Electronic device, control method, and control program

Abstract

This electronic device (100) comprises an acquisition unit (131) for acquiring point cloud data corresponding to points on a road surface, a cloth simulation unit (134) for outputting the shape of a point cloud represented by the point cloud data assessed on the basis of the shape of imaginary cloth when the point cloud data is covered by the imaginary cloth at a prescribed gravity, a clustering unit (135) for performing clustering of the point cloud on the basis of the shape of the point cloud, and a determining unit (136) for determining concave-convex features of the clustered point cloud. The cloth simulation unit (134) assesses the shape of the point cloud on the basis of the distances between first cloth lattice points when the point cloud data is covered by an imaginary cloth from a first direction and second cloth lattice points when the point cloud data is covered by an imaginary cloth from a second direction. The clustering unit (135) performs clustering on a point cloud excluding points assessed to be the road surface. The determining unit (136) determines concave-convex features of the point cloud on the basis of the number of concave-convex assessments given to the clustered point cloud.

Description

[Technical Field]

[0001] The present application relates to an electronic device, a control method, and a control program. [Background technology]

[0002] There are known techniques for recognizing surrounding obstacles. Patent Document 1 discloses a technique for selecting points on the floor surface using plane parameters of the detected floor surface and recognizing obstacles based on these points. Patent Document 2 discloses a technique for calculating the degree of unevenness of the road surface based on input information from the road surface to the vehicle. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2003-269937 [Patent Document 2] Japanese Patent Application Publication No. 2020-013537 Summary of the Invention [Problem to be solved by the invention]

[0004] It is desirable for autonomous vehicles to avoid not only surrounding objects but also small objects, steps, holes, etc. However, when using a camera mounted on a vehicle, it is difficult to determine the edges of unevenness from images, and there is room for improvement in the accuracy of detecting unevenness from images. [Means for solving the problem]

[0005] An electronic device according to one aspect includes an acquisition unit that acquires point cloud data corresponding to points on a road surface; a cloth simulation unit that outputs the shape of the point cloud indicated by the point cloud data, determined based on the shape of the virtual cloth when a virtual cloth having a predetermined tension is placed over the point cloud data at a predetermined gravity; a clustering unit that performs clustering of the point cloud based on the shape of the point cloud; and a determination unit that determines the unevenness of the clustered point cloud, wherein the cross simulation unit determines the shape of the point cloud based on the distance between a first cloth grid point when the virtual cloth is placed over the point cloud data from a first direction and a second cloth grid point when the virtual cloth is placed over the point cloud data from a second direction opposite to the first direction, the clustering unit performs clustering on the point cloud other than the points determined to be the road surface, and the determination unit determines the unevenness of the point cloud based on the number of unevenness determinations assigned to the clustered point cloud.

[0006] A control method according to one aspect includes an acquisition step in which an electronic device acquires point cloud data corresponding to points on a road surface; a cloth simulation step in which an electronic device outputs the shape of the point cloud indicated by the point cloud data, determined based on the shape of the virtual cloth when a virtual cloth having a predetermined tension is placed over the point cloud data at a predetermined gravity; a clustering step in which the point cloud is clustered based on the shape of the point cloud; and a determination step in which the unevenness of the clustered point cloud is determined. The cloth simulation step determines the shape of the point cloud based on the distance between a first cloth grid point when the virtual cloth is placed over the point cloud data from a first direction and a second cloth grid point when the virtual cloth is placed over the point cloud data from a second direction opposite to the first direction. The clustering step performs clustering on the point cloud other than the points determined to be the road surface. The determination step determines the unevenness of the point cloud based on the number of unevenness determinations assigned to the clustered point cloud.

[0007] A control program according to one aspect causes an electronic device to execute an acquisition process for acquiring point cloud data corresponding to points on a road surface; a cloth simulation process for outputting the shape of the point cloud indicated by the point cloud data, determined based on the shape of the virtual cloth when a virtual cloth having a predetermined tension is placed over the point cloud data at a predetermined gravity; a clustering process for clustering the point cloud based on the shape of the point cloud; and a determination process for determining the unevenness of the clustered point cloud. The cloth simulation process causes the electronic device to determine the shape of the point cloud based on the distance between a first cloth grid point when the virtual cloth is placed over the point cloud data from a first direction and a second cloth grid point when the virtual cloth is placed over the point cloud data from a second direction opposite to the first direction. The clustering process causes the electronic device to perform clustering on the point cloud other than the points determined to be the road surface. The determination process causes the electronic device to determine the unevenness of the point cloud based on the number of unevenness determinations assigned to the clustered point cloud. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating an overview of an electronic device according to an embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of the configuration of the electronic device according to the embodiment. [Figure 3] FIG. 3 is a flowchart illustrating an example of a processing procedure executed by the electronic device according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating an example of the detection data and the point cloud data. [Figure 5] FIG. 5 is a flowchart showing an example of the point cloud data generation process shown in FIG. [Figure 6] FIG. 6 is a flowchart showing an example of the cloth simulation shown in FIG. [Figure 7] FIG. 7 is a diagram for explaining the algorithm of the cloth simulation shown in FIG. [Figure 8] FIG. 8 is a diagram showing an example of a simulation result of the point cloud and the virtual cloth. [Figure 9] FIG. 9 is a diagram for explaining point cloud data used in the cross simulation shown in FIG. [Figure 10] FIG. 10 is a flowchart showing an example of the unevenness determination shown in FIG. [Figure 11] FIG. 11 is a diagram illustrating an example of the provided data of the electronic device according to the embodiment. [Figure 12] FIG. 12 is a diagram for explaining an example of detecting concaves and convexes from point cloud data of the electronic device according to the embodiment. [Figure 13] FIG. 13 is a diagram for explaining an example in which the cloth simulation shown in FIG. 12 is processed for each piece of virtual cloth. [Figure 14] FIG. 14 is a diagram for explaining the erroneous determination of the concave and convex portions in FIG. [Figure 15] FIG. 15 is a diagram for explaining an example of processing cloth simulation for each virtual cloth in FIG. DETAILED DESCRIPTION OF THE INVENTION

[0009] Several embodiments for implementing the electronic device, control method, control program, etc. according to the present application will be described in detail with reference to the drawings. Note that the present invention is not limited to the following description. Furthermore, the components in the following description include those that can be easily imagined by a person skilled in the art, those that are substantially the same, and those that are within the so-called equivalent range. In the following description, similar components may be assigned the same reference numerals. Furthermore, duplicated descriptions may be omitted.

