Movable body monitoring device, movable body monitoring method, and movable body monitoring program

Abstract

To appropriately determine whether a movable body exits even when illumination occurs or in the dark of a night.SOLUTION: A point group acquisition unit 21 acquires point group data obtained by applying light to a detection area and receiving light reflected at a reflection point. A difference identification unit 23 identifies a monitoring difference between background point group data acquired in a state where there is no movable body in the detection area and point group data acquired during a monitoring period. A movable body determination unit 25 determines whether a movable body exists on the basis of the monitoring difference included in a movement area where the movable body moves.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a technique for determining whether a moving object such as a train or a car is present. [Background technology]

[0002] There are systems that change control depending on whether or not a moving object is present. For example, a system for preventing collisions with trains at railroad crossings may be turned on when a train is approaching the crossing and turned off when there is no train near the crossing.Also, a system for detecting suspicious objects at tram or bus stops may be turned on when there is no tram approaching the tram stop or no bus approaching the bus stop, and turned off when there is a tram at the tram stop or a bus at the bus stop.

[0003] There is a technology that uses image data obtained by a camera to determine whether or not a moving object is present in a detection area (see Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2004-042777 Summary of the Invention [Problem to be solved by the invention]

[0005] When using image data obtained by a camera, there are issues such as the risk of false detection due to changes in lighting, and detection being difficult in the dark at night. The present disclosure aims to make it possible to appropriately determine whether or not a moving object is present even when there is a change in illumination or in the dark at night. [Means for solving the problem]

[0006] A mobile object monitoring device according to the present disclosure includes: a point cloud acquisition unit that acquires point cloud data obtained by irradiating a detection area with light and receiving reflected light reflected at a reflection point; a difference specifying unit that specifies a monitoring difference that is a difference between background point cloud data, which is point cloud data acquired by the point cloud acquiring unit when there is no moving object in the detection area, and point cloud data acquired by the point cloud acquiring unit during a monitoring period; a moving object determination unit that determines whether or not the moving object exists based on the monitoring difference identified by the difference identification unit, the monitoring difference being included in a moving area in which the moving object moves; Equipped with. [Effects of the Invention]

[0007] In the present disclosure, whether or not a moving object exists is determined using point cloud data, thereby enabling appropriate determination of whether or not a moving object exists even when there is a change in lighting or in the dark at night. In particular, in the present disclosure, the presence or absence of a moving object is determined based on the monitoring difference included in the moving area. Processing using point cloud data tends to require a larger amount of calculation than processing using image data. However, since the presence or absence of a moving object is determined based on the monitoring difference included in the moving area, rather than detecting a moving object based on point cloud data for the entire detection area, it is possible to reduce the amount of calculation. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a configuration diagram of a moving object monitoring device 10 according to a first embodiment. [Figure 2] 3 is a flowchart of the overall processing of the moving object monitoring device 10 according to the first embodiment. [Figure 3] 10 is a flowchart of a background point group generation process according to the first embodiment. [Figure 4] 10 is a flowchart of a movement area identification process according to the first embodiment. [Figure 5] FIG. 3 is an explanatory diagram of the appearance determination process according to the first embodiment. [Figure 6] FIG. 3 is an explanatory diagram of an area selection process according to the first embodiment. [Figure 7] 4 is a flowchart of a moving body determination process according to the first embodiment. [Figure 8] FIG. 10 is a configuration diagram of a moving object monitoring device 10 according to a second embodiment. [Figure 9] FIG. 10 is an explanatory diagram of a voxel definition 31 according to the second embodiment. [Figure 10] FIG. 10 is an explanatory diagram of a method for generating voxel data according to the second embodiment. [Figure 11] 10 is a flowchart of a background point group generation process according to the second embodiment. [Figure 12] 10 is a flowchart of a movement area identification process according to the second embodiment. [Figure 13] FIG. 10 is an explanatory diagram of a difference identification process according to the second embodiment. [Figure 14] 10 is a flowchart of a moving body determination process according to the second embodiment. [Figure 15] FIG. 11 is an explanatory diagram of a difference identification process according to Modification 3. DETAILED DESCRIPTION OF THE INVENTION

[0009] Embodiment 1 The mobile object monitoring device 10 according to the first embodiment determines whether or not a mobile object 60 is present in a detection area 50. The mobile object 60 may be a train, a vehicle, an aircraft, or the like. If the mobile object 60 is a train, the detection area 50 may be a railroad crossing area, a railroad track area near the crossing, or the like. If the mobile object 60 is a vehicle, the detection area 50 may be a bus stop area, an intersection area, or the like. If the mobile object 60 is an aircraft, the detection area 50 may be an area near a runway, an aircraft parking area, or the like. In the first embodiment, a case where the moving object 60 is a train will be described as an example.

[0010] ***Configuration Description*** The configuration of a moving object monitoring device 10 according to the first embodiment will be described with reference to FIG. The mobile object monitoring device 10 is a computer. The mobile object monitoring device 10 includes hardware such as a processor 11, a memory 12, a storage 13, and a communication interface 14. The processor 11 is connected to other hardware via signal lines and controls the other hardware.

