Detection device, detection method, and detection program

Abstract

To enable an object to be properly detected using point cloud data.SOLUTION: A point cloud acquisition unit 21 acquires point cloud data obtained by receiving reflected light of irradiation light obtained by irradiating a target region and reflected at a reflection point. A voxelization unit 22 correlates a point of interest which is each point in the point cloud data, to a voxel including the position of the point of interest among a plurality of voxels into which the target region is divided, and thereby generates voxel data corresponding to the point cloud data. A difference identification unit 23 identifies, for each of the plurality of voxels, a difference cell having a difference between the voxel data and reference data on the basis of the presence or absence of the correlated point. In this case, the difference identification unit 23 does not identify the voxel to which the point is correlated in the reference data, as a difference cell, and identifies the voxel to which the point is correlated in the voxel data, as the difference cell.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a technique for detecting an object using point cloud data. [Background technology]

[0002] There is a technology that uses a visible light camera to detect objects such as abandoned property. When using a visible light camera to detect objects, the object cannot be detected properly in an environment where there is insufficient light, such as at night, or in an environment where direct sunlight shines on the visible light camera and causes flare.

[0003] There are optical sensors such as LiDAR that collect point cloud data by emitting light and receiving the reflected light reflected at a reflection point. LiDAR is an abbreviation for Light Detection and Ranging. Optical sensors can obtain point cloud data even in environments with insufficient light, such as at night, and in environments where direct sunlight shines on a visible light camera and causes flare. Therefore, detection of objects based on point cloud data obtained using optical sensors is being considered.

[0004] Patent Document 1 describes a method of comparing background point cloud data stored in a memory with current point cloud data, and identifying the appearance of an object if a difference occurs. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2020-118619 A [Patent Document 2] JP 2015-014514 A [Patent Document 3] JP 2020-076580 A Summary of the Invention [Problem to be solved by the invention]

[0006] With an optical sensor, it is difficult to always irradiate the same spot with light, and some deviation occurs. As a result, the generated point cloud data is slightly different each time. Therefore, as described in Patent Document 1, when comparing the background point cloud data with the current point cloud data, there is a possibility that an object may be identified as appearing even in a part where an object does not actually appear. An object of the present disclosure is to enable appropriate detection of objects using point cloud data. [Means for solving the problem]

[0007] The detection device according to the present disclosure comprises: a point cloud acquisition unit that acquires point cloud data obtained by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point; a voxelization unit that generates voxel data corresponding to the point cloud data by associating each point of the point cloud data acquired by the point cloud acquisition unit with a voxel that includes a position of the target point among a plurality of voxels obtained by dividing the target region, the voxelization unit treating each point of the point cloud data acquired by the point cloud acquisition unit as a target point; a difference identification unit that identifies, for each of the plurality of voxels, a difference cell having a difference between the voxel data generated by the voxel generation unit and reference data based on the presence or absence of a corresponding point, the difference identification unit not identifying a voxel with a corresponding point in the reference data as the difference cell, and identifying a voxel with no corresponding point in the reference data and with a corresponding point in the voxel data as the difference cell; Equipped with. Effect of the Invention

[0008] In the present disclosure, each point is associated with a plurality of voxels obtained by dividing the target area, and a difference is identified based on the presence or absence of an associated point for each voxel. This makes it possible to appropriately detect an object even if there is a slight deviation in the irradiation position of the irradiation light. In particular, in the present disclosure, a voxel with a point associated in the reference data is not identified as a difference cell, and a voxel with no point associated in the reference data and with a point associated in the voxel data is identified as a difference cell. This makes it possible to prevent an object that passes in front of the optical sensor that irradiates the irradiation light when acquiring the reference data from being detected. [Brief description of the drawings]

[0009] [Figure 1] 1 is a configuration diagram of a detection device 10 according to a first embodiment. [Diagram 2] FIG. 3 is an explanatory diagram of a voxel definition 31 according to the first embodiment. [Diagram 3] 3 is a flowchart showing a process flow of the detection device 10 according to the first embodiment. [Figure 4] FIG. 3 is an explanatory diagram of a voxelization process according to the first embodiment. [Diagram 5] FIG. 4 is an explanatory diagram of a difference identification process according to the first embodiment. [Figure 6] FIG. 4 is an explanatory diagram of a first filtering process according to the first embodiment. [Figure 7] FIG. 4 is an explanatory diagram of a second filtering process according to the first embodiment. [Figure 8] FIG. 4 is a configuration diagram of a detection device 10 according to a first modified example. [Figure 9] 10 is a flowchart showing the flow of processing by the detection device 10 according to the first modification. [Figure 10] FIG. 11 is a configuration diagram of a detection device 10 according to a second embodiment. [Figure 11] 10 is a flowchart showing a process flow of the detection device 10 according to the second embodiment. [Figure 12] FIG. 11 is a configuration diagram of a detection device 10 according to a third embodiment. [Figure 13] 10 is a flowchart showing a process flow of the detection device 10 according to the third embodiment. [Figure 14] FIG. 11 is a configuration diagram of a detection device 10 according to a fourth embodiment. [Figure 15] FIG. 13 is an explanatory diagram of a mask definition 34 according to the fourth embodiment. [Figure 16] 10 is a flowchart showing a process flow of the detection device 10 according to the fourth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Embodiment 1 ***Configuration Description*** The configuration of a detection device 10 according to the first embodiment will be described with reference to FIG. The sensing device 10 is a computer. The detection device 10 includes the following hardware components: a processor 11, a memory 12, a storage 13, and a communication interface 14. The processor 11 is connected to other hardware components via signal lines and controls the other hardware components.

