WO2023233515A1

WO2023233515A1 - Object detection device, object detection method, and object detection program

Info

Publication number: WO2023233515A1
Application number: PCT/JP2022/022107
Authority: WO
Inventors: 道学吉田
Original assignee: 三菱電機株式会社
Priority date: 2022-05-31
Filing date: 2022-05-31
Publication date: 2023-12-07
Also published as: JPWO2023233515A1; JP7233627B1

Abstract

An object detection device (100) comprises a route planning unit (130). The route planning unit (130) determines whether there is the possibility that noise could be superimposed on an observation result of an image sensor on the basis of the orientation of the image sensor mounted on a mobile body (80), and an electromagnetic wave entry direction observed by the image sensor. If the route planning unit has determined that there is the possibility that noise could be superimposed on the observation result of the image sensor, the route planning unit calculates, as posture for the mobile body (80), a posture such that the orientation of the image sensor is diverted from the entry direction.

Description

Object detection device, object detection method, and object detection program

The present disclosure relates to an object detection device, an object detection method, and an object detection program.

There is a technique for estimating the position of a moving object based on data acquired by a camera mounted on the moving object. This technique has a problem in that estimation results become unstable due to light entering the camera near windows, skylights, etc. of the warehouse.
Patent Document 1 discloses a technique of switching a camera and a laser sensor in order to realize robust self-position estimation in this technique. In the technology disclosed in Patent Document 1, areas where the camera can be used and areas where the camera cannot be used are set in advance, and a laser sensor is used in the area where the camera cannot be used.

Japanese Patent Application Publication No. 2021-135580

When a camera mounted on a moving object is used to estimate the position of a moving object, the direction in which light enters the camera determines whether noise is likely to be included in the camera's observation results. Therefore, even if a moving object exists in the area disclosed in Patent Document 1 where the camera cannot be used, noise will not be added to the observation results of the camera mounted on the moving object depending on the attitude of the moving object.
Therefore, in the technology disclosed in Patent Document 1, in order to determine whether or not the camera mounted on a moving object can be used based only on the position of the moving object, the camera may be used depending on the attitude of the moving object. There is a problem in that even if it is possible to use the camera, it is determined that the camera cannot be used.

An object of the present disclosure is to avoid determining whether a camera mounted on a moving object can be used only based on the position of the moving object.

The object detection device according to the present disclosure includes:
Based on the orientation of an image sensor mounted on a moving body and the incoming direction of electromagnetic waves observed by the image sensor, it is determined whether there is a possibility that noise will be added to the observation result of the image sensor, and the image sensor is A path planning unit is provided that calculates, as the attitude of the moving body, an attitude in which the orientation of the image sensor deviates from the line input direction when it is determined that there is a possibility that noise may be added to the observation result of the sensor.

According to the present disclosure, when there is a possibility that noise may be added to the observation results of the image sensor mounted on the moving object, the attitude of the moving object is calculated such that the orientation of the image sensor deviates from the incoming line direction. Here, an image sensor can be used when the direction of the image sensor deviates from the incoming line direction, and a camera can be cited as a specific example of the image sensor. Therefore, by utilizing the present disclosure, it is possible to avoid determining whether or not a camera mounted on a moving object can be used based only on the position of the moving object.

1 is a diagram showing a configuration example of an object detection system 90 according to Embodiment 1. FIG. FIG. 3 is a diagram illustrating processing of the route planning unit 130 according to the first embodiment. FIG. 7 is a diagram showing how the vehicle body of the moving body 80 according to the first embodiment is swung from side to side. 1 is a diagram showing an example of the hardware configuration of an object detection device 100 according to Embodiment 1. FIG. 7 is a flowchart showing the operation of object detection system 90 according to the first embodiment. 7 is a flowchart showing the operation of object detection system 90 according to the first embodiment. 7 is a flowchart showing the operation of object detection system 90 according to the first embodiment. FIG. 3 is a diagram illustrating a hardware configuration example of an object detection device 100 according to a modification of the first embodiment.

