Information processing apparatus, information processing method, moving object, program product, storage medium, and system

Abstract

The invention provides an information processing device, an information processing method, a moving object, a program product, a storage medium, and a system, which can estimate the moving destination of a dynamic object with high precision. An information processing device for estimating a destination of movement of a moving target, the information processing device comprising: a selection unit that selects a plurality of geographic positions from the periphery of the moving target; a calculation unit that calculates, for each of the plurality of geographic positions, a likelihood of each geographic position as a destination using a likelihood function; a reselection section that reselects the plurality of geographic positions based on the likelihood at each geographic position; and an estimation unit that estimates a destination on the basis of the plurality of reselected geographic positions. The likelihood function is based on at least one of a state of the dynamic target and a position of an obstacle around the dynamic target.

Description

Technical Field

[0001] This invention relates to information processing apparatus, information processing method, mobile device, program product, storage medium, and system. Background Technology

[0002] To assist users, an autonomous mobile robot capable of following them is provided. In the technology described in Patent Document 1, a virtual following target, different from the following object, is set in order for the autonomous mobile robot to follow the user in a more natural manner.

[0003] Existing technical documents Patent documents Patent Document 1: Japanese Patent Application Publication No. 2021-77088 Summary of the Invention The problem that the invention aims to solve To control a moving body, the destination of its movement can be determined using dynamic targets in the vicinity of the moving body. One aspect of the invention provides a technique for accurately estimating the destination of a dynamic target's movement.

[0004] means for solving problems According to one embodiment, an information processing apparatus is provided for estimating the destination of a moving object, wherein the information processing apparatus comprises: a selection mechanism that selects a plurality of geographical locations from the vicinity of the moving object; a calculation mechanism that calculates, for each of the plurality of geographical locations, a likelihood function for each of the plurality of geographical locations as the destination; a reselection mechanism that reselects the plurality of geographical locations based on the likelihood at each geographical location; and an estimation mechanism that estimates the destination based on the reselected plurality of geographical locations, wherein the likelihood function is based on at least one of the state of the moving object and the positions of obstacles surrounding the moving object.

[0005] According to another embodiment, an information processing method is provided, which is performed by a computer to estimate the destination of a moving object, wherein the information processing method includes: a selection step, in which a plurality of geographical locations are selected from the vicinity of the moving object; a calculation step, in which a likelihood function is used to calculate the likelihood of each of the plurality of geographical locations as the destination; a reselection step, in which the plurality of geographical locations are reselected based on the likelihood at each geographical location; and an estimation step, in which the destination is estimated based on the reselected plurality of geographical locations, wherein the likelihood function is based on at least one of the state of the moving object and the positions of obstacles around the moving object.

[0006] According to another embodiment, a system is provided for estimating the destination of a moving object, wherein the system comprises: a selection mechanism that selects a plurality of geographical locations from the vicinity of the moving object; a calculation mechanism that calculates, for each of the plurality of geographical locations, a likelihood function for each of the plurality of geographical locations as the destination; a reselection mechanism that reselects the plurality of geographical locations based on the likelihood at each geographical location; and an estimation mechanism that estimates the destination based on the reselected plurality of geographical locations, wherein the likelihood function is based on at least one of the state of the moving object and the positions of obstacles surrounding the moving object.

[0007] Invention Effects According to one implementation method, the destination of a moving target object can be estimated with high accuracy. Attached Figure Description

[0008] Figure 1 This is a schematic diagram illustrating an example of the external configuration of a mobile body in one of the embodiments.

[0009] Figure 2 This is a block diagram illustrating an example of the functional configuration of a mobile body in one of the embodiments.

[0010] Figure 3 This is a flowchart illustrating a control method for a moving body in one of the embodiments.

[0011] Figure 4 This is a schematic diagram illustrating a method for determining the target location of a dynamic target object in one of the implementation methods.

[0012] Figure 5 This is a schematic diagram illustrating the method for determining the likelihood function in one of the implementation methods.

[0013] Explanation of reference numerals in the attached figures 100: Moving object; 201: Control unit; 501: Target object; 502: Static obstacle; 503: Geographic location. Detailed Implementation

[0014] The embodiments will now be described in detail with reference to the accompanying drawings. It should be noted that the following embodiments are not intended to limit the invention as defined in the claims. Furthermore, not all combinations of features described in the embodiments are necessarily essential to the invention. Any combination of two or more features described in the embodiments may be used. Additionally, identical or identical components are labeled with the same reference numerals, and repeated descriptions are omitted.