[0010] FIG. 1 is a diagram illustrating an overview of an electronic device according to an embodiment. The electronic device 100 shown in FIG. 1 is mounted on a moving body 1000 capable of at least one of autonomous driving and manual driving. The moving body 1000 includes, for example, a vehicle, a robot, a cart, a train, a flying object capable of taking off and landing, etc. The moving body 1000 is equipped with a detection unit 200 capable of detecting the surroundings in the moving direction. In FIG. 1, a direction Dx is the traveling direction of the moving body 1000. A direction Dy is the direction of gravity that intersects with the direction Dx and is the height direction.

[0011] The detection unit 200 includes, for example, a depth camera, a millimeter-wave radar, a LiDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging), etc. The detection unit 200, for example, irradiates a pulsed laser beam and measures the reflected light of the laser beam to measure the distance, direction, etc. of an object. The detection unit 200 irradiates a road surface 5000 on which the mobile object 1000 is traveling with the laser beam, measures the reflected light reflected by the road surface 5000 and targets on the road surface 5000, etc., and supplies detection data indicating the measurement results of the detection range 310 to the electronic device 100, the mobile object 1000, etc. The detection range 310 includes a range in which detection of unevenness on the road surface 5000 is required. The detection range 310 includes a range in which detection of the path ahead of the mobile object 1000, the surrounding environment, etc. is required. The detection range 310 includes a range calculated or set based on road surface conditions such as the speed of the mobile object 1000 and the slipperiness of the road surface 5000. The climbable angle 320 includes an angle at which the moving body 1000 can climb a slope without scraping the bottom. If the moving body 1000 is a vehicle, the climbable angle 320 can be calculated from the height from the road surface 5000 to the bottom of the vehicle and the wheelbase width. The minimum altitude 330 includes an altitude that is reached when the climbable angle 320 continues throughout the entire detection range 310. The minimum altitude 330 can be calculated from the detection range 310 and the climbable angle 320. The moving body 1000 can recognize obstacles, targets, etc. based on the detection results of the detection unit 200.

[0012] The electronic device 100 is configured to be able to communicate with the mobile object 1000 wirelessly or via a wired connection, and obtains the measurement results of the detection unit 200 via the mobile object 1000. The electronic device 100 has a function of estimating the self-position of the mobile object 1000 and a surrounding map using the measurement results of the detection unit 200, for example, by a technique such as SLAM (Simultaneous Localization and Mapping). The surrounding map can be expressed as point cloud data, for example. The point cloud data has information on three-dimensional coordinates and colors. The electronic device 100 can estimate the likelihood of the estimated self-position based on the estimated surrounding map and the amount of movement of the mobile object 1000.

[0013] In the example shown in FIG. 1 , the electronic device 100 acquires measurement results obtained by the detection unit 200 measuring the area ahead in the traveling direction M of the mobile object 1000. In this case, a slope 5100 exists on a road surface 5000 in the traveling direction M of the mobile object 1000. The electronic device 100 may determine that the slope 5100 is a "drivable road surface" because it is a slope at an angle that allows it to be climbed. The electronic device 100 may detect any depressions on the road surface 5000 or the slope 5100 that are large enough for a tire to get stuck in. If the depression is deep, the electronic device 100 may be unable to measure the reflected light, resulting in a defect, and may determine that the depression is highly likely to be a dangerous location for the mobile object 1000.

[0014] The road surface 5000 includes objects such as the surface of a road or a floor, over which the mobile object 1000 moves. The electronic device 100 acquires point cloud data capable of identifying the condition of the road surface 5000 from the measurement results obtained by the detection unit 200. If the slope 5100 of the road surface 5000 is deep, the detection unit 200 may not be able to measure reflected light. In this case, the point cloud data acquired by the electronic device 100 will be data lacking point clouds in deep parts of the slope 5100. In other words, deep parts of the slope 5100 are likely to be holes or the like in the road surface 5000, and are therefore likely to be dangerous areas on the road surface 5000 for the mobile object 1000 to move in. For this reason, the electronic device 100 according to this embodiment provides a technology for improving the accuracy of determining the slope 5100 around the mobile object 1000.

[0015] [Example of electronic device configuration] Fig. 2 is a diagram showing an example of the configuration of the electronic device 100 according to the embodiment. As shown in Fig. 2, the electronic device 100 includes a communication unit 110, a storage unit 120, and a control unit 130. The control unit 130 is electrically connected to the communication unit 110, the storage unit 120, etc. The electronic device 100 may also include other configurations.

[0016] The communication unit 110 can communicate with, for example, the mobile object 1000, other communication devices, etc. The communication unit 110 can support various communication standards. The communication unit 110 can send and receive various types of data via, for example, a wired or wireless network. The communication unit 110 can provide the received data to the control unit 130. The communication unit 110 can send data to a destination instructed by the control unit 130.

[0017] The storage unit 120 can store programs and data. The storage unit 120 is also used as a working area for temporarily storing processing results of the control unit 130. The storage unit 120 may include any non-transitory storage medium, such as a semiconductor storage medium or a magnetic storage medium. The storage unit 120 may include multiple types of storage media. The storage unit 120 may include a combination of a portable storage medium, such as a memory card, an optical disk, or a magneto-optical disk, and a storage medium reader. The storage unit 120 may include a storage device used as a temporary storage area, such as a RAM (Random Access Memory).

[0018] The storage unit 120 can store various data such as a program 121, a cloth simulation 122, point cloud data 123, posture data 124, and provided data 125. The program 121 causes the control unit 130 to execute a function for determining depressions in the road surface 5000 or the slope 5100. The cloth simulation 122 is a program capable of executing a cloth simulation method for estimating the shape of the point cloud data 123 from the shape of a cloth that has a predetermined tension and is placed over the point cloud data 123 under a predetermined gravity, and outputting the estimation result. The cloth simulation 122 may be included in the program 121 or may be stored in an external storage device or the like. The point cloud data 123 is data that indicates the shape of the road surface 5000, etc., detected by the detection unit 200, as a point cloud. The posture data 124 is data that can identify the posture of the detection unit 200 or the moving object 1000. The posture data 124 is data acquired at a predetermined timing, for example, at the time of setup or the start of processing. The provided data 125 includes the determination result made by the electronic device 100, data that can identify depressions in the road surface 5000 or the slope 5100, and the like.