In the description of the embodiments and the drawings, the same elements and corresponding elements are denoted by the same reference numerals. Descriptions of elements labeled with the same reference numerals will be omitted or simplified as appropriate. Arrows in the figure mainly indicate the flow of data or processing. Furthermore, "unit" may be read as "circuit," "process," "procedure," "process," or "circuitry" as appropriate.

Embodiment 1.
Hereinafter, this embodiment will be described in detail with reference to the drawings.

***Explanation of configuration***
FIG. 1 shows a configuration example of an object detection system 90 according to this embodiment. The object detection system 90 includes a false detection information DB (Database) 30 and a moving body 80, as shown in this figure. The object detection system 90 is a system that detects peripheral objects that are objects that exist around the moving body 80.

The false detection information DB 30 is a database that stores false detection information, and is typically held at air traffic control. The control is a base that controls the mobile object 80 as a specific example. Note that the moving body 80 or the object detection device 100 may hold the false detection information DB 30.
The false detection information is information regarding conditions that may cause false detection, that is, conditions that may cause the result of object detection by the object detection device 100 to be erroneous. The false detection information is information indicating a condition under which noise may be added to the observation result of the image sensor, and may be information indicating the position and orientation of the moving object 80. The false detection information may be information generated based on information indicating the source of the electromagnetic wave and information indicating structures around the moving body 80. As a specific example, false detection occurs when, among the surfaces of the image sensors constituting the sensor group 20, the surface that observes electromagnetic waves faces the source of the electromagnetic waves observed by the image sensors. Furthermore, as a specific example, false detection occurs when the image sensor receives direct light, indirect light, or reflected light. As another specific example, false detection is caused by an image sensor mounted on the moving object 80 capturing light passing through a transparent door, resulting in a matching error and a surrounding object existing near the image sensor. This is a phenomenon that appears to be the case. As another example, a false positive is a phenomenon caused by an image sensor capturing light that passes through a blind or sunshade instead of the transparent door. The image sensor is, for example, a camera. Specific examples of the electromagnetic waves include visible light or infrared light. When the electromagnetic waves are visible light, the source of the electromagnetic waves is, for example, the sun or lighting. When the electromagnetic waves are infrared light, the source of the electromagnetic waves is, for example, the sun or a heat source such as a heater. Sources of electromagnetic waves include objects that directly emit electromagnetic waves and objects that reflect electromagnetic waves emitted from the objects.
As a specific example, the false detection information is information indicating a combination of position, orientation, and time. In this example, when the moving body 80 exists at the position indicated by the false detection information at the time indicated by the false detection information, and the attitude of the moving body 80 is the attitude indicated by the false detection information, the object detection device 100 Object detection results may be incorrect. Furthermore, in this example, the false detection information is specifically a condition in which there is a possibility that noise may be added to the sensing information acquired by the sensors constituting the sensor group 20, and the position and orientation of the moving object 80, This is information indicating conditions related to time. The posture of the moving body 80 is, for example, a combination of the orientation of the moving body 80 and the shape of the moving body 80. The shape of the moving body 80 is determined by the shape of an arm included in the moving body 80, etc.
As another specific example, the false detection information includes information indicating the source of the electromagnetic wave and information used when calculating the path of light emitted from the source. As a specific example, the false detection information is generated based on weather forecast information for the time period in which the mobile object 80 is moving, and information on a map showing the structure of the building where the mobile object 80 is moving and the surrounding environment of the building. This is the information. The weather forecast and the map are used, as a specific example, when calculating the direction of light incidence to each sensor making up the sensor group 20.
When the false detection information DB 30 receives information indicating each of a start point, a way point, and a goal point from the object detection device 100 as information indicating the movement route of the moving object 80, noise may be added to the sensing information. If it is determined that the moving body 80 passes through a certain position in an attitude where there is a possibility that noise may be added to the sensing information, the object detection device 100 may be notified of the position and the attitude. At this time, the false detection information DB 30 may receive from the object detection device 100 information indicating the time period in which the moving object 80 passes through each point on the movement route.
Note that the false detection information DB 30 holds data for generating false detection information instead of false detection information, and may transmit the data to the object detection device 100 as appropriate.