[0011] The processor 11 is an IC that performs processing. IC stands for Integrated Circuit. Specific examples of the processor 11 include a CPU, a DSP, and a GPU. CPU stands for Central Processing Unit. DSP stands for Digital Signal Processor. GPU stands 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 stands for Static Random Access Memory. DRAM stands 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 external devices. Specific examples of the communication interface 14 include Ethernet (registered trademark), USB, and HDMI (registered trademark) ports. USB stands for Universal Serial Bus. HDMI stands for High-Definition Multimedia Interface.

[0015] The mobile object monitoring device 10 includes, as functional components, a point cloud acquisition unit 21, a background point cloud generation unit 22, a difference identification unit 23, a movement area identification unit 24, a moving object determination unit 25, and a notification unit 26. The movement area identification unit 24 includes an object detection unit 241, an appearance determination unit 242, a point cloud accumulation unit 243, and an area identification unit 244. The functions of each functional component of the mobile object monitoring device 10 are realized by software. The storage 13 stores programs that realize the functions of each functional component of the mobile object monitoring device 10. These programs are loaded into the memory 12 by the processor 11 and executed by the processor 11. As a result, the functions of each functional component of the mobile object monitoring device 10 are realized.

[0016] 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.

[0017] ***Explanation of Operation*** The operation of the moving object monitoring device 10 according to the first embodiment will be described with reference to FIGS. The operation procedure of the mobile object monitoring device 10 according to the embodiment 1 corresponds to the mobile object monitoring method according to the embodiment 1. Furthermore, the program that realizes the operation of the mobile object monitoring device 10 according to the embodiment 1 corresponds to the mobile object monitoring program according to the embodiment 1.

[0018] The overall processing of the moving object monitoring device 10 according to the first embodiment will be described with reference to FIG. (Step S1: Background point cloud generation process) The mobile object monitoring device 10 generates background point cloud data for the detection area 50. The background point cloud data is generated from point cloud data acquired when no mobile object 60 is present in the detection area 50.

[0019] (Step S2: Movement area identification process) The mobile object monitoring device 10 identifies a movement area 51 in the detection area 50 in which the mobile object 60 moves, based on the difference between the point cloud data acquired when the mobile object 60 is present in the detection area 50 and the background point cloud data.

[0020] (Step S3: Moving object determination process) The mobile object monitoring device 10 determines whether or not a mobile object 60 exists in the detection area 50 based on the difference between the point cloud data acquired during the monitoring period and the background point cloud data, and the movement area 51.

[0021] The background point cloud generation process (step S1 in FIG. 2) according to the first embodiment will be described with reference to FIG. The background point cloud generation process is executed during an absence period when there is no moving object 60 in the detection area 50. For example, if the moving object 60 is a train, the absence period is a period when the train is not running, such as at night. The absence period is a period of some length, such as 10 minutes.

[0022] (Step S101: Point cloud acquisition process) The point cloud acquisition unit 21 acquires point cloud data by irradiating the detection area 50 with light using an optical sensor and receiving light reflected at reflection points. A specific example of the optical sensor is a LiDAR. LiDAR stands for Light Detection and Ranging (LDAR). It is an abbreviation for Ranging.

[0023] (Step S102: End determination process) The background point cloud generating unit 22 determines whether or not the acquisition of point cloud data has been completed. Specifically, the background point cloud generating unit 22 determines whether the non-existence period has ended. If the non-existence period has ended, the background point cloud generating unit 22 determines that the acquisition of point cloud data has ended. If the background point cloud generation unit 22 has not yet completed acquisition of the point cloud data, the process proceeds to step S103. On the other hand, if the background point cloud generation unit 22 has completed acquisition of the point cloud data, the process ends.

[0024] (Step S103: Point cloud addition process) The background point cloud generation unit 22 adds the points included in the point cloud data acquired in step S101 to the background point cloud data.

[0025] That is, the background point cloud data is generated by superimposing multiple point cloud data acquired during the non-existence period. The point cloud data is acquired at a reference interval. The reference interval is set arbitrarily depending on the situation of the detection area 50, etc. The reference interval is, for example, 0.2 seconds. The background point cloud data is a superposition of multiple point cloud data acquired during periods of absence. Therefore, if there are plants in the detection area 50 and the plants are swaying, the background point cloud data will include point data for each position within the swaying range.

[0026] The travel area identification process (step S2 in FIG. 2) according to the first embodiment will be described with reference to FIG. The movement area identification process is executed for a set period that includes the time when the moving object 60 is in the detection area 50. For example, if the moving object 60 is a train, the set period is the period when the train is in operation, such as daytime. The set period is a period long enough for multiple moving objects to pass through. For example, if a train passes about once every five minutes, the set period is one or two hours.

[0027] (Step S201: Point cloud acquisition process) The point cloud acquisition unit 21 acquires point cloud data obtained by irradiating the detection area 50 with light using an optical sensor and receiving the light reflected at reflection points.

