Information processing method, mobile body, and program

Abstract

The present technology relates to an information processing method, a mobile body, and a program that make it possible to suitably track the position of an object. The information processing method according to the present technology comprises: identifying an object to be tracked from among objects recognized on the basis of sensor data of a sensor that detects an object in a surrounding three-dimensional space; tracking the position of the object to be tracked in the three-dimensional space; and controlling, on the basis of congestion around the sensor, a threshold used for identifying the object. The present technology can be applied to, for example, a robot that moves following a person.

Description

Information processing method, mobile device, and program

[0001] The present technology relates to an information processing method, a mobile object, and a program, and more particularly to an information processing method, a mobile object, and a program that enable the position of an object to be suitably tracked.

[0002] A system is known that recognizes objects such as people in images captured by a camera and provides users with information about the congestion situation in a three-dimensional space such as a commercial facility (see, for example, Patent Document 1). Technology for detecting people in images captured by a camera is also used to track the positions of people moving in three-dimensional space. In person tracking, a person to be tracked is identified from among people recognized in the image.

[0003] Japanese Patent Application Laid-Open No. 2020-173775

[0004] When the three-dimensional space is crowded, mismatches are likely to occur, such as when a person different from the person being tracked is identified as the person being tracked among the people recognized in the image. If the threshold value used for person identification is reduced to prevent mismatches, the person being tracked will likely be lost if the three-dimensional space is empty.

[0005] The present technology has been made in view of such circumstances, and makes it possible to suitably track the position of an object.

[0006] An information processing method according to a first aspect of the present technology includes identifying a tracked object from among objects recognized based on sensor data from a sensor that detects objects in a surrounding three-dimensional space, tracking the position of the tracked object in the three-dimensional space, and controlling a threshold value used to identify the object based on the degree of congestion around the sensor.

[0007] A moving body according to a second aspect of the present technology includes a sensor that detects objects in a surrounding three-dimensional space, a tracking unit that identifies a tracked object from among the objects recognized based on sensor data from the sensor and tracks the position of the tracked object in the three-dimensional space, a tracking control unit that controls a threshold used to identify the object based on the degree of congestion around the sensor, and a movement control unit that controls movement of the moving body based on the position of the tracked object.

[0008] A program according to a third aspect of the present technology causes a computer to execute a process of identifying a tracked object from among objects recognized based on sensor data from a sensor that detects objects in a surrounding three-dimensional space, tracking the position of the tracked object in the three-dimensional space, and controlling a threshold value used to identify the object based on the degree of congestion around the sensor.

[0009] In the first and third aspects of the present technology, a tracked object is identified from among objects recognized based on sensor data from a sensor that detects objects in a surrounding three-dimensional space, the position of the tracked object in the three-dimensional space is tracked, and a threshold value used to identify the object is controlled based on the degree of congestion around the sensor.

[0010] In a second aspect of the present technology, a sensor detects objects in the surrounding three-dimensional space, a tracking unit identifies the object to be tracked from among the objects recognized based on sensor data from the sensor, and tracks the position of the object to be tracked in the three-dimensional space, and a tracking control unit controls a threshold used to identify the object based on the degree of congestion around the sensor, and controls the movement of the aircraft based on the position of the object to be tracked.

[0011] 7 is a diagram illustrating an example of a configuration of the appearance of a robot according to an embodiment of the present technology. FIG. 8 is a block diagram illustrating an example of a configuration of a robot. FIG. 9 is a diagram illustrating an example of an object map. FIG. 10 is a diagram illustrating an example of a target range for congestion degree calculation. FIG. 11 is a diagram illustrating a method for identifying a person. FIG. 12 is a flowchart illustrating a process in which a calculation unit calculates a target range for congestion degree calculation. FIG. 13 is a flowchart illustrating a process in which a calculation unit calculates a congestion degree and tracks the position of a person to be tracked. FIG. 14 is a flowchart illustrating a person tracking process performed in step S25 of FIG. 7. FIG. 15 is a block diagram illustrating an example of a configuration of computer hardware.