The moving body 80 includes a sensor group 20 and an object detection device 100. The moving object 80 is an object that moves indoors or outdoors, and a specific example is an autonomous vehicle that moves within a factory. Note that the moving body 80 does not need to include the object detection device 100.

The sensor group 20 consists of at least one sensor included in the moving body 80. The sensor group 20 includes, as a specific example, at least one of a camera, a laser sensor, a millimeter wave radar, a sonar, and a GNSS (Global Navigation Satellite System) sensor. The sensor group 20 includes an image sensor. The image sensor is, for example, a sensor that observes visible light or infrared light.

The object detection device 100 includes a detection recognition section 110, a position/orientation detection section 120, a path planning section 130, a control section 140, and a false detection information collection section 150.

The detection recognition unit 110 receives sensing information from each sensor forming the sensor group 20, detects surrounding objects based on the received sensing information, and recognizes the detected surrounding objects. As a specific example, detecting a surrounding object means understanding the position, shape, size, etc. of the surrounding object, and recognizing a surrounding object means understanding the object's attributes in addition to detection. shall be. Specific examples of methods for recognizing peripheral objects include methods using voxelization, machine learning, or deep learning.

The position and orientation detection unit 120 receives sensing information from the sensors forming the sensor group 20, and calculates the position and orientation of the moving body 80 based on the received sensing information. When the object detection device 100 holds a detailed map in advance, the position and orientation detection unit 120 uses the absolute coordinates on the map and the roll angle of the mobile body 80 on the map as the position and orientation of the mobile body 80. , pitch angle, and yaw angle may be calculated. Further, the position and orientation detection unit 120 may calculate the latitude and longitude of the position where the mobile body 80 is located and the direction in which the mobile body 80 is facing, as the position and orientation of the mobile body 80, regardless of the map. .

The route planning unit 130 creates a travel route for the moving object 80. The route planning unit 130 may select one of a plurality of routes created in advance, or may create a free route connecting a given starting point, way point, and goal point. Furthermore, the route planning unit 130 determines that noise may be added to the observation results of the image sensor based on the orientation of the image sensor mounted on the moving body 80 and the incoming direction of the electromagnetic waves observed by the image sensor. If it is determined that there is a possibility that noise may be added to the observation result of the image sensor, the attitude of the moving object 80 is calculated such that the orientation of the image sensor deviates from the direction of entry. do. If the route planning unit 130 determines that noise is added to the observation result of the image sensor, it may calculate the position of the moving body 80 such that the direction of the image sensor deviates from the entry direction. The route planning unit 130 uses the false detection information DB 30 to determine whether there is a possibility that noise will be added to the observation result of the image sensor. The route planning unit 130 determines whether there is a possibility that noise will be added to the observation results of the image sensor at the target point on the set movement route of the moving object 80, and determines whether or not there is a possibility that noise will be added to the observation results of the image sensor at the target point. If it is determined that there is a possibility that noise may be added to the image sensor, the attitude of the moving body 80 at the target point may be calculated such that the orientation of the image sensor deviates from the incoming line direction. The route planning unit 130 determines whether there is a possibility that noise will be added to the observation results of the image sensor at the target point on the set movement route of the moving object 80, and determines whether or not there is a possibility that noise will be added to the observation results of the image sensor at the target point. If it is determined that there is a possibility that noise may be added to the target point, the moving route of the moving object 80 may be changed to a moving route that does not pass through the target point.
When selecting one of the plurality of routes created in advance, the route planning unit 130 refers to the false detection information DB 30 and moves at least one of the positions and postures where false detection may occur. Select a route that can be avoided at all times while the body 80 is moving.
On the other hand, when creating a free route, the path planning unit 130 takes measures such as virtually placing an object at a position where a false detection may occur, which may cause a false detection. A route is created in which the moving body 80 can always avoid at least one of the positions and postures that satisfy the conditions while moving.
In addition, when creating a travel route for the mobile object 80, the route planning unit 130 estimates the position of a light source such as the sun for each point on the travel route based on the season and time, and uses the estimated light source position and map. By calculating the incident direction of the light based on the information indicating and comparing the calculated incident direction with the direction of the image sensor, it is possible to determine whether the direction of the image sensor is in a direction that may cause noise to be added to the observation results of the image sensor. It may be determined whether or not there is one. At this time, if there is a point where the orientation of the image sensor is such that noise may be added to the observation result of the image sensor, the route planning unit 130 determines whether the orientation of the image sensor is such that noise may be added to the observation result of the image sensor. The moving route may be corrected in consideration of the position and orientation of the moving body 80 so as not to take a possible direction. In addition to an object that emits light, the light source may also be an object that reflects light emitted by another object. Note that the term "point" may be replaced with the term "region" as appropriate. In other words, the present embodiment also holds true even if the term "point" is appropriately replaced with a range having a certain area.