[0028] (Step S202: End determination process) The movement area specifying unit 24 determines whether or not the acquisition of point cloud data has been completed. Specifically, the movement area specifying unit 24 determines whether the set period has ended. If the set period has ended, the movement area specifying unit 24 determines that the acquisition of point cloud data has ended. If the movement area identification unit 24 has not yet completed acquisition of the point cloud data, the process proceeds to step S203. On the other hand, if the movement area identification unit 24 has completed acquisition of the point cloud data, the process proceeds to step S210.

[0029] (Step S203: Difference Identification Process) The difference identification unit 23 identifies a set difference, which is the difference between the background point cloud data generated in step S1 of FIG. 2 and the point cloud data acquired in step S201. The set difference includes points that are not in the background point cloud data but are in the point cloud data acquired in step S201, and points that are in the background point cloud data but not in the point cloud data acquired in step S201. Alternatively, the set difference may include only points that are not in the background point cloud data but are in the point cloud data acquired in step S201. The difference identification unit 23 stores the set difference in memory 12 as provisional data.

[0030] (Step S204: Object detection process) The object detection unit 241 detects an object from the set difference identified in step S203. Specifically, the object detection unit 241 clusters the points included in the set difference. The object detection unit 241 detects each set of points obtained by the clustering as an object existing in the detection area 50.

[0031] (Step S205: Period Determination Process) The appearance determination unit 242 determines whether or not a set difference for a reference period has been accumulated as provisional data. The reference period is a period of time that is approximately the length of time it takes for a moving object to pass through the detection area 50. If the set difference for the reference period has been accumulated, the occurrence determination unit 242 advances the process to step S206. On the other hand, if the set difference for the reference period has not been accumulated, the occurrence determination unit 242 returns the process to step S201.

[0032] (Step S206: Appearance determination process) The appearance determination unit 242 determines whether or not the moving object 60 to be processed appeared in the detection area 50 during the most recent reference period. If the moving object 60 to be processed did not appear in the detection area 50 during the most recent reference period, the appearance determination unit 242 proceeds to step S207. On the other hand, if the moving object 60 to be processed appeared in the detection area 50 during the most recent reference period, the appearance determination unit 242 proceeds to step S208.

[0033] This will be described in detail with reference to FIG. First, the appearance determination unit 242 sets each of the point cloud data acquired during the most recent reference period, excluding the most recently acquired point cloud data, as the target point cloud data. In Fig. 5, point cloud data P1 acquired at time t1, point cloud data P2 acquired at time t2, and point cloud data P3 acquired at time t3 exist as point cloud data acquired during the reference period. Therefore, point cloud data P1 and point cloud data P2 are each set as the target point cloud data.

[0034] Next, the appearance determination unit 242 determines the object detected in step S204 using the set difference identified in step S203 from the target point cloud data as the first object. The appearance determination unit 242 also determines the object detected in step S204 as the second object using the set difference identified in step S203 from the point cloud data next to the target point cloud data in time series. The appearance determination unit 242 determines whether the following conditions are met: the difference between the movement direction of the first object and the movement direction of the second object is within a reference difference, and the distance between the detection area of ​​the first object and the detection area of ​​the second object is within a reference distance. When the appearance determination unit 242 determines that the conditions are met when each point cloud data is the target point cloud data, it determines that a moving object 60 has appeared during the target period. The movement direction of the first object can be considered to be the direction from the center of the detection area of ​​the first object to the center of the detection area of ​​the second object. 5, when point cloud data P1 is used as the target point cloud data, the appearance determination unit 242 determines the object detected from point cloud data P1 as the first object and the object detected from point cloud data P2 as the second object. When point cloud data P2 is used as the target point cloud data, the appearance determination unit 242 determines the object detected from point cloud data P2 as the first object and the object detected from point cloud data P3 as the second object. When it is determined that the conditions are met both when point cloud data P1 is used as the target point cloud data and when point cloud data P2 is used as the target point cloud data, the appearance determination unit 242 determines that the moving object 60 has appeared in the target period.

[0035] A moving object 60 such as a train or car does not change its direction of movement significantly in a short time such as the interval between point cloud acquisitions. It also does not change its position significantly. In contrast, swaying plants and trees may move in the opposite direction over a short period of time. Furthermore, in the case of a falsely detected object, it may be detected in a different position at each time point. Therefore, by comparing the direction of movement and the detected position between adjacent point cloud data in a reference period, it is possible to determine whether a moving object 60 appeared during the reference period.

[0036] (Step S207: 1 data deletion process) The appearance determination unit 242 deletes the setting difference identified from the oldest point cloud data from among the setting differences stored as provisional data in the memory 12. Then, the appearance determination unit 242 returns the process to step S201.

[0037] (Step S208: Data accumulation process) The point cloud accumulation unit 243 accumulates the set difference for the most recent reference period in the memory 12 as main data.

[0038] (Step S209: Delete all data) The point cloud storage unit 243 deletes all the setting differences stored as provisional data in the memory 12. Then, the point cloud storage unit 243 returns the process to step S201.