[0015] <The Composition of a Moving Body> Reference Figure 1 Here, we will describe an example of the appearance configuration of the mobile body 100 according to a certain embodiment. Figure 1 In the diagram, arrow X indicates the forward / backward direction of the mobile body 100. F indicates forward, and R indicates backward. Arrows Y and Z indicate the width (left-right) and height (up-down) directions of the mobile body 100, respectively. The mobile body 100 is capable of autonomous movement. For example, the mobile body 100 is equipped with a battery and moves primarily through the power of a motor. The mobile body 100 can also be used in entertainment facilities, large commercial facilities, airports, parks, sidewalks, parking lots, and other similar locations. The mobile body 100 can be a vehicle that moves on the ground using wheels, an aircraft that moves in the air (e.g., a drone), or a robot that moves on the ground using two or more legs. Movement using wheels is also called driving. Movement in the air is also called flying. Movement using two or more legs is also called walking. The following description focuses on the case where the mobile body 100 is a vehicle, but the same description applies to other types of mobile bodies.

[0016] The mobile body 100 is capable of moving alongside its user (hereinafter referred to as "user"). Moving alongside the user means that the mobile body 100 moves based on the user's initiative. In the following examples, a configuration where no person rides on the mobile body 100 will be described, but the mobile body 100 can also accommodate persons other than the user. The mobile body 100 includes a travel unit 204 (… Figure 2 The vehicle includes, for example, a pair of front wheels 101 and a rear wheel 102. The driving unit 204 can also be in the form of a four-wheeled vehicle or a two-wheeled vehicle, or other forms. The mobile body 100 can also replace the user of the mobile body 100 as a companion or, based on this, be able to move autonomously to a pre-set destination.

[0017] The mobile body 100 has a frame 110 capable of storing goods. A cover 111 on the front surface of the frame, which can be opened and closed for storing goods, is provided with a locking mechanism. The locking mechanism is controlled by the mobile body 100. For example, the mobile body 100 unlocks upon successful user authentication. Alternatively, the mobile body 100 can also store goods as in other cases.

[0018] A touchscreen 120 is disposed on the upper surface 112 of the frame, allowing the user to change settings of the moving body 100 or confirm information related to the facility. The sensor box 130 contains detection units 206 such as cameras. Figure 2 The detection unit 206 observes the user and other targets in the environment surrounding the moving body 100 through the front surface 131, sides, and back of the sensor box 130.

[0019] <Example of the functional composition of a mobile entity> Reference Figure 2 The functional configuration example of the mobile body 100 will be described below. The mobile body 100 includes... Figure 2 The constituent elements shown. The moving body 100 may also include... Figure 2 The constituent elements not shown in the figure may also be omitted. Figure 2 A portion of the constituent elements shown.

[0020] The mobile body 100 includes a driving unit 204, which is an electric autonomous mobile body powered primarily by a battery 205. The battery 205 is, for example, a rechargeable battery such as a lithium-ion battery, and the driving unit 204 enables the mobile body 100 to move autonomously using the power supplied from the battery 205.

[0021] The driving unit 204 accelerates and decelerates the moving body 100 by changing the rotational speed of a pair of front wheels 101 using a motor as a drive source. The driving unit 204 may also include a braking mechanism for decelerating the moving body 100. The driving unit 204 steers the moving body 100 by differentiating the rotational speeds of the pair of front wheels 101. The driving unit 204 can detect and output physical quantities representing the motion of the moving body 100, such as its travel speed, acceleration, steering angle, angular velocity of the frame 110 of the moving body 100, and angular acceleration.

[0022] The mobile object 100 includes a detection unit 206 comprising one or more sensors. The detection unit 206 generates data for identifying targets (including objects and people) in the environment surrounding the mobile object 100. The detection unit 206 includes sensors such as a camera, radar, optical radar (Light Detection and Ranging), and ultrasonic sensors, which have the perimeter of the mobile object 100 as their detection range, and outputs sensor information. The camera may be a fisheye lens or a device capable of stereoscopic photography. Furthermore, the detection unit 206 includes a GNSS (Global Navigation Satellite System) sensor, which detects the current position of the mobile object 100 by receiving GNSS signals. The detection unit 206 can detect the current position using a Local Area Network (LAN) or Bluetooth signals. The camera may also be an RGB camera and may also have depth measurement capabilities. For example, the moving body 100 may have RGB cameras with depth measurement function on the front and rear sides of the sensor box 130, and RGB cameras without depth measurement function on the right and left sides of the sensor box 130.