[0011] The processor 11 is an IC that performs processing. IC is an abbreviation for Integrated Circuit. Specific examples of the processor 11 include a CPU, a DSP, and a GPU. CPU is an abbreviation for Central Processing Unit. DSP is an abbreviation for Digital Signal Processor. GPU is an abbreviation for Graphics Processing Unit.

[0012] The memory 12 is a storage device that temporarily stores data. Specific examples of the memory 12 include SRAM and DRAM. SRAM is an abbreviation for Static Random Access Memory. DRAM is an abbreviation for Dynamic Random Access Memory.

[0013] The storage 13 is a storage device that stores data. A specific example of the storage 13 is an HDD. HDD is an abbreviation for Hard Disk Drive. The storage 13 may also be a portable recording medium such as an SD (registered trademark) memory card, CompactFlash (registered trademark), NAND flash, a flexible disk, an optical disk, a compact disk, a Blu-ray (registered trademark) disk, or a DVD. SD is an abbreviation for Secure Digital. DVD is an abbreviation for Digital Versatile Disk.

[0014] The communication interface 14 is an interface for communicating with an external device. Specific examples of the communication interface 14 include an Ethernet (registered trademark), USB, and HDMI (registered trademark) port. USB stands for Universal Serial Bus. HDMI stands for High-Definition Multimedia Interface.

[0015] The detection device 10 is connected to an optical sensor 40 via the communication interface 14. The optical sensor 40 is a device that acquires point cloud data 41 by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point. The position of each point in the point cloud data 41 is identified based on the irradiation angle of the irradiation light and the time from when the irradiation light is irradiated to when the reflected light is received.

[0016] The detection device 10 includes, as functional components, a point cloud acquisition unit 21, a voxelization unit 22, a difference identification unit 23, and a detection unit 24. The functions of the functional components of the detection device 10 are realized by software. The storage 13 stores a program that realizes the function of each functional component of the detection device 10. The program is loaded into the memory 12 by the processor 11 and executed by the processor 11. In this way, the function of each functional component of the detection device 10 is realized.

[0017] The storage 13 also stores a voxel definition 31 .

[0018] 1 shows only one processor 11. However, there may be a plurality of processors 11, and the plurality of processors 11 may cooperate to execute programs that realize the respective functions.

[0019] ***Explanation of Operation*** The operation of the detection device 10 according to the first embodiment will be described with reference to FIGS. An operation procedure of the detection device 10 according to the embodiment 1 corresponds to a detection method according to the embodiment 1. Moreover, a program for realizing the operation of the detection device 10 according to the embodiment 1 corresponds to a detection program according to the embodiment 1.

[0020] The voxel definition 31 according to the first embodiment will be described with reference to FIG. The voxel definition 31 is data that defines the voxels 32. The voxels 32 are three-dimensional regions obtained by dividing the target region 33. The size of the voxels 32 can be set arbitrarily according to the position in the target region 33. In the first embodiment, the larger the distance from the optical sensor 40, the larger the size of the voxels 32 defined.

[0021] The flow of processing performed by the detection device 10 according to the first embodiment will be described with reference to FIG. In the first embodiment, the detection device 10 detects objects left behind in the target area 33 as targets.

[0022] (Step S11: Point cloud acquisition process) The point cloud acquisition unit 21 acquires point cloud data 41 collected by the optical sensor 40 .

[0023] (Step S12: Voxelization process) The voxelization unit 22 generates voxel data 42 from the point cloud data 41 acquired in step S11. A specific description will be given with reference to Fig. 4. The voxelization unit 22 sets each point of the point cloud data 41 acquired in step S11 as a target point. The voxelization unit 22 associates the target point with a voxel that includes the position of the target point among a plurality of voxels obtained by dividing the target region 33. In this way, voxel data 42 corresponding to the point cloud data 41 is generated.