[0012] Hereinafter, an embodiment of the present technology will be described. The description will be made in the following order: 1. Configuration of the robot 2. Operation of the calculation unit

[0013] 1. Configuration of Robot> FIG. 1 is a diagram illustrating an example of the external configuration of a robot 1 according to an embodiment of the present technology.

[0014] As shown in Fig. 1, the robot 1 is a mobile object with a humanoid upper body and a wheeled mobility mechanism. A flattened spherical head 12 is provided on top of a torso 11. Two visual sensors 12A, which mimic human eyes, are provided on the front of the head 12. The visual sensors 12A are sensors that detect objects in the surrounding three-dimensional space and are composed of a camera and LiDAR (Light Detection and Ranging, Laser Imaging Detection and Ranging).

[0015] An arm section formed by a manipulator with multiple degrees of freedom is provided at the upper end of the body section 11, and a hand section is provided at each tip of the arm section.

[0016] A carriage-like mobile unit 13 is provided at the lower end of the body 11. The robot 1 can move by rotating wheels provided on the left and right sides of the mobile unit 13 or by changing the direction of the wheels.

[0017] For example, a calculation unit that controls the operation of the robot 1 is provided inside the body 11. The calculation unit tracks the position of a tracking target object (e.g., a person) based on sensor data from the visual sensor 12A, and controls the moving body unit 13 so that the robot 1 moves while following the object.

[0018] The robot 1 may be configured as a single-arm robot (with one hand) or armless robot, instead of the dual-arm robot shown in Fig. 1. Also, instead of the cart (moving body 13), the body 11 may be provided on the legs.

[0019] FIG. 2 is a block diagram showing an example of the configuration of the robot 1.

[0020] As shown in FIG. 2, the robot 1 includes a visual sensor 12A, a memory unit 21, a calculation unit 22, and an actuator 23.

[0021] The storage unit 21 stores an object map that three-dimensionally shows the shapes, positions, and moving speeds of people and other objects that exist around the robot 1 .

[0022] FIG. 3 is a diagram showing an example of an object map.

[0023] 3, the object map is, for example, a map that shows a bird's-eye view of the surroundings of the robot 1. In the example of FIG. 3, people P1 to P3 and objects Obj1 to Obj4 are present in front of the robot 1.

[0024] The object map is generated based on, for example, sensor data from the visual sensor 12A. Specifically, the shapes, positions, and movement speeds of people and objects captured in images captured by a camera serving as the visual sensor 12A are registered in the object map. Also, the shapes, positions, and movement speeds of people and objects detected by a LiDAR serving as the visual sensor 12A are registered in the object map.

[0025] The object map also registers, for example, the camera's imaging range (sensing range) and the LiDAR's scanning range (sensing range). In Figure 3, the camera's imaging range is indicated by a solid line, and the LiDAR's scanning range is indicated by a dotted line. The camera's imaging range includes person P1, person P2, and object Obj3, and the LiDAR's scanning range includes people P1 to P3 and objects Obj1 to Obj4.

[0026] Returning to Figure 2, the calculation unit 22 identifies a person to be tracked from among the people recognized based on the sensor data of the visual sensor 12A, tracks the position of the person to be tracked in three-dimensional space, and controls the threshold value used for person identification based on the degree of congestion around the robot 1 (visual sensor 12A).

[0027] The calculation unit 22 has a recognition unit 41 , a tracking unit 42 , a target range calculation unit 43 , a congestion level calculation unit 44 , a threshold calculation unit 45 , a UI (User Interface) control unit 46 , a behavior planning unit 47 , and a movement control unit 48 .

[0028] The recognition unit 41 recognizes the position of a person present around the robot 1 based on the sensor data of the visual sensor 12A, and supplies the person recognition result to the tracking unit 42.

[0029] The tracking unit 42 tracks the position of the person to be tracked based on the person recognition result by the recognition unit 41, and updates the position and movement speed of the person to be tracked in the object map. Specifically, the tracking unit 42 identifies the person to be tracked from among the people recognized by the recognition unit 41 (people included in the camera's shooting range or the LiDAR's scanning range), and registers the current position and movement speed of the person to be tracked in the object map.