[0019] The control unit 130 is an arithmetic processing device. Examples of the arithmetic processing device include, but are not limited to, a central processing unit (CPU), a system-on-a-chip (SoC), a micro control unit (MCU), a field-programmable gate array (FPGA), and a coprocessor. The control unit 130 can comprehensively control the operation of the electronic device 100 to realize various functions.

[0020] Specifically, the control unit 130 can execute instructions included in the program 121 stored in the storage unit 120 while referring to the data stored in the storage unit 120 as needed. The control unit 130 then controls the functional units in accordance with the data and instructions, thereby realizing various functions. The functional units include, but are not limited to, a communication unit 230, for example.

[0021] The control unit 130 has functional units such as an acquisition unit 131, a discrimination unit 132, a completion unit 133, a cross-simulation unit 134, a clustering unit 135, a determination unit 136, and a provision unit 137. The control unit 130 executes the program 121 to realize the functions of the acquisition unit 131, the discrimination unit 132, the completion unit 133, the cross-simulation unit 134, the clustering unit 135, the determination unit 136, and the provision unit 137. The program 121 is a program for causing the control unit 130 of the electronic device 100 to function as the acquisition unit 131, the discrimination unit 132, the completion unit 133, the cross-simulation unit 134, the clustering unit 135, the determination unit 136, and the provision unit 137.

[0022] The acquisition unit 131 acquires point cloud data 123 that indicates a collection of points that can identify the state of the road surface 5000. The acquisition unit 131 acquires the point cloud data 123 generated based on a depth image of the detection data from the detection unit 200, posture data of the moving object 1000, etc. The acquisition unit 131 may acquire point cloud data 123 that is generated outside the electronic device 100. The acquisition unit 131 stores the acquired point cloud data 123 in the storage unit 120.

[0023] The discrimination unit 132 discriminates missing portions of the point cloud data 123 that are at or below the height of the road surface 5000 within the detection range 310 of the point cloud data 123. The missing portions of the point cloud data 123 are depressions in the road surface 5000 or the slope 5100. The discrimination unit 132 discriminates whether or not a point cloud is missing in the point cloud data 123. The discrimination unit 132 associates missing data that can identify missing portions in the point cloud data 123 with the point cloud data 123.

[0024] The complementing unit 133 complements missing portions of the point cloud data 123 with pseudo point clouds at positions at a predetermined depth from the road surface 5000. The complementing unit 133 calculates the minimum altitude 330 calculated from the detection range 310 of the point cloud data 123 and the climbable angle 320 of the mobile object 1000. The complementing unit 133 complements missing portions with pseudo point clouds at heights lower than the calculated minimum altitude 330. The complementing unit 133 complements missing portions with pseudo point clouds at a depth that can be sufficiently regarded as depressions in the road surface 5000 or the slope 5100.

[0025] The cloth simulation unit 134 outputs information that enables identification of the shape of the point cloud data 123 from the cloth shape when a virtual cloth with a predetermined tension is draped over the point cloud data 123 under a predetermined gravity. By executing the cross simulation 122, the cloth simulation unit 134 connects, among the multiple points included in the point cloud indicated by the point cloud data 123, a first point that is the highest in the height direction and a second point that is around the first point and is lower in the height direction (direction Dy shown in FIG. 1 ), so as to form a shape of a bendable virtual cloth, thereby outputting the shape of the point cloud data 123. By executing the cross simulation 122, the cloth simulation unit 134 may also connect, among the multiple points included in the point cloud indicated by the point cloud data 123, a certain point and its surrounding points so as to form a shape of a bendable virtual cloth, thereby outputting the shape of the point cloud data 123.

[0026] The clustering unit 135 performs clustering of the point clouds based on the shapes of the point clouds indicated by the generated point cloud data 123, the point cloud data 123 supplemented with the pseudo point cloud, etc. For example, the clustering unit 135 performs clustering on the point clouds other than the points determined to be the road surface 5000. For example, the clustering unit 135 clusters the point cloud data 123 into clusters of irregularities, etc. For example, if the point cloud data 123 has multiple irregularities, the clustering unit 135 clusters the point cloud data 123 into clusters for each of the multiple irregularities. The clustering unit 135 stores the clustering determination results in the storage unit 120 in association with the point cloud data 123.

[0027] The determination unit 136 determines the unevenness of the clustered point cloud. The determination unit 136 determines the unevenness of the point cloud based on, for example, the number of unevenness determinations assigned to the clustered point cloud. The determination unit 136 stores the determination result of the unevenness of the point cloud in the storage unit 120 in association with the point cloud data 123.

[0028] The providing unit 137 provides the provided data 125 capable of identifying the concave and convex portions 5200 of the road surface 5000 or the slope 5100 determined by the determining unit 136. The providing unit 137 can provide the provided data 125 to the moving object 1000 via the communication unit 110. This enables the moving object 1000 to change its route so as to avoid the concave portions of the road surface 5000 or the slope 5100 indicated by the provided data 125. The providing unit 137 may provide the provided data 125 to a display device, a smartphone, or the like.

[0029] An example of the functional configuration of the electronic device 100 according to this embodiment has been described above. Note that the configuration described above using Fig. 2 is merely an example, and the functional configuration of the electronic device 100 according to this embodiment is not limited to this example. The functional configuration of the electronic device 100 according to this embodiment can be flexibly modified according to specifications and operations.

[0030] [Example of disposal procedure for electronic devices] FIG. 3 is a flowchart showing an example of a processing procedure executed by the electronic device 100 according to the embodiment. FIG. 4 is a diagram showing an example of detection data and point cloud data. FIG. 5 is a flowchart showing an example of a process for generating the point cloud data 123 shown in FIG. 3. FIG. 6 is a flowchart showing an example of the cloth simulation shown in FIG. 3. FIG. 7 is a diagram for explaining an algorithm of the cross simulation shown in FIG. 6. FIG. 8 is a diagram showing an example of a simulation result of the point cloud 123P and the virtual cloth 40. FIG. 9 is a diagram for explaining point cloud data used in the cross simulation shown in FIG. 6. FIG. 10 is a flowchart showing an example of the unevenness determination shown in FIG. 3. The processing procedures shown in FIGS. 3, 5, 6, and 10 are implemented by the control unit 130 of the electronic device 100 executing the program 121.