[0039] (Step S210: Clustering process) The area specifying unit 244 clusters the points included in the set difference accumulated as the main data in step S208.

[0040] (Step S211: Area selection process) 6, the area identification unit 244 sets each of the one or more areas identified by clustering in step S210 as a target area. The area identification unit 244 divides the target area into unit areas, and only retains multi-point areas where the number of points included in the set difference accumulated as main data per unit area is equal to or greater than a reference number, and deletes the remaining areas. The area identification unit 244 identifies the remaining multi-point area as a movement area 51 through which the moving object moves. As shown in FIG. 6, the movement area 51 may not be a continuous area but may be a discrete area. In reality, a moving object does not move in a discrete manner, and the area through which the moving object passes should be a continuous area. However, due to the influence of factors such as the interval at which point cloud data is acquired, the movement area 51 may be a discrete area. However, even if the movement area 51 becomes a discrete area, this does not affect the process of determining whether or not a moving object exists, which will be described later. The unit area is, for example, a cube with a side length of approximately 1 cm to 5 cm.

[0041] The area specifying unit 244 may specify the range between the multi-point area and the ground as the movement area 51. Point cloud data is acquired only for the surface of the moving body 60. Therefore, for example, if point cloud data is acquired using an optical sensor installed at a high position, point cloud data is acquired only for the roof portion of the moving body 60. Then, if the multi-point region is set as the moving area 51, only the region at the height of the roof portion of the moving body 60 becomes the moving area 51. In this case, in the process of determining whether or not a moving body exists, which will be described later, it may not be possible to properly determine a moving body 60 that is low in height. Therefore, by setting the range between the multi-point region and the ground as the moving area 51, it becomes possible to properly determine even a moving body 60 that is low in height. The ground surface can be identified by existing plane estimation processing based on point cloud data.

[0042] (Step S212: Bird's-eye view conversion process) The area identification unit 244 performs a bird's-eye view transformation on the movement area 51 identified in step S211 to identify the movement trajectory of the moving object on the plan view. Here, the data of each point used to identify the movement area 51 is data obtained using an optical sensor and includes three-dimensional position information. Therefore, the area identification unit 244 can perform the bird's-eye view transformation by using the three-dimensional position information. As shown in FIG. 6, if the movement area 51 is a discrete region, the area identification unit 244 may interpolate between the movement areas 51 to identify a continuous movement trajectory. The width of the movement trajectory obtained varies depending on the installation position of the optical sensor. For example, if the optical sensor is installed at a high position, point cloud data of the top surface of the moving object can be acquired, and a movement trajectory with a width roughly the same as the width of the moving object can be obtained. However, if the optical sensor is installed at a low position, point cloud data can only be acquired of the side surface of the moving object, and a movement trajectory that resembles a line thinner than the width of the moving object can be obtained.

[0043] (Step S213: Save processing) The area identification unit 244 stores in the memory 12 the movement area 51 identified in step S211 and the movement trajectory identified in step S212.

[0044] The moving object determination process (step S3 in FIG. 2) according to the first embodiment will be described with reference to FIG. The moving object determination process is executed during a monitoring period to monitor whether or not a moving object 60 is present in the detection area 50. For example, if the moving object 60 is a train, the monitoring period is a period during which the train is in operation, such as daytime. Note that the entire period may be specified as the monitoring period after the movement area identification process is completed.

[0045] (Step S301: Initial setting process) The moving body determination unit 25 reads the moving area 51 identified in step S2 of FIG.

[0046] (Step S302: Point cloud acquisition process) The point cloud acquisition unit 21 acquires point cloud data obtained by irradiating the detection area 50 with light using an optical sensor and receiving the light reflected at reflection points.

[0047] (Step S303: End determination process) The moving body determination unit 25 determines whether or not the acquisition of point cloud data has been completed. Specifically, the moving body determination unit 25 determines whether or not the monitoring period has ended. If the monitoring period has ended, the moving body determination unit 25 determines that the acquisition of point cloud data has ended. If the acquisition of point cloud data has not been completed, the moving body determination unit 25 advances the process to step S304. On the other hand, if the acquisition of point cloud data has been completed, the moving body determination unit 25 ends the process.

[0048] (Step S304: Difference Identification Process) 2, the difference identification unit 23 identifies a monitoring difference, which is the difference between the background point cloud data generated in step S1 and the point cloud data acquired in step S302. Similar to the set difference, the monitoring difference includes points that are not in the background point cloud data but are in the point cloud data acquired in step S302, and points that are in the background point cloud data but not in the point cloud data acquired in step S302. Alternatively, the monitoring difference may include only points that are not in the background point cloud data but are in the point cloud data acquired in step S302.

