WO2021192357A1 - Dispositif d'inspection automatique - Google Patents

Dispositif d'inspection automatique Download PDF

Info

Publication number
WO2021192357A1
WO2021192357A1 PCT/JP2020/034430 JP2020034430W WO2021192357A1 WO 2021192357 A1 WO2021192357 A1 WO 2021192357A1 JP 2020034430 W JP2020034430 W JP 2020034430W WO 2021192357 A1 WO2021192357 A1 WO 2021192357A1
Authority
WO
WIPO (PCT)
Prior art keywords
map
inspection
automatic inspection
sensor
inspection device
Prior art date
Application number
PCT/JP2020/034430
Other languages
English (en)
Japanese (ja)
Inventor
範安 長谷島
和哉 室谷
勇太 江阪
悠一 五十嵐
健人 緒方
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Publication of WO2021192357A1 publication Critical patent/WO2021192357A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M99/00—Subject matter not provided for in other groups of this subclass
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions

Definitions

  • the present invention relates to an automatic inspection device that autonomously inspects infrastructure equipment.
  • Patent Document 1 Japanese Patent Application Laid-Open No. 2015-161577
  • the control unit switches the means for estimating the position information of the device main body between the first position estimation unit and the second position estimation unit, and based on the acquired position information of the device main body, A self-propelled inspection device that advances the main body of the device along a predetermined route by controlling the traveling unit and inspects the production equipment to be inspected by controlling the inspection unit is described. See summary).
  • Patent Document 1 describes a self-propelled inspection device (automatic inspection device) that executes inspection of production equipment (infrastructure equipment) to be inspected.
  • Patent Document 1 does not describe that an automatic inspection device autonomously inspects infrastructure equipment and therefore generates an inspection map necessary for autonomous driving.
  • an expensive map generator that generates a highly accurate inspection map required for autonomous driving is made independent as an attachment, and the map generator is attached to the automatic inspection device only at the time of map generation. At the time of automatic inspection, it is removed from the automatic inspection device to provide an inexpensive and highly efficient automatic inspection device.
  • the automatic inspection device of the present invention includes a camera that captures an image of the surrounding inspection target, a first external world recognition sensor that acquires three-dimensional information of the surroundings, and a first position that acquires a position.
  • the information storage unit that stores the first sensor information and the inspection map of the first external world recognition sensor and the first position estimation sensor, and the first sensor information
  • the arithmetic processing for executing the automatic inspection is executed. It is characterized by having an arithmetic processing unit for performing an operation and a control unit for controlling autonomous traveling, and having a attachment / detachment mechanism of a map generation device for generating an inspection map.
  • an expensive map generator that generates a highly accurate inspection map required for autonomous driving is made independent as an attachment, and the map generator is attached to the automatic inspection device only at the time of map generation. At the time of automatic inspection, it is possible to provide an inexpensive and highly efficient automatic inspection device by removing it from the automatic inspection device.
  • FIG. It is explanatory drawing explaining the structure at the time of automatic inspection of the automatic inspection apparatus 200 described in Example 1.
  • FIG. It is explanatory drawing explaining the inspection map described in Example 1.
  • FIG. It is explanatory drawing explaining the inspection map generation part 105 in the map generation apparatus 100 described in Example 1.
  • FIG. It is a flowchart explaining the storage of the sensor information in the map generation described in Example 1.
  • FIG. It is explanatory drawing explaining the automatic inspection using the inspection map described in Example 1.
  • FIG. It is a flowchart explaining the automatic inspection described in Example 1.
  • FIG. It is explanatory drawing explaining the inspection map generation part 105 in the map generation apparatus 100 described in Example 2.
  • FIG. It is explanatory drawing explaining the map generation described in Example 2.
  • FIG. It is a flowchart explaining the map generation described in Example 2.
  • FIG. 1 is an explanatory diagram illustrating a configuration at the time of map generation of the automatic inspection device 200 described in the first embodiment.
  • the automatic inspection device 200 can be equipped with the map generation device 100, and the map generation device 100 can be removed.
  • the map generator 100 is attached to and removed from the automatic inspection device 200. That is, the automatic inspection device 200 has a attachment / detachment mechanism (not shown) of the map generation device 100.
  • the attachment / detachment mechanism mechanically and electrically connects the automatic inspection device 200 and the map generation device 100.
  • the attachment / detachment mechanism is a mechanism for attaching / detaching the housing of the map generation device 100, and the housing of the map generation device 100 includes a camera 101, an outside world recognition sensor 102, a position estimation sensor 103, and information, which will be described later.
  • a storage unit 104 and an inspection map generation unit 105 are installed.
