Apparatus and method for performing a path finding

Abstract

The present specification describes a robot for performing pathfinding, comprising: an input unit for acquiring images through a camera; a sensing unit for detecting the position of an object through a LiDAR; a memory for storing data acquired by the pathfinding and a reduced map generated based on the data; and a processor functionally connected to the input unit, the sensing unit, and the memory to control the overall operation of the robot, wherein the processor controls the generation of the reduced map by scanning the surrounding environment of the robot using the camera and the LiDAR, and including an object placement structure that arranges the object detected as a result of the scan in a specific form and a path structure in a coded form using connecting lines.

Description

Technology Field

[0001] This specification relates to a method for a robot to recognize its approximate location without using a high-precision map in a record-replay robot navigation system. More specifically, it relates to a method for estimating the robot's current location by recognizing the shape of the path and surrounding objects, and utilizing this to enable robot navigation without complex sensing or high-performance processing. Background Technology

[0002] Modern autonomous driving technology uses sensors such as Lidar, cameras, and GPS to generate precise maps and predict the location of robots or vehicles based on them. Most representative research in autonomous driving technology utilizes Simultaneous Localization and Mapping (SLAM) technology, which enables robots to scan their surrounding environment, create maps in real time, and track their location. SLAM is widely used in applications ranging from autonomous vehicles to various robot platforms.

[0003] SLAM systems perform high-accuracy environmental perception and position estimation simultaneously through sensors, but because they require advanced computational power and the processing of large amounts of data, high-performance computers and large-capacity storage devices are essential. In particular, autonomous vehicles require highly precise maps by processing data collected from multiple Lidar sensors in real time, which is essential for accurate position tracking in complex environments such as highways and urban areas.

[0004] However, for robot systems, particularly small robots, significant constraints on power and weight make it difficult to use high-performance computers or process large amounts of data like those used in autonomous vehicles. Since small robots mostly rely on battery power, using high-performance sensors and computer systems can shorten battery life and significantly limit the robot's range of movement. Consequently, lightweight location tracking and navigation methods utilizing embedded systems are required. Unlike conventional autonomous vehicles, robots must use embedded systems instead of high-performance computers, making it practically difficult to generate and store complex overall maps.

[0005] Furthermore, unlike autonomous vehicles, many robot applications frequently travel along fixed routes, such as commuting paths or specific work paths. In such cases, it is more efficient to use simplified maps based only on route information and major landmarks rather than complex 3D maps. For example, for guide dog robots or robots used in logistics warehouses, storing simple 1D maps or route information is more suitable than storing a full 3D map, as the path to the destination is predetermined.

[0006] Existing autonomous driving robot systems require highly precise maps and primarily track the robot's position in real time by combining SLAM technology with sensors such as GPS, Lidar, and high-performance cameras. For example, Google's self-driving car project, Waymo, is designed to enable vehicles to drive autonomously accurately in road environments by utilizing SLAM and precise pre-built maps. The Waymo system detects all objects around the vehicle in real time and generates high-resolution maps based on this to calculate the path required for autonomous driving.

[0007] However, such systems require a significant amount of computational resources and power, making them difficult to apply, particularly to small robots with limited computational capabilities or battery-powered robots. Furthermore, high-precision sensors like Lidar are expensive and heavy, making it unsuitable for small robots to use multiple units. Additionally, due to limited computing resources, it is practically impossible for these robots to construct a 3D map of the entire environment in real time and track their position based on it. Therefore, small robots utilizing limited resources require a new system capable of recognizing location using simpler data and methods. The problem to be solved

[0008] Accordingly, the present specification aims to provide a method that enables autonomous robots to perform path navigation without the need for high-precision 3D maps, and enables efficient path recognition and re-searching even in small robots with limited computing resources and storage space.

[0009] To this end, the present specification aims to recognize the environment through object detection and semantic segmentation based on a record-replay method, and to provide a path structure through a simplified one-dimensional map and symbolization.

[0010] Therefore, this specification is applicable to embedded system-based robots by efficiently compressing data and excluding unnecessary information to save storage resources.

