

[0088] The upper part of Fig. 11 shows that an area where the accuracy of the point cloud data has deteriorated, that is, the error has worsened, has occurred between the reference points PQ. The middle part of Fig. 11 shows that boundary reference points S1 and S2 have been set (generated) at the boundary between the area indicated by the new point cloud data (area to be updated) and the existing point cloud data (existing driving trajectory). These boundary reference points may be set at any point in the boundary area using the aerial photograph data described in the first to third embodiments. Note that the area between the boundary reference points S1 and S2 is new point cloud data (new driving trajectory). In other words, the new point cloud data is inserted as is between the boundary reference points S1 and S2.

[0089] The point cloud matching unit 34 performs point cloud matching between the existing point cloud data and the new point cloud data at the boundary portion between the existing point cloud data and the new point cloud data, including the boundary reference points S1 and S2. For point cloud matching, for example, ICP is used to match points at the same position at the point level.

[0090] The trajectory correction unit 35 corrects the traveling trajectory between the boundary reference point S1, which has been subjected to matching processing by the point cloud matching unit 34, and the reference point P, and similarly corrects the traveling trajectory between the boundary reference point S2 and the reference point Q (the dashed line portion in the lower part of FIG. 11). That is, the absolute position of the point cloud of the area adjacent to the area to be updated is corrected. In this embodiment, since the absolute position of the boundary reference point S1 is shifted between the existing point cloud data and the new point cloud data due to the new point cloud data, the traveling trajectory is corrected based on the amount of shift in the same manner as in the first embodiment.

[0091] The point cloud correction unit 36 ​​corrects the absolute positions of the existing point cloud data based on the traveling trajectory corrected by the trajectory correction unit 35, and generates updated point cloud data 42. Note that the updated point cloud data 42 may also include the traveling trajectory corrected by the trajectory correction unit 35.

[0092] As is clear from the above explanation, as new reference points, reference point S1 (third reference point) is set on the boundary on the reference point P (first reference point) side, and reference point S2 (fourth reference point) is set on the boundary on the reference point Q (second reference point) side, and the boundary reference point generation unit 33, point cloud matching unit 34, trajectory correction unit 35, and point cloud correction unit 36 ​​function as update units that correct the absolute position of the point cloud between reference point P (first reference point) and reference point S1 (third reference point) and the absolute position of the point cloud between reference point Q (second reference point) and reference point S2 (fourth reference point).

[0093] Next, the operation (information processing method) of the server device 10C described above will be described with reference to the flowchart of Fig. 12. First, the data acquisition unit 31 acquires new point cloud data (step S21), and the error deterioration determination unit 32 judges whether the error of the corrected point cloud data 24 has deteriorated based on the corrected point cloud data 24, the point cloud data acquired by the data acquisition unit 31, and the error DB 41 (step S22). If the error has deteriorated as a result of the judgment (if YES in step S22), the boundary reference point generation unit 33 generates a boundary reference point (step S23). On the other hand, if the error has not deteriorated (if NO in step S22), the process returns to step S21.

[0094] Next, after generating the boundary reference points, the point cloud matching unit 34 performs point cloud matching between the existing point cloud data and the new point cloud data (step S24), and the trajectory correction unit 35 corrects the traveling trajectory between the boundary reference point S1, which has been subjected to the matching process in the point cloud matching unit 34, and the reference point P, and between the boundary reference point S2 and the reference point Q (step S25). Then, the point cloud correction unit 36 ​​corrects the traveling trajectory between the boundary reference point S1, which has been subjected to the matching process in the point cloud matching unit 34, and the reference point P, and similarly corrects the traveling trajectory between the boundary reference point S2 and the reference point Q (step S26).

[0095] That is, step S21 functions as a second acquisition step, step S22 functions as a first acquisition step, and steps S23 to S26 function as an update step.

[0096] In the explanation of Figure 11, only the parts where the point cloud accuracy has deteriorated are updated with new point cloud data, but if data between reference points PQ has been acquired as new point cloud data, the parts between reference points PQ may be replaced with the new point cloud data and updated.

[0097] Also, between the boundary reference points S1 and S2, the travel trajectory and point cloud may be corrected based on aerial photograph data as in the first embodiment.