  • the automatic inspection device 200 is mounted on a vehicle and autonomously executes an inspection (automatic inspection: automatic shooting) of an inspection target (for example, a meter or a marker in infrastructure equipment). That is, the automatic inspection device 200 autonomously travels and executes automatic inspection of the inspection target.
  • an inspection target for example, a meter or a marker in infrastructure equipment
  • the automatic inspection device 200 autonomously travels on a plane in two dimensions, but may autonomously move in space in three dimensions.
  • the map generator 100 includes a camera 101 capable of capturing an image of an inspection target around the automatic inspection device 200 and a highly accurate external world recognition capable of acquiring detailed three-dimensional information around the automatic inspection device 200. It is possible to store the sensor 102, the highly accurate position estimation sensor 103 capable of acquiring the position of the automatic inspection device 200 in detail, and the sensor information (second sensor information) of the external world recognition sensor 102 and the position estimation sensor 103. It has an information storage unit 104 capable of generating an inspection map, and an inspection map generation unit 105 capable of autonomously traveling and generating an inspection map for executing an automatic inspection of an inspection target, and generates an inspection map. (Execute map generation).
  • the inspection map is generated for the automatic inspection device 200 to autonomously travel and for executing the automatic inspection.
  • the camera 101 captures an image of the inspection target around the automatic inspection device 200.
  • Camera 101 is a camera used for map generation.
  • the camera 101 rotates 360 degrees in the horizontal direction and 90 degrees in the vertical direction, and captures an image of an inspection target in all directions without significantly changing the orientation of the automatic inspection device 200.
  • the camera faces only one direction, it is possible to take an image of the inspection target in all directions by using the turning operation of the automatic inspection device 200 together.
  • a spherical camera that can capture an image of an inspection target around the automatic inspection device 200 by capturing one image using a fisheye lens may be used.
  • the outside world recognition sensor 102 is a sensor that uses laser light for remote sensing, and is, for example, LiDAR (Laser Imaging Detection and Ringing).
  • the external world recognition sensor 102 is a high-precision sensor that remotely senses 360 degrees in the horizontal direction and ⁇ 20 degrees in the vertical direction by 32 to 48 scan lines.
  • the external world recognition sensor 102 can acquire the relationship (for example, distance) from the automatic inspection device 200 to the inspection target as point cloud information.
  • the mounting position and posture of the outside world recognition sensor 102 are known and are stored in the information storage unit 104. From the stored mounting position and posture, the distance from the automatic inspection device 200 to the relatively existing inspection target can be accurately acquired.
  • the camera 101 is calibrated with the outside world recognition sensor 102, the distance to the inspection target in the image can be grasped.
  • the position estimation sensor 103 is a sensor that acquires the position of the automatic inspection device 200 with high accuracy of 1 to 5 cm by combining GNSS (Global Navigation Satellite System), IMU (Inertial Measurement Unit), and the like.
  • GNSS Global Navigation Satellite System
  • IMU Inertial Measurement Unit
  • the position estimation sensor 103 can acquire the position of the automatic inspection device 200 as position information.
  • the position of the automatic inspection device 200 can also be obtained by using the external world recognition sensor 102 and using SLAM (Simultaneous Localization And Mapping) technology.
  • SLAM Simultaneous Localization And Mapping
  • the automatic inspection device 200 preferably travels at a low speed (1 to 2 km / s). As a result, the position of the automatic inspection device 200 can be acquired with high accuracy.
  • the information storage unit 104 is, for example, a memory such as a ROM or RAM, which acquires the second sensor information, converts it into a physical quantity, and stores it.
  • the inspection map generation unit 105 autonomously travels in the inspection site (the place where the automatic inspection is executed) by the automatic inspection device 200, and generates an inspection map for executing the automatic inspection.
  • the inspection map generation unit 105 generates an inspection map composed of point clouds based on the point cloud information acquired from the outside world recognition sensor 102 and the position information acquired from the position estimation sensor 103.
  • This map generation may be executed online (in real time) or offline.
  • the second sensor information stored in the information storage unit 104 is temporarily written to an external storage device to generate an inspection map.
  • an inspection map When executing online, use a combination of CPU, ROM and RAM, FPGA (Field Programmable Gate Array) which is a rewritable logic circuit, and ASIC (Application Specific Integrated Circuit) which is an integrated circuit for specific applications. Generate an inspection map.
  • the inspection map generation unit 105 is composed of a combination of CPU, ROM and RAM, a combination of CPU, ROM, RAM and FPGA, and the like.
  • the automatic inspection device 200 has a control unit 206.