[0023] The mobile body 100 includes a control unit (ECU, Electronic Control Unit) 201. The control unit 201 functions as a control device for the mobile body 100. The control unit 201 includes one or more processors 202, such as a CPU (Central Processing Unit), and a memory 203, such as a semiconductor memory. Therefore, the control unit 201 can also be referred to as an information processing device or a computer. The memory 203 stores the programs executed by the processors 202, the data used by the processors 202 during processing, etc. The processors 202 and the memory 203 can also be configured as multiple sets according to the functions of the mobile body 100 and capable of communicating with each other.

[0024] The control unit 201 acquires physical quantities representing motion output by the driving unit 204, detection results from the detection unit 206, input information from the touchscreen 120, and sound information input from the sound input device 207, and performs corresponding processing. For example, the control unit 201 controls the motor of the driving unit 204, controls the display of the touchscreen 120, and reports to the surrounding environment via sound.

[0025] The sound input device 207 collects ambient sounds from the surrounding environment of the mobile body 100. The control unit 201 can recognize the input sound and perform corresponding processing. The storage device 208 is a non-volatile, high-capacity storage device that stores map information, including information such as the travel routes that the mobile body 100 can travel on, restricted areas, landmarks, and shops. The storage device 208 can also store programs executed by the processor 202 and data used by the processor 202 in processing. The communication device 209 is, for example, a communication device capable of connecting to an external network via wireless communication such as fifth-generation mobile communication or wireless local area network.

[0026] The prompting device 210 displays (prompts) a user interface screen on the touchscreen 120, or outputs (prompts) spoken voice to the surrounding environment of the mobile body 100 via a microphone. The input device 211 includes, for example, a touch panel, and can be integrated with the touchscreen 120. The input device 211 accepts operation input from the user via the touch panel.

[0027] The processor 202 of the control unit 201 executes programs stored in the memory 203 or storage device 208 to perform the functions of the information acquisition unit 221, the target determination unit 222, the driving control unit 223, and the destination estimation unit 224. The information acquisition unit 221 acquires various information used in the processing of other components. The target determination unit 222 determines the target position of the mobile body 100 relative to the user based on the information acquired by the information acquisition unit 221. The target position refers to the location of the target that becomes the destination of the mobile body 100. The driving control unit 223 supplies control signals to the driving unit 204 to move the mobile body 100. The driving control unit 223 moves the mobile body 100 toward the target position. The destination estimation unit 224 estimates the destination of the moving target. Details regarding the processing of the information acquisition unit 221, the target determination unit 222, the driving control unit 223, and the destination estimation unit 224 will be described later.

[0028] <Control Method of Mobile Unit 100> Reference Figure 3 The method by which the control unit 201 controls the moving body 100 will be described. As described above, the control unit 201 can control the moving body 100 in a manner that allows it to move alongside the user. Figure 3 The steps of the method can be executed by processor 202 executing a program stored in memory 203 or storage device 208. Instead, Figure 3 At least some steps of the method can be executed by an application-specific integrated circuit (ASIC). Figure 3The method can also begin when the user instructs the moving body 100 to travel alongside the user. The control unit 201 repeatedly (e.g., at 100ms cycles) executes the processes of S301 to S303.

[0029] Figure 3 The method is executed with the user of the mobile body 100 as the target. The control unit 201 in Figure 3 The method determines the user at the start time and controls the mobile body 100 to travel alongside the user. The control unit 201 can also identify a person located in front of the mobile body 100 in an image captured by the detection unit 206 (e.g., a camera) as a user, or a person who has previously registered as a user. During execution... Figure 3 During the process, the control unit 201 continuously determines whether a user can be detected. If a user cannot be detected within a predetermined time from the last time a user was detected, the control unit 201 determines that the user has been lost and interrupts the process. Figure 3 This is the method. Afterwards, the control unit 201 can restart based on the detected user situation. Figure 3 The method.

[0030] In S301, the control unit 201 (e.g., information acquisition unit 221) acquires information used in subsequent processing. This information may include user information, moving body information, and environmental information. The control unit 201 may store at least a portion of the acquired information in the memory 203 or storage device 208 for use in subsequent processing.

[0031] User information refers to information related to a user. User information may include the user's current geographical location and current posture. Control unit 201 may acquire user information based on the detection results of detection unit 206 (e.g., images captured by imaging device).

[0032] Mobile body information refers to information related to mobile body 100. Mobile body information may include the current speed of mobile body 100, the current geographical location of mobile body 100, and the current angular velocity of mobile body 100. Control unit 201 may also acquire mobile body information based on outputs from driving unit 204 and detection results from detection unit 206 (e.g., GNSS positioning data, inertial sensor data).