[0024] (Step S13: Difference identification process) The difference identification unit 23 identifies a difference cell 44 having a difference between the voxel data 42 and the reference data 43. In the first embodiment, the reference data 43 is the voxel data 42 generated from the background point cloud data. That is, the reference data 43 is the voxel data 42 generated by performing the process of step S12 on the background point cloud data. The background point cloud data is the point cloud data 41 acquired by the optical sensor 40 in a state where no object or the like is present in the target region 33. Specifically, the difference identification unit 23 sets each of the multiple voxels 32 as a target voxel 32. The difference identification unit 23 determines whether or not there is a difference between the voxel data 42 and the reference data 43 for the target voxel 32, based on whether or not there is a corresponding point, that is, based on the presence or absence of a point in the voxel 32. If there is a difference, the difference identification unit 23 identifies the voxel 32 as a difference cell 44. 5, the difference identification unit 23 identifies a target voxel 32 as a difference cell 44 when no points are associated with the target voxel 32 in the reference data 43 and points are associated with the target voxel 32 in the voxel data 42. In other words, the difference identification unit 23 does not identify a voxel 32 with points associated with the target voxel 32 in the reference data 43 as a difference cell 44 even if the number of associated points is different.

[0025] (Step S14: First filtering process) The difference identification unit 23 excludes from the difference cells 44 those voxels 32 for which the number of associated points in the voxel data 42 is equal to or less than a reference number, among the difference cells 44 identified in step S13. For example, the number of reference points is set to 2. In this case, as shown in Fig. 6, voxels 32 for which the number of associated points in the voxel data 42 is two or less are excluded from the difference cell 44. In Fig. 6, the middle voxel 32 has two associated points and is therefore excluded from the difference cell 44. On the other hand, the bottom voxel 32 has three associated points and is therefore left as the difference cell 44.

[0026] (Step S15: Second filtering process) The difference identification unit 23 clusters the remaining difference cells 44 that were not excluded in step S14 to form a cluster 45. The clustering method may use an existing clustering technique, such as grouping adjacent difference cells 44 into the same cluster. The difference identification unit 23 excludes from the difference cells 44 the voxels 32 that constitute a cluster 45 in which the number of voxels 32 included in the obtained cluster 45 is equal to or less than a first voxel number. For example, the first voxel number is set to 5. In this case, as shown in Fig. 7, the voxels 32 constituting cluster 45B, which has 4 voxels 32, are excluded from the difference cell 44. On the other hand, the voxels 32 constituting cluster A, which has 10 voxels 32, are left as difference cells 44.

[0027] (Step S16: Detection process) The detection unit 24 detects an object based on the difference cells 44 identified by the difference identification unit 23. That is, the detection unit 24 detects an object based on the difference cells 44 that remain without being excluded in step S15. For example, the detection unit 24 detects a cluster 45 constituted by the difference cells 44 that are not excluded in step S15 as an object. In the process flow described using Fig. 3, similar to the size of the voxel 32, the number of first box accesses and the number of reference points may be set to be smaller as the distance from the optical sensor 40 increases. In this case, the setting of the number of first box accesses and the number of reference points may or may not be linked to the setting of the size of the voxel 32. Only one of the number of first box accesses or the number of reference points may be set. Also, the number of first box accesses and the number of reference points may be set to be larger as the distance from the optical sensor 40 increases. That is, the size of the voxel 32, the number of first box accesses, and the number of reference points are determined based on the distance from the optical sensor 40. A value according to the distance can be set.

[0028] ***Advantages of the First Embodiment*** As described above, the detection device 10 according to embodiment 1 detects the target object by matching each point of the point cloud data 41 to a plurality of voxels 32 obtained by dividing the target area 33, and identifying differences based on the presence or absence of a matched point for each voxel 32. It is difficult for the optical sensor 40 to always irradiate the same spot with the irradiated light, and some deviation occurs. As a result, the generated point cloud data 41 also differs slightly each time. Therefore, if the point cloud data 41 is simply compared with the reference data 43, differences will occur in various places, making it difficult to properly detect the target object. The detection device 10 according to the first embodiment performs the comparison for each voxel 32. This makes it difficult for a slight deviation in the irradiation position of the irradiation light to appear as a difference. As a result, it is possible to appropriately detect the object.

[0029] The farther away from the optical sensor 40 is the reflection point, the greater the deviation in position will be even if the angle at which the optical sensor 40 irradiates light is slightly shifted. Therefore, the farther away from the optical sensor 40 is the reflection point, the more likely deviation in the irradiated light will occur. In the first embodiment, the size of the voxel 32 increases as the distance from the optical sensor 40 increases. Therefore, even in an area far from the optical sensor 40, the deviation of the irradiation position of the irradiated light is unlikely to appear as a difference, and the object can be appropriately detected.

[0030] When acquiring the background point cloud data, an object may pass in front of the optical sensor 40. In this case, no point cloud exists in the voxel data 42, and many voxels 32 exist in the reference data 43 generated from the background point cloud data. The detection device 10 according to the first embodiment does not identify a voxel 32 with which a point is associated in the reference data 43 as a difference cell 44. The detection device 10 according to the first embodiment identifies a voxel 32 with which a point is not associated in the reference data 43 and which is associated in the voxel data 42 as a difference cell 44. Therefore, it is possible to prevent an object passing in front of the optical sensor 40 from being detected.