[0030] More specifically, the tracking unit 42 estimates the current position of the person to be tracked using a particle filter or a Kalman filter. The tracking unit 42 searches for the person to be tracked from among the people recognized by the recognition unit 41 who are within a predetermined distance from the estimated position of the person to be tracked. Hereinafter, the range in which the person to be tracked is searched for is also referred to as a search range, and the size of the search range is also referred to as a person identification threshold. The person identification threshold indicates, for example, the radius of the search range.

[0031] The target range calculation unit 43 acquires information indicating the position and movement speed of the person to be tracked and the camera's shooting range from the storage unit 21, and calculates the target range for the congestion degree calculation based on this information.

[0032] Fig. 4 is a diagram showing an example of the target range for calculating the congestion degree. In Fig. 4, the explanation will be given assuming that eight people P11 to P18 are present around the robot 1. Furthermore, person P15, indicated by a black dot in Fig. 4, is the person to be tracked.

[0033] In the example shown in A in Fig. 4, the target range calculation unit 43 sets a rectangular area A1 as the target range for calculating the congestion level when the area around the robot 1 is viewed from above. Area A1 is the rectangular area with the smallest area among the rectangular areas that surround all of the people P14 to P16 that are included in the camera's shooting range. Note that the target range for calculating the congestion level may also be the circular area with the smallest area among the circular areas that surround all of the people that are included in the camera's shooting range.

[0034] 4B, the target range calculation unit 43 sets a circular area A2, centered on the tracking target person P15, as the target range for calculating the congestion level when the area around the robot 1 is viewed from above. The radius a of the area A2 is set in advance, for example. The area A2 includes four people P14 to P17.

[0035] In the example shown in FIG. 4C, the target range calculation unit 43 sets an elliptical area A3 centered on the person P15 to be tracked when viewing the area around the robot 1 from above as the target range for calculating the congestion level. Area A3 is an elliptical area formed taking into account the movement speed of the person P15 to be tracked. In the example shown in FIG. 3, person P15 is moving in the direction indicated by the outline arrow. For example, the radius a of area A3 perpendicular to the movement direction of the person P15 to be tracked is set in advance, and the radius b of area A3 parallel to the movement direction of the person P15 to be tracked is set to (a + movement speed × 1 second). Area A3 includes four people P14 to P17.

[0036] The target range calculation unit 43 supplies information indicating the target range for the congestion degree calculation as described above to the congestion degree calculation unit 44 and the UI control unit 46 in Fig. 3. Note that the target range for the congestion degree calculation is not limited to the area having the shape described above, and may be an area having another shape, such as a polygonal shape or a shape with a part of a circle or ellipse cut out when the area around the robot 1 is viewed from above.

[0037] The congestion degree calculation unit 44 calculates the degree of congestion around the robot 1 based on, for example, the number of people included in the target range calculated by the target range calculation unit 43. Specifically, the congestion degree calculation unit 44 counts the number of people included in the target range by referring to the object map stored in the storage unit 21, and determines the value obtained by dividing the number of people included in the target range by the area of ​​the target range as the degree of congestion around the robot 1. The congestion degree calculation unit 44 supplies the degree of congestion around the robot 1 to the threshold calculation unit 45.

[0038] A range of the congestion degree (maximum and minimum values) is set in advance in the robot 1. If the congestion degree calculated by the congestion degree calculation unit 44 exceeds the maximum value, the congestion degree value is treated as the maximum value, and if the calculated congestion degree is less than the minimum value, the congestion degree value is treated as the minimum value.

[0039] The threshold calculation unit 45 calculates the person identification threshold based on the congestion degree calculated by the congestion degree calculation unit 44. In the robot 1, a range of the person identification threshold (maximum and minimum values) is set in advance. The threshold calculation unit 45 calculates the person identification threshold by linearly mapping the congestion degree. If the congestion degree and the person identification threshold are in a proportional relationship, for example, the threshold calculation unit 45 calculates the person identification threshold using the following formula (1):

[0040]

[0041] For example, if the congestion level ranges from 0 to 4 and the person identification threshold is from 0.7 m to 0.4 m, then the threshold value = 0.7 - (congestion level - 0) x 0.3 / 4. Here, when the congestion level is at its maximum, the person identification threshold is at its minimum, and when the congestion level is at its minimum, the person identification threshold is at its maximum. In other words, the higher the congestion level, the smaller the person identification threshold.

