Autonomous driving device for calibrating sensor parameter and calibration method thereof

Abstract

An autonomous driving device is disclosed. The present autonomous device includes: a communication unit; a sensor for performing sensing based on at least one sensor parameter; a memory; and a processor, wherein the processor: stores, in the memory, first observation information acquired by observing objects existing in a space in which the autonomous driving device is located on the basis of a sensing value of the sensor; receives second observation information and driving history information of at least one external autonomous driving device from the at least one external autonomous driving device among the objects existing in the space through the communication unit and stores same in the memory; and calibrates at least one sensor parameter by assigning a weight to the first observation information and the second observation information stored in the memory on the basis of driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device.

Description

Autonomous driving device for calibrating sensor parameters and calibration method thereof

[0001] The present disclosure relates to an autonomous driving device for calibrating sensor parameters and a calibration method thereof, and more particularly, to an autonomous driving device for calibrating sensor parameters by acquiring various observation information from the outside and a calibration method thereof.

[0002] Autonomous driving devices are currently being used in a variety of environments. Autonomous driving devices are devices that can autonomously determine their own path and drive without the need for a human driver or remote control. Technologies can be used to create a map of the space in which the device is located using sensors such as lidar and cameras, and to identify its location on the map.

[0003] For accurate mapping and localization, sensor parameters must be properly adjusted. Sensor parameters can be various variables that influence the sensing behavior of sensors installed in the device. However, when using an autonomous driving device, it may encounter obstacles or be subjected to other external forces. In these cases, the position and orientation of the autonomous driving device's sensors may become misaligned, leading to inaccurate localization and map generation.

[0004] At this time, sensor parameters must be properly calibrated. In the past, methods were used to calibrate sensor parameters using surrounding objects as landmarks or to calibrate solely using the relationship with the opposing autonomous driving device. However, when using surrounding objects as landmarks, accurate sensor parameter calibration required a large number of landmarks or the landmarks had to be photographed from various angles, resulting in many inconveniences for users operating autonomous driving devices. Furthermore, the method of calibrating sensor parameters solely using the relationship with the opposing autonomous driving device encountered the problem that if the sensor parameters of two or more autonomous driving devices were not all accurate, it would be difficult to accurately calibrate the sensor parameters even if the relationship with the opposing autonomous driving device was utilized. Therefore, there has been a growing need for a technology that can automatically calibrate the sensor parameters of autonomous driving devices and accurately calibrate the sensor parameters using multiple autonomous driving devices even if there is a problem with the sensor parameters of one of the multiple autonomous driving devices.

[0005] An autonomous driving device according to at least one embodiment of the present disclosure includes a communication unit, a sensor for performing sensing based on at least one sensor parameter, a memory, and a processor for storing first observation information obtained by observing objects existing in a space where the autonomous driving device is located based on a sensing value of the sensor in the memory, receiving second observation information from at least one external autonomous driving device among objects existing in the space and driving history information of the at least one external autonomous driving device through the communication unit and storing the second observation information in the memory, and for correcting the at least one sensor parameter by assigning weights to the first observation information and the second observation information stored in the memory based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device.

[0006] A method for calibrating an autonomous driving device according to at least one embodiment of the present disclosure includes: storing first observation information obtained by observing objects existing in a space where the autonomous driving device is located based on a sensing value of a sensor; receiving and storing second observation information from at least one external autonomous driving device among the objects existing in the space and driving history information of the at least one external autonomous driving device; and calibrating at least one sensor parameter of the sensor by assigning weights to the stored first observation information and the second observation information based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device.

[0007] A computer-readable recording medium including a program for executing a method for calibrating an autonomous driving device according to at least one embodiment of the present disclosure, wherein the method for calibrating an autonomous driving device includes: a step of storing first observation information obtained by observing objects existing in a space where the autonomous driving device is located based on a sensing value of a sensor; a step of receiving and storing second observation information from at least one external autonomous driving device among the objects existing in the space and driving history information of the at least one external autonomous driving device; and a step of calibrating at least one sensor parameter of the sensor by assigning weights to the stored first observation information and the second observation information based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device.

[0008] FIG. 1 is a diagram schematically illustrating the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0009] FIG. 2 is a block diagram illustrating a configuration of an autonomous driving device according to one or more embodiments of the present disclosure.

[0010] FIG. 3 is a diagram illustrating the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0011] FIG. 4 is a diagram for explaining a method for generating location information of an autonomous driving device according to one or more embodiments of the present disclosure.

[0012] FIG. 5 is a diagram illustrating a method for calibrating sensor parameters of an autonomous driving device according to one or more embodiments of the present disclosure.

[0013] FIG. 6 is a diagram for explaining the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0014] FIG. 7 is a diagram illustrating a weighting method of an autonomous driving device according to one or more embodiments of the present disclosure.

[0015] FIG. 8 is a flowchart illustrating a method for calibrating sensor parameters of an autonomous driving device according to one or more embodiments of the present disclosure.

[0016] Hereinafter, the present disclosure will be described in detail with reference to the attached drawings.

[0017] The terms used in the embodiments of this disclosure have been selected from widely used, current terms, taking into account the functions of this disclosure. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, and in such cases, their meanings will be described in detail in the description of the relevant disclosure. Therefore, the terms used in this disclosure should not be defined simply as names of terms, but rather based on the meanings of the terms and the overall content of this disclosure.

[0018] In this specification, expressions such as “has,” “can have,” “includes,” or “may include” indicate the presence of a feature (e.g., a number, function, operation, or component such as a part), and do not exclude the presence of additional features.

[0019] The expression "at least one of A and / or B" should be understood to mean either "A" or "B" or "A and B".

[0020] As used herein, the expressions “first,” “second,” “first,” or “second,” etc., may describe various components, regardless of order and / or importance, and are only used to distinguish one component from another, but do not limit the components.

[0021] When it is said that a component (e.g., a first component) is “(operatively or communicatively) coupled with / to” or “connected to” another component (e.g., a second component), it should be understood that the component may be directly coupled to the other component, or may be connected through another component (e.g., a third component).

[0022] Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, terms such as "comprise" or "consist of" are intended to indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the specification, but should be understood not to preclude the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0023] In the present disclosure, a "module" or "part" performs at least one function or operation and may be implemented as hardware or software, or as a combination of hardware and software. Furthermore, multiple "modules" or multiple "parts" may be integrated into at least one module and implemented as at least one processor (not shown), excluding any "modules" or "parts" that need to be implemented as specific hardware.

[0024] In this specification, the term user may refer to a person using an electronic device or a device using an electronic device (e.g., an artificial intelligence electronic device).

[0025] An embodiment of the present disclosure will be described in more detail with reference to the attached drawings below.

[0026] FIG. 1 is a diagram schematically illustrating the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0027] An autonomous driving device (100) can be a device capable of autonomous driving without human intervention. The autonomous driving device (100) can be implemented as various types of devices, such as an autonomous driving robot, an autonomous driving car, an autonomous driving drone, a vacuum cleaner, a serving robot, a mobile projector, or a mobile speaker.