[0031] As shown in FIG. 3, the control unit 130 of the electronic device 100 acquires detection data 210 from the detection unit 200 (step S100). For example, the control unit 130 acquires the detection data 210 from the detection unit 200 of the moving object 1000 via the communication unit 110. In this embodiment, the control unit 130 acquires the detection data 210 having a depth image capturing an image of the area ahead of the moving object 1000, as shown in FIG. 4. In the example shown in FIG. 4, the detection data 210 has a depth image capturing an image of a recess 5100a present on a road surface 5000 near the moving object 1000 and two convex portions 5200 present on the road surface 5000 ahead of the recess 5100a. Returning to FIG. 3, the control unit 130 stores the acquired detection data 210 in the storage unit 120. If the control unit 130 is unable to acquire the detection data 210, the control unit 130 stores in the storage unit 120 a message indicating that the detection data 210 could not be acquired. When the process of step S100 ends, control unit 130 advances the process to step S200.

[0032] The control unit 130 determines whether the acquisition of the detection data 210 was successful (step S200). For example, if the storage unit 120 stores the latest detection data 210, the control unit 130 determines that the acquisition of the detection data 210 was successful. If the control unit 130 determines that the acquisition of the detection data 210 was not successful (No in step S200), the control unit 130 ends the processing procedure shown in Fig. 3. Note that the processing procedure shown in Fig. 3 may be a processing procedure in which, if the acquisition of the detection data 210 was not successful, the processing to acquire the detection data 210 again in step S100 is executed.

[0033] Furthermore, if the control unit 130 determines that the acquisition of the detection data 210 has been successful (Yes in step S200), the control unit 130 proceeds to step S300. The control unit 130 executes a process for generating the point cloud data 123 (step S300). For example, the control unit 130 converts the depth image indicated by the acquired detection data 210 into the point cloud data 123 by performing a known coordinate transformation on the depth image.

[0034] For example, as shown in FIG. 5, the control unit 130 generates a point cloud in the camera coordinate system from pixel values ​​and pixel coordinates (x, y) of the depth image (step S301). Then, the control unit 130 calculates a rotation matrix of the point cloud data 123 from the orientation data 124 stored in the storage unit 120 and converts the point cloud data 123 to a world coordinate system (step S302). The world coordinate system is one of the coordinate systems used in fields such as three-dimensional computer graphics. To represent the position and movement of an object in three-dimensional space, the entire space is defined as a coordinate system using X, Y, and Z axes. The control unit 130 then removes point clouds outside the range and noise from the point cloud data 123 (step S303). For example, the control unit 130 reduces the number of point clouds to be processed by removing point clouds outside the detection range 310 and noise. In the example shown in FIG. 4, the point cloud data 123 is a perspective view of a road surface 5000 having a recess 5100a and two protrusions 5200. In this case, the point cloud of the recess 5100a is missing from the point cloud data 123. Returning to Fig. 5, the control unit 130 stores the generated point cloud data 123 in the storage unit 120 (step S304). When the processing procedure shown in Fig. 5 ends, the control unit 130 returns to step S300 shown in Fig. 3 and proceeds to step S400.

[0035] The control unit 130 executes a cross simulation (step S400). For example, the control unit 130 executes the cross simulation 122 in the storage unit 120 to perform a cross simulation on the point cloud data 123.

[0036] For example, as shown in Fig. 6, the control unit 130 generates a virtual cloth and associates the point cloud with the cloth grid points (step S401). The virtual cloth is a virtual cloth of a size that can cover the point cloud of the point cloud data 123, and has a plurality of grid points arranged in a matrix. After generating the virtual cloth, the control unit 130 sets the surface hardness of the point cloud corresponding to each cloth grid point of the virtual cloth.

[0037] The control unit 130 determines a missing portion 123D of the point cloud data 123 (step S402). For example, when the point cloud of the point cloud data 123 is covered with a virtual cloth, the control unit 130 determines a portion where the cloth grid points of the virtual cloth are not linked to the point cloud as the missing portion 123D.

[0038] For example, as shown in scene C0 in Fig. 7, the point cloud data 123 has a plurality of point clouds 123P. For simplicity of explanation, Fig. 7 shows a cross section of a portion of the plurality of point clouds 123P of the point cloud data 123. In this case, the control unit 130 places the virtual cloth 40 so as to cover the plurality of point clouds 123P, and associates the cloth grid points 400 with the point clouds 123P of the point cloud data 123. For example, when the virtual cloth 40 is brought into contact with the point cloud data 123, the control unit 130 determines that a portion of the cloth grid point 400a that is not in contact with the point cloud 123P is a missing portion 123D of the point cloud data 123.

[0039] Returning to FIG. 6, the control unit 130 associates the determination result with the point cloud data 123 and stores it in the storage unit 120, and then proceeds to step S403. The control unit 130 executes a process of interpolating the point cloud data 123 (step S403). For example, the control unit 130 executes an interpolation process to interpolate the pseudo point cloud 123S into the missing portion 123D of the point cloud data 123. For example, when the missing portion 123D of the point cloud data 123 has been determined, the control unit 130 determines the height of the point cloud to be interpolated based on the detection range 310, the climbable angle 320, and the minimum altitude 330 shown in FIG. 1. The height of the point cloud to be interpolated is determined based on detection requirements, vehicle requirements, etc., so that it is lower than the minimum altitude 330.

[0040] For example, as shown in scene C0 in Fig. 7, when the point cloud data 123 has a missing portion 123D, the control unit 130 determines the height of the point cloud to be complemented at the missing portion 123D, and complements the point cloud data 123 with a pseudo point cloud 123S at that height. As a result, as shown in scene C1 in Fig. 7, the control unit 130 complements the point cloud data 123 with a pseudo point cloud 123S of a depth that allows the missing portion 123D to be sufficiently determined as a slope 5100 (a recess 5100a).

[0041] 6, the control unit 130 advances the process to step S404 after complementing the missing portion 123D of the point cloud data 123 with the pseudo point cloud 123S. Note that, if it is determined in the above-described step S402 that the missing portion 123D does not exist in the point cloud data 123, the control unit 130 skips the process of step S403 and advances the process to step S404.