[0042] The threshold calculation unit 45 supplies the calculated threshold to the tracking unit 42. The threshold calculation unit 45 functions as a tracking control unit that controls the threshold for person identification by the tracking unit 42 based on the degree of congestion around the robot 1. The tracking unit 42 identifies the person to be tracked using the threshold calculated by the threshold calculation unit 45.

[0043] FIG. 5 is a diagram illustrating a method for identifying a person.

[0044] First, the tracking unit 42 estimates the position of person P15 at time t+1, indicated by the dashed dot in Fig. 5, based on the position of person P15 at time t, shown in the lower part of Fig. 5. Next, the tracking unit 42 acquires the positions of the people recognized by the recognition unit 41 based on the sensor data acquired at time t+1. In the example of Fig. 5, the positions of people P14' to P16' are recognized.

[0045] Next, the tracking unit 42 searches for person P15 from among the people P14' to P16' recognized by the recognition unit 41 who are included in a search range A11 that is centered around the estimated position of person P15 at time t+1 and is within the threshold (distance) calculated by the threshold calculation unit 45. In the example of Fig. 5, since person P15' is the only person included in search range A11, the tracking unit 42 identifies person P15' as person P15 to be tracked.

[0046] It should be noted that instead of one person being the tracking target, multiple people may be the tracking targets and the positions of each person may be tracked simultaneously.

[0047] Returning to FIG. 2, the UI control unit 46 displays the UI on a display unit provided in the controller of the robot 1, for example, and receives user operations via the UI.

[0048] For example, the UI control unit 46 receives, via the UI, the designation of a person to be followed by the robot 1. The UI control unit 46 notifies the tracking unit 42 of the person to be followed designated by the user as the person to be tracked.

[0049] Furthermore, for example, the UI control unit 46 presents the target range of the congestion degree calculation to the user. The user can specify the shape of the target range (e.g., rectangle, circle, or ellipse) while viewing the presented target range of the congestion degree calculation.

[0050] The behavior planning unit 47 acquires the position of the person to be followed (tracked) by referring to the object map stored in the storage unit 21, and creates a behavior plan for the robot 1 based on the position of the person to be followed. For example, the behavior planning unit 47 acquires a route for following the person to be followed, and creates a behavior plan for the robot 1 to move along the route. The behavior planning unit 47 supplies the created behavior plan to the movement control unit 48.

[0051] The movement control unit 48 controls the movement of the robot 1 (its own machine) to realize the behavior plan created by the behavior planning unit 47. For example, the movement control unit 48 controls the actuator 23 provided in the moving body unit 13 so that the robot 1 moves along the path acquired by the behavior planning unit 47.

[0052] Note that part or all of the configuration of the calculation unit 22 may be realized by a controller provided separately from the robot 1. For example, the recognition unit 41, tracking unit 42, target range calculation unit 43, congestion degree calculation unit 44, threshold calculation unit 45, and UI control unit 46 are realized by the controller. Also, the storage unit 21 may be provided in a device other than the robot 1, such as a cloud.

[0053] 2. Operation of the Calculation Unit Next, with reference to the flowchart in FIG. 6, a process of the calculation unit 22 having the above configuration calculating the target range for congestion degree calculation will be described.

[0054] In step S1, the UI control unit 46 displays candidates for the person to be tracked on, for example, the display unit of the controller, and presents the candidates to the user. Here, the UI control unit 46 presents, for example, people recognized by the recognition unit 41 to the user as candidates for the person to be tracked, and accepts the user's selection of the person to be tracked. The user can select the person to be tracked from the presented candidates.

[0055] In step S2, the UI control unit 46 determines whether or not the user has selected a person to be followed up.

[0056] If it is determined in step S2 that the user has not selected a person to follow, the process returns to step S1, and candidates for people to follow continue to be displayed until the user selects a person to follow.