[0028] According to FIG. 1, the autonomous driving device (100) can observe various external objects existing within the space in which the autonomous driving device (100) is located. Specifically, the autonomous driving device (100) can observe external autonomous driving devices (200-1, 200-2) existing within the same space, and can observe structures such as walls, doors, ceilings, and pillars constituting the space, as well as various objects existing within the space. Structures or objects existing at fixed locations within the space can be used as landmarks (300).

[0029] In the present disclosure, the space may refer to the surrounding environment in which the autonomous driving device (100) is located. If the autonomous driving device is a car or a drone, various external environments such as roads and the sky may be included in the space. If the autonomous driving device is implemented in the form of a home appliance such as a vacuum cleaner, a portable projector, or a portable speaker, the space may refer to a general household environment or an indoor environment. A landmark (300) may refer to an identifiable point or object that represents the surrounding environment of the autonomous driving device (100) and helps the autonomous driving device estimate its own location. For example, a landmark (300) may refer to various objects having characteristics that distinguish it from the surrounding environment, such as a wall, door, pillar, or object fixed at a specific location within the space.

[0030] An autonomous driving device (100) can drive while avoiding collisions with other external objects within a space based on the sensing values ​​of the sensors. The autonomous driving device (100) can generate a map of the internal structure of the space based on the sensing values ​​of the sensors, determine a driving path based on the map, and drive along the driving path. In order to avoid collisions with external objects and appropriately set a driving path to a target point, the autonomous driving device (100) must be able to accurately generate a map of the space in which it is located and accurately determine its own location information. Sensor parameters can be classified into extrinsic parameters and intrinsic parameters of the sensor. Extrinsic parameters may include the position and direction of the sensor, and intrinsic parameters may include the FOV of the lidar sensor, the focal length and principal point of the camera, etc.

[0031] The position, direction, principal point, focal length, etc. of the sensor of the autonomous driving device (100) may change due to various forces such as inertia and gravity that occur when the autonomous driving device collides with an external object, an external force applied by the user, rapid driving or sudden stopping, or aging of the parts. In this case, the position, direction, etc. of the sensor change, but if the user does not directly change the sensor parameter value by the amount of change in the position and direction of the sensor, the sensor parameter will maintain the existing value. In this case, the autonomous driving device (100) cannot help but generate an inaccurate map and determine the location, and thus requires correction of the sensor parameter.

[0032] Existing autonomous driving devices have been calibrating sensor parameters by having the user directly calibrate them or by comparing sensing data from multiple sensors equipped on a single autonomous driving device to calibrate the sensor parameters.

[0033] An autonomous driving device (100) according to one or more embodiments of the present disclosure can comprehensively consider not only observation information observed by the autonomous driving device (100) itself about its surroundings but also observation information observed by external autonomous driving devices (200-1, 200-2) about its surroundings. Accordingly, the autonomous driving device can automatically correct the sensor parameters of its own sensors. The observation information may be information about the detection result of sensing the surroundings using at least one sensor. Alternatively, the observation information may be referred to as sensing information, detection information, etc., but is described as observation information in the present disclosure. In addition, for the convenience of explanation, the observation information observed by the autonomous driving device itself is referred to as first observation information, and the observation information observed by the external autonomous driving device is referred to as second observation information.

[0034] Specifically, the autonomous driving device (100) can obtain second observation information from at least one external autonomous driving device (200-1, 200-2), assign weights to the first observation information and the second observation information observed by the autonomous driving device, and automatically correct the sensor parameters of the autonomous driving device based on the weighted information. Accordingly, even if the user does not regularly check whether there is a problem with the sensor of the autonomous driving device (100), the autonomous driving device (100) can perform accurate map generation and location identification, thereby enabling efficient work performance.

[0035] In Fig. 1, the external autonomous driving device (200-1, 200-2) is illustrated as an autonomous driving device of the same type as the autonomous driving device (100), but this is only one example, and even if the external autonomous driving device (200-1, 200-2) is an autonomous driving car and the autonomous driving device (100) is an autonomous driving robot, and the autonomous driving devices of different types are present, it goes without saying that they can exchange observation information with each other and automatically correct sensor parameters.

[0036] FIG. 2 is a block diagram illustrating a configuration of an autonomous driving device according to one or more embodiments of the present disclosure.

[0037]

[0038] *Referring to FIG. 2, the autonomous driving device (100) may include a communication unit (110), a sensor (120), a memory (130), and a processor (140).

[0039] The communication unit (110) is configured to communicate with various external devices. The communication unit (110) can receive observation information, i.e., second observation information, from at least one external autonomous driving device. In addition, the communication unit (110) can transmit observation information, i.e., first observation information, of the autonomous driving device (100) to at least one external autonomous driving device under the control of the processor (140).

[0040] The communication unit (110) can be connected to at least one external autonomous driving device using a communication method such as Bluetooth, AP-based Wi-Fi (Wireless LAN network), Zigbee, wired / wireless LAN (Local Area Network), WAN (Wide Area Network), Ethernet, IEEE 1394, HDMI (High-Definition Multimedia Interface), USB (Universal Serial Bus), MHL (Mobile High-Definition Link), AES / EBU (Audio Engineering Society / European Broadcasting Union), optical, coaxial, etc.

[0041] In addition to observation information, the communication unit (110) can also transmit and receive driving history information. Driving history information may include various information related to the driving of the autonomous driving device.

[0042] The sensor (120) is a configuration for performing sensing based on at least one sensor parameter with respect to the space in which the autonomous driving device (100) is located. The sensor (120) may include various sensors such as a camera, an IMU sensor, an encoder, a lidar, etc. that enable the autonomous driving device (100) to create a map of the space or determine the location of the autonomous driving device (100).

[0043] In FIG. 2, one sensor (120) is illustrated, but a plurality of different sensors may be used depending on the type, size, shape, usage environment, etc. of the autonomous driving device (100).

[0044] The memory (130) is configured to store commands, an operating system, and application programs or related data for controlling the overall operation of the autonomous driving device (100). The memory (130) can store observation information acquired based on the sensing values ​​and sensor parameters of the sensor (120), observation information acquired from an external autonomous driving device through the communication unit (110), and driving history information of the external autonomous driving device.

[0045] Here, the observation information may include location information that can be obtained by performing calculations using sensor parameters on the sensing values ​​of the sensor (120). The driving history information of the external autonomous driving device may include information such as the number of collisions with obstacles and the driving time in space. The observation information and driving history information will be described in detail in the following sections.