[0049] (Step S305: Existence determination process) The moving body determination unit 25 determines whether or not a moving body is present in the detection area 50 based on the monitoring difference included in the moving area 51 and identified in step S304. Specifically, the moving body determination unit 25 identifies the number of monitoring difference points included in the moving area 51. If the number of identified points is greater than a threshold, the moving body determination unit 25 determines that a moving body is present in the detection area 50. On the other hand, if the number of identified points is less than the threshold, the moving body determination unit 25 determines that a moving body is not present in the detection area 50.

[0050] (Step S306: Notification process) The notification unit 26 notifies the object monitoring device of the determination result in step S305. The object monitoring device is a device that switches whether or not to monitor an object depending on whether or not there is a moving object. For example, the object monitoring device is a device for preventing collision accidents with trains at railroad crossings. In this object monitoring device, monitoring is turned on when a train is approaching the railroad crossing, and monitoring is turned off when there is no train near the crossing. The object monitoring device can switch whether or not to monitor an object depending on the notification.

[0051] The notification unit 26 may transmit the movement trajectory identified in step S212 of Fig. 4 to the object monitoring device. This allows the user of the object monitoring device to refer to the movement trajectory and set which area of ​​the detection area 50 to monitor.

[0052] ***Effects of the First Embodiment*** As described above, the mobile object monitoring device 10 according to the first embodiment uses point cloud data to determine whether or not the mobile object 60 exists. This makes it possible to appropriately determine whether or not the mobile object 60 exists even when there is a change in lighting or in the dark at night.

[0053] The mobile object monitoring device 10 according to the first embodiment determines whether or not a mobile object 60 exists based on the monitoring difference included in the movement area 51. Processing using point cloud data tends to require a larger amount of calculation compared to processing using image data. However, the mobile object monitoring device 10 does not detect a mobile object based on point cloud data for the entire detection area 50, but determines whether or not a mobile object 60 exists based on the monitoring difference included in the movement area 51. This makes it possible to reduce the amount of calculation.

[0054] The mobile object monitoring device 10 according to the first embodiment uses point cloud data to identify the movement area 51. By identifying the movement area 51 using point cloud data similar to the point cloud data for determining whether or not the mobile object 60 exists, it is possible to appropriately determine whether or not the mobile object 60 exists.

[0055] ***Other Configurations*** <Variation 1> In the first embodiment, each functional component is realized by software. However, as a first modification, each functional component may be realized by hardware. The differences between the first embodiment and the first modification will be described below.

[0056] When each functional component is realized by hardware, the mobile object monitoring device 10 includes an electronic circuit 15 instead of the processor 11, memory 12, and 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.

[0057] 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. GA stands for Gate Array. ASIC stands for Application Specific Integrated Circuit. FPGA stands for Field-Programmable Gate Array. Each functional component may be realized by one electronic circuit 15, or each functional component may be realized by distributing it among a plurality of electronic circuits 15.

[0058] <Variation 2> As a second modification, some of the functional components may be realized by hardware, and other functional components may be realized by software.

[0059] The processor 11, memory 12, storage 13, and 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.

[0060] Furthermore, the term "unit" in the above description may be read as a "circuit," "step," "procedure," "process," or "processing circuit."

[0061] Embodiment 2 The second embodiment differs from the first embodiment in that processing is performed after the point cloud data is converted into voxels. In the second embodiment, this difference will be explained, and explanation of the same points will be omitted.

[0062] ***Configuration Description*** The configuration of a moving object monitoring device 10 according to the second embodiment will be described with reference to FIG. 1 in that the mobile object monitoring device 10 includes a voxelization unit 27 as a functional component. The function of the voxelization unit 27 is realized by software or hardware, like the other functional components. Furthermore, the mobile object monitoring device 10 stores a voxel definition 31 in the storage 13. The voxel definition 31 is data that defines a voxel 32.

[0063] ***Explanation of Operation*** A voxel definition 31 according to the second embodiment will be described with reference to FIG. The voxel definition 31 defines voxels 32, which are three-dimensional regions obtained by dividing the detection area 50. The size of the voxels 32 can be set arbitrarily depending on the position in the detection area 50. In the second embodiment, the larger the distance from the optical sensor, the larger the size of the defined voxels 32.

[0064] The voxelization unit 27 generates voxel data from the point cloud data. 10, the voxelization unit 27 sets each point of the point cloud data as a target point. The voxelization unit 27 associates the target point with a voxel 32 that includes the position of the target point among a plurality of voxels 32 obtained by dividing the detection area 50. In this way, voxel data corresponding to the point cloud data is generated.

[0065] The background point cloud generation process (step S1 in FIG. 2) according to the second embodiment will be described with reference to FIG. The processes from step S101A to step S103A are the same as the processes from step S101 to step S103 in Fig. 3. However, in step S102A, if the background point cloud generation unit 22 has finished acquiring point cloud data, the process proceeds to step S104A.

[0066] (Step S104A: Voxelization process) The voxelization unit 27 voxels the background point cloud data generated by adding the point cloud data in step S103A to generate background voxel data. That is, the voxelization unit 27 sets each point of the background point cloud data to a target point. The voxelization unit 27 associates the target point with a voxel 32 that includes the position of the target point among the multiple voxels 32 obtained by dividing the detection area 50. In this way, background voxel data corresponding to the background point cloud data is generated.