[0031] Wind or vibrations can cause the detected positions of vegetation or structures to shift, and such shifts may not be absorbed by voxelization alone. The detection device 10 according to the first embodiment excludes voxels 32 whose number of points is equal to or less than a reference number of points from the difference cell 44. Furthermore, the detection device 10 according to the first embodiment excludes voxels 32 that constitute a cluster 45 whose number of voxels 32 is equal to or less than a first voxel number from the difference cell 44. This makes it difficult for a point group that appears due to wind or vibration to be identified as a difference cell 44. As a result, it is possible to appropriately detect an object.

[0032] ***Other configurations*** <Variation 1> In the first embodiment, the detection unit 24, which is a functional component of the detection device 10, detects the final object. However, the detection of the final object may be configured to be performed by a specific type of object detection device that is an external device to the detection device 10. In other words, the detection device 10 may be configured to perform processing up to generating the difference cell 44, which is information for detecting the final object, and the detection of the final object may be performed by an external device.

[0033] The configuration of a detection device 10 according to the first modification will be described with reference to FIG. Unlike the detection device 10 shown in FIG. 1, the detection device 10 does not include a detection unit 24 as a functional component. Furthermore, detection of an object is performed by a specific type of object detection device. The specific type is a person, a vehicle, etc. In the third embodiment, the specific type is a person, and the detection of a person is performed by a specific type of object detection device. An example of outputting the difference cell 44 to the detection device will be described.

[0034] The flow of processing performed by the detection device 10 according to the first modification will be described with reference to FIG. The processing from step S21 to step S25 is the same as the processing from step S11 to step S15 in FIG.

[0035] (Step S26: Detection result output process) The difference specifying unit 23 passes the information of the difference cell 44 to the human detection device. The human detection device receives the information in the difference cell 44 as input information, judges whether or not the object is a human, and if it judges that the object is a human, judges the position of the human. As described above, in the detection device 10 according to the first modification, detection of an object and determination of the type are performed by a specific type of object detection device. Therefore, an effect is obtained in that the processing load on the specific type of object detection device, for example, the human detection device, is reduced. In addition, an effect is expected in that erroneous detection on the human detection device side is suppressed.

[0036] Embodiment 2 The second embodiment differs from the first embodiment in that the target object is detected after the identification of the differential cells 44 is repeated for a certain period of time. In the second embodiment, this difference will be described, and a description of the same points will be omitted.

[0037] ***Configuration Description*** The configuration of a detection device 10 according to the second embodiment will be described with reference to FIG. 1 in that the detection device 10 includes, as a functional component, a storage processing unit 25. The function of the storage processing unit 25, like the other functions, is realized by software.

[0038] ***Explanation of Operation*** The flow of processing performed by the detection device 10 according to the second embodiment will be described with reference to FIG. 11 is repeatedly executed for each reference interval. The processes from step S31 to step S35 are the same as the processes from step S11 to step S15 in FIG.

[0039] (Step S36: Storage process) The accumulation processing unit 25 records in the memory 12 the voxels 32 that were not excluded in step S35 and remain as difference cells 44.

[0040] (Step S37: High-proportion cell identification process) The accumulation processing unit 25 identifies as target cells 46 voxels 32 whose proportion of voxels 32 recorded as differential cells 44 is equal to or greater than a reference proportion. Specifically, the number of times the process shown in FIG. 11 was executed in the most recent reference period is set as T. The accumulation processing unit 25 sets each voxel 32 as a target voxel 32. The accumulation processing unit 25 identifies the number of times C that the target voxel 32 was recorded as a difference cell 44 in the most recent reference period. The accumulation processing unit 25 divides the number of times C by the number of times T to calculate a proportion R of the target voxel 32 that was recorded as a difference cell 44 in the most recent reference period. If the proportion R is equal to or greater than the reference proportion α, the accumulation processing unit 25 identifies the target voxel 32 as a target cell 46.

[0041] The reference rate α may be set to a value according to the position in the target area 33. Here, the reference rate α is set to a lower value as the distance from the optical sensor 40 in the target area 333 increases.

[0042] (Step S38: Clustering process) The detection unit 24 performs clustering on the target cells 46 identified in step S37 to form a cluster 47. The clustering method may use an existing clustering technique, such as grouping adjacent difference cells 44 into the same cluster.