[0046] The memory (130) according to one or more embodiments of the present disclosure may be implemented as an internal memory such as a ROM (e.g., an electrically erasable programmable read-only memory (EEPROM)) or RAM included in the processor (140), or may be implemented as a separate memory from the processor (140). In this case, the memory (130) may be implemented as a memory embedded in the autonomous driving device (100) or as a memory detachable from the autonomous driving device (100) depending on the purpose of data storage. For example, data for driving the autonomous driving device (100) may be stored in a memory embedded in the autonomous driving device (100), and data for expanding functions of the autonomous driving device (100) may be stored in a memory detachable from the autonomous driving device (100).

[0047] Meanwhile, in the case of memory embedded in the autonomous driving device (100), it is implemented as at least one of volatile memory (e.g., dynamic RAM (DRAM), static RAM (SRAM), or synchronous dynamic RAM (SDRAM)), non-volatile memory (e.g., one time programmable ROM (OTPROM), programmable ROM (PROM), erasable and programmable ROM (EPROM), electrically erasable and programmable ROM (EEPROM), mask ROM, flash ROM, flash memory (e.g., NAND flash or NOR flash), hard drive, or solid state drive (SSD)), and in the case of memory that can be attached or detached to the autonomous driving device (100), it is implemented as a form of memory card (e.g., compact flash (CF), secure digital (SD), micro secure digital (Micro-SD), mini secure digital (Mini-SD), extreme digital (xD), multi-media card (MMC), etc.), external memory that can be connected to a USB port (e.g., USB memory), etc. Can be.

[0048] The processor (140) controls the overall operation of the autonomous driving device (100). Specifically, the processor (140) is connected to a communication unit (110), a sensor (120), and a memory (130), and can perform various operations by executing at least one command stored in the memory (130).

[0049] The processor (140) may be implemented as a digital signal processor (DSP) that processes digital signals, a microprocessor, but is not limited thereto, and may include one or more of a central processing unit (CPU), a micro controller unit (MCU), a micro processing unit (MPU), a controller, an application processor (AP), a communication processor (CP), an ARM processor, and an artificial intelligence (AI) processor, or may be defined by the relevant terms. In addition, the processor (140) may be implemented as a system on chip (SoC) or large scale integration (LSI) having a built-in processing algorithm, or may be implemented in the form of a field programmable gate array (FPGA). The processor (140) may perform various functions by executing computer executable instructions stored in the memory (130).

[0050] When a method according to various embodiments of the present disclosure includes multiple operations, the multiple operations may be performed by one processor or by multiple processors. The processor (140) may be implemented as a single core processor including one core, or may be implemented as one or more multicore processors including multiple cores (e.g., homogeneous multicores or heterogeneous multicores).

[0051] The processor (140) can obtain observation information by observing objects existing within the space where the autonomous driving device (100) is located based on the sensing value of the sensor (120). Here, the objects existing within the space may include an external autonomous driving device, objects fixed at a specific location within the space (e.g., walls, furniture, home appliances, etc.), objects whose locations change within the space (e.g., people or animals), etc.

[0052] For example, if the sensor (120) includes a camera, the processor (140) can capture an object through the camera and obtain observation information, including object location information, through the captured image of the object. In this case, the object location information can be obtained through calculations based on the camera's external and internal parameters.

[0053] The external parameters of the camera include the position where the camera is installed on the autonomous driving device, the direction in which the camera is installed, etc., and the processor (140) can perform an operation to convert from the world coordinate system to the camera coordinate system based on the external parameters of the camera, and an operation to convert from the camera coordinate system to the global coordinate system. Here, the global coordinate system is a coordinate system that is generally used as a reference when expressing the position of an object, and means a coordinate system that calculates the coordinate values ​​of a specific position based on the X-axis, Y-axis, and Z-axis by taking any point in space as the origin. In addition, the camera coordinate system means a coordinate system that calculates the coordinate values ​​based on the camera, and means a coordinate system that calculates the coordinate values ​​of a specific position based on the X-axis, Y-axis, and Z-axis by taking the location of the camera as the origin. If the processor (140) accurately identifies the external parameter values ​​of the camera, the coordinate values ​​in the camera coordinate system can be accurately converted to coordinate values ​​in the world coordinate system, and the coordinate values ​​in the world coordinate system can also be accurately converted to coordinate values ​​in the camera coordinate system.

[0054] The internal parameters of the camera include the principal point, focal length, etc. of the camera, and the processor (140) can perform an operation to convert from the image coordinate system to the camera coordinate system and an operation to convert from the camera coordinate system to the image coordinate system based on the internal parameters of the camera. Here, the image coordinate system refers to a coordinate system that calculates the coordinate values ​​of a specific position based on the X-axis and Y-axis by taking a specific point as the origin in a 2D image captured by the camera. If the processor (140) accurately understands the internal parameters of the camera, the coordinate values ​​in the image coordinate system can be accurately converted into coordinate values ​​in the camera coordinate system, and the coordinate values ​​in the camera coordinate system can also be accurately converted into coordinate values ​​in the image coordinate system.

[0055] Accordingly, the processor (140) can obtain the coordinate values ​​(a, b) in the image coordinate system of a specific object existing in an image captured by the camera, and then convert (a, b) into coordinate values ​​(a', b', c') in the camera coordinate system through an operation based on the internal parameters of the camera. In addition, the processor (140) can convert (a', b', c') into coordinate values ​​(a'', b'', c'') in the global coordinate system through an operation based on the external parameters of the camera. By obtaining the coordinate values ​​(a'', b'', c'') in the global coordinate system of the specific object, the processor (140) can obtain observation information about the specific object.

[0056] The above description only describes a method of obtaining observation information on objects existing in space through a camera, but this is only one example, and observation information can also be obtained using various other sensors. For example, when the sensor (120) includes a lidar sensor, observation information on external objects can be obtained based on external parameters such as the position at which the lidar sensor is mounted on the device, the direction in which the lidar sensor is mounted, and internal parameters such as the field of view (FOV) of the lidar sensor, the firing angle of the laser, and the reception intensity of the laser.

[0057] Although the above description only describes a method for obtaining location information about an external object existing in space, the processor (140) may also estimate location information of the autonomous driving device (100) based on an external object whose location information is already known. For example, if the coordinate values ​​of the external object are already known, the distance between the autonomous driving device (100) and the external object can be determined through the sensing value of the sensor (120), and the direction in which the autonomous driving device (100) is looking at the external object can also be determined based on the external parameters and internal parameters of the sensor (120). Therefore, the coordinate values ​​of the autonomous driving device (100) in the global coordinate system can be calculated based on the coordinate values ​​of the external object through an operation based on the sensor parameters.

[0058] The processor (140) can store the observation information obtained in the same manner as described above, i.e., the first observation information, in the memory (130).

[0059] In addition, the processor (140) can receive observation information of an external autonomous driving device existing in space, i.e., second observation information, through the communication unit (110) and store it in the memory (130). The second observation information of the external autonomous driving device is observation information acquired by the external autonomous driving device through a sensor in the same manner as described above.