[0033] Environmental information refers to information related to the environment surrounding the mobile body 100. Environmental information may include the number, type, location, and size of objects in the environment surrounding the mobile body 100. Objects may include static objects and dynamic objects. Static objects may include structures such as walls, railings, pillars, and steps. Static objects are those directly or indirectly fixed to the ground and do not move. Static objects that may become obstacles to the movement of the mobile body 100 can be called static obstacles. Dynamic objects may include pedestrians, cyclists, autonomous mobile bodies, animals, etc. Dynamic objects are those capable of moving relative to the ground. Dynamic objects that may become obstacles to the movement of the mobile body 100 can be called dynamic obstacles. The control unit 201 may acquire environmental information based on the detection results of the detection unit 206 (e.g., images captured by the imaging device).

[0034] In S302, the control unit 201 (e.g., target determination unit 222) determines the target position of the moving body 100. For example, the control unit 201 may first predict the direction of movement of the user based on various information acquired in S301. The user's movement direction may be predicted based on, for example, the user's body orientation or the user's past movement paths. Then, the control unit 201 may determine the position at a predetermined angle to the predicted movement path and at a predetermined distance from the user as the target position. The control unit 201 may determine the target position in a manner that does not overlap with obstacles (e.g., static obstacles).

[0035] In S303, the control unit 201 (e.g., the driving control unit 223) moves the moving body 100 toward the target position determined in S302. Specifically, the control unit 201 generates a trajectory from the current position of the moving body 100 toward the target position, and moves the moving body 100 along this trajectory. Depending on the current attitude and speed of the moving body 100, sometimes the trajectory passes through the target position, and sometimes the trajectory does not pass through the target position. Afterward, the control unit 201 transfers the processing to S301, and repeats the above processing.

[0036] Control unit 201 can also execute Figure 3 During the process, the distance between the moving body 100 and dynamic obstacles around it is monitored. If this distance falls within a threshold (e.g., 1.5m), actions are taken to make the dynamic obstacles give way (e.g., sound broadcast). Furthermore, the control unit 201 can also execute... Figure 3During the process, if the distance between the moving body 100 and the dynamic obstacles around the moving body 100 becomes within another threshold (e.g., 1m), the moving body 100 is stopped. Furthermore, the control unit 201 can also generate a trajectory in S303 to avoid dynamic obstacles.

[0037] <Methods for obtaining the location of dynamic targets> Reference Figure 4 The method for presuming the destination of a moving object is described. The moving object can be a user of the mobile body 100, or other moving objects contained in the environment surrounding the mobile body 100. The destination of the moving object refers to the geographical location to which the moving object intends to move. The process of presuming the destination of the moving object can also be included in... Figure 3 In the processing of user information and environmental information by S301. That is, Figure 4 The method can also be used as Figure 3 This is part of S301. As described above, S301 is executed repeatedly. Figure 4 The method can be executed either each time S301 is executed, or at a predetermined frequency relative to the execution of S301. Instead, Figure 4 The method can also be executed under different circumstances than the execution of S301.

[0038] Figure 4 Each step can also be executed by the control unit 201 (e.g., the destination estimation unit 224). When multiple dynamic targets are present around the moving body 100, the control unit 201 can execute the steps for each dynamic target. Figure 4 The method can also be implemented for specific dynamic targets (e.g., the user of the mobile body 100, pedestrians around the mobile body 100, autonomous mobile robots around the mobile body 100). Figure 4 The method. In the following Figure 4 In the description, the dynamic target object of the presumed destination is represented as the object target object. Figure 4 In this method, the destination is estimated by performing the same processing as the particle filter.

[0039] In S401, the control unit 201 acquires information about the target object and its surrounding environment. The target object information may include, for example, its geographical location and direction of movement. The geographical location can be represented using two-dimensional coordinates on a horizontal plane. Figure 4The method of starting the movement is defined by the relative position of the moving object 100 at the time of its commencement, or it can be defined by latitude and longitude. The direction of movement of the target object is the direction in which the target object intends to move. The direction of movement of the target object can also be determined, for example, based on the orientation of the target object's body or its past movement path. For example, in the case of a human, the orientation of its waist can be used as the orientation of the target object's body (i.e., the direction of movement of the target object). The orientation of the target object's body can also be determined by parsing data observed by the detection unit 206 (e.g., images captured by a camera). Information about the environment surrounding the target object can also include information about static obstacles existing around the target object. Information about static obstacles can also include the area occupied by the static obstacles.