[0042] The control unit 130 reflects the external force (gravity) on the virtual cloth 40 (step S404). For example, as shown in scene C2 of FIG. 7, the control unit 130 moves the cloth grid points 400 of the virtual cloth 40 from the starting point 410 in the direction M1 (gravity direction) below the point cloud 123P. Then, as shown in scene C3 of FIG. 7, when the cloth grid points 400 of the virtual cloth 40 become lower than the point cloud 123P, the control unit 130 moves the cloth grid points 400A of the virtual cloth 40 that are associated with the point cloud 123P that indicates the road surface 5000 from the starting point 410 in the direction M2 to above the point cloud 123P. In this case, the control unit 130 does not move the cloth grid points 400 of the virtual cloth 40 that are not associated with the point cloud 123P that indicates the road surface 5000. Returning to FIG. 6, when the process of step S404 ends, control unit 130 advances the process to step S405.

[0043] The control unit 130 reflects the internal force (tension) on the virtual cloth 40 (step S405). For example, the control unit 130 reflects the pulling force between the multiple cloth grid points 400 of the virtual cloth 40. The tension is fixed at a value that allows the slope 5100, which is the object to be detected, to be detected. For example, as shown in scene C4 of FIG. 7, the control unit 130 moves the cloth grid points 400 of the virtual cloth 40 that are not linked to the point cloud 123P from the start point 410 toward the direction M2 according to the tension. As a result, the control unit 130 obtains the cloth grid points 400B of the virtual cloth 40 from one side (the sky side) that are not attached to the point cloud 123P, i.e., the cloth grid points 400B that are not attached to the road surface 5000 indicated by the point cloud 123P. The control unit 130 distinguishes the multiple cloth grid points 400 into cloth grid points 400A that are attached to the road surface 5000 and cloth grid points 400B that are not attached to the road surface 5000. Returning to FIG. 6, when the processing of step S405 ends, the control unit 130 advances the processing to step S406.

[0044] The control unit 130 determines whether the predetermined number of times has been reached (step S406). For example, in the case of a steep slope or a wall, the virtual cloth 40 may not fall all the way down to the road surface 5000 during cloth simulation. For this reason, the control unit 130 sets a predetermined number of times for executing cloth simulation from two directions, and determines that the predetermined number of times has been reached when the number of executions matches the predetermined number of times. If the control unit 130 determines that the predetermined number of times has not been reached (No in step S406), the control unit 130 returns the process to step S404, which has already been described, and continues the process. If the control unit 130 determines that the predetermined number of times has been reached (Yes in step S406), the control unit 130 proceeds to step S407.

[0045] The control unit 130 pastes the virtual cloth 40, of the virtual cloth 40 that is not pasted on the road surface 5000, at a location that has no elevation difference from the location already pasted (step S407). For example, assume that the control unit 130 has obtained a simulation result of the point cloud 123P and the virtual cloth 40, as shown in FIG. 8. In this case, the control unit 130 determines that the cloth grid point 400B-2 adjacent to the already fixed cloth grid point 400A-1 may be fixed to the actual measurement value because the point clouds 123P-1 and 123P-2 indicating the heights of the road surfaces being compared are almost the same height. The control unit 130 determines that the cloth grid point 400B-4 adjacent to the already fixed cloth grid point 400A-3 should not be fixed to the actual measurement value because there is a difference between the point clouds 123P-3 and 123P-4 indicating the heights of the road surfaces being compared. Returning to FIG. 6, once the control unit 130 stores the processing results for the underground side of the road surface 500 in the storage unit 120, the control unit 130 advances the processing to step S408.

[0046] The control unit 130 determines whether the specified number of times has been reached (step S408). If the control unit 130 determines that the specified number of times has not been reached (No in step S408), the process returns to step S407, which has already been described, and the process continues. If the control unit 130 determines that the specified number of times has been reached (Yes in step S408), the process proceeds to step S409.

[0047] The control unit 130 flips the point cloud data 123 upside down (step S409). For example, the control unit 130 flips the sky side and the underground side of the point cloud 123P (road surface 5000) indicated by the point cloud data 123. When the process of step S409 ends, the control unit 130 advances the process to step S410.

[0048] The control unit 130 reflects the external force (gravity) on the virtual cloth 40 (step S410). For example, similar to step S404 already described, the control unit 130 performs a process of overlaying the cloth grid points 400 of the virtual cloth 40 onto the point cloud data 123 obtained by inverting the point cloud 123P and the pseudo point cloud 123S, as shown in Fig. 9. Returning to Fig. 6, when the process of step S410 ends, the control unit 130 advances the process to step S411.

[0049] The control unit 130 reflects the internal force (tension) in the virtual cloth 40 (step S411). For example, the control unit 130 reflects the pulling force between the multiple cloth grid points 400 of the virtual cloth 40, as in step S405 already described. As a result, the control unit 130 obtains a result in which, among the cloth grid points 400 of the virtual cloth 40 from the other side (underground side), the points that are not attached to the point cloud 123P, i.e., the points that are not attached to the road surface 5000 indicated by the point cloud 123P. When the process of step S411 ends, the control unit 130 proceeds to step S412.

[0050] The control unit 130 determines whether the predetermined number of times has been reached (step S412). For example, similar to the already-described step S408, if the control unit 130 determines that the predetermined number of times has not been reached (No in step S412), the control unit 130 returns the process to the already-described step S410 and continues the process. On the other hand, if the control unit 130 determines that the predetermined number of times has been reached (Yes in step S412), the control unit 130 proceeds to step S413.

[0051] The control unit 130 pastes the virtual cloth 40 that is not attached to the road surface 5000 at a location that has no difference in elevation with the already attached location (step S413). The control unit 130 makes the same determination as in step S407 already described, and stores the processing result for the sky side of the road surface 5000 in the storage unit 120, and then proceeds to step S414.

[0052] The control unit 130 determines whether the predetermined number of times has been reached (step S414). If the control unit 130 determines that the predetermined number of times has not been reached (No in step S414), the process returns to step S413, which has already been described, and continues the process. If the control unit 130 determines that the predetermined number of times has been reached (Yes in step S414), the control unit 130 ends the process procedure shown in Fig. 6, returns to step S400 shown in Fig. 3, and proceeds to step S500.