[0060] Specifically, the first observation information may include location information of the autonomous driving device (100), location information of a fixed object within a space observed by the autonomous driving device (100), location information of an external autonomous driving device observed by the autonomous driving device (100), etc. The second observation information may include location information of an external autonomous driving device, location information of a fixed object within a space observed by the external autonomous driving device, location information of the autonomous driving device (100) observed by the external autonomous driving device, etc.

[0061] In addition, the processor (140) can obtain driving history information of the external autonomous driving device from the external autonomous driving device through the communication unit (110). Here, the driving history information can include various information related to the driving of the external autonomous driving device, such as the driving distance of the external autonomous driving device, the driving time, the number of collisions with obstacles, the intensity of collisions with obstacles, and the history of parameter correction of sensors mounted on the external autonomous driving device. Here, obstacles can include all objects that impede the driving of the autonomous driving device, such as walls, people, and other autonomous driving devices.

[0062] In addition, the processor (140) may assign a weight to the first observation information based on the driving history information of the autonomous driving device (100), and may assign a weight to the second observation information acquired from the external autonomous driving device based on the driving history information of the external autonomous driving device. Taking Fig. 1 as an example, the processor (140) may obtain observation information (A1) and driving history information (A2) from the external autonomous driving device (200-1) through the communication unit (110), and may obtain observation information (B1) and driving history information (B2) from the external autonomous driving device (200-2).

[0063] Typically, as driving time increases, the likelihood that the sensor status of the autonomous driving device will deviate from its initial settings increases. Accordingly, the processor (140) may compare driving time among driving history information and assign greater weight to observation information from devices with shorter driving periods than to observation information from devices with longer driving periods.

[0064] For example, if the driving time in the driving history information of the autonomous driving device (100) is 10 days, the driving time in the driving history information (A2) is 10 days, and the driving time in the driving history information (B2) is 40 days, the processor (140) may assign a greater weight to the observation information and the observation information (A1) of the autonomous driving device (100) than to the observation information (B1). The weight is a numerical value assigned to the observation information, and may be a numerical value indicating how much importance the corresponding observation information has in performing the correction task of the sensor parameter. The weight may be expressed in units of percentage or decimal. For example, the observation information and the observation information (A1) of the autonomous driving device (100) may be assigned a weight of 44.5% or 0.445, respectively, and the observation information (B1) may be assigned a weight of 11% or 0.11. In the above example, the weight is set to approximately 4 times the size considering that the driving time is 4 times longer. However, the difference in driving time and the weight do not necessarily need to be proportional and can be determined at various ratios.

[0065] In addition, although the above description explains that the observation information of each external autonomous driving device can be weighted based on the driving time of each external autonomous driving device, the driving time of each external autonomous driving device is only an example of the driving history information of the external autonomous driving device, and the weight can also be assigned based on driving history information such as the number of times the external autonomous driving device collided with an obstacle existing in the space, the intensity of the collision with the obstacle, etc., and of course, the weight can be assigned using all of the information such as the driving time of the external autonomous driving device, the number of times the external autonomous driving device collided with an obstacle existing in the space, and the intensity of the collision with the obstacle. The processor (140) can assign a lower weight as the number of collisions increases or the intensity of the collision increases.

[0066] The processor (140) may estimate the intensity of the collision arithmetically by considering the speed and direction of movement of the autonomous driving device at the time of collision with an external object, the speed and direction of movement of the external object, etc., or may directly sense the intensity of the collision based on the sensing value of the pressure sensor, or may estimate the intensity of the collision based on the change in the position of the autonomous driving device (100) before and after the collision.

[0067] The processor (140) can correct the sensor parameters based on the first observation information and the second observation information to which weights are assigned by the weight assignment method described above. As described above, the processor (140) must accurately identify the sensor parameter values ​​corresponding to the state of the sensor (120), such as the current position and direction of the sensor (120), in order to determine accurate location information, and therefore performs the task of correcting the sensor parameters. The processor (140) can obtain more reliable information through the weight assignment method described above and can correct the sensor parameters based on the highly reliable information. The method for correcting the sensor parameters will be described in detail in the following section.

[0068] FIG. 3 is a diagram illustrating the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0069] According to FIG. 3, when the autonomous driving device (100) and the external autonomous driving device (200-1, 200-2) are observing the same landmark (300), the autonomous driving device (100) can share observation information with the external autonomous driving device (200-1, 200-2).

[0070] The autonomous driving device (100) can obtain observation information about a landmark (300) based on the sensing value and sensor parameters of the sensor (120), and the external autonomous driving devices (200-1, 200-2) can also obtain observation information about the landmark (300) through sensors mounted on each external autonomous driving device.

[0071] At this time, whenever the autonomous driving device (100) and the external autonomous driving device (200-1, 200-2) exist together within a certain range from the landmark (300), or whenever the autonomous driving device (100) and the external autonomous driving device (200-1, 200-2) encounter each other, or whenever the external autonomous driving device (200-1, 200-2) exists within a certain distance range from the autonomous driving device (100), the autonomous driving device (100) and the external autonomous driving device (200-1, 200-2) can share observation information with each other through the communication unit (110).

[0072] The processor (140) can generate observation information that the landmark (300) exists at (10, 0). In addition, the first external autonomous driving device (200-1) can generate observation information that the landmark (300) exists at (10, 4) through the installed sensor, and the second external autonomous driving device (200-2) can generate observation information that the landmark (300) exists at (8, 3). The processor (140) can generate the coordinate value of the landmark (300) based on the observation information (10, 0) of the autonomous driving device (100), the observation information (10, 4) of the first external autonomous driving device (200-1) received through the communication unit (110), and the observation information (8, 3) of the second external autonomous driving device (200-2). An example of the process of generating the coordinate value of the landmark (300) is as follows.

[0073] First, the processor (140) analyzes the driving history information of the first external autonomous driving device (200-1), the driving history information of the second external autonomous driving device (200-2), and the driving history information of the autonomous driving device (100) and assigns weights to them. For example, the driving history of the first external autonomous driving device (200-1) includes 1 number of collisions with obstacles and a driving time of 2 days, the driving history of the second external autonomous driving device (200-2) includes 10 number of collisions with obstacles and a driving time of 4 days, and the driving history of the autonomous driving device (100) includes 15 number of collisions with obstacles and a driving time of 10 days. The processor (140) determines that the first external autonomous driving device has the lowest number of collisions and the shortest driving time, and thus is the most reliable information, and assigns a weight of 70% to the observation information of the first external autonomous driving device (200-1), then assigns a weight of 20% to the observation information of the second external autonomous driving device (200-2), and then assigns a weight of only 10% to the observation information of the autonomous driving device (100), which is determined to be the least reliable information. The processor (140) can simply add up the weighted observation information to identify that the location of the landmark (300) is at (9.6, 3.4).

[0074] The processor (140) compares the weighted result (9.6, 3.4, 0.2) for the observation information on the landmark (300) with its own observation information (10, 0, 0) to calculate an offset value. The processor (140) corrects the existing sensor parameters so that the observation result can be compensated for by the calculated offset value.