[0011] The technical problems to be solved by the present invention are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0012] The present specification describes a robot for performing pathfinding, comprising: an input unit for acquiring images through a camera; a sensing unit for detecting the position of an object through a lidar; a memory for storing data acquired by the pathfinding and a reduced map generated based on the data; and a processor functionally connected to the input unit, the sensing unit, and the memory to control the overall operation of the robot, wherein the processor controls the generation of the reduced map, which includes an object placement structure that arranges objects detected as a result of the scan in a specific form and a path structure in a coded form using connecting lines, by scanning the surrounding environment of the robot using the camera and the lidar.

[0013] Additionally, in this specification, the processor is characterized by setting the detected object as a landmark and arranging a figure representing the set landmark in the specific shape.

[0014] Additionally, in this specification, the processor is characterized by setting the current position of the robot as a center coordinate and arranging a shape related to the detected object in the specific shape based on the point in time when the robot looks at the object.

[0015] In addition, the specific form described in this specification is characterized as being radial, grid-like, or tree-like.

[0016] In addition, the processor in this specification is characterized by recognizing the path of the surrounding environment through semantic segmentation and converting the recognized result into a BEV (Bird's Eye View) form to generate the path structure in the symbolized form.

[0017] In addition, the path described in this specification is characterized by including a pedestrian walkway, an intersection, and a crosswalk.

[0018] Additionally, in this specification, the processor is characterized by creating a new node from the node detected by the event occurrence when it detects a difference between the nodes detected at the time of the event occurrence and the nodes on the robot's movement path already stored in the memory.

[0019] In addition, the processor in this specification is characterized by updating the robot's movement path by reflecting the newly generated node on the robot's movement path.

[0020] In addition, the occurrence of the event in this specification is characterized as a change in the position of an object, detection of a new object, or a change in the path of the surrounding environment.

[0021] Additionally, the present specification describes a method for a robot to perform pathfinding, comprising: scanning the surrounding environment of the robot through a camera and a lidar; generating a path structure in a coded form through semantic segmentation and Bird's Eye View (BEV) transformation for a path of the surrounding environment through the scan; generating an object placement structure in which objects detected through the scan are placed in a specific form; and generating a reduced map including the coded path structure and the object placement structure.

[0022] Additionally, the step of generating the object placement structure in this specification is characterized by including: a step of setting the current position of the robot as a center coordinate; and a step of arranging a shape related to the detected object in the specific shape based on the viewpoint where the robot looks at the object.

[0023] Additionally, the step of generating the path structure in the symbolized form in this specification is characterized by including: a step of recognizing the path of the surrounding environment through semantic segmentation; and a step of converting the recognized result into a BEV (Bird's Eye View) form.

[0024] Additionally, the method described herein is characterized by further including the step of creating a new node from the node detected by the event occurrence when a difference is detected between the nodes detected at the time of the event occurrence and the nodes on the robot's movement path already stored in the memory.

[0025] Additionally, the method described herein is characterized by further including the step of updating the robot's movement path by reflecting the newly generated node on the robot's movement path.

[0026] Additionally, the step of generating the object placement structure in the present specification is characterized by including: the step of setting the detected object as a landmark; and the step of placing a figure representing the set landmark in the specific shape.

[0027] Additionally, the present specification describes an autonomous driving device for performing path search, comprising: a memory for storing path search-related data; and a processor that scans a surrounding environment through a camera and a lidar, generates a symbolized path structure using connecting lines through semantic segmentation and a Bird's Eye View (BEV) transformation for a path recognized in the surrounding environment, generates an object placement structure that arranges objects detected through the scan in a specific form, and generates a reduced map including the symbolized path structure and the object placement structure. Effects of the invention

[0028] This specification has the effect of preventing redundant storage of unnecessary data and significantly saving storage space by symbolizing only important landmarks and path structures and storing them in a simplified form, and maximizing data efficiency by using odometry data to condense and record travel distances in repetitive or featureless sections.

[0029] In addition, the present specification has the effect of enabling a robot to quickly refer to information during pathfinding by representing the path structure of each section with simplified symbols, creating new nodes only when important landmarks or walkway structures change during movement between nodes, and efficiently utilizing existing data to improve pathfinding speed and increase accuracy.