FIG. 2 is a diagram illustrating the process by which the route planning unit 130 corrects the travel route. The left side of FIG. 2 corresponds to the process of selecting a route, and the right side of FIG. 2 corresponds to the process of generating a free route.
On the left side of FIG. 2, the route planning unit 130 first selects a first travel route from the candidate travel routes, and takes into account the orientation of the image sensor mounted on the mobile object 80 and the direction of incoming sunlight. Then, it is determined that noise is added to the observation result of the image sensor at the target point on the selected first travel route. At this time, the route planning unit 130 refers to the weather forecast and map. Next, the route planning unit 130 selects a second travel route, which is a travel route such that the orientation of the image sensor deviates from the entry direction at the target point, from the travel route candidates.
On the right side of FIG. 2, the route planning unit 130 first generates a travel route based on the given start point and goal point, and similarly to the left side of FIG. It is determined that noise is added to the observation results. Next, the route planning unit 130 generates another travel route that avoids the target point.

The control unit 140 controls the movement of the moving body 80. As a specific example, the control unit 140 calculates the traveling speed of the moving body 80, the angular velocity when the moving body 80 turns, etc. based on the route created by the route planning unit 130. The control unit 140 may control each sensor forming the sensor group 20, such as changing the sensor used for observing surrounding objects.

The false positive information collection unit 150 collects or generates false positive information, and stores the collected or generated false positive information in the false positive information DB 30. As a specific example, the false detection information collection unit 150 extracts conditions related to the position and orientation of the moving body 80 as conditions under which noise may be added to the observation results of the image sensor. At this time, as a specific example, when the detection recognition unit 110 detects a surrounding object, the false detection information collection unit 150 shakes the vehicle body of the moving body 80 from side to side to determine if there is a position and orientation at which the detected surrounding object disappears. It may be determined whether or not the observation result contains noise by detecting whether or not the noise is present. Further, the false detection information collection unit 150 collects the observation results of the image sensor mounted on the moving body 80 and the image sensors mounted on other moving bodies 80 coming from the opposite direction to the traveling direction of the moving body 80. It may be determined whether or not the observation results include noise by comparing the observation results with the observation results. In addition, the false detection information collection unit 150 collects observation results that are considered to include noise and observations made when the mobile object 80 passes through the position where the observation result was observed at the same time as the observation result. It may be determined whether or not the observation results include noise by comparing the results.
When the image sensor acquires a noise observation result that is an observation result containing noise, the false detection information collection unit 150 collects the time point at which the noise observation result was acquired and the moving object 80 at that time point as false detection information. , and stores the generated information in the false detection information DB 30.

FIG. 3 shows how the vehicle body of the moving body 80 is shaken from side to side when the detection recognition unit 110 detects a surrounding object.
In FIG. 3, first, the detection recognition unit 110 detects surrounding objects based on the observation results of the image sensor when the moving body 80 is facing in the front direction. However, the observation results of the image sensor are actually noise.
Next, the control unit 140 turns the vehicle body of the moving body 80 diagonally to the left, and checks whether the detection recognition unit 110 detects a surrounding object. At this time, the detection recognition unit 110 does not detect surrounding objects. Note that the diagonal right direction is the same as the diagonal left direction.
Therefore, although the detection recognition unit 110 detects surrounding objects when the moving body 80 is facing forward, the detection recognition unit 110 does not detect the surrounding objects when the vehicle body of the moving body 80 is shaken from side to side. Therefore, the false detection information collection unit 150 determines that the observation result of the image sensor when the moving object 80 is facing forward is contaminated with noise. Therefore, the false detection information collection unit 150 generates, as false detection information, information indicating each of the position and orientation of the moving body 80 when the moving body 80 faces in the front direction, and the time when the noise was observed, The generated false detection information is stored in the false detection information DB 30.