[0075] In the above description, the location of the landmark (300) is calculated by simply adding up the coordinate values ​​of the landmark observed by each autonomous driving device by multiplying the above weights, but this is only an example, and the coordinates of the landmark can be calculated through various calculation methods, such as calculating the location of the landmark through a calculation process that takes into account the probability that the landmark exists at a specific location by considering Gaussian distribution, etc.

[0076] The above description describes a method in which three autonomous driving devices that observed a landmark (300) share observation information about the landmark (300) to calculate the coordinates of the landmark. However, this is only one example, and even if three or more autonomous driving devices observe the same landmark (300), they can share observation information with each other and calculate the coordinates of the landmark (300). In addition, even after calculating the coordinates of the landmark (300) once, the coordinates of the landmark (300) can be calculated again whenever observation information is shared with other external autonomous driving devices, and thus the sensor parameters can be re-calibrated.

[0077] The processor (140) can not only calculate the coordinate values ​​of the landmark (300) generated through the above-described weighting method and perform correction of the sensor parameters based on the calculated coordinate values ​​of the landmark (300), but also generate the location information of the autonomous driving device (100) based on the landmark (300). The processor (140) can share the generated location information with other autonomous driving devices and correct the sensor parameters through the shared observation information, which will be described in detail in the description of FIGS. 4 and 5 described below.

[0078] FIG. 4 is a diagram for explaining a method for generating location information of an autonomous driving device according to one or more embodiments of the present disclosure.

[0079] According to FIG. 4, the autonomous driving device (100) and the external autonomous driving device (200-1) can generate their own location information (501, 502) based on the landmark (300).

[0080] The autonomous driving device (100) can use the coordinate values ​​of the landmark (300) to determine its own location. For example, if the coordinate values ​​of the landmark (300) are calculated as (9.6, 3.4) in the manner described in FIG. 3, the processor (140) determines the distance and direction that the autonomous driving device (100) is away from the landmark based on the sensor (120) and the sensor parameters, and determines the location of the autonomous driving device (100) as (X i )(501) can be calculated. The external autonomous driving device (200-1) can also calculate its position (X) using the same method. j )(502) can be calculated. At this time, if the external autonomous driving device (200-1) is within a certain range from the autonomous driving device (100) or the external autonomous driving device (200-1) encounters the autonomous driving device (100), the two devices can observe each other and share observation information.

[0081] External autonomous driving device (200-1) (X j )(502), when the autonomous driving device (100) is observed from the position of the external autonomous driving device (200-1), the position of the autonomous driving device (100) observed by the external autonomous driving device (200-1) is the position (X) of the autonomous driving device (100) calculated by the processor (140) based on the sensing value and sensor parameters of the sensor (120) with respect to the landmark. i)(501). If the position information of the autonomous driving device (100) observed by the external autonomous driving device (200-1) received by the processor (140) from the external autonomous driving device (200-1) through the communication unit (110) is different from the position information of the autonomous driving device (100) calculated by the processor (140), the processor (140) can use the received position information for the correction of the sensor parameters. The operation of correcting the sensor parameters based on the observation information received from the external autonomous driving device (200-1) will be described in detail in the description of FIG. 5.

[0082] Although FIG. 4 illustrates only one external autonomous driving device (200-1), this is only one example, and the processor (140) may receive multiple location information from multiple external autonomous driving devices (200-1) and compare the multiple location information with the location information of the autonomous driving device (100) calculated by the processor (140) itself as described above.

[0083] FIG. 5 is a diagram illustrating a method for calibrating sensor parameters of an autonomous driving device according to one or more embodiments of the present disclosure.

[0084] According to Fig. 5, (X j )(502) The external autonomous driving device (200-1) present at the location can observe the autonomous driving device (100), and the processor (140) can receive information about the autonomous driving device (100) observed from the external autonomous driving device (200-1) through the communication unit (110).

[0085] The processor (140) is the autonomous driving device (100) itself (X i )(501), while the external autonomous driving device (200-1) is assumed to exist at the location of the autonomous driving device (100) (X i')(503). The location information estimated by the autonomous driving device (100) and the location information observed by the external autonomous driving device (200-1) of the autonomous driving device (100) may not match, and in the case where they do not match, the error occurring between the two pieces of information is E i (504) can be defined.

[0086] Location information of autonomous driving device (100) received from external autonomous driving device (200-1) (X i ')(503) is correct, the processor (140) slightly changes the sensor parameter values ​​and the above error E i (504) It is possible to perform a sensor parameter correction operation by finding a sensor parameter value that makes it 0. However, the position information (X) of the autonomous driving device (100) received from the external autonomous driving device (200-1) i ')(503) may not be accurate, and the autonomous driving device (100) cannot find out by itself whether the information is accurate unless the user directly checks it. Therefore, the processor (140) does not correct the sensor parameters based only on the observation information received from one external autonomous driving device (200-1), but stores the position information received from multiple external autonomous driving devices in the memory (130) and performs the sensor parameter correction operation only when sufficient position information is stored to perform the sensor parameter correction operation by assigning weights to each of the position information received from multiple external autonomous driving devices.

[0087] For example, an autonomous driving device (100) may sequentially receive location information from three external autonomous driving devices. At this time, since the processor (140) cannot directly know which of the three external autonomous driving devices received location information from which external autonomous driving device is accurate, it compares the driving history information of each of the multiple external autonomous driving devices and assigns weights to it, thereby assigning importance to the observation information (e.g., location information) received from each external autonomous driving device.

[0088] Here, importance is a measure of how much influence each observation information received from each of multiple autonomous driving devices will have when performing a calibration task for sensor parameters. For example, if observation information with a weight of 100% is received by the autonomous driving device (100), the processor (140) compares only its own observation information with the observation information with a weight of 100% without utilizing other observation information, calculates an offset value, and calibrates the sensor parameters so that the observation results can be compensated for by the calculated offset value.

[0089] For example, assuming that the number of collisions of the first external autonomous driving device is 1, the number of collisions of the second external autonomous driving device is 2, and the number of collisions of the third external autonomous driving device is 7, the processor (140) can assign a weight of 55% to the observation information received from the first external autonomous driving device, a weight of 35% to the observation information received from the second external autonomous driving device, and a weight of 10% to the observation information received from the third external autonomous driving device.

[0090] The processor (140) may place the greatest importance on minimizing the error (E1) generated by comparing the position information (X1') received from the first external autonomous driving device with the position information (X1) of the autonomous driving device (100) estimated by the processor (140), according to the size of each assigned weight, and may place the lowest importance on minimizing the error (E3) generated by comparing the position information (X3') received from the third external autonomous driving device with the position information (X3) of the autonomous driving device (100) estimated by the processor (140).