  • the control unit 206 controls the autonomous driving of the automatic inspection device 200 based on the control signal (command value) input from the controller 210.
  • the controller 210 outputs a control signal for controlling the autonomous driving of the automatic inspection device 200 to the control unit 206 from the outside of the automatic inspection device 200.
  • the automatic inspection device 200 autonomously travels and generates an inspection map.
  • the control unit 206 operates the drive mechanism (not shown) of the automatic inspection device 200 based on the control signal.
  • the automatic inspection device 200 It is also possible to manually operate the automatic inspection device 200 to generate an inspection map.
  • the inspection map When the inspection map is generated by manual operation, the inspection map may be accompanied by the automatic inspection device 200, the automatic inspection device 200 may be operated and the automatic inspection device 200 may be driven, or a camera may be used to remotely control the automatic inspection device 200.
  • the automatic inspection device 200 may be operated and the automatic inspection device 200 may be driven.
  • controller 210 is a wired type or a wireless type, it suffices if the controller 210 can output the speed and the direction to the control unit 206.
  • a handle-type controller may be used to input the speed and direction separately, or a joystick-type controller may be used to input the tilt amount as the speed and the tilt direction as the direction.
  • controller 210 may be used to accompany the automatic inspection device 200 and operate the automatic inspection device 200, or may be used to remotely control the automatic inspection device 200 from a remote location. ..
  • the map generation device 100 is attached to the automatic inspection device 200, and the control unit 206 is used.
  • the control unit 206 is used.
  • the map generation device 100 is attached to the automatic inspection device 200 at the time of map generation, and the map generation device 100 is removed from the automatic inspection device 200 at the time of automatic inspection. Then, one map generation device 100 is used for a plurality of automatic inspection devices 200. As a result, it is possible to provide an inexpensive automatic inspection device 200 with a reduced unit price.
  • FIG. 2 is an explanatory diagram illustrating a configuration at the time of automatic inspection of the automatic inspection device 200 described in the first embodiment.
  • the automatic inspection device 200 includes a camera 203 capable of capturing an image of an inspection target around the automatic inspection device 200, and an inexpensive external world recognition sensor 201 capable of acquiring three-dimensional information around the automatic inspection device 200. , An inexpensive position estimation sensor 202 capable of acquiring the position of the automatic inspection device 200, and information storage capable of storing the sensor information (first sensor information) of the external world recognition sensor 201 and the position estimation sensor 202 and the inspection map.
  • the arithmetic processing unit 205 that can execute the arithmetic processing for executing the automatic inspection by using the first sensor information, and the control signal. It has a control unit 206 that can perform automatic inspection.
  • the camera 203 captures an image of the inspection target around the automatic inspection device 200. Then, the camera 203 reads, for example, a meter or a marker to be inspected at the time of automatic inspection.
  • the image to be inspected taken by the camera 203 is transmitted to the center or the like via the network. Then, the image of the inspection target transmitted to the center or the like is analyzed, and the normality or abnormality of the inspection target is determined.
  • a person may analyze the transmitted image to be inspected and determine whether it is normal or abnormal.
  • the transmitted image to be inspected may be automatically analyzed to determine whether it is normal or abnormal, and a person may be contacted only when the image is abnormal.
  • Camera 203 is a camera used for automatic inspection.
  • the camera 203 rotates 360 degrees in the horizontal direction and 90 degrees in the vertical direction, and captures an image of an inspection target in all directions without significantly changing the orientation of the automatic inspection device 200.
  • the camera faces only one direction, it is possible to take an image of the inspection target in all directions by using the turning operation of the automatic inspection device 200 together.
  • a spherical camera that can capture an image of an inspection target around the automatic inspection device 200 by capturing one image using a fisheye lens may be used.
  • the same camera 203 and the camera 101 can be used.
  • the inspection map can be generated by using the camera 203 without mounting the camera 101 on the map generation device 100.
  • the outside world recognition sensor 201 is a sensor that uses laser light for remote sensing, and is, for example, LiDAR (Laser Imaging Detection and Ringing).
  • the external world recognition sensor 201 is a sensor that remotely senses 180 degrees in the horizontal direction and ⁇ 20 degrees in the vertical direction by 1 to 16 scan lines.
  • the external world recognition sensor 201 is cheaper and inferior in accuracy (equal or less) than the external world recognition sensor 102.
  • the external world recognition sensor 201 can acquire the relationship (for example, distance) from the automatic inspection device 200 to the inspection target as point cloud information.
  • the mounting position and posture of the outside world recognition sensor 201 are known and are stored in the information storage unit 204.