[0057] On the other hand, if it is determined in step S2 that the user has selected a person to be followed up, then in step S3, the target range calculation unit 43 calculates a target range for the congestion degree calculation.

[0058] In step S4, the UI control unit 46 displays the target range for the congestion degree calculation on the display unit of the controller to present the target range to the user. After the target range for the congestion degree calculation is displayed, the process returns to step S3, and the target range for the congestion degree calculation is calculated again, for example, if a person present around the robot 1 or a person to be followed moves.

[0059] Next, the process in which the calculation unit 22 calculates the congestion degree and tracks the position of the person to be tracked will be described with reference to the flowchart of FIG.

[0060] In step S21, the recognition unit 41 recognizes the position of a person present around the robot 1 based on the sensor data of the visual sensor 12A. For example, the position of the person in an image captured by a camera serving as the visual sensor 12A is recognized, and the position of the person in the image is projected onto a position in three-dimensional space. Each frame of the image captured by the camera is assigned a frame number, for example, starting from 0.

[0061] In step S22, the tracking unit 42 determines whether the frame number of the frame in which the person is recognized (current frame) is 1 or greater.

[0062] If it is determined in step S22 that the frame number of the current frame is 0, the position in three-dimensional space of the person appearing in the current frame is registered in the object map, and the process returns to step S21.

[0063] On the other hand, if it is determined in step S22 that the frame number of the current frame is 1 or greater, then in step S23 the congestion degree calculation unit 44 calculates the congestion degree around the robot 1. Specifically, the congestion degree calculation unit 44 refers to an object map in which the positions in three-dimensional space of people shown in past frames are registered, counts the number of people included in the target area, and calculates the congestion degree by dividing the number of people by the area of ​​the target area.

[0064] In addition, when the target range for the congestion degree calculation is calculated based on the camera's shooting range, the congestion degree calculation unit 44 can also count the number of people included in the target range by referring to the positions of people recognized in the current frame.

[0065] In step S24, the threshold calculation unit 45 calculates a threshold for identifying a person based on the degree of congestion around the robot 1.

[0066] In step S25, the tracking unit 42 performs a person tracking process. The person tracking process uses the threshold calculated in step S24 to identify a person to be tracked from among the people recognized in the current frame in step S21. Details of the person tracking process will be described later with reference to FIG. 8.

[0067] After the person tracking process is performed in step S25, the process returns to step S21, and the subsequent processes are repeated.

[0068] Next, the person tracking process performed in step S25 of FIG. 7 will be described with reference to the flowchart of FIG.

[0069] In step S41, the tracking unit 42 acquires the person recognition result obtained by the recognition unit 41.

[0070] In step S42, the tracking unit 42 estimates the movement speed and current position of the tracked person. Specifically, the tracking unit 42 creates a motion model of the tracked person based on the position of the tracked person at the time of the past frame, and uses the motion model to estimate the position of the tracked person at the time of the current frame (current position), etc. For example, to estimate the position of the tracked person at time t+1 of the current frame, a motion model created based on the position of the tracked person at time t-1 and the position of the tracked person at time t is used.

[0071] In step S43, the tracking unit 42 identifies the person to be tracked at the time of the current frame. Specifically, the tracking unit 42 searches for the person to be tracked from among the people recognized in the current frame who exist within a range within a threshold centered on the estimated position of the person to be tracked.

[0072] In step S44, the tracking unit 42 registers the current position and moving speed of the person to be tracked in the object map. Here, for example, the positions of people other than the person to be tracked are also registered in the object map.

[0073] After the current position and moving speed of the person to be tracked are registered in the object map, the process returns to step S25 in FIG. 7, and the subsequent processes are carried out.

[0074] As described above, in the computing unit 22 of the present technology, a tracking target object is identified from among objects such as a person recognized based on sensor data from the visual sensor 12A that detects objects in the surrounding three-dimensional space, and the position of the tracking target object in the three-dimensional space is tracked. Furthermore, in the computing unit 22 of the present technology, a threshold value used for identifying the object is controlled based on the degree of congestion around the visual sensor 12A (robot 1).