[0091] As described above, the processor (140) repeatedly performs a task of changing the sensor parameter values ​​in order to minimize each given error according to each importance, and when a sensor parameter value capable of minimizing the error is found, the found sensor parameter value can be stored as a corrected sensor parameter value to perform a sensor parameter correction task.

[0092] The above sensor parameter correction work can be performed by various algorithms that minimize errors by considering the given observation information and the importance of the information, such as the weighted least square method, which finds specific parameter values ​​that minimize the target function that is the sum of the squared values ​​of each error and the product of each weight.

[0093] Meanwhile, the autonomous driving device (100) may be set to share observation information only when it encounters an external autonomous driving device, or may be set to share observation information only when an external autonomous driving device is present within a certain range (e.g., 1 m) of the autonomous driving device (100). For example, if observation information is set to be shared whenever an external autonomous driving device approaches within 1 m of the autonomous driving device (100), observation information may be received from the external autonomous driving device whenever the distance between the autonomous driving device (100) and the external autonomous driving device becomes less than 1 m.

[0094] The above description only exemplifies a case where an autonomous driving device (100) encounters an external autonomous driving device or exists together within a certain range. However, the autonomous driving device (100) can also receive observation information through various methods, such as a method of transmitting information on an external object observed at regular intervals (e.g., every hour) to an external server, so that if there is an autonomous driving device that has observed the same external object, the external server can transmit the observation information to the autonomous driving device.

[0095] FIG. 6 is a diagram for explaining the operation of an autonomous driving device according to one or more embodiments of the present disclosure.

[0096] According to FIG. 6, the actual driving direction (601) of the autonomous driving device (100) and the driving directions (602, 603) recognized by the processor (140) may match or differ. For example, if the autonomous driving device (100) was in a state immediately after the sensor parameters were accurately calibrated, the actual driving direction (601) and the driving direction (602) recognized by the processor (140) may match. However, if the autonomous driving device (100) collides with an obstacle or carries a heavy object while driving in a space, an external force may be applied to the autonomous driving device (100), causing the position or direction of the sensor (120) to be different from the existing position or direction. In this case, the driving direction (603) recognized by the processor (140) based on the sensing value and sensor parameters of the sensor (120) may be different from the actual driving direction (601). Therefore, if the position and direction of the sensor are misaligned, the observed information may not be accurate.

[0097] Accordingly, if the autonomous driving device (100) obtains observation information from an external autonomous driving device whose sensor position and direction are misaligned, the observation information will not be accurate, so the weight of the observation information obtained therefrom is set to be small. Conversely, the observation information obtained from an external autonomous driving device whose sensor parameters are accurately calibrated will contain accurate information, so the weight of the observation information obtained therefrom can be set to be large. These weights can be assigned based on the driving history information of each external autonomous driving device that provided the observation information.

[0098] For example, if an external autonomous driving device collides with obstacles in space a large number of times, there is a high probability that the position or direction of the sensor of the external autonomous driving device is misaligned due to the external force caused by the collision, and therefore, a low weight may be given to the observation information obtained from the external autonomous driving device with a large number of collisions. Specifically, if the number of collisions of the first external autonomous driving device with obstacles is 20, the number of collisions of the second external autonomous driving device with obstacles is 30, and the number of collisions of the third external autonomous driving device with obstacles is 50, the processor (140) may give a 50% weight to the observation information obtained from the first external autonomous driving device, a 30% weight to the observation information obtained from the second external autonomous driving device, and a 20% weight to the observation information obtained from the third external autonomous driving device.

[0099] As another example, if the intensity of the collision between the external autonomous driving device and an obstacle existing in space is large, there is a high probability that the position or direction of the sensor of the external autonomous driving device is misaligned due to the large impact, and therefore, a low weight may be given to the observation information obtained from the external autonomous driving device with a large collision intensity. At this time, as described above, the processor (140) may arithmetically estimate the collision intensity by considering the speed and moving direction of the autonomous driving device at the time of the collision with the external object, the speed and moving direction of the external object, etc., or may directly sense the collision intensity based on the sensing value of the pressure sensor, or may estimate the collision intensity based on the change in the position of the autonomous driving device (100) before and after the collision.

[0100] Specifically, if the first external autonomous driving device drives in space and collides with an obstacle and the accumulated collision strength is 100 Ns, the accumulated collision strength of the second external autonomous driving device is 200 Ns, and the accumulated collision strength of the third external autonomous driving device is 300 Ns, the processor (140) can give a weight of 60% to the observation information obtained from the first external autonomous driving device, a weight of 30% to the observation information obtained from the second external autonomous driving device, and a weight of 10% to the observation information obtained from the third external autonomous driving device.

[0101] As another example, the processor (140) compares the time that the external autonomous driving device has driven in the space, and determines that the longer the time that the device has driven in the space, the higher the probability that the direction or position of the sensor is misaligned due to various types of driving operations. Therefore, a lower weight may be given to observation information obtained from an external autonomous driving device with a longer driving time, and a higher weight may be given to observation information obtained from an external autonomous driving device with a shorter driving time.

[0102] As another example, the processor (140) may assign weights based on the calibration history of the parameters of the sensors mounted on the external autonomous driving device. If the sensor parameters of the external autonomous driving device have been recently calibrated, accurate location information can be generated based on the sensing values, whereas if a lot of time has passed since the sensor parameters were calibrated for the external autonomous driving device, the location information generated based on the sensing values ​​may not be accurate. Therefore, the processor (140) may perform calibration by comparing the difference between the current time and the time at which the sensor parameters of each external autonomous driving device were calibrated, and assigning a smaller weight to observation information obtained from an external autonomous driving device with a larger difference, and assigning a larger weight to observation information obtained from the external autonomous driving device with a smaller difference.

[0103] The above description describes a method of assigning weights based on one piece of driving history information of an external autonomous driving device, such as the number of times the external autonomous driving device collided with an obstacle existing in space, the intensity of the collision, the driving time of the external autonomous driving device in space, and the calibration history of the sensor parameters mounted on the external autonomous driving device. However, this is only an example, and weights can be assigned to observation information obtained from an external autonomous driving device in various ways, such as by assigning weights by considering all of the number of times the external autonomous driving device collided with an obstacle, the intensity of the collision, the driving time of the external autonomous driving device in space, and the calibration history of the sensor parameters mounted on the external autonomous driving device, or by combining two or three pieces of information from the above.

[0104] The processor (140) may also assign weights according to time order to the observation information stored in the memory (130). This will be described in detail in the description of FIG. 7.

[0105] FIG. 7 is a diagram illustrating a weighting method of an autonomous driving device according to one or more embodiments of the present disclosure.

[0106] According to FIG. 7, the memory (130) can store observation information (701, 702, 703, 704) acquired from an external autonomous driving device in the order in which the information was generated.