[0043] (Step S39: Detection process) The detection unit 24 detects, as the object, a cluster 47 that satisfies the condition among the clusters 47 formed in step S38. Specifically, the detection unit 24 detects a cluster 47 that satisfies both of the following conditions 1 and 2 as an object. Condition 1 is a condition that the number of voxels 32 included in the cluster 47 is greater than a second voxel count. Condition 2 is a condition that the size of the cluster 47 is greater than a reference size. The size of the cluster 47 is any one of the vertical direction, horizontal direction, and depth direction, or a combination of at least two or more of the vertical direction, horizontal direction, and depth direction. Here, the second number of voxels may be the same value as the first number of voxels, or may be a different value. Also, like the first number of voxels, the second number of voxels may be set to be smaller the farther the distance from the optical sensor 40. In this case, the setting of the second number of voxels may or may not be linked to the setting of the size of the voxels 32. Also, the second number of voxels may be set to be larger the farther the distance from the optical sensor 40. In other words, the second number of voxels can be set to a value according to the distance from the optical sensor 40.

[0044] ***Effects of the second embodiment*** As described above, the detection device 10 according to the second embodiment detects an object after repeatedly identifying the differential cells 44 for a certain period of time. This makes it possible to appropriately detect an abandoned object, which is an object.

[0045] The detection device 10 according to the second embodiment detects an object using voxels 32 in which the proportion R recorded as the difference cells 44 is equal to or greater than the reference proportion α. ​​By using the proportion recorded as the difference cells 44, the object can be detected even if it is temporarily obscured, for example, by being hidden behind an object.

[0046] In the second embodiment, the reference ratio α is set to a lower value as the distance from the optical sensor 40 in the target region 333 increases. As a result, the more distant the position, the lower the ratio R will be for it to be detected as an object. There is a tendency for point clouds to be less accurately acquired because the more distant the position, the more likely it is to be occluded, making it easier to prevent missed detection of objects at distant positions. On the other hand, when an object is located close by, the reference ratio α is set to be high, making it easier to prevent it from being erroneously detected as an object.

[0047] Embodiment 3 The third embodiment differs from the first embodiment in that the movement trajectory of the object is specified. In the third embodiment, this difference will be described, and a description of the same points will be omitted.

[0048] ***Configuration Description*** The configuration of a detection device 10 according to the third embodiment will be described with reference to FIG. 1 in that the detection device 10 includes a trajectory identification unit 28 and a notification unit 29 instead of the detection unit 24 as functional components. The functions of the trajectory identification unit 28 and the notification unit 29 are realized by software, similar to other functions.

[0049] ***Explanation of Operation*** The flow of processing performed by the detection device 10 according to the third embodiment will be described with reference to FIG. In the third embodiment, the detection device 10 detects a moving object moving in the target area 33 as a target object.

[0050] The process from step S41 to step S45 is repeated at each reference interval. The process from step S41 to step S45 is the same as the process from step S11 to step S15 in Fig. 3. However, in step S45, the identified cluster 45 is recorded in the memory 12 for a certain period of time.

[0051] (Step S46: Trajectory identification process) The trajectory identification unit 28 links the cluster 45 identified in the most recent step S45 with the corresponding cluster 45 among the clusters 45 identified in step S45 executed one cycle before, to identify a movement trajectory. Specifically, the trajectory identification unit 28 compares the position of the cluster 45 identified in the most recent step S45 with the cluster 45 identified in the previously executed step S45. If the positional relationship is as expected from the moving direction and speed of the object, the trajectory identification unit 28 determines that both clusters 45 are the same object. At this time, the trajectory identification unit 28 may determine whether or not both clusters 45 are the same object, taking into consideration the condition that the difference in size between both clusters 45 is less than an upper limit value.

[0052] (Step S47: Notification process) The notification unit 29 notifies the user of the movement trajectory identified in step S46. Specifically, the notification unit 29 identifies a movement trajectory that satisfies the conditions of having characteristics of a moving object, such as the same moving direction and a constant moving speed. Then, the notification unit 29 notifies the user of the identified movement trajectory. For example, the notification unit 29 notifies the user by displaying the identified movement trajectory.

[0053] ***Effects of the Third Embodiment*** As described above, the detection device 10 according to the third embodiment identifies and notifies the movement trajectory of the target object. This makes it possible to identify the movement trajectory while detecting the moving object.

[0054] The detection device 10 according to the third embodiment notifies only the movement trajectory that satisfies the condition of having the characteristics of a moving object. This allows the notification unit 29 to notify after excluding objects that continue to move in irregular directions in the same area, such as swaying grass and trees.

[0055] ***Other configurations*** <Variation 2> In the first embodiment, voxel data 42 generated from background point cloud data is used as reference data 43. In the third embodiment, voxel data 42 generated in step S42 one cycle before may be used as reference data 43. This makes it possible to omit processes such as obtaining background point cloud data in advance and periodically updating the background point cloud data.

[0056] Embodiment 4 The fourth embodiment differs from the first to fourth embodiments in that an area in which no object is detected is set within the target area 33. In the fourth embodiment, this difference will be described, and a description of the same points will be omitted. In the fourth embodiment, a case will be described in which a function is added to the first embodiment. However, it is also possible to add a function to any of the second to fourth embodiments.