  • the distance from the automatic inspection device 200 to the relatively existing inspection target can be obtained from the stored mounting position and posture.
  • the camera 203 is calibrated with the outside world recognition sensor 201, the distance to the inspection target in the image can be grasped.
  • the position estimation sensor 202 is a GPS (Global Positioning System) or DGPS (Differential GPS) generally used for car navigation systems and smartphones, and is a sensor that acquires the position of the automatic inspection device 200 with an accuracy of 50 to 100 cm. be.
  • the position estimation sensor 202 is cheaper than the position estimation sensor 103 and is inferior in accuracy (equal or less).
  • the position estimation sensor 202 can acquire the position of the automatic inspection device 200 as position information.
  • the information storage unit 204 is, for example, a memory such as a ROM or RAM, which acquires the first sensor information, converts it into a physical quantity, and stores it. Then, the information storage unit 204 also stores the inspection map and the control signal for controlling the autonomous traveling of the automatic inspection device 200.
  • the information storage unit 204 may store the second sensor information and the first sensor information by using the information storage unit 204 without mounting the information storage unit 104 on the map generation device 100.
  • the first sensor information stored in the information storage unit 204 is transmitted to the center or the like via the network, and the first sensor information is collected. It is also possible to mount the storage device on the automatic inspection device 200 and collect the first sensor information without using the network.
  • the arithmetic processing unit 205 generates a control signal for executing the following actions.
  • the arithmetic processing unit 205 generates a control signal for executing the following actions. -The act of operating the camera 203 based on the position of the inspection target and the position of the automatic inspection device 200 of the inspection map to take an image of the inspection target.
  • the arithmetic processing unit 205 generates a control signal for the automatic inspection device 200 to execute the automatic inspection, that is, executes the arithmetic processing for executing the automatic inspection.
  • the arithmetic processing unit 205 For the arithmetic processing unit 205, a combination of a CPU, ROM and RAM, an FPGA which is a rewritable logic circuit, an ASIC which is an integrated circuit for a specific application, and the like can be used. Further, the arithmetic processing unit 205 is composed of a combination of a CPU, a ROM and a RAM, a combination of a CPU, a ROM, a RAM and an FPGA, and the like.
  • the control unit 206 operates the drive mechanism (not shown) of the automatic inspection device 200 based on the control signal generated by the arithmetic processing unit 205.
  • the automatic inspection device 200 described in the first embodiment has a attachment / detachment mechanism of the map generation device 100, and the map generation device 100 is attached at the time of map generation, and the map generation device 100 is removed at the time of automatic inspection.
  • the automatic inspection device 200 includes a camera 203 that captures an image of the surrounding inspection target, an external world recognition sensor (first external world recognition sensor) 201 that acquires three-dimensional information of the surroundings, and a position estimation sensor (1st external world recognition sensor) 201 that acquires the position.
  • the first position estimation sensor) 202, the information storage unit 204 that stores the first sensor information and the inspection map of the external world recognition sensor 201 and the position estimation sensor 202, and the first sensor information are used to perform an automatic inspection. It has an arithmetic processing unit 205 that executes arithmetic processing, and a control unit 206 that controls autonomous travel of the automatic inspection device 200 based on a control signal.
  • the map generation device 100 acquires in detail the camera 101 that captures an image of the inspection target around the automatic inspection device 200 and the three-dimensional information around the automatic inspection device 200, and has higher accuracy than the external world recognition sensor 201.
  • External world recognition sensor (second external world recognition sensor) 102 and position estimation sensor (second position estimation sensor) 103 that acquires the position of the automatic inspection device 200 in detail and is more accurate than the position estimation sensor 202, and external world recognition. It has an information storage unit 104 that stores the second sensor information of the sensor 102 and the position estimation sensor 103, and an inspection map generation unit 105 that generates an inspection map.
  • an expensive map generator 100 that generates a highly accurate inspection map required for autonomous driving is made independent as an attachment, and the map generator 100 is attached to the automatic inspection device 200 only at the time of map generation.
  • the map generator 100 is attached to the automatic inspection device 200 only at the time of map generation.
  • FIG. 3 is an explanatory diagram for explaining the inspection map described in the first embodiment.
  • the inspection map is installed on the paved road 303 and the paved road 303 in the inspection site where pedestrians, automobiles, etc. walk and travel, and exists in the standard traveling route 304 and the inspection site for the automatic inspection device 200 to autonomously travel.