[0040] In S402, control unit 201 selects multiple geographic locations from the vicinity of the dynamic target. Control unit 201 can select multiple geographic locations randomly or regularly (e.g., in a grid pattern). Multiple geographic locations can also be selected from specific ranges within a map (e.g., a range within 100 units of the moving object or a predetermined distance from the target object). In the following description, we assume that N geographic locations are selected. N can be, for example, 100, 1000, etc. Each of the multiple geographic locations corresponds to a particle in a particle filter.

[0041] In S403, the control unit 201 uses a likelihood function to calculate the likelihood of each of the multiple geographic locations selected in S402 (or reselected in S404) as the destination. The likelihood function is a function that outputs the likelihood of a specific geographic location as the destination. Specific examples of the likelihood function will be described later.

[0042] In S404, the control unit 201 reselects multiple geographic locations based on the likelihood of each location. This reselection of multiple geographic locations is equivalent to resampling in a particle filter. For example, the control unit 201 selects new N geographic locations by repeating the process of selecting one of the existing N geographic locations based on the likelihood of each location N times. As a result, geographic locations with low likelihood among the existing locations are discarded with a high probability (i.e., not selected as new geographic locations). On the other hand, geographic locations with high likelihood among the existing locations are selected with a high probability, and the same geographic location is selected multiple times depending on the situation. In this way, some of the new N geographic locations can be the same location.

[0043] In S405, control unit 201 determines whether the condition for ending repeated execution (the termination condition) is met. If the termination condition is met ("Yes" in S405), control unit 201 transfers the process to S406; otherwise ("No" in S405), control unit 201 transfers the process to S403. Thus, S403 to S405 are repeated until the termination condition is met. The termination condition can also be that S403 to S405 are repeated a predetermined number of times (e.g., five times). Alternatively, or based on this, the termination condition can also be based on a reselected convergence condition. For example, control unit 201 can determine that reselected convergence (i.e., the termination condition is met) is achieved if the change in the centroid of multiple geographical locations before and after S404 is less than a threshold.

[0044] In S406, the control unit 201 estimates the destination based on multiple geographical locations at that point in time (i.e., multiple geographical locations that were just reselected). For example, the control unit 201 estimates the centroid of multiple geographical locations as the destination. As described above, multiple geographical locations that are determined to have high likelihood by the likelihood function are reselected. Therefore, the destination is determined to be the location with the high likelihood of being the destination.

[0045] The control unit 201 can store the destination estimated in S406 in the memory 203 or storage device 208 for subsequent processing. The control unit 201 can also, as described above... Figure 3 In S303, the movement of the mobile body 100 is controlled based on the destination of the dynamic target. For example, if the dynamic target is the user of the mobile body 100, the control unit 201 can also control the mobile body 100 to move along a predicted trajectory leading to the user's destination. If the dynamic target is a dynamic obstacle, the control unit 201 can also control the movement of the mobile body 100 to avoid intersecting with the predicted trajectory leading to the destination of the dynamic obstacle.

[0046] <Example of Likelihood Function> exist Figure 4 The likelihood function used in S403 can also be based on at least one of the state of the target object and the position of obstacles around the target object. Specifically, the likelihood function can be based on at least one of the following: the direction of movement of the target object, whether each geographical location overlaps with a static obstacle, and the predicted trajectory of the target object moving toward each geographical location, or it can be based on all of them.

[0047] For example, the likelihood function p(i) is given by the following formula.

[0048]

[0049] In this formula, i represents the number of multiple geographic location labels, which is an integer greater than 1 and less than N. The numerator consists of three terms W. d （i） W o （i） and W t （i） The sum of terms. Item W d （i） This is the value calculated for the i-th geographical location based on the direction of movement of the target object. Term W o （i） This is the value calculated for the i-th geographical location, based on whether each geographical location overlaps with a static obstacle. Term W t （i） This value is calculated for the i-th geographical location based on the predicted trajectory of the target object moving towards each geographical location. The larger the values ​​of the terms in the numerator, the larger the value output by the likelihood function p(i). The denominator on the right is the term used to normalize the likelihood. For simplicity, the superscript (i) of each term is omitted below.

[0050] Reference Figure 5 For item W d W o and W t An example of the calculation method is given below. Map 500 shows the environment surrounding the target object 501. Static obstacles 502 exist around the target object 501. There are three static obstacles 502 in map 500, with only one labeled. Multiple geographic locations 503 are selected in map 500. In this example of map 500, six geographic locations 503 are selected, with only one labeled. The multiple geographic locations 503 can be those selected in S402 or those reselected in S404.