[0098] According to this embodiment, in the server device 10C, the error deterioration determination unit 32 acquires existing point cloud data including at least information on the first reference point and the second reference point, and the data acquisition unit 31 acquires new point cloud information. Then, the error deterioration determination unit 32 judges the deterioration of the error, the boundary reference point generation unit generates a boundary reference point at the boundary of the update target area, and the point cloud matching unit 34 matches the existing point cloud at the boundary reference point with the new point cloud. Then, the trajectory correction unit 35 corrects the running trajectory between the boundary reference point and the existing reference point, and the point cloud correction unit 36 ​​corrects the point cloud based on the correction of the running trajectory. In this way, the point cloud can be updated based on the reference point. Therefore, it is possible to update only the necessary area without performing a wide-ranging update such as replacing all the point clouds at once.

[0099] Furthermore, the error deterioration determination unit 32 updates the existing point cloud data with the new point cloud data if it determines that the error has worsened as a result of comparing the existing point cloud data with the new point cloud data. In this way, it is possible to detect a large difference between the existing point cloud data and the new point cloud data and update the data.

[0100] The update unit also sets boundary reference points S1 and S2 on the boundaries of the update target area included between the two reference points PQ, and corrects the travel trajectory (absolute position of the point cloud) of the point cloud between the reference point P and the boundary reference point S1 and the travel trajectory (absolute position of the point cloud) of the point cloud between the reference point Q and the boundary reference point S2 based on the reference point PQ and the boundary reference points S1 and S2. In this way, it is possible to correct the absolute position of the adjacent areas affected by the update of the update target area.

[0101] Furthermore, the boundary reference point generating unit inserts new point cloud data into the update target area. In this way, it is possible to insert new point cloud information into the update target area and update it. EXAMPLES

[0102] Next, an information processing device according to a fifth embodiment of the present invention will be described with reference to Figures 13 and 14. Note that the same parts as those in the first to fourth embodiments described above are given the same reference numerals and the description thereof will be omitted.

[0103] Fig. 13 shows a functional configuration of a server device 10D as an information processing device according to this embodiment. In the server device 10D shown in Fig. 13, the error deterioration determination unit 32 in Fig. 10 is changed to an update time determination unit 37, and the error DB 41 is omitted.

[0104] In this embodiment, the timing of the update is not due to a deterioration in error (precision) caused by new point cloud data acquired by the data acquisition unit, but is performed at a predetermined time, for example, every month or every season. The update time determination unit 37 determines whether or not it is such a predetermined time (update time), and if it is determined that it is time to update, causes the boundary reference point generation unit 33 and the like to perform update processing. That is, updating is performed based on a predetermined time. The processing after the boundary reference point generation unit 33 is the same as in the fourth embodiment.

[0105] Fig. 14 is a flowchart of the operation (information processing method) of the server device 10D according to the present embodiment. In Fig. 14, step S31 is the same as step S21 in Fig. 12, step S33 is the same as step S23 in Fig. 12, step S34 is the same as step S24 in Fig. 12, step S35 is the same as step S25 in Fig. 12, and step S36 is the same as step S26 in Fig. 12.

[0106] In step S32, the update time determination unit 37 determines whether it is time to update or not, and if it is time to update (YES), the process proceeds to step S33, and if it is not time to update (NO), the process returns to step S31.

[0107] According to this embodiment, the update time determination unit 37 updates the point cloud data based on a predetermined update time, so that updating can be performed at a specific time or at regular intervals.

[0108] Furthermore, the present invention is not limited to the above-mentioned embodiment. In other words, a person skilled in the art can implement the present invention by modifying it in various ways according to the conventionally known knowledge without departing from the gist of the present invention. As long as the information processing device of the present invention is still included even after such modification, it is of course included in the scope of the present invention. [Explanation of symbols]

[0109] 10, 10A, 10B, 10C, 10D Server device (information processing device) 11 Point cloud integration unit (point cloud information acquisition unit) 12 Trajectory extraction part 13 Reference point acquisition part 14. Displacement calculation unit (overhead view information acquisition unit, displacement calculation unit) 15 Trajectory correction section (correction section) 16 Point cloud correction section (correction section) 17 Point cloud processing section (conversion section) 18 Error evaluation section (judgment section) 21 Sensor data (sensor information) 22 Absolute position data (information about absolute position) 23 Aerial photograph data (bird's-eye view information) 24 Corrected point cloud data 31 Data acquisition unit (second acquisition unit) 32 Error deterioration determination unit (first acquisition unit) 33 Boundary reference point generation section (update section) 34 Point cloud matching section (update section) 35 Trajectory correction section (update section) 36 Point cloud correction section (update section) 37 Update timing determination section (update section) 41 Error DB 42 Updated point cloud data P reference point (1st reference point) Q reference point (second reference point) R reference point (3rd reference point) S1 Boundary reference point (3rd reference point) S2 Boundary reference point (4th reference point) V Vehicle (moving object)