  • the point group ( ⁇ : black circle) 301 at the time of map creation and the inspection target ( ⁇ : white circle) 302 in the infrastructure equipment 300 for taking a picture with the camera 203, which are acquired from the infrastructure equipment 300 and the outside world recognition sensor 102, are displayed. Will be done.
  • the automatic inspection device 200 uses this inspection map to execute the automatic inspection.
  • the inspection target 302 is displayed on the inspection map. As a result, automatic inspection can be performed with high efficiency.
  • FIG. 4 is an explanatory diagram illustrating an inspection map generation unit 105 in the map generation device 100 described in the first embodiment.
  • the inspection map generation unit 105 generates an inspection map composed of a point cloud 301, and has a route generation function 106 and an inspection target setting function 107.
  • the route generation function 106 and the inspection target setting function 107 are functions that are used after the map generation device 100 generates an inspection map.
  • the route generation function 106 sets a normative travel route 304 for the automatic inspection device 200 to autonomously travel on the inspection map.
  • the automatic inspection device 200 When setting the standard travel route 304, the automatic inspection device 200 may actually be used, the paved road 303 may be traveled by the automatic inspection device 200, and the traveled trajectory may be used as the standard travel route 304.
  • a person may manually generate a node and a link on the inspection map to use the normative travel route 304.
  • the inspection target setting function 107 sets the position of the inspection target 302 with respect to the inspection map.
  • the position of the inspection target 302 may be read from an existing sketch or design drawing in the inspection site and the position may be set on the inspection map, or the camera 101 may take a picture.
  • the image to be inspected may be used and the position may be manually set on the inspection map by a person.
  • the map generator 100 displays the point cloud on the inspection map on which the paved road is displayed, based on the position acquired from the position estimation sensor 103 and the point cloud acquired from the outside world recognition sensor 102.
  • An inspection map is generated by plotting.
  • the map generation device 100 has a route generation function 106 for setting a normative travel route for autonomous driving with respect to the inspection map, and an inspection target setting for setting the position of the inspection target with respect to the inspection map. It has a function 107.
  • the camera 101, the external world recognition sensor 102, and the position estimation sensor 103 are preferably calibrated.
  • the inspection target is specified in the image to be inspected by the camera 101. You can get the coordinates of the target.
  • FIG. 5 is a flowchart illustrating the storage of sensor information in the map generation described in the first embodiment.
  • the automatic inspection device 200 confirms that the map generation device 100 is attached (S101). ⁇ If the installation cannot be confirmed (No), the flow is terminated. -If the mounting can be confirmed (Yes), the second sensor information is collected (S102). Next, the second sensor information is stored (S103). -Finally, it is confirmed whether or not the storage completion button is pressed by the user (S104). -If pressed (Yes), the flow ends. -If not pressed (No), the process returns to S102.
  • FIG. 6 is a flowchart illustrating map generation described in the first embodiment.
  • the position of the automatic detection device 200 (the position of the own vehicle) is acquired from the position estimation sensor 103 (S201).
  • a corresponding point cloud acquired from the outside world recognition sensor 102 and having the same time is extracted (S202).
  • the coordinates are aligned based on the position of the automatic detection device 200 and the extracted point cloud, and the point cloud 301 is plotted on the inspection map on which the paved road 303 is displayed (S203).
  • S204 it is confirmed whether or not the map generation is completed (S204). -If it is not completed (No), the process returns to S201.
  • S201 to S203 are executed after executing S103. Then, the confirmation of S204 and the confirmation of S104 are the same conditions.
  • the flow shown in FIG. 6 is executed after the flow shown in FIG. 5 is completed.
  • S204 plots all the point clouds 301 and confirms whether or not the map generation is completed.
  • FIG. 7 is an explanatory diagram for explaining the automatic inspection using the inspection map described in the first embodiment.
  • the outside world recognition sensor 201 acquires the relationship (for example, distance) from the automatic inspection device 200 to the inspection target as a point cloud (gray circle) 401 at the time of automatic inspection.
  • the automatic inspection device 200 acquires (estimates) the position and autonomously travels. Further, when there is an inspection target around the automatic inspection device 200, the camera 203 is pointed at the inspection target in the direction 400 where the inspection target is, and an image of the inspection target is taken.
  • FIG. 8 is a flowchart illustrating the automatic inspection described in the first embodiment. -First, in the automatic inspection, it is determined whether or not the accuracy of the position estimation sensor 202 is sufficient (S301). Whether or not the accuracy is sufficient is determined based on a predetermined threshold value set in advance. -If the accuracy is sufficient, the position of the automatic detection device 200 is acquired (estimated) from the position estimation sensor 202 (S302). If the accuracy is not sufficient, the position of the automatic inspection device 200 is acquired (estimated) based on the inspection map and the point cloud 401 acquired from the outside world recognition sensor 201 (S303).