[0051] Referring to map 510, the term W is calculated based on the movement direction 511 of the target object 501. d The method will be explained below. Generally speaking, there is a high probability that object 501 is moving in a manner close to its destination. Therefore, item W d The angle θ formed by the moving direction 511 of the target object 501 and the direction 512 from the target object 501 toward each geographical location 503 d The value is calculated by using the method that the smaller the value (above 0 degrees and below 180 degrees), the larger the value.

[0052] For example, item W d It can be calculated using the following formula.

[0053]

[0054] A is a predetermined positive constant, during execution Figure 5 The method is determined beforehand and stored in memory 203 or storage device 208. Area 513 visually displays items W related to each geographical location 503. d The size of region 513. The larger the region, the larger the term W. d The larger.

[0055] Referring to map 520, calculate the term W based on whether each geographical location 503 overlaps with a static obstacle 502. o The method is explained below. The target object 501 may move towards a static obstacle 502 (e.g., a shop, a meeting place). Therefore, the term W is calculated in a way that it has a larger value when each geographical location 503 overlaps with the static obstacle 502 than when each geographical location 503 does not overlap with the static obstacle 502. o .

[0056] For example, item W o It can be calculated using the following formula.

[0057]

[0058] B is a predetermined positive constant, during execution Figure 5 The method is previously determined and stored in memory 203 or storage device 208. Area 521 visually displays items W related to each geographical location 503. o The size of region 521. The larger the region, the larger the term W. o The larger.

[0059] The term W is calculated on the reference map 530 based on the predicted trajectory 531 of the target object 501 moving toward various geographical locations 503. t The method will be explained as follows. The target object 501 may be moving in a direction different from its destination in order to avoid a static obstacle 502. Therefore, the angle θ between the trajectory 531 and the moving direction 511 of the target object 501 will be used to explain this. t The term W is calculated by using the method that the smaller the value, the larger the value. (Above 0 degrees and below 180 degrees) t The trajectory 531 is determined based on the current position of the target object 501, the geographical locations 503, and the positions of static obstacles 502. Angle θ t It can also be the angle between the tangent direction of the trajectory 531 at the current position of the target object 501 and the movement direction 511.

[0060] For example, item W t It can be calculated using the following formula.

[0061]

[0062] C is a predetermined positive constant, during execution Figure 5 The method is previously determined and stored in memory 203 or storage device 208. In the case where geographical location 503 overlaps with static obstacle 502, item W... t The value is zero. Region 532 visually displays the item W related to each geographic location 503. t The size of region 532. The larger the region, the larger the term W. t The larger.

[0063] Map 540 represents the likelihood calculated using the likelihood function p(i) for each geographic location 503. Region 541 visually shows the likelihood for each geographic location 503. The larger the region 541, the greater the likelihood.

[0064] As mentioned above, the likelihood function p(i) is based on at least one of the state of the target object and the positions of the obstacles surrounding the target object, therefore, through Figure 4 This method can accurately estimate the destination of the moving target 501. In the example above, the likelihood function p(i) is represented by three terms W. d （i） W o （i） and W t （i） The normalized sum. Alternatively, the likelihood function p(i) can also be given by other operations on these three terms. Furthermore, the likelihood function p(i) may contain only one or two of these three terms.

[0065] In the above embodiments, Figure 4 The method is executed by control unit 201. Instead, Figure 4 At least one step of the method can also be performed by an external server connected to the mobile unit 100 via a network. In other words, it can also be performed in cooperation with the control unit 201 via an external server. Figure 4 The method. In this case, the server and control unit 201 constitute an information processing system for controlling the mobile body 100. Additionally, in Figure 4 If all steps of the method are executed by an external server, the external server acts as the execution... Figure 4 The information processing device functions by employing the method of [unclear].

[0066] <Summary of Implementation Methods> (Project 1) An information processing device (201) is used to estimate the destination of a moving target (501), wherein, The information processing device includes: The selection mechanism selects multiple geographic locations from the vicinity of the dynamic target (503). The computing agency uses a likelihood function (p(i)) to calculate the likelihood of each of the plurality of geographic locations as the destination; The institution is reselected based on the likelihood at each geographic location; and The presumption agency estimates the destination based on the reselected plurality of geographical locations. The likelihood function is based on at least one of the state of the dynamic target and the position of the obstacles (502) around the dynamic target.