FIG. 4 shows an example of the hardware configuration of the object detection device 100 according to this embodiment. Object detection device 100 consists of a computer. Object detection device 100 may include multiple computers.

As shown in this figure, the object detection device 100 is a computer that includes hardware such as a processor 11, a memory 12, an auxiliary storage device 13, an input/output IF (Interface) 14, and a communication device 15. These pieces of hardware are appropriately connected via signal lines 19.

The processor 11 is an IC (Integrated Circuit) that performs arithmetic processing, and controls hardware included in the computer. The processor 11 is, for example, a CPU (Central Processing Unit), a DSP (Digital Signal Processor), or a GPU (Graphics Processing Unit).
The object detection device 100 may include a plurality of processors in place of the processor 11. A plurality of processors share the role of the processor 11.

The memory 12 is typically a volatile storage device, and a specific example is a RAM (Random Access Memory). Memory 12 is also called main storage or main memory. The data stored in the memory 12 is stored in the auxiliary storage device 13 as necessary.

The auxiliary storage device 13 is typically a nonvolatile storage device, and specific examples include a ROM (Read Only Memory), an HDD (Hard Disk Drive), or a flash memory. Data stored in the auxiliary storage device 13 is loaded into the memory 12 as needed.
The memory 12 and the auxiliary storage device 13 may be configured integrally.

The input/output IF 14 is a port to which an input device and an output device are connected. The input/output IF 14 is, for example, a USB (Universal Serial Bus) terminal. Specific examples of the input device include a keyboard and a mouse. A specific example of the output device is a display.

The communication device 15 is a receiver and a transmitter. The communication device 15 is, for example, a communication chip or a NIC (Network Interface Card).

Each part of the object detection device 100 may use the input/output IF 14 and the communication device 15 as appropriate when communicating with other devices.

The auxiliary storage device 13 stores an object detection program. The object detection program is a program that causes a computer to realize the functions of each part of the object detection device 100. The object detection program is loaded into memory 12 and executed by processor 11. The functions of each part included in the object detection device 100 are realized by software.

Data used when executing the object detection program, data obtained by executing the object detection program, etc. are appropriately stored in the storage device. Each part of the object detection device 100 uses a storage device as appropriate. The storage device includes, as a specific example, at least one of the memory 12, the auxiliary storage device 13, a register within the processor 11, and a cache memory within the processor 11. Note that the terms "data" and "information" may have the same meaning. The storage device may be independent of the computer.
The functions of the memory 12 and the auxiliary storage device 13 may be realized by other storage devices.

The object detection program may be recorded on a computer-readable nonvolatile recording medium. Specific examples of the nonvolatile recording medium include an optical disk or a flash memory. The object detection program may be provided as a program product.

***Operation explanation***
The operation procedure of the object detection device 100 corresponds to an object detection method. Further, a program that realizes the operation of the object detection device 100 corresponds to an object detection program.

FIG. 5 is a flowchart showing an example of the operation of the object detection system 90. The operation of the object detection system 90 will be explained with reference to FIG. Note that the object detection system 90 may execute the process shown in FIG. 5 while the moving object 80 is moving along the set moving route, and in a simulation in which the moving object 80 moves along the set moving route. The process shown in FIG. 5 may also be executed. Further, in FIG. 5, broken lines indicate data transmission and reception.

(Step S101)
The detection recognition unit 110 receives sensing information from each sensor making up the sensor group 20.