[0057] ***Configuration Description*** The configuration of a detection device 10 according to the fourth embodiment will be described with reference to FIG. The detection device 10 differs from the detection device 10 shown in FIG. 1 in that it includes a mask processing unit 30 as a functional component. The function of the mask processing unit 30, like other functions, is controlled by software. This is achieved by: The storage 13 also stores a mask definition 34 .

[0058] ***Explanation of Operation*** The mask definition 34 according to the fourth embodiment will be described with reference to FIG. The mask definition 34 indicates an area to be excluded from detection within the target area 33. For example, if the target is an object placed on the ground, an area in the air that is a certain distance from the ground is set as an area to be excluded from detection. The mask definition 34 may be set by accepting a designation of an area to be excluded from the detection target. Alternatively, the mask definition 34 may accept a designation of an area to be the target area 33 within a certain area, and the area not designated may be set as an area to be excluded from the detection target.

[0059] The flow of processing performed by the detection device 10 according to the fourth embodiment will be described with reference to FIG. The process of step S51 is the same as the process of step S11 in Fig. 3. The processes of steps S55 to S57 are the same as the processes of steps S14 to S16 in Fig. 3.

[0060] (Step S52: Masking) The mask processing unit 30 sets the area of ​​the target area 33 indicated by the mask definition 34 as a mask area.

[0061] (Step S53: Voxelization process) 3, the voxelization unit 22 generates voxel data 42 from the point cloud data 41. At this time, the voxelization unit 22 associates points only with the voxels 32 in the target region 33 excluding the mask region set in step S52.

[0062] (Step S54: Difference identification process) 3, the difference identification unit 23 identifies difference cells 44 where there is a difference between the voxel data 42 and the reference data 43. At this time, the difference identification unit 23 identifies the difference cells 44 by comparing only the voxels 32 that constitute the target region 33 excluding the mask region set in step S52.

[0063] ***Effects of the Fourth Embodiment*** As described above, the detection device 10 according to the fourth embodiment sets a mask area and detects an object only in the target area 33 excluding the mask area. This makes it possible to reduce the calculation load. It is also possible to suppress unnecessary erroneous detection.

[0064] ***Other configurations*** <Modification 3> In the first embodiment, each functional component is realized by software. However, as a third modification, each functional component may be realized by hardware. The following describes the third modification, focusing on the differences from the first embodiment.

[0065] When each functional component is realized by hardware, the detection device 10 includes an electronic circuit 15 instead of the processor 11, the memory 12, and the storage 13. The electronic circuit 15 is a dedicated circuit for realizing the functions of each functional component, the memory 12, and the storage 13.

[0066] The electronic circuit 15 may be a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a logic IC, a GA, an ASIC, or an FPGA. is an abbreviation for Gate Array. ASIC is an abbreviation for Application Specific Integrated Circuit. FPGA is an abbreviation for Field-Programmable Gate Array. Each functional component may be realized by one electronic circuit 15, or each functional component may be distributed across multiple electronic circuits 15.

[0067] <Variation 4> As a fourth modification, some of the functional components may be realized by hardware, and other functional components may be realized by software.

[0068] The processor 11, the memory 12, the storage 13, and the electronic circuit 15 are collectively referred to as a processing circuit. In other words, the functions of the respective functional components are realized by the processing circuit.

[0069] Furthermore, the term "part" in the above description may be read as a "circuit," "step," "procedure," "processing," or "processing circuit."