[0053] The control unit 130 determines unevenness (step S500). For example, the control unit 130 determines whether the point cloud 123P in the point cloud data 123 is uneven by using a determination condition. The determination condition includes a condition that determines the point cloud 123P as a road surface 5000 when the difference in elevation between the virtual cloth 40 on the sky side and the virtual cloth 40 on the underground side is equal to or less than a specified value, and determines the point cloud 123P as unevenness when the difference in elevation between the virtual cloth 40 on the sky side and the virtual cloth 40 on the underground side is not equal to or less than the specified value. The control unit 130 then clusters the point cloud 123P separated from the road surface 5000, and performs the following process on each cluster of unevenness, etc. A cluster is, for example, a collection of points in the point cloud 123P at locations such as unevenness that have not been determined to be the road surface 5000. The control unit 130 determines the unevenness as a convex portion when the virtual cloth 40 on the sky side has already been attached and half or more of the virtual cloth 40 on the underground side is not attached. Furthermore, when no virtual cloth 40 on the sky side is attached and the number of virtual cloths 40 attached on the underground side is half or more, the control unit 130 determines that the unevenness is a recess 5100a. An example of the unevenness determination process is shown below.

[0054] As shown in FIG. 10, the control unit 130 acquires the results of cloth simulation from two directions (step S501). For example, the control unit 130 acquires the results of performing cloth simulation from two directions, the sky side and the underground side, with respect to the road surface 5000 in the above-mentioned step S400. Then, the control unit 130 executes a loop process for separating the road surface 5000 from the unevenness (step S502). The loop process for separating the road surface 5000 from the unevenness is, for example, a loop process for separating all point clouds in the point cloud data 123 into the road surface 5000 and the unevenness. The termination condition for the loop process for separating the road surface 5000 from the unevenness includes, for example, execution for all point clouds. After executing the loop process for separating the road surface 5000 from the unevenness, the control unit 130 determines whether the difference in elevation of the virtual cloth 40 in two directions is equal to or greater than a specified value (step S503). For example, the control unit 130 calculates the difference in elevation between the cloth grid points 400 of the two virtual cloths 40 and compares the difference in elevation with a specified value. If the control unit 130 determines that the difference in elevation is equal to or greater than the specified value (Yes in step S503), the control unit 130 proceeds to step S504.

[0055] The control unit 130 determines that the point cloud 123P is either a convex or concave portion (step S504). For example, the control unit 130 determines that the point cloud 123P to be extracted is either a convex or concave portion, and associates data capable of identifying the convex or concave portion with the point cloud 123P. When the process of step S504 ends and the termination condition of the loop process of separating the road surface 5000 from the convex or concave portion in step S502 is not satisfied, the control unit 130 continues the loop process of separating the road surface 5000 from the convex or concave portion for the next point cloud 123P. When the process of step S504 ends and the termination condition of the loop process of separating the road surface 5000 from the convex or concave portion in step S502 is satisfied, the control unit 130 advances the process to step S506, which will be described later.

[0056] Furthermore, if the control unit 130 determines in step S503 that the difference in elevation between the virtual cloth 40 in two directions is not equal to or greater than a specified value (No in step S503), the control unit 130 proceeds to step S505. The control unit 130 determines the point cloud 123P to be the road surface 5000 (step S505). For example, the control unit 130 determines the point cloud 123P to be extracted to be the road surface 5000, and associates data that can identify the road surface 5000 with the point cloud 123P. When the process of step S505 ends and the termination condition of the loop process for separating the road surface 5000 and the irregularities in step S502 is not satisfied, the control unit 130 continues the loop process for separating the road surface 5000 and the irregularities for the next point cloud 123P. Furthermore, when the process of step S505 ends and the condition for ending the loop process of separating the road surface 5000 and the irregularities in step S502 is satisfied, the control unit 130 advances the process to step S506, which will be described later.

[0057] The control unit 130 clusters the points determined to be uneven (step S506). For example, the control unit 130 clusters the uneven point group 123P based on the separation result of the loop process for separating the road surface 5000 and the unevenness in step S502. When the process of step S506 ends, the control unit 130 advances the process to step S507.

[0058] The control unit 130 executes a loop process for determining whether a cluster is concave or convex (step S507). The loop process for determining whether a cluster is concave or convex is, for example, a loop process for determining the cluster point cloud 123P of the clustered concave / convex cluster. The termination condition for the loop process for determining whether a cluster is concave or convex includes, for example, execution for all clusters. The control unit 130 determines whether the number of points where the virtual cloth 40 on the sky side has been attached and the virtual cloth 40 on the underground side has not been attached is half or more (step S508). For example, the control unit 130 determines whether the number of points where the virtual cloth 40 on the sky side has been attached and the virtual cloth 40 on the underground side has not been attached is half or more based on the contact state between the clustered concave / convex cluster point cloud 123P and the cloth grid points 400. If the control unit 130 determines that half or more of the points have the virtual cloth 40 on the sky side already attached and the virtual cloth 40 on the underground side not yet attached (Yes in step S508), the control unit 130 proceeds to step S509.

[0059] The control unit 130 determines that the cluster is convex (step S509). For example, when the control unit 130 determines that the cluster is convex, it associates information indicating that the cluster has been determined to be convex with the point cloud data 123. When the process of step S509 ends and the termination condition of the cluster concave / convex determination loop process of step S507 is not satisfied, the control unit 130 continues the cluster concave / convex determination loop process for the next cluster. When the process of step S509 ends and the termination condition of the cluster concave / convex determination loop process of step S507 is satisfied, the control unit 130 ends the processing procedure shown in FIG. 10, returns to step S500 shown in FIG. 3, and proceeds to step S600.

[0060] Furthermore, if the control unit 130 determines in step S508 that the number of points where the virtual cloth 40 on the sky side is already attached and the virtual cloth 40 on the underground side is not yet half (No in step S508), the control unit 130 proceeds to step S510. The control unit 130 determines whether the number of points where the virtual cloth 40 on the sky side is not attached and the virtual cloth 40 on the underground side is already attached is yet half (step S510). For example, the control unit 130 determines whether the number of points where the virtual cloth 40 on the sky side is not attached and the virtual cloth 40 on the underground side is already attached is yet half based on the contact state between the clustered unevenness cluster point cloud 123P and the cloth grid points 400.

[0061] When the control unit 130 determines that half or more of the points have the virtual cloth 40 on the sky side unattached and the virtual cloth 40 on the underground side attached (Yes in step S510), the control unit 130 proceeds to step S511. The control unit 130 determines that the cluster is concave (step S511). For example, when the control unit 130 determines that the cluster is concave, the control unit 130 associates information indicating that the cluster has been determined to be concave with the point cloud data 123. When the process of step S511 ends and the termination condition of the cluster concave / convex determination loop process of step S507 is not satisfied, the control unit 130 continues the cluster concave / convex determination loop process for the next cluster. When the process of step S511 ends and the termination condition of the cluster concave / convex determination loop process of step S507 is satisfied, the control unit 130 ends the processing procedure shown in FIG. 10, returns to step S500 shown in FIG. 3, and proceeds to step S600.