Claims

1. a point cloud information acquisition unit that acquires point cloud information generated based on sensor information acquired by a sensor mounted on a moving object and information relating to an absolute position of the moving object when the sensor information is acquired; an overhead information acquisition unit that acquires overhead information of an earth's surface from the sky; a calculation unit that calculates a deviation amount between an absolute position of the overhead view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; a correction unit that corrects absolute positions of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit, the calculation unit calculates a deviation amount between the two set reference points, that is, a first reference point and a second reference point, an absolute position of the overhead view information corresponding to the first reference point and an absolute position of the overhead view information corresponding to the second reference point, and an absolute position of a point cloud included in the point cloud information corresponding to the first reference point and the second reference point, respectively; The information processing apparatus according to claim 1, wherein the correction unit corrects the position of the point cloud information in the range between the first reference point and the second reference point based on the amount of deviation calculated by the calculation unit.

2. An information processing method executed by an information processing device that performs a predetermined process, a point cloud information acquisition step of acquiring point cloud information generated based on sensor information acquired by a sensor mounted on a moving object and information relating to an absolute position of the moving object when the sensor information was acquired; an overhead information acquisition step of acquiring overhead information of an earth's surface from the sky; a calculation step of calculating a deviation amount between an absolute position of the overhead view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; a correction step of correcting absolute positions of the point cloud included in the point cloud information based on the deviation amount calculated in the calculation step, In the calculation step, a deviation amount is calculated between the two set reference points, that is, a first reference point and a second reference point, an absolute position of the overhead view information corresponding to the first reference point and an absolute position of the overhead view information corresponding to the second reference point, and an absolute position of a point cloud included in the point cloud information corresponding to the first reference point and the second reference point, respectively; An information processing method, characterized in that in the correction process, the position of the point cloud information in the range between the first reference point and the second reference point is corrected based on the amount of deviation calculated in the calculation process.

3. A sensor information acquisition unit that acquires sensor information obtained by a sensor mounted on a moving object; an absolute position information acquisition unit that acquires information regarding an absolute position of the moving object when the sensor information is acquired; a point cloud information generating unit that generates point cloud information based on the sensor information and information regarding the absolute position; an overhead information acquisition unit that acquires overhead information of an earth's surface from the sky; a calculation unit that calculates a deviation amount between an absolute position of the overhead view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; a correction unit that corrects absolute positions of the point cloud included in the point cloud information based on the amount of deviation calculated by the calculation unit, the calculation unit calculates a deviation amount between the two set reference points, that is, a first reference point and a second reference point, an absolute position of the overhead view information corresponding to the first reference point and an absolute position of the overhead view information corresponding to the second reference point, and an absolute position of a point cloud included in the point cloud information corresponding to the first reference point and the second reference point, respectively; The information processing apparatus according to claim 1, wherein the correction unit corrects the position of the point cloud information in the range between the first reference point and the second reference point based on the amount of deviation calculated by the calculation unit.

4. An information processing method executed by an information processing device that performs a predetermined process, a sensor information acquiring step of acquiring sensor information obtained by a sensor mounted on the moving object; an absolute position information acquisition step of acquiring information regarding an absolute position of the moving object when the sensor information is acquired; a point cloud information generating step of generating point cloud information based on the sensor information and information regarding the absolute position; an overhead information acquisition step of acquiring overhead information of an earth's surface from the sky; a calculation step of calculating a deviation amount between an absolute position of the overhead view information corresponding to a set reference point and an absolute position of a point cloud included in the point cloud information corresponding to the reference point; a correction step of correcting absolute positions of the point cloud included in the point cloud information based on the deviation amount calculated in the calculation step, In the calculation step, a deviation amount is calculated between the two set reference points, that is, a first reference point and a second reference point, an absolute position of the overhead view information corresponding to the first reference point and an absolute position of the overhead view information corresponding to the second reference point, and an absolute position of a point cloud included in the point cloud information corresponding to the first reference point and the second reference point, respectively; An information processing method, characterized in that in the correction process, the position of the point cloud information in the range between the first reference point and the second reference point is corrected based on the amount of deviation calculated in the calculation process.

5. An information processing program characterized by causing a computer to execute the information processing method described in claim 2 or 4.

6. A computer-readable storage medium characterized by storing the information processing program described in claim 5.

Images