(Step S102)
The detection recognition unit 110 detects surrounding objects based on the sensing information received in step S101.
When the detection recognition unit 110 detects a surrounding object, the detection recognition unit 110 recognizes the detected surrounding object, and then the object detection device 100 proceeds to step S103. In other cases, the object detection device 100 proceeds to step S101.

(Step S103)
The position and orientation detection unit 120 detects the current position and orientation of the moving object 80, and transmits information indicating the detected position and orientation to the false detection information DB 30.

(Step S104)
The route planning unit 130 receives false detection information corresponding to the position and orientation detected in step S103 from the false detection information DB 30, and determines the possibility that the surrounding object detected in step S102 is noise based on the received false detection information. Determine whether or not there is.
Note that the route planning unit 130 does not need to receive false detection information from the false detection information DB 30. If the route planning unit 130 does not receive false detection information from the false detection information DB 30, the route planning unit 130 determines that the surrounding object detected in step S102 is not noise.
If there is a possibility that the surrounding object detected in step S102 is noise, the object detection device 100 proceeds to step S105. In other cases, the object detection device 100 proceeds to step S108.

(Step S105)
First, the route planning unit 130 calculates the attitude of the moving body 80 such that noise is not added to the observation result of the image sensor mounted on the moving body 80 based on the false detection information received in step S104.
Next, the control unit 140 changes the attitude of the moving body 80 to the attitude calculated by the route planning unit 130.
Next, the detection recognition unit 110 receives sensing information from each sensor forming the sensor group 20.

(Step S106)
The detection recognition unit 110 detects surrounding objects based on the sensing information received in step S105.
When the detection recognition unit 110 detects a surrounding object, that is, when the surrounding object detected in step S102 is considered not to be noise, the detection recognition unit 110 recognizes the detected surrounding object, and then the object detection device 100 The process advances to step S107. In other cases, the object detection device 100 proceeds to step S108 and step S109.

(Step S107)
The route planning unit 130 changes the travel route of the moving body 80.

(Step S108)
The moving body 80 continues moving along the set moving route.

(Step S109)
The false detection information collection unit 150 extracts information indicating the conditions under which noise is detected as false detection information, and registers the extracted false detection information in the false detection information DB 30.

FIG. 6 is a flowchart showing another example of the operation of the object detection system 90. The operation of the object detection system 90 will be explained with reference to FIG. Note that the object detection system 90 may execute the process shown in FIG. 6 while the moving body 80 is moving along the set moving route, and in a simulation in which the moving body 80 moves along the set moving route. The process shown in FIG. 6 may also be executed.

(Step S121)
The route planning unit 130 receives false detection information corresponding to the position and orientation detected in step S103 from the false detection information DB 30.
Note that the object detection system 90 repeats the processing from step S103 to step S121 N times (N is an integer of 2 or more) for the purpose of improving the accuracy of determining whether or not the observation result of the image sensor is noise. Execute.

(Step S122)
First, the route planning unit 130 determines whether there is a possibility that noise will be added to the observation result of the image sensor mounted on the mobile object 80 based on the false detection information received in step S121.
Next, if it is determined that there is a possibility that noise may be added to the observation result of the image sensor, the route planning unit 130 determines whether noise may be added to the observation result of the image sensor based on the false detection information received in step S121. The posture of the moving body 80 that does not exist is calculated. Note that when the route planning unit 130 determines that there is no possibility of noise being added to the observation results of the image sensor, the route planning unit 130 sets the attitude of the moving body 80 to a different attitude from the current attitude of the moving body 80, A posture in which there is no possibility that noise will be added to the observation result of the image sensor may be calculated based on the false detection information received in step S121. After that, the control unit 140 changes the attitude of the moving body 80 to the attitude calculated by the route planning unit 130.
Next, the detection recognition unit 110 receives sensing information from each sensor forming the sensor group 20.

(Step S123)
The detection recognition unit 110 detects surrounding objects based on the sensing information received in step S122.
When the detection recognition unit 110 detects a peripheral object, the detection recognition unit 110 recognizes the detected peripheral object.
Furthermore, based on the result of executing step S123 N times, it is determined whether the surrounding object detected by the detection recognition unit 110 is noise. If the detection recognition unit 110 determines that the detected peripheral object is noise, the object detection device 100 proceeds to steps S108 and S109. In other cases, the object detection device 100 proceeds to step S107.