[0062] Furthermore, if the control unit 130 determines in step S510 that the number of points where the virtual cloth 40 on the sky side is not yet attached and the virtual cloth 40 on the underground side is not half or more (No in step S510), the control unit 130 proceeds to step S512. The control unit 130 determines that the cluster is abnormal (step S512). For example, when the control unit 130 determines that the cluster is abnormal, the control unit 130 associates information indicating that the cluster has been determined to be abnormal with the point cloud data 123. When the process of step S512 ends and the termination condition of the cluster concavity / convexity determination loop process of step S507 is not satisfied, the control unit 130 continues the cluster concavity / convexity determination loop process for the next cluster. Furthermore, when the processing of step S512 is completed and the termination condition of the loop processing for determining whether a cluster is concave or convex in step S507 is satisfied, the control unit 130 terminates the processing procedure shown in FIG. 10, returns to step S500 shown in FIG. 3, and proceeds to step S600.

[0063] The control unit 130 provides the provided data 125 indicating the result of the determination of the concave and convex portions (step S600). For example, the control unit 130 generates the provided data 125 capable of identifying the concave portions 5100a and the convex portions 5200 in the point cloud data 123, and executes a process of providing the provided data 125. The providing process includes, for example, a process of providing (transmitting) the provided data 125 to the mobile object 1000 via the communication unit 110, a process of providing (transmitting) the provided data 125 to another electronic device via the communication unit 110, a process of displaying the provided data on a display device via the communication unit 110, etc.

[0064] FIG. 11 is a diagram illustrating an example of the provided data 125 of the electronic device 100 according to the embodiment. As illustrated in FIG. 11, the control unit 130 generates the provided data 125 indicating the result of determining the unevenness of the point cloud data 123 illustrated in FIG. 4. The provided data 125 is data that enables identification of the determination result of the road surface 5000, the recessed portion 5100a, and the protruding portion 5200. By providing the provided data 125 that enables identification of the recessed portion 5100a and the protruding portion 5200 on the road surface 5000, the control unit 130 can assist in improving the accuracy of detecting the surrounding recessed portion 5100a. This allows the electronic device 1000 to accurately identify the recessed portion 5100a on the road surface 5000 and move while avoiding the recessed portion 5100a, thereby assisting in improving safety. Furthermore, the electronic device 100 can accurately detect the recessed portion 5100a and the protruding portion 5200 even on the road surface 5000 on which the mobile object 1000 is not traveling, thereby contributing to improving safety.

[0065] Returning to FIG. 3, when the processing of step S600 ends, control unit 130 determines whether to end (step S700). For example, control unit 130 determines to end if an end instruction has been received, it is time to end, or the like. If control unit 130 determines not to end (No in step S700), it returns the processing to step S100 already described and continues the processing. On the other hand, if control unit 130 determines to end (Yes in step S700), it ends the processing procedure shown in FIG. 3.

[0066] Fig. 12 is a diagram for explaining an example of detecting unevenness in the point cloud data 123 of the electronic device 100 according to the embodiment. Fig. 13 is a diagram for explaining an example of processing the cloth simulation shown in Fig. 12 for each piece of virtual cloth. Fig. 14 is a diagram for explaining erroneous determination of unevenness in Fig. 13. Fig. 15 is a diagram for explaining an example of processing the cloth simulation shown in Fig. 12 for each piece of virtual cloth.

[0067] 12, the electronic device 100 acquires point cloud data 123. The point cloud data 123 includes a point cloud 123P that can identify the state of the road surface 5000, including the depressions 5100a and the protrusions 5200 of the road surface 5000. In this case, the electronic device 100 moves the virtual cloth 40 closer to the point cloud data 123 from the sky side MA and the underground side MB so as to cover the virtual cloth 40.

[0068] First, a case will be described in which the electronic device 100 overlays the point cloud data 123 one by one with the virtual cloth 40. In scene C11 of Fig. 13, the electronic device 100 performs a cloth simulation in which the virtual cloth 40 is overlaid from the underground side MB, and determines the point cloud 123X of the point cloud data 123 to be a convex portion based on the cloth grid points 400. Then, in scene C12, the electronic device 100 performs a cloth simulation in which the virtual cloth 40 is overlaid from the sky side MA, and determines the point cloud 123Y of the point cloud data 123 to be a concave portion based on the cloth grid points 400. Then, in scene C13, the electronic device 100 performs clustering of the point clouds 123X and 123Y, and detects one convex portion and two concave portions.

[0069] Scene G1 in FIG. 14 shows a perspective view of the point cloud data 123, and scene G2 shows the top surface of the point cloud data 123. As shown in FIG. 14, when a cloth grid point 400X of the virtual cloth 40 is set at a boundary portion of the point cloud data 123 between convex and concave portions, as with the cloth grid point 400Y, the expected height may not be achieved due to a small difference in the coordinate values ​​of the point cloud 123P. If the virtual cloth 40 is placed over the point cloud data 123 in this state from above, the central portion of the virtual cloth 40 will appear to be floating. This may satisfy the condition for determining a concave portion, and the point cloud 123P that should be determined to be a convex portion may be determined to be a concave portion.

[0070] 15 shows an example in which the electronic device 100 simultaneously covers two pieces of virtual cloth 40 over the point cloud data 123. In scene C21 of FIG. 15, the electronic device 100 performs a cloth simulation in which two pieces of virtual cloth 40 are covered from the sky side MA and the underground side MB, and provisionally determines a point cloud 123Z of the point cloud data 123 as unevenness based on the cloth grid points 400. Then, in scene C22, the electronic device 100 clusters the point cloud 123P, and determines the cluster with the largest number of provisionally determined unevenness as unevenness.

[0071] As described above, the electronic device 100 clusters the point cloud 123P based on the shape of the virtual cloth 40 when the virtual cloth 40 is placed over the point cloud data 123 from two directions with a predetermined gravity, and can determine the unevenness of the point cloud 123P based on the number of unevenness determinations of the clustered point cloud 123P. As a result, the electronic device 100 can extract the unevenness without being affected by the road surface conditions by separating and determining the point cloud data 123 into the road surface 5000 and the unevenness. As a result, the electronic device 100 can improve the accuracy of detecting unevenness from an image by improving the determination accuracy of the edges of the unevenness, etc.