[0070] Various aspects of the present disclosure are summarized below as appendices. (Supplementary Note 1) A point cloud acquisition unit that acquires point cloud data obtained by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point; a voxelization unit that generates voxel data corresponding to the point cloud data by associating each point of the point cloud data acquired by the point cloud acquisition unit with a voxel that includes a position of the target point among a plurality of voxels obtained by dividing the target region, the voxelization unit treating each point of the point cloud data acquired by the point cloud acquisition unit as a target point; a difference identification unit that identifies a difference cell having a difference between the voxel data generated by the voxel generation unit and reference data based on the presence or absence of a corresponding point for each of the plurality of voxels; A detection device comprising: (Appendix 2) The detection device further comprises: a detection unit that detects an object based on the differential cell identified by the differential identification unit; 2. The detection device of claim 1, comprising: (Appendix 3) The difference specification unit transfers the specified difference cell to a specific type of object detection device. 2. A detection device as described in claim 1. (Appendix 4) The difference identification unit does not identify a voxel to which a point is associated in the reference data as the difference cell, and identifies a voxel to which a point is not associated in the reference data and which a point is associated in the voxel data as the difference cell. 2. A detection device as described in claim 1. (Appendix 5) The difference identification unit excludes voxels having a number of corresponding points in the voxel data equal to or less than a reference number from the difference cells. 5. A detection device as described in claim 4. (Appendix 6) The number of reference points is set to a value according to the distance from an optical sensor that acquires the point cloud data in the target area. 6. A detection device as described in claim 5. (Appendix 7) The difference identification unit excludes voxels that constitute a cluster obtained by clustering the difference cells and in which the number of voxels included in the cluster is equal to or less than a first number of voxels from the difference cells. do 5. A detection device as described in claim 4. (Appendix 8) The first number of voxels is set to a value according to a distance from an optical sensor that acquires the point cloud data in the target region. 8. The detection device according to claim 7. (Appendix 9) The point cloud acquisition unit repeatedly acquires point cloud data, The detection device further comprises: an accumulation processing unit that identifies, as target cells, voxels whose ratio of identified differential cells is equal to or greater than a reference ratio among the voxels identified as the differential cells based on the point cloud data acquired during a detection period; Equipped with The detection unit detects an object based on the target cell identified by the storage processing unit. 2. A detection device as described in claim 1. (Appendix 10) The detection device according to claim 9, wherein the detection unit detects, as the target object, a cluster obtained by clustering the target cell, the cluster having a number of voxels contained therein that is greater than a second number of voxels. (Appendix 11) At least one of the reference ratio and the second number of voxels is set to a value according to a distance from an optical sensor that acquires the point cloud data in the target region. 11. The detection device of claim 10. (Appendix 12) The point cloud acquisition unit repeatedly acquires point cloud data, The detection device further comprises: a trajectory identification unit that identifies a movement trajectory by linking each of the repeatedly acquired point cloud data as target point cloud data, one or more clusters obtained by clustering the difference cells identified based on the target point cloud data, with a corresponding cluster among one or more clusters obtained by clustering the difference cells identified based on the point cloud data acquired before the target point cloud data; 2. The detection device of claim 1, comprising: (Appendix 13) The detection device further comprises: a notification unit that notifies, as a notification target, a movement trajectory that satisfies a condition defined based on at least one of a movement direction and a movement speed among the movement trajectories identified by the trajectory identification unit. 13. The detection device of claim 12, comprising: (Appendix 14) The reference data is voxel data corresponding to point cloud data acquired in a state where the object does not exist in the target region. 2. A detection device as described in claim 1. (Appendix 15) The reference data is voxel data corresponding to point cloud data acquired immediately before the point cloud data of the target. 13. The detection device of claim 12. (Appendix 16) The detection device further comprises: A masking processing unit that sets a mask area that is to be excluded from detection within the target area Equipped with The difference identification unit identifies a difference cell having a difference between the voxel data and the reference data for voxels constituting a region of the target region excluding the mask region set by the mask processing unit. 2. A detection device as described in claim 1. (Appendix 17) The computer acquires point cloud data obtained by irradiating the target area with irradiation light and receiving reflected light reflected at reflection points; a computer generates voxel data corresponding to the point cloud data by associating each point of the point cloud data with a voxel that includes a position of the target point among a plurality of voxels obtained by dividing the target region, the voxel data being corresponding to the point cloud data; A detection method in which a computer identifies, for each of the plurality of voxels, a difference cell in which there is a difference between the voxel data and reference data, based on the presence or absence of a corresponding point. (Appendix 18) A point cloud acquisition process for acquiring point cloud data obtained by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point; a voxelization process in which each point of the point cloud data acquired by the point cloud acquisition process is regarded as a target point, and the target point is associated with a voxel that includes the position of the target point among a plurality of voxels obtained by dividing the target area, thereby generating voxel data corresponding to the point cloud data; a difference identification process for identifying a difference cell having a difference between the voxel data generated by the voxelization process and reference data, based on the presence or absence of a corresponding point for each of the plurality of voxels; A detection program that causes a computer to function as a detection device that performs the above steps.

[0071] The above describes the embodiments and modifications of the present disclosure. Some of these embodiments and modifications may be combined and implemented. Also, one or some of them may be partially implemented. Note that the present disclosure is not limited to the above embodiments and modifications, and various modifications are possible as necessary. [Explanation of symbols]

[0072] 10 detection device, 11 processor, 12 memory, 13 storage, 14 communication interface, 21 point cloud acquisition unit, 22 voxelization unit, 23 difference identification unit, 24 detection unit, 25 accumulation processing unit, 26 position acquisition unit, 27 type determination unit, 28 trajectory identification unit, 29 notification unit, 30 mask processing unit, 31 voxel definition, 32 voxel, 33 target area, 34 mask definition, 40 optical sensor, 41 point cloud data, 42 voxel data, 43 reference data, 44 difference cell, 45 cluster, 46 target cell, 47 cluster.