[0067] According to this project, the destination of a moving target can be estimated with high accuracy.

[0068] (Project 2) According to the information processing apparatus of Project 1, the likelihood function is based on the moving direction (511) of the dynamic target.

[0069] According to this project, since the intention of a dynamic target can be understood based on its direction of movement, the destination of the target's movement can be estimated with high accuracy.

[0070] (Project 3) According to the information processing device described in Project 2, the direction of movement of the dynamic target is determined based on the orientation of the dynamic target's body.

[0071] According to this project, the direction of movement of dynamic targets can be determined with high precision.

[0072] (Project 4) According to the information processing apparatus described in Project 2 or 3, the angle (θ) between the moving direction (511) of the dynamic target and the direction (512) from the dynamic target toward each geographical location is... d The smaller the value, the larger the output value of the likelihood function.

[0073] According to this project, since the intention of a dynamic target can be understood based on its direction of movement, the destination of the target's movement can be estimated with high accuracy.

[0074] (Project 5) The information processing apparatus according to any one of items 1 to 4, wherein the likelihood function is based on whether each geographical location overlaps with a static obstacle (502).

[0075] According to this project, it is possible to grasp the intention of a dynamic target moving toward a static obstacle, and thus to accurately estimate the destination of the dynamic target's movement.

[0076] (Project 6) According to the information processing device described in Project 5, when each geographical location overlaps with a static obstacle, the likelihood function outputs a larger value than when each geographical location does not overlap with a static obstacle.

[0077] According to this project, it is possible to grasp the intention of a dynamic target moving toward a static obstacle, and thus to accurately estimate the destination of the dynamic target's movement.

[0078] (Project 7) The information processing apparatus according to any one of items 1 to 6, wherein the likelihood function is based on the predicted trajectory (531) of the dynamic target object moving toward each geographical location.

[0079] According to this project, it is possible to grasp the intention of a dynamic target that moves to avoid static obstacles, and thus to accurately estimate the destination of the dynamic target's movement.

[0080] (Project 8) According to the information processing device described in Project 7, the angle (θ) between the trajectory and the moving direction (511) of the dynamic target object is... d The smaller the value, the larger the output value of the likelihood function.

[0081] According to this project, it is possible to grasp the intention of a dynamic target that moves to avoid static obstacles, and thus to accurately estimate the destination of the dynamic target's movement.

[0082] (Project 9) The information processing apparatus according to any one of items 1 to 8, wherein the likelihood function is based on the moving direction of the dynamic target, whether each geographical location overlaps with a static obstacle, and the predicted trajectory of the dynamic target moving toward each geographical location.

[0083] According to this project, by calculating the likelihood of each geographical location from various viewpoints, it is possible to estimate the destination of the movement of dynamic objects with high accuracy.

[0084] (Project 10) The information processing apparatus according to any one of items 1 to 9, wherein the information processing apparatus further comprises a control mechanism that controls the movement of a moving body (100) equipped with a sensor (206) for observing the dynamic target based on the estimated geographical location.

[0085] According to this project, it is possible to properly control moving objects.

[0086] (Project 11) A mobile body (100), wherein, The mobile body has: The information processing apparatus (201) described in any one of items 1 to 10; and Sensor (206) for observing dynamic targets.

[0087] According to this project, a moving body capable of accurately estimating the destination of a moving target is provided.

[0088] (Project 12) A program product for enabling a computer (201) to function as a mechanism of an information processing apparatus as described in any one of items 1 to 10, or a storage medium storing a program for enabling a computer (201) to function as a mechanism of an information processing apparatus as described in any one of items 1 to 10.

[0089] According to this project, a program product or storage medium capable of accurately estimating the destination of a moving target object is provided.

[0090] (Project 13) An information processing method is an information processing method executed by a computer (201) for estimating the destination of a moving target (501), wherein, The information processing method has the following characteristics: In the selection step (S402), multiple geographical locations are selected from the vicinity of the dynamic target (503). Calculation step (S403), in which a likelihood function (p(i)) is used to calculate the likelihood of each of the plurality of geographical locations as the destination; A reselection step (S404) is performed, in which the plurality of geographical locations are reselected based on the likelihood at each geographical location; and In the estimation step (S406), the destination is estimated based on the reselected plurality of geographical locations. The likelihood function is based on at least one of the state of the dynamic target and the position of the obstacles (502) around the dynamic target.

[0091] According to this project, the destination of a moving target can be estimated with high accuracy.