FIG. 7 is a flowchart showing another example of the operation of the object detection system 90. The operation of the object detection system 90 will be explained with reference to FIG. According to this example, the detection recognition unit 110 acquires false detection information before detecting surrounding objects, and changes the attitude of the moving object 80 based on the acquired false detection information, thereby reducing the number of times the movement route is changed. can be reduced. Note that the object detection system 90 may execute the process shown in FIG. 7 while the moving body 80 is moving along the set moving route, and in a simulation in which the moving body 80 moves along the set moving route. The process shown in FIG. 7 may also be executed.

(Step S141)
First, the route planning unit 130 receives false detection information corresponding to the position and orientation detected in step S103 from the false detection information DB 30, and based on the received false detection information, the route planning unit 130 Determine whether there is a possibility that noise may be added to the observation results.
Next, when it is determined that there is a possibility that noise will be added to the observation result of the image sensor, the route planning unit 130 calculates a posture of the moving object 80 in which there is no possibility that noise will be added to the observation result of the image sensor. do.
Next, the control unit 140 changes the attitude of the moving body 80 to the attitude calculated by the route planning unit 130.
Next, the detection recognition unit 110 receives sensing information from each sensor forming the sensor group 20.

(Step S142)
The detection recognition unit 110 detects surrounding objects based on the sensing information received in step S141.
When the detection recognition unit 110 detects a surrounding object, the detection recognition unit 110 recognizes the detected surrounding object, and then the object detection device 100 proceeds to step S107. In other cases, the object detection device 100 proceeds to step S108.

***Explanation of effects of Embodiment 1***
Existing technology that estimates the position of a moving object based on data acquired by a camera mounted on the moving object has a problem in that autonomous driving stops due to false detection at a location where a false detection occurs.
On the other hand, according to the present embodiment, since the state and direction of the light source and the position of the moving body 80 are known due to the presence of the false detection information DB 30, the observation results of the camera can be obtained using direct light, indirect light, or reflected light. It is possible to calculate whether there is a possibility that noise will be added to the image, and to avoid at least one of the positions and postures where noise may be added.
Further, according to the present embodiment, since the false detection information DB 30 exists, the position and orientation where false detection is likely to occur are known in advance. Therefore, according to the present embodiment, it is possible to generate or select in advance a movement route in which there is no possibility of noise being introduced, and to move in a posture that does not cause false detection at a position where there is a possibility of noise being introduced. Travel routes can be calculated in advance.

***Other configurations***
<Modification 1>
If the route planning unit 130 determines that the observation result of the image sensor contains noise, instead of changing the travel route, it generates information indicating that the image sensor will not be used.
According to this modification, even if it is determined that noise is added to the observation result of the image sensor, there is no need to change the movement route of the moving body 80.

<Modification 2>
FIG. 8 shows an example of the hardware configuration of an object detection device 100 according to this modification.
The object detection device 100 includes a processing circuit 18 instead of the processor 11, the processor 11 and the memory 12, the processor 11 and the auxiliary storage device 13, or the processor 11, the memory 12, and the auxiliary storage device 13.
The processing circuit 18 is hardware that implements at least a portion of each section included in the object detection device 100.
Processing circuit 18 may be dedicated hardware or may be a processor that executes a program stored in memory 12.

When the processing circuit 18 is dedicated hardware, the processing circuit 18 may be, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), or an FPGA (Field Programmable Gate Array) or a combination thereof.
The object detection device 100 may include a plurality of processing circuits that replace the processing circuit 18. The plurality of processing circuits share the role of the processing circuit 18.

In the object detection device 100, some functions may be realized by dedicated hardware, and the remaining functions may be realized by software or firmware.

The processing circuit 18 is implemented, for example, by hardware, software, firmware, or a combination thereof.
The processor 11, memory 12, auxiliary storage device 13, and processing circuit 18 are collectively referred to as a "processing circuitry." That is, the functions of each functional component of the object detection device 100 are realized by processing circuitry.