[0030] In addition, the present specification creates a new node only when there is an environmental change during path search and modifies the path to reflect it when there is a significant difference from the existing path, thereby enabling an autonomous driving robot to continuously maintain an accurate path in a dynamic environment and enhancing real-time response capabilities.

[0031] The effects obtainable from the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0032] The accompanying drawings, which are included as part of the detailed description to aid in understanding the present invention, provide embodiments of the present invention and explain the technical features of the present invention together with the detailed description. FIG. 1 is a diagram showing an example of an internal block diagram of a robot to which the method proposed in this specification can be applied. FIG. 2 is a figure showing an example of a simplified map structure proposed in this specification. FIG. 3 is a figure showing an example of the form of an object arrangement structure limited in the present specification. Figure 4 is a figure showing an example of recording to a new node according to the movement of the robot. Figure 5 illustrates the process of a robot navigating a path in an urban environment and recording it. FIG. 6 is a flowchart illustrating an example of a robot path search method using a simplified map structure proposed in this specification. Specific details for implementing the invention

[0033] It should be noted that technical terms used in this specification are used merely to describe specific embodiments and are not intended to limit the scope of the technology disclosed herein. Furthermore, unless specifically defined otherwise in this specification, technical terms used in this specification shall be interpreted in the sense generally understood by those skilled in the art to which the technology disclosed herein belongs, and shall not be interpreted in an overly broad or overly narrow sense. Additionally, if a technical term used in this specification is an incorrect technical term that fails to accurately express the concept of the technology disclosed herein, it shall be understood as being replaced by a technical term that can be correctly understood by those skilled in the art to which the technology disclosed herein belongs. Furthermore, general terms used in this specification shall be interpreted according to their prior definitions or according to the context, and shall not be interpreted in an overly narrow sense.

[0034] Terms including ordinal numbers, such as first, second, etc., as used in this specification may be used to describe various components, but said components should not be limited by said terms. Such terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component.

[0035] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are given the same reference number regardless of the drawing symbols, and redundant descriptions thereof will be omitted.

[0036] Furthermore, in describing the technology disclosed in this specification, detailed descriptions of related prior art are omitted if it is determined that such descriptions could obscure the essence of the technology disclosed in this specification. Additionally, it should be noted that the attached drawings are intended merely to facilitate an understanding of the concept of the technology disclosed in this specification, and should not be interpreted as limiting the concept of the technology.

[0038] FIG. 1 is a diagram showing an example of an internal block diagram of a robot to which the method proposed in this specification can be applied.

[0039] Referring to FIG. 1, the robot (10) may include an input unit (100), a sensing unit (200), a communication unit (300), a processor (400), a memory (500), and an output unit (600), and may be referred to as a robot device, a mobile device, a terminal, an AI device, an intelligent device, an autonomous driving device, etc.

[0040] The components shown in FIG. 1 are not essential, so a robot with more or fewer components may be implemented.

[0041] Below, we will examine each component of the robot in turn.

[0042] The input unit (100) is for inputting audio signals or video signals or user input. For inputting audio signals or video signals, a camera and a microphone, etc., may be included.

[0043] The camera processes image frames, such as still images or videos, obtained by an image sensor in shooting mode. The processed image frames may be displayed on a display unit and may include an RGB camera, an AI camera, etc.

[0044] The sensing unit (200) detects the current state of the robot and the surrounding environment, such as the robot's position, the robot's orientation, and the robot's acceleration / deceleration, and generates a sensing signal to control the robot's operation, and may include a lidar, a laser, etc.

[0045] The communication unit (300) may include one or more modules that enable wired / wireless communication between the robot and an external device. That is, the communication unit (300) may include a short-range communication module, a mobile communication module, etc.

[0046] A short-range communication module refers to a module for short-range communication, and short-range communication technologies such as Bluetooth, RFID (Radio Frequency Identification), infrared communication (IrDA: Infrared Data Association), UWB (Ultra Wideband), and ZigBee can be used.