[0092] (Project 14) A system for estimating the destination of a moving target (501), wherein, The system has the following features: The selection mechanism selects multiple geographic locations from the vicinity of the dynamic target (503). The computing agency uses a likelihood function (p(i)) to calculate the likelihood of each of the plurality of geographic locations as the destination; The institution is reselected based on the likelihood at each geographic location; and The presumption agency estimates the destination based on the reselected plurality of geographical locations. The likelihood function is based on at least one of the state of the dynamic target and the position of the obstacles (502) around the dynamic target.

[0093] According to this project, the destination of a moving target can be estimated with high accuracy.

[0094] This invention is not limited to the embodiments described above, and various modifications and alterations can be made within the scope of the spirit of this invention.

Claims

1. An information processing apparatus that is for inferring a destination of movement of a dynamic target object, in which the information processing apparatus includes: a selection mechanism that selects a plurality of geographical locations from a vicinity of the dynamic target object; a calculation mechanism that calculates, for each of the plurality of geographical locations, a likelihood that each geographical location is the destination using a likelihood function; a re-selection mechanism that re-selects the plurality of geographical locations based on the likelihood at each geographical location; and an inference mechanism that infers the destination based on the re-selected plurality of geographical locations, the likelihood function is based on at least one of a state of the dynamic target object and a position of an obstacle in the vicinity of the dynamic target object. the likelihood function is based on a moving direction of the dynamic target object.

2. The information processing apparatus according to claim 1, wherein the moving direction of the dynamic target object is determined based on an orientation of a body of the dynamic target object.

3. The information processing apparatus according to claim 2, wherein the smaller an angle between the moving direction of the dynamic target object and a direction from the dynamic target object toward each geographical location, the larger a value that the likelihood function outputs.

4. The information processing apparatus according to claim 2, wherein the likelihood function is based on whether each geographical location overlaps with a static obstacle.

5. The information processing apparatus according to claim 1, wherein the likelihood function outputs a larger value in a case where each geographical location overlaps with a static obstacle than in a case where each geographical location does not overlap with a static obstacle.

6. The information processing apparatus according to claim 5, wherein the likelihood function is based on a trajectory that is predicted to be moved by the dynamic target object in order to move toward each geographical location.

7. The information processing apparatus according to claim 1, wherein the smaller an angle between the trajectory and the moving direction of the dynamic target object, the larger a value that the likelihood function outputs.

8. The information processing apparatus according to claim 7, wherein the likelihood function is based on the moving direction of the dynamic target object, whether each geographical location overlaps with a static obstacle, and the trajectory that is predicted to be moved by the dynamic target object in order to move toward each geographical location.

9. The information processing apparatus according to claim 1, wherein the information processing apparatus further includes a control mechanism that controls movement of a mobile body that includes a sensor that observes the dynamic target object based on the inferred geographical location.

10. The information processing apparatus according to claim 1, wherein 11.A mobile body in which the mobile body includes: the information processing apparatus according to any one of claims 1 to 10; and a sensor that observes a dynamic target object. The program product is for causing a computer to function as each mechanism of the information processing apparatus according to any one of claims 1 to 10.

12. A program product, wherein, The storage medium stores a program for causing a computer to function as each mechanism of the information processing apparatus according to any one of claims 1 to 10.

13. A storage medium, wherein, 14.An information processing method that is for inferring a destination of movement of a dynamic target object and that is executed by a computer, in which the information processing method includes: a selection step of selecting a plurality of geographical locations from a vicinity of the dynamic target object; a calculation step of calculating, for each of the plurality of geographical locations, a likelihood that each geographical location is the destination using a likelihood function; a re-selection step of re-selecting the plurality of geographical locations based on the likelihood at each geographical location; and an inference step of inferring the destination based on the re-selected plurality of geographical locations. ​ an inferring step of inferring the destination based on the reselected plurality of geographical positions, the likelihood function is based on at least one of a state of the dynamic target and a position of an obstacle surrounding the dynamic target.

15. A system for inferring a destination of movement of a dynamic target, wherein, the system comprises: a selecting mechanism that selects a plurality of geographical positions from a surrounding of the dynamic target; a calculating mechanism that calculates, for each of the plurality of geographical positions, a likelihood that the respective geographical position is the destination using a likelihood function; a reselecting mechanism that reselects the plurality of geographical positions based on the likelihood at each geographical position; and an inferring mechanism that infers the destination based on the reselected plurality of geographical positions, the likelihood function is based on at least one of a state of the dynamic target and a position of an obstacle surrounding the dynamic target. ​

Images