[0067] The travel area identification process (step S2 in FIG. 2) according to the second embodiment will be described with reference to FIG. The processes in steps S201A and S202A are the same as those in steps S201 and S202 in Fig. 4. The processes in steps S206A to S210A are the same as those in steps S205 to S209 in Fig. 4. The processes in steps S212A to S214A are the same as those in steps S211 to S213 in Fig. 4.

[0068] (Step S203A: Voxelization process) The voxelization unit 27 voxels the point cloud data acquired in step S201A to generate voxel data. That is, the voxelization unit 27 sets each point of the point cloud data to a target point. The voxelization unit 27 associates the target point with a voxel 32 that includes the position of the target point among the multiple voxels 32 obtained by dividing the detection area 50. In this way, voxel data corresponding to the point cloud data is generated.

[0069] (Step S204A: Difference Identification Process) The difference specifying unit 23 specifies a set difference indicating a difference cell where there is a difference between the background voxel data generated in step S104A of FIG. 11 and the voxel data generated in step S203A. 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 corresponding point for the target voxel 32, that is, whether or not there is a difference between the background voxel data and the voxel data, 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. 13, the difference identification unit 23 may identify a target voxel 32 as a difference cell when no points are associated with the target voxel 32 in the background voxel data, but points are associated with the target voxel 32 in the target voxel data. In other words, the difference identification unit 23 may not identify a voxel 32 that has points associated with it in the background voxel data as a difference cell even if the number of associated points is different.

[0070] (Step S205A: Object detection process) The object detection unit 241 detects an object from the set difference identified in step S204A. Specifically, the object detection unit 241 clusters the difference cells indicated by the set difference. The object detection unit 241 detects each set of cells obtained by the clustering as an object present in the detection area 50.

[0071] (Step S211A: Clustering process) The area specifying unit 244 clusters the difference cells indicated by the set difference accumulated as the main data in step S210A.

[0072] The moving object determination process (step S3 in FIG. 2) according to the second embodiment will be described with reference to FIG. The processes from step S301A to step S303A are the same as the processes from step S301 to step S303 in Fig. 7. The process from step S307A is the same as the process from step S306 in Fig. 7.

[0073] (Step S304A: Voxelization process) The voxelization unit 27 generates voxel data by voxelizing the point cloud data acquired in step S302A. The method of generating voxel data is the same as step S203A in FIG.

[0074] (Step S305A: Difference Identification Process) The difference identification unit 23 identifies a monitoring difference that indicates a difference cell that has a difference between the background voxel data generated in step S104A and the voxel data generated in step S304A in Fig. 11. The method for identifying the monitoring difference is the same as the method for identifying the set difference in step S204A in Fig. 12.

[0075] (Step S306A: Existence determination process) The moving body determination unit 25 determines whether or not a moving body is present in the detection area 50 based on the monitoring difference included in the moving area 51 and identified in step S305A. Specifically, the moving object determination unit 25 identifies the number of differential cells indicated by the monitoring differential included in the moving area 51. If the number of identified differential cells is greater than a threshold, the moving object determination unit 25 determines that a moving object is present in the detection area 50. On the other hand, if the number of identified differential cells is less than the threshold, the moving object determination unit 25 determines that a moving object is not present in the detection area 50.

[0076] ***Effects of the Second Embodiment*** As described above, the moving object monitoring device 10 according to the second embodiment performs processing after converting the point cloud data into voxels. It is difficult for an optical sensor to always project light onto the same spot, resulting in slight deviations. As a result, the point cloud data generated varies slightly each time. Therefore, simply comparing point cloud data with background point cloud data can result in differences in various locations, making it difficult to properly detect the target object. The moving object monitoring device 10 according to the second embodiment performs a 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 properly detect the target object.

[0077] The farther away from the optical sensor is the larger the position of the reflection point will be even if the angle at which the optical sensor emits light is slightly off. Therefore, the farther away from the optical sensor is the more likely it is that the emitted light will be off. In the second embodiment, the size of the voxels 32 increases as the distance from the optical sensor increases. Therefore, even in areas far from the optical sensor, deviation in the irradiation position of the irradiated light is less likely to appear as a difference, making it possible to properly detect the object.

[0078] When acquiring background point cloud data, an object may cross in front of the optical sensor. In this case, in the moving area identification process (step S2 in FIG. 2) or the moving object determination process (step S3 in FIG. 2), no point cloud exists in the voxel data, and many voxels 32 that exist in the point cloud appear in the background voxel data generated from the background point cloud data. In the moving object monitoring device 10 according to the second embodiment, the voxel 32 associated with a point in the background voxel data may not be identified as the difference cell 44. In this way, it is possible to prevent an object passing in front of the optical sensor from being detected.