[0075] For example, when the area around the robot 1 is crowded, the calculation unit 22 can prevent a mismatch from occurring, such as a person different from the person to be tracked being identified as the person to be tracked from among the people recognized in the image, by reducing the threshold for person identification. Also, when the area around the robot 1 is empty, for example, the calculation unit 22 can prevent the tracking target person from being lost by increasing the threshold for person identification.

[0076] In this way, the calculation unit 22 controls the threshold for person identification according to the degree of congestion, thereby preventing mismatches and preventing the person being tracked from losing sight of the person, and making it possible to optimally track the position of the person being tracked.

[0077] <Example of Computer Configuration> The above-described series of processes can be executed by hardware or software. When the series of processes is executed by software, the program constituting the software is installed from a program recording medium into a computer incorporated in dedicated hardware, or into a general-purpose personal computer, etc.

[0078] FIG. 9 is a block diagram showing an example of the hardware configuration of a computer that executes the above-described series of processes by a program.

[0079] A CPU (Central Processing Unit) 501 , a ROM (Read Only Memory) 502 , and a RAM (Random Access Memory) 503 are interconnected by a bus 504 .

[0080] An input / output interface 505 is also connected to the bus 504. An input unit 506 including a keyboard, a mouse, etc., and an output unit 507 including a display, a speaker, etc. are connected to the input / output interface 505. Also connected to the input / output interface 505 are a storage unit 508 including a hard disk, a nonvolatile memory, etc., a communication unit 509 including a network interface, etc., and a drive 510 that drives removable media 511.

[0081] In a computer configured as described above, the CPU 501 performs the above-described series of processes by, for example, loading a program stored in the storage unit 508 into the RAM 503 via the input / output interface 505 and the bus 504 and executing it.

[0082] The program executed by the CPU 501 is installed in the storage unit 508 by being recorded on, for example, a removable medium 511 or provided via a wired or wireless transmission medium such as a local area network, the Internet, or digital broadcasting.

[0083] The program executed by the computer may be a program that processes in chronological order according to the order described in this specification, or may be a program that processes in parallel or at the required timing, such as when called.

[0084] The effects described in this specification are merely examples and are not limiting, and other effects may also be present.

[0085] The embodiments of the present technology are not limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the present technology.

[0086] For example, the present technology can be configured as a cloud computing system in which a single function is shared and processed collaboratively by a plurality of devices via a network.

[0087] Furthermore, each step described in the above flowchart can be executed by one device, or can be shared and executed by a plurality of devices.

[0088] Furthermore, when one step includes multiple processes, the multiple processes included in that one step can be executed by one device or can be shared and executed by multiple devices.

[0089] <Examples of Combinations of Configurations> The present technology can also have the following configurations.

[0090] (1) An information processing method comprising: identifying a tracking target object from among objects recognized based on sensor data of a sensor that detects objects in a three-dimensional space, and tracking the position of the tracking target object in the three-dimensional space; and controlling a threshold used for identifying the object based on a congestion level around the sensor. (2) The information processing method described in (1), including identifying the tracking target object from among objects recognized based on the sensor data and included in a search range, the threshold indicating the size of the search range. (3) The information processing method described in (2), wherein the search range is a range centered on an estimated position of the object, and the threshold indicates a radius of the search range. (4) The information processing method described in any of (1) to (3), wherein the threshold is calculated by linearly mapping the congestion level. (5) The information processing method described in any of (1) to (4), further comprising calculating the congestion level based on the number of objects included in a range targeted for calculation of the congestion level. (6) The information processing method according to (5), wherein the target range for the calculation of the congestion degree is a polygonal, circular, elliptical, or partially cut-out region of a circle or an ellipse when the three-dimensional space is viewed from above. (7) The information processing method according to (5) or (6), further comprising calculating the target range for the calculation of the congestion degree based on the position of the object to be tracked, the movement speed of the object to be tracked, or the sensing range of the sensor. (8) The information processing method according to any of (5) to (7), further comprising presenting the target range for the calculation of the congestion degree to a user. (9) The information processing method according to any of (1) to (8), further comprising presenting the object recognized based on the sensor data to a user as a candidate object to be tracked, and accepting selection of the object to be tracked by the user. (10) The information processing method according to any of (1) to (9), further comprising tracking the position of a person as the object in the three-dimensional space, and controlling the threshold used to identify the person based on the congestion degree.(11) The information processing method according to any one of (1) to (10), wherein the sensor includes at least one of a camera and LiDAR. (12) The information processing method according to any one of (1) to (11), wherein the sensor is provided on a mobile body that moves based on the position of the tracked object in the three-dimensional space. (13) A mobile body comprising: a sensor that detects objects in three-dimensional space; a tracking unit that identifies the tracked object from among the objects recognized based on sensor data of the sensor, and tracks the position of the tracked object in the three-dimensional space; a tracking control unit that controls a threshold used to identify the object based on a congestion level around the sensor; and a movement control unit that controls movement of the mobile body based on the position of the tracked object. (14) A program that causes a computer to execute processes of identifying the tracked object from among the objects recognized based on sensor data of a sensor that detects objects in three-dimensional space, and tracking the position of the tracked object in the three-dimensional space, and controlling a threshold used to identify the object based on a congestion level around the sensor.