  • the amount of deviation between the acquired position of the automatic inspection device 200 and the standard traveling path 304 is calculated based on the amount of deviation between the point cloud 301 and the point group 401 (S304).
  • the control amount for controlling the autonomous driving of the automatic inspection device 200 is calculated so that the automatic inspection device 200 travels along the standard traveling path 304, and the control amount is used as a control signal.
  • Output (S305).
  • it is confirmed whether or not there is an inspection target in the vicinity based on the position of the automatic inspection device 200 and the inspection map in which the inspection target is stored (S306). -If there is an inspection target (Yes), the inspection target is photographed with the camera 203 (S307).
  • the process proceeds to the step of confirming whether or not the automatic inspection is completed. -Next, when there is no inspection target or when the inspection target is photographed, it is confirmed whether or not the automatic inspection is completed (whether or not all the inspection targets are photographed) (S308). -When the automatic inspection is completed (Yes), the automatic inspection device 200 automatically returns to the original position such as the standby place (S309). -If the automatic inspection is not completed (No), the process returns to S301.
  • the expensive map generation device 100 that generates the highly accurate inspection map required for autonomous driving is made independent as an attachment, and the map generation device 100 is automatically inspected only at the time of map generation. It is possible to provide an inexpensive and highly efficient automatic inspection device 200 by attaching it to the device 200 and removing it from the automatic inspection device 200 at the time of normal automatic inspection.
  • FIG. 9 is an explanatory diagram illustrating an inspection map generation unit 105 in the map generation device 100 described in the second embodiment.
  • the inspection map generation unit 105 described in the second embodiment is different from the inspection map generation unit 105 described in the first embodiment in that it has a map synthesis function 108.
  • the map composition function 108 extracts a nearby point cloud (data) from the data acquired from a certain data acquisition position and the data acquired from a certain data acquisition position, and if there is a nearby point cloud, there is a nearby point cloud. , These are collectively set as a point cloud 301.
  • the map generation device 100 has a map composition function 108 that acquires a plurality of point clouds acquired from a plurality of data acquisition positions, extracts nearby point clouds from these point clouds, and sets the point cloud 301.
  • FIG. 10 is an explanatory diagram for explaining the map generation described in the second embodiment.
  • the map generation device 100 is installed at the data acquisition position 500, and data is acquired within the data acquisition range 501. Then, the map generation device 100 acquires data at a plurality of installation locations.
  • the map generator 100 extracts, for example, a point cloud 3010 as a nearby point cloud in the data acquired from the data acquisition position 500A and the data acquired from the data acquisition position 500B, and puts them together. It is set as a point cloud 301.
  • the number of data to be extracted increases by increasing the number of places where the data acquisition position 500 is installed and increasing the range of the data acquisition range 501.
  • FIG. 11 is a flowchart illustrating map generation described in the second embodiment.
  • data is acquired from the outside world recognition sensor 102 (S401). For example, a point cloud acquired at the data acquisition position 500B.
  • S401 a point cloud acquired at the data acquisition position 500B.
  • S407 it is confirmed whether or not there is data acquired in the vicinity (S402).
  • S407 the data acquired in S401 is plotted on the inspection map (S407).
  • the data acquired in the vicinity is acquired (S403).
  • the data acquired in S401 and the data acquired in S403 are searched (S404).
  • the inspection map can be generated by installing the map generation device 100 on the paved road 303 and acquiring the data without autonomously driving the automatic inspection device 200. ..
  • the present invention is not limited to the above-described examples, and includes various modifications.
  • the above-described embodiment has been specifically described in order to explain the present invention in an easy-to-understand manner, and is not necessarily limited to those having all the described configurations.
  • 100 ... Map generator, 101 ... Camera, 102 ... External recognition sensor, 103 ... Position estimation sensor, 104 ... Information storage unit, 105 ... Inspection map generator, 106 ... -Route generation function, 107 ... Inspection target setting function, 108 ... Map composition function, 200 ... Automatic inspection device, 201 ... External recognition sensor, 202 ... Position estimation sensor, 203 ... Camera, 204 ... Information storage unit, 205 ... Arithmetic processing unit, 206 ... Control unit, 210 ... Controller, 300 ... Infrastructure equipment, 301 ... Point group, 302 ... Inspection Target, 303 ... paved road, 304 ... traveling route, 400 ... direction, 401 ... point group, 500 ... data acquisition position, 501 ... data acquisition range