[0047] In addition, the communication unit may additionally include a location information module. The location information module is a module for acquiring the position of the robot, and a representative example thereof is a GPS (Global Position System) module.

[0048] The above processor (400) refers to a module that controls the overall operation of a robot, and may be referred to as a control unit, control module, controller, microcontroller, microprocessor, etc., and may be implemented by hardware, firmware, software, or a combination thereof.

[0049] The above processor may include an application-specific integrated circuit (ASIC), other chipsets, logic circuits, and / or data processing devices.

[0050] The memory (500) can store a program for the operation of the processor (400) and can also temporarily store input / output data. The memory is a medium for storing various information of the robot and is connected to the processor to store a program for the operation of the processor, an application, general files, and input / output data. The memory may include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (ReadOnly Memory), EEPROM (Electrically Erasable Programmable ReadOnly Memory), PROM (Programmable ReadOnly Memory), magnetic memory, magnetic disk, and optical disk.

[0051] The output unit (1070) is intended to generate output related to sight, hearing, or touch, and may include a display unit, a sound output module, etc.

[0052] Although not illustrated in FIG. 1, the robot may further include a power supply unit representing a module that receives external power and internal power under the control of the processor and supplies power necessary for the operation of each component.

[0054] Next, we examine the method proposed in this specification for an autonomous driving robot to recognize a path in a complex environment, simplify and store the recognized path, and navigate the path. The main technical feature of the method proposed in this specification is the use of a simplified map structure for path navigation through object detection, semantic segmentation, and Bird's Eye View (BEV) transformation.

[0055] First, the robot stores recorded (or collected) data using a record-replay method.

[0056] When a robot navigates a path, it scans the surrounding environment based on the Lidar of the sensing unit and the camera of the input unit (e.g., RGB camera) and records the path based on this. That is, the robot analyzes the scanned surrounding environment through object detection and semantic segmentation methods, recognizes path structures such as walkways and roads around the robot, and encodes the recognized path structure information to generate a simplified (or simplified) one-dimensional map. Additionally, the robot records the walkway structure between each node for the data explored in Record mode, and the robot stores additional data included in the one-dimensional map based on the distance traveled between nodes. A node as used in this specification is a point where the robot scans the surrounding environment and generates a simplified one-dimensional map based on the scanned results; it may include a pre-set node, a node that changes due to the occurrence of events, or a newly added node. The events may include changes in the robot's movement path or the recognition of a new object.

[0057] When the robot detects an event, such as a change in path or the recognition of a new object, it creates a new node on the movement path corresponding to the point where the event occurs, records it, and stores data including distance and time for each segment. The recorded and stored data is used in the robot's replay mode to refer to the data when moving back along a previously explored path in the future.

[0059] FIG. 2 is a figure showing an example of a simplified map structure proposed in this specification.

[0060] Referring to FIG. 2, FIG. 2b shows an example of a result obtained by semantic segmentation of a robot's movement path and converted into a BEV form, and FIG. 2a shows an example of a map structure simplified using connecting lines for the complex data of FIG. 2b. In the connecting lines of FIG. 2a, black (210) represents a sidewalk, blue (220) represents a roadway, and red (230) represents an uncertain area.

[0061] As previously described, when the robot searches for a path, it scans the surrounding environment using Lidar and RGB cameras. During the scanning process, the robot recognizes the path structure of the searched surrounding environment (walkway, intersection, crosswalk, etc.) through semantic segmentation and converts the recognized path structure into a BEV form as shown in FIG. 2b. Then, the robot encodes the path structure of each segment of the result converted into the BEV form to generate a simplified map structure including multiple connecting lines (210, 220, 230) as shown in FIG. 2a.

[0062] In FIG. 2b, the position and direction of the robot are indicated by a black arrow (250), and the symbolization (240) of the path structure of each section can be indicated as a pedestrian path (or sidewalk, 210), a roadway (220), etc. relative to the robot, as shown in FIG. 2a. The path structure information regarding the symbolized path structure can be usefully utilized when the robot later re-searches the path.