[0072] The electronic device 100 clusters the point group 123P into clusters of irregularities other than the points determined to be the road surface 5000. This allows the electronic device 100 to separate the point group 123P into clusters of the road surface 5000 and clusters of irregularities, and therefore, it is possible to determine the irregularities without being affected by the condition of the road surface 5000.

[0073] The electronic device 100 can determine whether a cluster of clustered unevenness is concave or convex based on the attachment state between the point cloud 123P and the virtual cloth 40 from two directions. As a result, the electronic device 100 can suppress omissions and erroneous determinations by using the results of cloth simulation from two directions, and can determine unevenness without being affected by the condition of the road surface 5000.

[0074] In the present embodiment, the electronic device 100 has been described as complementing the point cloud data 123 with the pseudo point cloud 123S, but the present invention is not limited to this. For example, the electronic device 100 can obtain the above-described advantageous effects by performing cross simulation from two directions even on the point cloud data 123 that does not have the missing portion 123D.

[0075] In the present embodiment, the electronic device 100 is described as being mounted on the mobile object 1000, but the present invention is not limited to this. For example, the electronic device 100 may be implemented as a server, a roadside device, or the like, outside the mobile object 1000.

[0076] Characteristic embodiments have been described to fully and clearly disclose the technology claimed in the appended claims. However, the appended claims should not be limited to the above-described embodiments, but should be construed to embody all modifications and alternative configurations that may be conceived by those skilled in the art within the scope of the basic concepts set forth herein. Those skilled in the art can make various modifications and alterations to the content of the present disclosure based on the present disclosure. Therefore, these modifications and alterations are within the scope of the present disclosure. For example, in each embodiment, each functional unit, each means, each step, etc. may be added to other embodiments without logical contradiction, or may be replaced with each functional unit, each means, each step, etc. of other embodiments. Furthermore, in each embodiment, multiple functional units, each means, each step, etc. may be combined or divided into one. Furthermore, each embodiment of the present disclosure described above is not limited to faithful implementation of each described embodiment, and each feature may be combined or partially omitted as appropriate. The method of the present disclosure may be implemented by a device equipped with a CPU and memory, in which the CPU executes a program stored in the memory. [Explanation of symbols]

[0077] 100 Electronic equipment 110 Communications Department 120 Storage section 121 Programs 122 Cloth Simulation 123 point cloud data 123D Missing part 123P point cloud 123S pseudo point cloud 124 posture data 125 Provided Data 130 Control Unit 131 Acquisition Department 132 Discrimination part 133 Complementary Section 134 Cross Simulation Section 135 Clustering Department 136 Decision Section 137 Provision Department 200 Detector 310 Detection Range 320 Climbing angle 330 minimum altitude 400 cloth grid points 1000 Mobile 5000 road surface 5100 Slope 5100a Recess 5200 convex part

Claims

1. an acquisition unit that acquires point cloud data corresponding to points on a road surface; a cloth simulation unit that outputs a shape of a point cloud represented by the point cloud data, the shape being determined based on a shape of the virtual cloth when the virtual cloth having a predetermined tension is placed over the point cloud data with a predetermined gravity; a clustering unit that performs clustering of the point cloud based on the shape of the point cloud; a determination unit that determines the unevenness of the clustered point cloud; Equipped with the cloth simulation unit determines a shape of the point cloud based on a distance between a first cloth grid point when the virtual cloth is overlaid on the point cloud data from a first direction and a second cloth grid point when the virtual cloth is overlaid on the point cloud data from a second direction opposite to the first direction; the clustering unit performs clustering on the point cloud other than the points determined to be the road surface; The electronic device wherein the determination unit determines whether the point cloud is concave or convex based on the number of concave / convex determinations assigned to the clustered point cloud.

2. the clustering unit clusters the point cloud other than the points determined to be the road surface into clusters of irregularities; The electronic device according to claim 1 .

3. the determining unit determines whether the cluster of clustered concave and convex portions is concave or convex based on a state of attachment between the point cloud and the virtual cloth from two directions. The electronic device according to claim 2 .

4. a providing unit that provides provision data capable of identifying the unevenness of the road surface determined by the determining unit, The electronic device according to claim 3 .

5. the acquisition unit acquires the point cloud data based on a depth image obtained by a detection unit mounted on a moving body measuring the road surface; the providing unit provides the provided data to the mobile object.

5. The electronic device according to claim 4.

6. Electronic devices, an acquisition step of acquiring point cloud data corresponding to points on a road surface; a cloth simulation step of outputting a shape of a point cloud represented by the point cloud data, the shape being determined based on a shape of the virtual cloth when the virtual cloth having a predetermined tension is placed over the point cloud data under a predetermined gravity; a clustering step of clustering the point cloud based on the shape of the point cloud; a determining step of determining the unevenness of the clustered point cloud; Including, the cloth simulation step includes determining a shape of the point cloud based on a distance between a first cloth grid point when the virtual cloth is overlaid on the point cloud data from a first direction and a second cloth grid point when the virtual cloth is overlaid on the point cloud data from a second direction opposite to the first direction; the clustering step performs clustering on the point cloud other than the points determined to be the road surface; A control method, wherein the determining step determines whether the point cloud is concave or convex based on the number of concave / convex judgments assigned to the clustered point cloud.

7. For electronic devices, an acquisition step of acquiring point cloud data corresponding to points on a road surface; a cloth simulation step of outputting a shape of a point cloud represented by the point cloud data, the shape being determined based on a shape of the virtual cloth when the virtual cloth having a predetermined tension is placed over the point cloud data under a predetermined gravity; a clustering step of clustering the point cloud based on the shape of the point cloud; a determining step of determining the unevenness of the clustered point cloud; Execute the cloth simulation step causes the electronic device to determine a shape of the point cloud based on a distance between a first cloth grid point when the virtual cloth is overlaid on the point cloud data from a first direction and a second cloth grid point when the virtual cloth is overlaid on the point cloud data from a second direction opposite to the first direction; the clustering step causes the electronic device to perform clustering on the point cloud other than the points determined to be the road surface; The determination step causes the electronic device to determine whether the point cloud is concave or convex based on the number of concave / convex judgments assigned to the clustered point cloud.

Images