[0107] The processor (140) can assign weights to observation information based on the generation time of the observation information stored in the memory (130) in sorted order according to time. When accumulating observation information in the memory (130) for a certain period of time, there is a high possibility that the position and direction of the sensor will slightly shift as time passes. Therefore, the processor (140) can assign the lowest weight to the observation information generated by the most recent observation by sorting the observation information stored in the memory (130) in chronological order, and assign a greater weight to observation information that has been observed for a long time. In order to distinguish the weight assigned in chronological order as described above from the weight assigned based on the driving history information described above, the term “time-series weight” will be used in the following description.

[0108] For example, the autonomous driving device (100) can first share observation information with the first external autonomous driving device (200-1) and then share the observation information with the second external autonomous driving device (200-2). The autonomous driving device (100) can receive information (701) that the first external autonomous driving device (200-1) observed the autonomous driving device (100) from the first external autonomous driving device (200-1), and can obtain information (702) that the first external autonomous driving device (200-1) observed an object existing in space. Thereafter, the autonomous driving device (100) can obtain information (703) that the second external autonomous driving device (200-2) observed the autonomous driving device (100), and can receive information (704) that the second external autonomous driving device (200-2) observed the autonomous driving device (100).

[0109] The above information (701, 702, 703, 704) is sorted by time and stored in memory, and the processor (140) assigns a time series weight to each piece of information. The processor (140) may assign the largest time series weight to the observation information (701) having the largest difference between the time the information was collected and the current time, while assigning the smallest time series weight to the observation information (704) having the smallest difference. For example, the information (701) observing the first external autonomous driving device may be assigned a time series weight of 40%, the information (702) observing the object of the first external autonomous driving device may be assigned a time series weight of 25%, the information (703) observing the object of the second external autonomous driving device may be assigned a time series weight of 20%, and the information (704) observing the second external autonomous driving device may be assigned a time series weight of 15%, thereby allowing the importance of each piece of observation information to the task of correcting the sensor parameters to be set differently.

[0110] The time series weighting task according to the time order of the above-described observation information may be performed after the weighting task according to the driving history information described in the description section for FIGS. 3 to 6 is performed, or may be performed first without the weighting task according to the driving history information being performed.

[0111] Although the above explanation assumes a situation in which all observation information is generated at different times, the same time series weight may be applied to observation information generated at the same time, and the time series weight may be assigned to the observation information through various methods depending on the algorithm applied to the autonomous driving device (100).

[0112] FIG. 8 is a flowchart illustrating a method for calibrating sensor parameters of an autonomous driving device according to one or more embodiments of the present disclosure.

[0113] According to FIG. 8, the autonomous driving device can share observation information and driving history information with an external autonomous driving device within a certain range from the autonomous driving device (S810).

[0114] Here, the observation information may include location information of an external autonomous driving device observed by the autonomous driving device through a sensor, location information of an object existing in space, location information of an autonomous driving device observed by the external autonomous driving device through a sensor mounted on the external autonomous driving device, and location information of an object existing in space.

[0115] Additionally, the driving history information may include the number of times the external autonomous driving device collided with an obstacle existing in space, the intensity with which the external autonomous driving device collided with an obstacle existing in space, the driving time of the external autonomous driving device in space, and the time at which the sensor parameters of the sensors mounted on the external autonomous driving device were most recently calibrated.

[0116] Next, the autonomous driving device identifies whether the observation information received from the external autonomous driving device is sufficient to perform a correction operation of the sensor parameters (S820).

[0117] If the received observation information is not sufficient, the autonomous vehicle drives through the space and receives and stores observation information from the external autonomous vehicle whenever it encounters another external autonomous vehicle or comes within a certain distance of the other autonomous vehicle or observes the same landmark.

[0118] If the received observation information is sufficient, the autonomous driving device assigns weight to the received observation information based on the driving history information of the external autonomous driving device received (S830).

[0119] According to one embodiment, the autonomous driving device may compare the number of times each of at least one external autonomous driving device collides with an obstacle, and assign the smallest weight to observation information received from an external autonomous driving device with the largest number of collisions with an obstacle among the at least one external autonomous driving device, and assign a larger weight to observation information received from an external autonomous driving device with a smaller number of collisions with an obstacle among the at least one external autonomous driving device.

[0120] According to another embodiment, the autonomous driving device may compare the intensity with which each of at least one external autonomous driving device collides with an obstacle existing in space, and may assign the smallest weight to observation information received from an external autonomous driving device with the greatest intensity with respect to the obstacle among the at least one external autonomous driving device, and may assign a larger weight to observation information received from an external autonomous driving device with a smaller intensity with respect to the obstacle among the at least one external autonomous driving device.

[0121] According to another embodiment, the autonomous driving device may compare the time each of at least one external autonomous driving device has driven in space, and may assign the smallest weight to observation information received from an external autonomous driving device with the longest driving time among the at least one external autonomous driving device, and may assign a larger weight to observation information received from an external autonomous driving device with a shorter driving time among the at least one external autonomous driving device.

[0122] According to another embodiment, the autonomous driving device may identify a difference between a time at which a sensor parameter of each of the external autonomous driving devices was most recently calibrated and a current time based on sensor parameter calibration history information of a sensor mounted on each of the at least one external autonomous driving device, and may assign the smallest weight to observation information received from the external autonomous driving device having the largest difference among the at least one external autonomous driving device, and may assign a larger weight to observation information received from the external autonomous driving device among the at least one external autonomous driving device as the difference becomes smaller.

[0123] Next, the autonomous driving device can sort the observation information in the order in which the observation information was collected, assign the smallest weight to the most recently collected observation information, and assign a greater weight to the observation information as more time has passed since the observation information was collected (S840).

[0124] Next, the autonomous driving device can calibrate the sensor parameters of the autonomous driving device based on the weighted observation information (S850).

[0125] While various embodiments have been described individually or in combination above, each embodiment is not necessarily implemented independently. That is, the various embodiments described above may be implemented together in whole or in part with at least one other embodiment in a single autonomous driving device.

[0126] Meanwhile, the methods according to the various embodiments of the present disclosure described above may be implemented in the form of an application that can be installed on an existing autonomous driving device.

[0127] Additionally, the methods according to the various embodiments of the present disclosure described above can be implemented only with a software upgrade or a hardware upgrade for an existing autonomous driving device.

[0128] Additionally, the various embodiments of the present disclosure described above can also be performed through an embedded server provided in an autonomous driving device, or at least one external server.

[0129] Meanwhile, according to a temporary example of the present disclosure, the various embodiments described above can be implemented as software including instructions stored in a machine-readable storage medium that can be read by a machine (e.g., a computer). The device is a device that can call instructions stored in the storage medium and operate according to the called instructions, and may include an autonomous driving device according to the disclosed embodiments. When the instruction is executed by the processor, the processor can perform a function corresponding to the instruction directly or under the control of the processor by using other components. The instruction may include code generated or executed by a compiler or interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' means that the storage medium does not contain a signal and is tangible, but does not distinguish between whether data is stored semi-permanently or temporarily in the storage medium.