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

L'invention concerne un dispositif d'inspection automatique à haut rendement et à faible coût, dans lequel un dispositif de génération de carte coûteux qui génère une carte d'inspection précise requise pour un déplacement autonome est fourni sous la forme d'un accessoire séparé. Le dispositif de génération de carte est fixé au dispositif d'inspection automatique uniquement pendant la génération de carte et détaché du dispositif d'inspection automatique pendant une inspection automatique normale. Le dispositif d'inspection automatique selon la présente invention est caractérisé en ce qu'il comprend : une caméra qui capture une image d'une cible d'inspection à proximité ; un premier capteur de reconnaissance d'environnement qui acquiert des informations tridimensionnelles concernant l'environnement ; un premier capteur d'estimation d'emplacement qui acquiert un emplacement ; une unité de stockage d'informations qui stocke une carte d'inspection et des premières informations de capteur concernant le premier capteur de reconnaissance d'environnement et le premier capteur d'estimation d'emplacement ; une unité de traitement de calcul qui utilise les premières informations de capteur pour réaliser un traitement de calcul pour une inspection automatique ; et une unité de commande qui commande le déplacement autonome. Le dispositif d'inspection automatique est également caractérisé en ce qu'il comprend un mécanisme de d'attache/détachement pour un dispositif de génération de carte qui génère une carte d'inspection.
PCT/JP2020/034430 2020-03-23 2020-09-11 Dispositif d'inspection automatique WO2021192357A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2020050822A JP7235691B2 (ja) 2020-03-23 2020-03-23 自動検査装置
JP2020-050822 2020-03-23

Publications (1)

Publication Number Publication Date
WO2021192357A1 true WO2021192357A1 (fr) 2021-09-30

Family

ID=77848494

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/034430 WO2021192357A1 (fr) 2020-03-23 2020-09-11 Dispositif d'inspection automatique

Country Status (2)