***Other embodiments***
Although Embodiment 1 has been described, a plurality of parts of this embodiment may be implemented in combination. Alternatively, this embodiment may be partially implemented. In addition, this embodiment may be modified in various ways as necessary, and may be implemented as a whole or in part in any combination.
Note that the embodiments described above are essentially preferable examples, and are not intended to limit the present disclosure, its applications, and the scope of use. The procedures described using flowcharts and the like may be changed as appropriate.

11 Processor, 12 Memory, 13 Auxiliary storage device, 14 Input/output IF, 15 Communication device, 18 Processing circuit, 19 Signal line, 20 Sensor group, 30 False detection information DB, 80 Mobile object, 90 Object detection system, 100 Object detection Device, 110 detection recognition unit, 120 position and orientation detection unit, 130 route planning unit, 140 control unit, 150 false detection information collection unit.

Claims

Based on the orientation of an image sensor mounted on a moving body and the incoming direction of electromagnetic waves observed by the image sensor, it is determined whether there is a possibility that noise will be added to the observation result of the image sensor, and the image sensor is Object detection comprising a path planning unit that calculates a posture of the moving body such that the orientation of the image sensor deviates from the line entry direction when it is determined that there is a possibility that noise may be added to the observation results of the sensor. Device.
The object detection device according to claim 1, wherein the route planning unit generates information indicating that the image sensor is not used when it is determined that noise is added to the observation result of the image sensor.
3. The route planning unit calculates a position of the moving object such that the orientation of the image sensor deviates from the line entry direction when it is determined that noise is added to the observation result of the image sensor. The object detection device described.
The route planning unit stores false detection information, which is information indicating conditions where noise may be added to the observation results of the image sensor, and is information indicating the position and orientation of the moving body. The object detection device according to any one of claims 1 to 3, which uses an information database to determine whether there is a possibility that noise will be added to the observation results of the image sensor.
The object detection device according to claim 4, wherein the false detection information is information generated based on information indicating a source of electromagnetic waves and information indicating structures around the moving object.
The object detection device further includes:
When the image sensor acquires a noise observation result that is an observation result containing noise, the false detection information includes the time when the noise observation result was acquired, and the position and orientation of the moving object at the time. The object detection device according to claim 4, further comprising a false detection information collection unit that generates information indicating the false detection information and stores the generated information in the false detection information database.
The route planning unit determines whether or not there is a possibility that noise will be included in the observation results of the image sensor at a target point on the set movement route of the moving body, and determines whether or not there is a possibility that noise will be included in the observation results of the image sensor at the target point. 7. If it is determined that there is a possibility that noise may be added to the result, an attitude is calculated as the attitude of the moving body at the target point such that the orientation of the image sensor deviates from the line input direction. The object detection device according to any one of the items.
The route planning unit determines whether or not there is a possibility that noise will be included in the observation results of the image sensor at a target point on the set movement route of the moving body, and determines whether or not there is a possibility that noise will be included in the observation results of the image sensor at the target point. The object detection device according to any one of claims 1 to 6, wherein when it is determined that there is a possibility that noise may be included in the result, the moving route of the moving object is changed to a moving route that does not pass through the target point.
A computer determines whether or not there is a possibility that noise will be added to the observation result of the image sensor based on the direction of the image sensor mounted on the moving body and the incoming direction of electromagnetic waves observed by the image sensor. , an object detection method that calculates, as the attitude of the moving object, an attitude in which the orientation of the image sensor deviates from the line entry direction when it is determined that there is a possibility that noise may be added to the observation result of the image sensor.
Based on the orientation of an image sensor mounted on a moving body and the incoming direction of electromagnetic waves observed by the image sensor, it is determined whether there is a possibility that noise will be added to the observation result of the image sensor, and the image sensor is If it is determined that there is a possibility that noise may be added to the observation results of the sensor, the computer performs a route planning process of calculating an orientation of the moving object such that the orientation of the image sensor deviates from the line entry direction. An object detection program that is executed by an object detection device.