[0064] Additionally, when the robot detects an object (or entity) while scanning the surrounding environment during pathfinding, the robot sets the detected object as a characteristic landmark. As illustrated in FIG. 3, the robot creates and stores an object placement structure for the detected object, which places the detected object in one of various structural forms, such as a radial, grid, or tree structure. The object placement method of FIG. 3 sets the robot's current position as the center coordinate and places the detected object in one of the structural forms (radial, grid, or tree) based on the viewpoint of the robot. FIG. 3 is a figure illustrating an example of the form of an object placement structure limited by the present specification.

[0065] That is, Fig. 3a shows an example of arranging objects in a radial structure, whereby detected objects are arranged radially around the robot and each object is placed considering the distance. Fig. 3b shows an example of arranging objects in a grid structure, whereby the space is divided into grids of a certain size and objects are placed in the divided cells of the grid. Fig. 3c shows an example of arranging objects in a tree structure, whereby the relationships between objects are represented as a tree structure based on major nodes and arranged accordingly.

[0066] In FIGS. 3a to 3c, the triangle (310) and the square (320) each represent characteristic landmarks set by the robot.

[0068] Next, we will examine how the robot creates new nodes during pathfinding by referring to the relevant diagram.

[0069] When a difference arises between the nodes included in the explored path and the previously recorded nodes while the robot is exploring a path, the robot determines or creates a new node that is not included in the previously recorded nodes. As previously discussed, the reasons for the creation of a new node may include changes in the position of an object, the discovery of a new object, or changes in the path structure. Additionally, the new node may be created over a certain distance or after the passage of time. The purpose of determining or creating the new node is to allow the robot to accurately reflect the changed surrounding environment during path exploration, update the path, and move accurately using the updated path. Figure 4 illustrates an example of recording a new node based on the robot's movement. Figure 4a represents an existing node, and Figure 4b represents a new node. In the case of Figure 4a, two objects are far apart from each other, while in the case of Figure 4b, the distance between the two objects becomes closer or changes due to the robot's movement; in this case, the robot can record the changed object as a new node.

[0070] The landmarks set by the robot may be new nodes or waypoints, and these landmarks are used as important reference points that the robot can refer to while moving.

[0071] The conditions for creating a new node are determined or created when a new object is detected, the location of an existing object changes, or when distance and time change beyond a certain threshold. Additionally, the robot sets important points between nodes as waypoints, and the type of object, location information, pedestrian path, and road structure are recorded and stored together as descriptive information. FIG. 5 is a diagram illustrating an example of a waypoint proposed in this specification. That is, FIG. 5 illustrates the process of a robot navigating a path in an urban environment and recording it. Referring to FIG. 5, it can be seen that the robot recognizes landmarks on the path it traveled, sets them as key nodes, and records and stores information regarding them.

[0072] While navigating the path, the robot scans the surrounding environment using Lidar and RGB cameras at each major section (e.g., sidewalks, intersections, traffic lights, etc.). 510 and 520 in FIG. 5 represent the scanned screen of the surrounding environment. These major sections can be set as waypoints. Through the scanning process, the robot recognizes path structures such as roads, sidewalks, and intersections through semantic segmentation, and the recognized information is stored in the form of a simplified map structure (511, 521, 530) through BEV conversion. In FIG. 5, the path (540) indicated by the red line represents the main path the robot travels, and the red dots (550) on the travel path (540) represent the major nodes recorded by the robot.

[0073] The robot records the path structure and landmarks (511, 521, 530) of the corresponding section at each node.

[0074] For example, landmarks such as buildings, trees, bus stops, traffic lights, etc. are indicated by various symbols (570) such as squares, triangles, and circles, as shown in FIG. 5. Additionally, the landmarks are important points referenced on the path where the robot moves, and depending on their respective locations, they can be stored in a radial, grid, or tree form, as seen in FIG. 3.

[0075] Referring to FIG. 5, landmarks (traffic lights, trees, bus stops) at each node from the robot's perspective are arranged in a radial pattern, and the path structure between nodes is stored and displayed in a simplified map structure form by symbolizing the distinction between roads and pedestrian paths (511, 521, 530). In the path structure of FIG. 5, black represents the sidewalk, blue represents the roadway, and red represents an uncertain area.