Country Link
JP (1) JP7235691B2 (fr)
WO (1) WO2021192357A1 (fr)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116224462A (zh) * 2022-10-28 2023-06-06 同方威视技术股份有限公司 检查设备的自主检查方法、装置和电子设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4980606B2 (ja) * 2005-11-15 2012-07-18 株式会社岩根研究所 移動型自動監視装置
US20180188026A1 (en) * 2016-12-30 2018-07-05 DeepMap Inc. Visual odometry and pairwise alignment for high definition map creation
JP2018133010A (ja) * 2017-02-17 2018-08-23 三菱重工業株式会社 屋内空間の検査方法
JP2018160228A (ja) * 2017-03-21 2018-10-11 株式会社東芝 経路生成装置、経路制御システム、及び経路生成方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100823739B1 (ko) * 2006-12-08 2008-04-21 한국전자통신연구원 주변 환경 변화에 신속하게 적응하여 환경 지도를 작성할수 있는 이동체의 환경 지도 작성 장치 및 그 방법
KR101591471B1 (ko) * 2008-11-03 2016-02-04 삼성전자주식회사 물체의 특징 정보를 추출하기 위한 장치와 방법, 및 이를 이용한 특징 지도 생성 장치와 방법
JP6272460B2 (ja) * 2014-03-31 2018-01-31 株式会社日立産機システム 三次元地図生成システム
KR102373926B1 (ko) * 2016-02-05 2022-03-14 삼성전자주식회사 이동체 및 이동체의 위치 인식 방법
EP4029372A1 (fr) * 2016-12-15 2022-07-20 Positec Power Tools (Suzhou) Co., Ltd. Procédé de retour de dispositif automoteur, dispositif automoteur, support de stockage et serveur

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4980606B2 (ja) * 2005-11-15 2012-07-18 株式会社岩根研究所 移動型自動監視装置
US20180188026A1 (en) * 2016-12-30 2018-07-05 DeepMap Inc. Visual odometry and pairwise alignment for high definition map creation
JP2018133010A (ja) * 2017-02-17 2018-08-23 三菱重工業株式会社 屋内空間の検査方法
JP2018160228A (ja) * 2017-03-21 2018-10-11 株式会社東芝 経路生成装置、経路制御システム、及び経路生成方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
SUJIWO, ADI ET AL.: "Localization Based on Multiple Visual-Metric Maps", 2017 IEEE INTERNATIONAL CONFERENCE ON MULTISENSOR FUSION AND INTEGRATION FOR INTELLIGENT SYSTEMS (MFI 2017, 16 November 2017 (2017-11-16), pages 212 - 219, XP033264835, DOI: 10.1109/MFI.2017.8170431 *

Also Published As

Publication number Publication date
JP2021148698A (ja) 2021-09-27
JP7235691B2 (ja) 2023-03-08

Similar Documents

Publication Publication Date Title
US11927960B2 (en) Control device, control method, and computer program
JP7078093B2 (ja) 情報処理装置、情報処理方法及び情報処理プログラム
JP7452743B2 (ja) 飛行情報制御方法、制御装置及びコンピュータプログラム
CN109664301B (zh) 巡检方法、装置、设备及计算机可读存储介质
JP6975513B2 (ja) カメラベースの自動化された高精度道路地図の生成システム及び方法
CN111813130A (zh) 一种输变电站智能巡检机器人自主导航避障系统
CN111837136A (zh) 基于本地感测的自主导航以及相关联的系统和方法
CN110618436A (zh) 基于即时定位与地图构建的巡检方法、装置及设备
CN112447058B (zh) 泊车方法、装置、计算机设备及存储介质
CN115793649A (zh) 一种电缆沟自动巡检装置及巡检方法
WO2021192357A1 (fr) Dispositif d'inspection automatique
JP2024020389A (ja) 制御方法、制御装置及びプログラム
JP2020201850A (ja) 飛行体の制御方法
JP2018174002A (ja) 移動体
Sokolov et al. Coordinated use of visual odometry and landmarks for navigation of mobile ground vehicles
US20240184290A1 (en) Control device, control method, and computer program
KR20230101586A (ko) 농업용 자율 주행 차량 관제 장치 및 방법
JP2023016267A (ja) 保守支援システムおよび保守支援装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20927077

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20927077

Country of ref document: EP

Kind code of ref document: A1