[0079] ***Other Configurations*** <Variation 3> In steps S204A and S305A, the difference identification unit 23 may exclude, from the difference cells, voxels 32 in which the number of associated points in the voxel data is equal to or less than the reference number of points. For example, the number of reference points is set to 2. In this case, as shown in Fig. 15, voxels 32 in which the number of associated points in the voxel data is 2 or less are excluded from the difference cell. In Fig. 15, the middle voxel 32 has 2 associated points and is therefore excluded from the difference cell. On the other hand, the bottom voxel 32 has 3 associated points and is therefore left as a difference cell.

[0080] Various aspects of the present disclosure are summarized below as appendices. (Appendix 1) a point cloud acquisition unit that acquires point cloud data obtained by irradiating a detection area with light and receiving reflected light reflected at a reflection point; a difference specifying unit that specifies a monitoring difference that is a difference between background point cloud data, which is point cloud data acquired by the point cloud acquiring unit when there is no moving object in the detection area, and point cloud data acquired by the point cloud acquiring unit during a monitoring period; a moving object determination unit that determines whether or not the moving object is present in the detection area based on the monitoring difference identified by the difference identification unit, the monitoring difference being included in a movement area in which the moving object moves within the detection area; A mobile object monitoring device comprising: (Appendix 2) the difference specifying unit specifies a set difference which is a difference between the background point cloud data and each of a plurality of point cloud data acquired by the point cloud acquiring unit during a set period; The mobile object monitoring device further includes: a movement area specifying unit that specifies a movement area of ​​a moving object from the set difference for each of the plurality of point cloud data; Equipped with The moving object determination unit determines whether or not the moving object exists by using the moving area specified by the moving area specification unit. 2. The mobile object monitoring device of claim 1. (Appendix 3) The movement area identification unit identifies, as the movement area, a multi-point area in which the number of points included in the set difference for each of the plurality of point cloud data is equal to or greater than a reference number per unit area. 3. A mobile object monitoring device as described in appendix 2. (Appendix 4) The movement area specifying unit specifies the range between the multi-point area and the ground as the movement area. 4. A mobile object monitoring device as described in appendix 3. (Appendix 5) The movement area identification unit an appearance determination unit that determines that the moving object has appeared in the target period when a difference in movement direction between a first object detected using the set difference identified from the target point cloud data and a second object detected using the set difference identified from the point cloud data next to the target point cloud data in time series is within a reference difference, with each reference period extracted from the set period being a target period, and each of the point cloud data excluding the last acquired point cloud data among the plurality of point cloud data acquired in the target period being target point cloud data; a point cloud storage unit that stores a point cloud included in the set difference identified from point cloud data acquired during a period in which the appearance determination unit has determined that the object is the moving body; and an area identification unit that identifies the movement area by clustering the point cloud accumulated by the point cloud accumulation unit; 5. A mobile object monitoring device according to any one of appendices 2 to 4, comprising: (Appendix 6) The appearance determination unit determines that the moving object has appeared during the target period when a distance between the detection area of ​​the first object and the detection area of ​​the second object is within a reference distance. 6. A mobile object monitoring device as described in appendix 5. (Appendix 7) The mobile object monitoring device further includes: a voxelization unit that generates voxel data corresponding to the point cloud data by treating each point of the point cloud data as a target point and associating the target point with a voxel that includes the position of the target point among a plurality of voxels obtained by dividing the detection area; Equipped with The difference identification unit identifies, for each of the plurality of voxels, a difference cell having a difference between the voxel data generated from the background point cloud data by the voxelization unit and the voxel data generated from the point cloud data acquired during the monitoring period by the voxelization unit, as the monitoring difference, based on the presence or absence of a corresponding point. 7. A mobile object monitoring device according to any one of appendices 1 to 6. (Appendix 8) The mobile object monitoring device further includes: a notification unit that notifies the result of the determination by the moving object determination unit to an object monitoring device that switches whether or not to monitor an object depending on the presence or absence of the moving object; 8. A mobile object monitoring device according to any one of appendices 1 to 7, comprising: (Appendix 9) The computer acquires point cloud data obtained by irradiating the detection area with light and receiving the light reflected at the reflection points, The computer identifies a monitoring difference that is a difference between background point cloud data, which is point cloud data acquired when there is no moving object in the detection area, and point cloud data acquired during a monitoring period; A moving object monitoring method in which a computer determines whether or not the moving object is present in the detection area based on the monitoring difference included in a movement area in which the moving object moves within the detection area. (Appendix 10) a point cloud acquisition process for acquiring point cloud data obtained by irradiating the detection area with light and receiving reflected light reflected at reflection points; a difference identification process for identifying a monitoring difference, which is a difference between background point cloud data, which is point cloud data acquired by the point cloud acquisition process when there is no moving object in the detection area, and point cloud data acquired during a monitoring period by the point cloud acquisition process; a moving object determination process for determining whether or not the moving object is present in the detection area based on the monitoring difference identified by the difference identification process, the monitoring difference being included in a moving area in the detection area where the moving object moves; A mobile object monitoring program that causes a computer to function as a mobile object monitoring device that performs the above.