Claims

1. a point cloud acquisition unit that acquires point cloud data obtained by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point; a voxelization unit that generates voxel data corresponding to the point cloud data by associating each point of the point cloud data acquired by the point cloud acquisition unit with a voxel that includes a position of the target point among a plurality of voxels obtained by dividing the target region, the voxelization unit treating each point of the point cloud data acquired by the point cloud acquisition unit as a target point; a difference identification unit that identifies, for each of the plurality of voxels, a difference cell having a difference between the voxel data generated by the voxel generation unit and reference data based on the presence or absence of a corresponding point, the difference identification unit not identifying a voxel with a corresponding point in the reference data as the difference cell, and identifying a voxel with no corresponding point in the reference data and with a corresponding point in the voxel data as the difference cell; A detection device comprising:

2. The detection device further comprises: a detection unit that detects an object based on the differential cell identified by the differential identification unit; The detection device of claim 1 .

3. The difference specification unit transfers the specified difference cell to a specific type of object detection device. The detection device of claim 1 .

4. The difference identification unit excludes voxels having a number of corresponding points in the voxel data equal to or less than a reference number from the difference cells. The detection device of claim 1 .

5. The number of reference points is set to a value according to the distance from an optical sensor that acquires the point cloud data in the target area.

5. The detection device of claim 4.

6. The difference identification unit excludes voxels that constitute clusters obtained by clustering the difference cells and in which the number of voxels included in the clusters is equal to or less than a first voxel number from the difference cells. The detection device of claim 1 .

7. The first number of voxels is set to a value according to a distance from an optical sensor that acquires the point cloud data in the target region. The detection device of claim 6.

8. The point cloud acquisition unit repeatedly acquires point cloud data, The detection device further comprises: an accumulation processing unit that identifies, as target cells, voxels whose ratio of identified differential cells is equal to or greater than a reference ratio among the voxels identified as the differential cells based on the point cloud data acquired during a detection period; Equipped with The detection unit detects an object based on the target cell identified by the storage processing unit. The detection device of claim 2.

9. The detection device according to claim 8 , wherein the detection unit detects, as the object, a cluster obtained by clustering the target cell, the cluster including a number of voxels greater than a second number of voxels.

10. At least one of the reference ratio and the second number of voxels is set to a value according to a distance from an optical sensor that acquires the point cloud data in the target region.

10. The detection device of claim 9.

11. The point cloud acquisition unit repeatedly acquires point cloud data, The detection device further comprises: a trajectory identification unit that identifies a movement trajectory by linking each of the repeatedly acquired point cloud data as target point cloud data, one or more clusters obtained by clustering the difference cells identified based on the target point cloud data, with a corresponding cluster among one or more clusters obtained by clustering the difference cells identified based on the point cloud data acquired before the target point cloud data; The detection device of claim 1 .

12. The detection device further comprises: a notification unit that notifies, as a notification target, a movement trajectory that satisfies a condition defined based on at least one of a movement direction and a movement speed among the movement trajectories identified by the trajectory identification unit. The sensing device of claim 11 .

13. The reference data is voxel data corresponding to point cloud data acquired in a state where no object is present in the target region. The detection device of claim 1 .

14. The reference data is voxel data corresponding to point cloud data acquired immediately before the point cloud data of the target. The sensing device of claim 11.

15. The detection device further comprises: A masking processing unit that sets a mask area that is to be excluded from detection within the target area Equipped with The difference identification unit identifies a difference cell having a difference between the voxel data and the reference data for voxels constituting a region of the target region excluding the mask region set by the mask processing unit. The detection device of claim 1 .

16. The computer acquires point cloud data obtained by irradiating the target area with irradiation light and receiving reflected light reflected at reflection points; a computer generates voxel data corresponding to the point cloud data by associating each point of the point cloud data with a voxel that includes a position of the target point among a plurality of voxels obtained by dividing the target region, the voxel data being corresponding to the point cloud data; A detection method in which a computer identifies, for each of the multiple voxels, difference cells in which there is a difference between the voxel data and reference data based on the presence or absence of corresponding points, does not identify voxels with corresponding points in the reference data as difference cells, and identifies voxels with no corresponding points in the reference data and with corresponding points in the voxel data as difference cells.

17. A point cloud acquisition process for acquiring point cloud data obtained by irradiating a target area with irradiation light and receiving reflected light reflected at a reflection point; a voxelization process in which each point of the point cloud data acquired by the point cloud acquisition process is regarded as a target point, and the target point is associated with a voxel that includes the position of the target point among a plurality of voxels obtained by dividing the target area, thereby generating voxel data corresponding to the point cloud data; a difference identification process for identifying a difference cell having a difference between the voxel data generated by the voxelization process and reference data based on the presence or absence of a corresponding point for each of the plurality of voxels, the difference identification process not identifying a voxel with a corresponding point in the reference data as the difference cell, and identifying a voxel with no corresponding point in the reference data and with a corresponding point in the voxel data as the difference cell; A detection program that causes a computer to function as a detection device that performs the above steps.

Images