[0076] Additionally, when the robot is exploring a path, if it detects that a certain distance or time has passed or that there is a significant change in the path structure, the robot creates a new node (560, 570). The nodes (560, 570) corresponding to the red dots on the robot's movement path in FIG. 5 are created when landmarks on the path are different from before or when new objects are recognized. For example, in the case of the node corresponding to 560, it can be seen that the robot created a new node by detecting a new traffic light and a tree as it passed through an intersection.

[0077] Through the process described above, the robot stores information about objects detected during pathfinding—specifically, detected object information (including landmarks, nodes, etc.)—in the form of a simplified map structure represented by connecting lines, and utilizes this simplified map structure during pathfinding. In other words, instead of recording all data obtained through the scanning of the surrounding environment in detail, the robot simplifies characteristic landmarks and path structure information to record only the necessary information, thereby saving storage space and maximizing data efficiency. For example, the robot can reduce data throughput by recording only the relative position information between objects or by storing only the changes between major nodes along the path.

[0079] FIG. 6 is a flowchart illustrating an example of a robot path search method using a simplified map structure proposed in this specification.

[0080] First, the robot scans the surrounding environment of the robot through a camera and lidar (S610).

[0081] And, the robot generates a path structure in a symbolized form through semantic segmentation and BEV (Bird's Eye View) transformation on the path of the surrounding environment through the scan (S620).

[0082] And, the robot generates an object placement structure that arranges the object detected through the scan in a specific shape (S630).

[0083] And, the robot generates a reduced map including the symbolized path structure and the object placement structure (S640).

[0084] Step S620, that is, the step of generating a path structure in a symbolized form, may include a process of recognizing the path of the surrounding environment through semantic segmentation and converting the recognized result into a BEV (Bird's Eye View) form.

[0085] Step S630, that is, the step of generating an object placement structure, may include the process of setting the detected object as a landmark and placing a figure representing the set landmark in the specific shape.

[0086] In step S630, the robot sets the robot's current position as the center coordinates and can arrange the shape related to the detected object in the specific shape based on the point of view where the robot looks at the object.

[0087] Here, the specific form may be radial, grid-like, or tree-like, and the path may include pedestrian walkways, intersections, and crosswalks.

[0088] Additionally, if the robot detects a difference between the nodes detected at the time of the event and the nodes on the robot's movement path already stored in the memory, it may create the node detected by the event as a new node and update the robot's movement path by reflecting the newly created node on the robot's movement path.

[0089] The occurrence of the above event may be a change in the position of an object, the detection of a new object, or a change in the path of the surrounding environment.

[0091] The embodiments described above are combinations of the components and features of the present invention in a specific form. Each component or feature should be considered optional unless otherwise explicitly stated. Each component or feature may be implemented in a form not combined with other components or features. Additionally, it is possible to construct embodiments of the present invention by combining some components and / or features. The order of operations described in the embodiments of the present invention may be changed. Some components or features of one embodiment may be included in another embodiment, or may be replaced with corresponding components or features of another embodiment. It is obvious that embodiments may be constructed by combining claims that do not have an explicit citation relationship in the claims, or that they may be included as new claims through amendments made after filing.

[0092] Embodiments according to the present invention may be implemented by various means, for example, hardware, firmware, software, or a combination thereof. In the case of implementation by hardware, an embodiment of the present invention may be implemented by one or more ASICs (application specific integrated circuits), DSPs (digital signal processors), DSPDs (digital signal processing devices), PLDs (programmable logic devices), FPGAs (field programmable gate arrays), processors, controllers, microcontrollers, microprocessors, etc.

[0093] In the case of implementation by firmware or software, an embodiment of the present invention may be implemented in the form of a module, procedure, function, etc., that performs the functions or operations described above. The software code may be stored in memory and executed by a processor. The memory may be located inside or outside the processor and may exchange data with the processor by various known means.