[0081] The embodiments and modifications of the present disclosure have been described above. Some of these embodiments and modifications may be combined and implemented. Also, one or more of them may be implemented partially. Note that the present disclosure is not limited to the above embodiments and modifications, and various modifications are possible as needed. [Explanation of symbols]

[0082] 10 Mobile object monitoring device, 11 Processor, 12 Memory, 13 Storage, 14 Communication interface, 21 Point cloud acquisition unit, 22 Background point cloud generation unit, 23 Difference identification unit, 24 Moving area identification unit, 25 Moving object determination unit, 26 Notification unit, 27 Voxelization unit, 241 Object detection unit, 242 Appearance determination unit, 243 Point cloud accumulation unit, 244 Area identification unit, 31 Voxel definition, 32 Voxel, 50 Detection area, 51 Moving area, 60 Moving object.

Claims

1. a point cloud acquisition unit that acquires point cloud data obtained by irradiating a detection area with light and receiving reflected light reflected at a reflection point; a difference specifying unit that specifies a monitoring difference that is a difference between background point cloud data, which is point cloud data acquired by the point cloud acquiring unit when there is no moving object in the detection area, and point cloud data acquired by the point cloud acquiring unit during a monitoring period; a moving object determination unit that determines whether or not the moving object is present in the detection area based on the monitoring difference identified by the difference identification unit, the monitoring difference being included in a movement area in the detection area where the moving object moves; A mobile object monitoring device comprising:

2. the difference specifying unit specifies a set difference which is a difference between the background point cloud data and each of a plurality of point cloud data acquired by the point cloud acquiring unit during a set period; The mobile object monitoring device further includes: a movement area specifying unit that specifies a movement area of ​​a moving object from the set difference for each of the plurality of point cloud data; Equipped with The moving object determination unit determines whether or not the moving object exists by using the moving area specified by the moving area specification unit. The mobile object monitoring device according to claim 1.

3. The movement area identification unit identifies, as the movement area, a multi-point area in which the number of points included in the set difference for each of the plurality of point cloud data is equal to or greater than a reference number per unit area. The mobile object monitoring device according to claim 2.

4. The movement area specifying unit specifies the range between the multi-point area and the ground as the movement area. The mobile object monitoring device according to claim 3.

5. The movement area identification unit an appearance determination unit that determines that the moving object has appeared in the target period when a difference in movement direction between a first object detected using the set difference identified from the target point cloud data and a second object detected using the set difference identified from the point cloud data next to the target point cloud data in time series is within a reference difference, the first object being detected using the set difference identified from the target point cloud data, and the second object being detected using the set difference identified from the point cloud data next to the target point cloud data in time series, is within a reference difference; a point cloud storage unit that stores a point cloud included in the set difference identified from point cloud data acquired during a period in which the appearance determination unit has determined that the object is the moving body; and an area identification unit that identifies the movement area by clustering the point cloud accumulated by the point cloud accumulation unit; The mobile object monitoring device according to claim 2 .

6. The appearance determination unit determines that the moving object has appeared during the target period when a distance between the detection area of ​​the first object and the detection area of ​​the second object is within a reference distance. The mobile object monitoring device according to claim 5.

7. The mobile object monitoring device further includes: a voxelization unit that generates voxel data corresponding to the point cloud data by treating each point of the point cloud data as a target point and associating the target point with a voxel that includes the position of the target point among a plurality of voxels obtained by dividing the detection area; Equipped with The difference identification unit identifies, for each of the plurality of voxels, a difference cell having a difference between the voxel data generated from the background point cloud data by the voxelization unit and the voxel data generated from the point cloud data acquired during the monitoring period by the voxelization unit, as the monitoring difference, based on the presence or absence of a corresponding point. The mobile object monitoring device according to claim 1.

8. The mobile object monitoring device further includes: a notification unit that notifies the result of the determination by the moving object determination unit to an object monitoring device that switches whether or not to monitor an object depending on the presence or absence of the moving object; The mobile object monitoring device according to claim 1 .

9. The computer acquires point cloud data obtained by irradiating the detection area with light and receiving the light reflected at the reflection points, The computer identifies a monitoring difference that is a difference between background point cloud data, which is point cloud data acquired when there is no moving object in the detection area, and point cloud data acquired during a monitoring period; A moving object monitoring method in which a computer determines whether or not the moving object is present in the detection area based on the monitoring difference included in a movement area in which the moving object moves within the detection area.

10. a point cloud acquisition process for acquiring point cloud data obtained by irradiating the detection area with light and receiving reflected light reflected at reflection points; a difference identification process for identifying a monitoring difference, which is a difference between background point cloud data, which is point cloud data acquired by the point cloud acquisition process when there is no moving object in the detection area, and point cloud data acquired during a monitoring period by the point cloud acquisition process; a moving object determination process for determining whether or not the moving object is present in the detection area based on the monitoring difference identified by the difference identification process, the monitoring difference being included in a moving area in the detection area where the moving object moves; A mobile object monitoring program that causes a computer to function as a mobile object monitoring device that performs the above.

Images