[0130] Furthermore, according to one embodiment of the present disclosure, the method according to the various embodiments described above may be provided as included in a computer program product. The computer program product may be traded as a product between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or online through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily generated in a storage medium, such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0131] In addition, each of the components (e.g., modules or programs) according to the various embodiments described above may be composed of a single or multiple entities, and some of the corresponding sub-components described above may be omitted, or other sub-components may be further included in various embodiments. Alternatively or additionally, some components (e.g., modules or programs) may be integrated into a single entity, which may perform the same or similar functions as those performed by each of the corresponding components prior to integration. Operations performed by modules, programs or other components according to various embodiments may be executed sequentially, in parallel, iteratively or heuristically, or at least some operations may be executed in a different order, omitted, or other operations may be added.

[0132] Although the preferred embodiments of the present disclosure have been illustrated and described above, the present disclosure is not limited to the specific embodiments described above, and various modifications may be made by a person having ordinary skill in the art to which the present disclosure pertains without departing from the gist of the present disclosure as claimed in the claims, and such modifications should not be understood individually from the technical idea or prospect of the present disclosure.

Claims

1. In autonomous driving devices, Department of Communications; A sensor for performing sensing based on at least one sensor parameter; memory; and a processor; including; The above processor, Based on the sensing value of the above sensor, first observation information obtained by observing objects existing in the space where the autonomous driving device is located is stored in the memory, The second observation information obtained by observing at least one external autonomous driving device among objects existing in the space and the driving history information of the at least one external autonomous driving device are received through the communication unit and stored in the memory, An autonomous driving device, which assigns a weight to each of the first observation information and the second observation information stored in the memory based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device, and compares the first and second observation information to which the weights are assigned to correct the at least one sensor parameter.

2. In paragraph 1, The above first observation information includes location information of the autonomous driving device estimated by the autonomous driving device, The second observation information includes position information of the autonomous driving device estimated by the at least one external autonomous driving device, The above processor, Based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device, the weight is assigned to each of the location information of the autonomous driving device estimated by the autonomous driving device and the location information of the autonomous driving device estimated by the external autonomous driving device. An autonomous driving device that compares the weighted position information and compensates for the difference by correcting at least one sensor parameter.

3. In paragraph 1, Driving history information of at least one external autonomous driving device, An autonomous driving device, comprising at least one of the number of times the at least one external autonomous driving device collided with an obstacle present in the space, the intensity with which the at least one external autonomous driving device collided with an obstacle present in the space, the driving time of the at least one external autonomous driving device in the space, and the time at which a sensor parameter of a sensor mounted on the at least one external autonomous driving device was most recently calibrated.

4. In paragraph 3, The above processor, The autonomous driving device and the at least one external autonomous driving device each compare the number of collisions that occurred while driving within the space, An autonomous driving device that sequentially assigns greater weights to devices with a higher number of collisions than devices with a lower number of collisions.

5. In paragraph 3, The above processor, Comparing the impact strength of each of the autonomous driving device and the at least one external autonomous driving device while driving within the space, An autonomous driving device that sequentially increases the weight from a device with a large collision intensity to a device with a small collision intensity.

6. In paragraph 3, The above processor, Comparing the driving time of each of the autonomous driving device and the at least one external autonomous driving device in the space; An autonomous driving device that sequentially assigns greater weights to devices with longer driving times than devices with shorter driving times.

7. In paragraph 3, The above processor, By identifying the difference between the latest correction time and the current time of each of the autonomous driving device and the at least one external autonomous driving device, An autonomous driving device that sequentially increases the weight from the device with the largest difference to the device with the smallest difference.

8. In paragraph 1, The above processor, By identifying the difference between the observation information collection time and the current time of each of the autonomous driving device and the at least one external autonomous driving device, An autonomous driving device that sequentially assigns greater weights to observation information from the observation information with the largest difference to the observation information with the smallest difference.

9. In a method for calibrating an autonomous driving device, A step of storing first observation information obtained by observing objects existing within a space where the autonomous driving device is located based on the sensing value of the sensor; A step of receiving second observation information obtained by observing at least one external autonomous driving device among objects existing in the space and driving history information of the at least one external autonomous driving device; and A correction method, comprising: a step of assigning a weight to each of the stored first observation information and the stored second observation information based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device, and comparing the weighted first and second observation information to correct at least one sensor parameter of the sensor.

10. In paragraph 9, The above first observation information includes location information of the autonomous driving device estimated by the autonomous driving device, The second observation information includes position information of the autonomous driving device estimated by the at least one external autonomous driving device, The step of calibrating at least one sensor parameter of the above sensor comprises: A correction method comprising: a step of assigning a weight to each of position information of the autonomous driving device estimated by the autonomous driving device and position information of the autonomous driving device estimated by the external autonomous driving device based on driving history information of the autonomous driving device and driving history information of the at least one external autonomous driving device; comparing the weighted position information and correcting the at least one sensor parameter to compensate for the difference; 11. In paragraph 9, The step of calibrating at least one sensor parameter of the above sensor comprises: A step of comparing the number of collisions that occurred while each of the autonomous driving device and the at least one external autonomous driving device was driving within the space; and A compensation method, comprising: a step of sequentially increasing the weight from a device with a high number of collisions to a device with a low number of collisions.

12. In paragraph 9, The step of calibrating at least one sensor parameter of the above sensor comprises: A step of comparing the collision intensity of each of the autonomous driving device and the at least one external autonomous driving device while driving within the space; and A compensation method comprising: a step of sequentially increasing the weight from a device having a large collision intensity to a device having a small collision intensity; 13. In paragraph 9, The step of calibrating at least one sensor parameter of the above sensor comprises: A step of comparing the driving time of each of the autonomous driving device and the at least one external autonomous driving device in the space; and A correction method comprising: a step of sequentially increasing the weight from a device having a longer driving time to a device having a shorter driving time; 14. In paragraph 9, The step of calibrating at least one sensor parameter of the above sensor comprises: A correction method, comprising: a step of identifying the difference between the observation information collection time and the current time of each of the autonomous driving device and the at least one external autonomous driving device, and sequentially assigning a greater weight to observation information from the observation information with the largest difference to the observation information with the smallest difference.

15. A non-transitory computer-readable recording medium storing computer instructions that, when executed by a processor of an autonomous driving device, cause the autonomous driving device to perform an operation, wherein the operation is: A step of storing first observation information obtained by observing objects existing within a space where the autonomous driving device is located based on the sensing value of the sensor; A step of receiving and storing second observation information from at least one external autonomous driving device among objects existing in the space and driving history information of the at least one external autonomous driving device; and A non-transitory computer-readable recording medium comprising: a step of correcting at least one sensor parameter of the sensor by assigning weights to the stored first observation information and the second observation information based on the driving history information of the autonomous driving device and the driving history information of the at least one external autonomous driving device;

Images