[0094] It is obvious to those skilled in the art that the present invention may be embodied in other specific forms without departing from the essential features of the invention. Accordingly, the foregoing detailed description should not be interpreted restrictively in all respects but should be considered exemplary. The scope of the invention shall be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are included within the scope of the invention. Explanation of the symbols

[0095] 10: Robot

Claims

Claim 1 A robot that performs pathfinding, comprising: an input unit that acquires images through a camera; a sensing unit that detects the position of an object through a lidar; a memory that stores data acquired by the pathfinding and a reduced map generated based on the data; and a processor that is functionally connected to the input unit, the sensing unit, and the memory to control the overall operation of the robot, wherein the processor controls the generation of the reduced map, which includes an object placement structure that arranges objects detected as a result of the scan in a specific form and a path structure in a coded form using connecting lines, by scanning the surrounding environment of the robot using the camera and the lidar. Claim 2 A robot according to claim 1, wherein the processor sets the detected object as a landmark and arranges a figure representing the set landmark in the specific shape. Claim 3 A robot according to claim 1, wherein the processor sets the current position of the robot as a center coordinate and arranges a shape related to the detected object in the specific shape based on the point in time when the robot looks at the object. Claim 4 A robot characterized in that, in claim 3, the specific form is radial, grid-shaped, or tree-shaped. Claim 5 A robot according to claim 1, wherein the processor recognizes the path of the surrounding environment through semantic segmentation and converts the recognized result into a BEV (Bird's Eye View) form to generate the path structure in the symbolized form. Claim 6 A robot according to claim 5, characterized in that the path includes a pedestrian walkway, an intersection, and a crosswalk. Claim 7 A robot characterized in that, in claim 5, the processor, when detecting a difference between the nodes detected at the time of event occurrence and the nodes on the robot’s movement path already stored in the memory, creates the node detected by the event occurrence as a new node. Claim 8 A robot according to claim 7, wherein the processor updates the robot's movement path by reflecting the newly generated node onto the robot's movement path. Claim 9 A robot according to claim 8, characterized in that the occurrence of the event is a change in the position of an object, detection of a new object, or a change in the path of the surrounding environment. Claim 10 A method for a robot to perform pathfinding, comprising: scanning the surrounding environment of the robot through a camera and lidar; generating a path structure in a coded form through semantic segmentation and Bird's Eye View (BEV) transformation for a path of the surrounding environment through the scan; generating an object placement structure in which objects detected through the scan are placed in a specific form; and generating a reduced map including the coded path structure and the object placement structure. Claim 11 A method according to claim 10, wherein the step of generating the object placement structure comprises: a step of setting the current position of the robot as a center coordinate; and a step of arranging a shape related to the detected object in the specific shape based on the point of view where the robot looks at the object. Claim 12 A robot characterized in that, in claim 11, the specific form is radial, grid-shaped, or tree-shaped. Claim 13 A method according to claim 10, wherein the step of generating the path structure in the symbolized form comprises: a step of recognizing the path of the surrounding environment through semantic segmentation; and a step of converting the recognized result into a BEV (Bird's Eye View) form. Claim 14 A robot according to claim 13, characterized in that the path includes a pedestrian walkway, an intersection, and a crosswalk. Claim 15 A method characterized in that, in claim 13, when a difference is detected between the nodes detected at the time of the event and the nodes on the robot’s movement path already stored in the memory, the method further comprises the step of creating the node detected by the event as a new node. Claim 16 A method according to claim 15, further comprising the step of updating the robot's movement path by reflecting the newly generated node onto the robot's movement path. Claim 17 A method according to claim 16, characterized in that the occurrence of the event is a change in the position of an object, detection of a new object, or a change in the path of the surrounding environment. Claim 18 A method according to claim 10, wherein the step of generating the object placement structure comprises: a step of setting the detected object as a landmark; and a step of placing a figure representing the set landmark in the specific shape. Claim 19 An autonomous driving device for performing path search, comprising: a memory for storing path search-related data; and a processor that scans a surrounding environment through a camera and a lidar, generates a symbolized path structure using connecting lines through semantic segmentation and a Bird's Eye View (BEV) form transformation for a path recognized in the surrounding environment, generates an object placement structure that arranges objects detected through the scan in a specific form, and generates a reduced map including the symbolized path structure and the object placement structure.

Images