[0091] DESCRIPTION OF SYMBOLS 1 Robot, 11 Body, 12 Head, 12A Visual sensor, 13 Moving body, 21 Memory, 22 Calculation unit, 23 Actuator, 41 Recognition unit, 42 Tracking unit, 43 Target range calculation unit, 44 Congestion degree calculation unit, 45 Threshold calculation unit, 46 UI control unit, 47 Action planning unit, 48 Movement control unit

Claims

1. Identifying the object to be tracked from among the objects recognized based on the sensor data of a sensor that detects objects in a three-dimensional space, and tracking the position of the object to be tracked in the three-dimensional space; and controlling a threshold value used for identifying the object based on the degree of congestion around the sensor. An information processing method including these steps.

2. Identifying the object to be tracked from among the objects included in a search range among the objects recognized based on the sensor data, wherein the threshold value indicates the size of the search range. The information processing method according to claim 1.

3. The search range is a range centered on the estimated position of the object, and the threshold value indicates the radius of the search range. The information processing method according to claim 2.

4. Calculating the threshold value by linearly mapping the degree of congestion. The information processing method according to claim 1.

5. Further including calculating the degree of congestion based on the number of objects included in the target range for calculating the degree of congestion. The information processing method according to claim 1.

6. The target range for calculating the degree of congestion is a region having a polygonal shape, circular shape, elliptical shape, or a shape obtained by cutting out a part of a circle or an ellipse when viewing the three-dimensional space from above. The information processing method according to claim 5.

7. Further including calculating the target range for calculating the degree of congestion based on the position of the object to be tracked, the moving speed of the object to be tracked, or the sensing range of the sensor. The information processing method according to claim 5.

8. Further including presenting the target range for calculating the degree of congestion to the user. The information processing method according to claim 5.

9. Further including presenting the object recognized based on the sensor data to the user as a candidate for the object to be tracked, and accepting the selection of the object to be tracked by the user. The information processing method according to claim 1.

10. Tracking the position of a person in the three-dimensional space as the object, and controlling the threshold value used for identifying the person based on the degree of congestion. The information processing method according to claim 1.

11. The sensor includes at least one of a camera and LiDAR. The information processing method according to claim 1.

12. The sensor is provided on a moving body that moves based on the position of the object to be tracked in the three-dimensional space. The information processing method according to claim 1.

13. A mobile body comprising: a sensor that detects an object in a three-dimensional space; a tracking unit that identifies the object to be tracked from among the objects recognized based on the sensor data of the sensor and tracks the position of the object to be tracked in the three-dimensional space; a tracking control unit that controls a threshold value used for the identification of the object based on the degree of congestion around the sensor; and a movement control unit that controls the movement of the own vehicle based on the position of the object to be tracked.

14. A program for causing a computer to execute a process of identifying an object to be tracked from among the objects recognized based on sensor data of a sensor that detects an object in a three-dimensional space, tracking the position of the object to be tracked in the three-dimensional space, and controlling a threshold value used for the identification of the object based on the degree of congestion around the sensor.

Images