Ai-based walking robot for determining walkable area and the controling method thereof

Abstract

A walking robot that determines a walking area based on AI is disclosed. The walking robot that determines a walking area based on AI includes a sensor unit for acquiring point cloud data, an image capturing unit for acquiring images, and a processor that determines a walking area based on the point cloud data acquired through the sensor unit and the images acquired through the image capturing unit. Accordingly, the accuracy of determining the walking area is increased, thereby enabling rapid determination of the walking area by the walking robot and enhancing autonomous walking capabilities. Furthermore, such a walking robot can replace human labor, leading to increased productivity and enhanced user convenience.

Description

Technology Field

[0001] The present invention relates to a walking robot that determines a walking area based on AI and a control method thereof, and more specifically, to a walking robot that determines a walking area based on AI and a control method thereof that determines a walking area based on image and geometry information. Background Technology

[0002] With the recent development of autonomous driving technology, many autonomous driving-related technologies are being developed for devices capable of walking or driving, such as robots and vehicles.

[0003] In particular, regarding autonomous robot locomotion, various technologies are being developed to estimate the robot's current position and to determine and recognize areas within the surrounding region where locomotion is possible based on the estimated position; this has been made possible by the development of various sensors and the diversification of AI-based sensing data processing methods.

[0004] Meanwhile, in the case of conventional technology, terrain was analyzed by utilizing point cloud data measured by LiDAR or RGBD sensors to analyze slope, roughness, and obstacles; however, because only terrain information based on point cloud data was used without utilizing camera image information, it was not possible to distinguish texture information, danger zone information, or the presence or absence of specific substances, resulting in low accuracy in determining walkable areas.

[0005] In addition, when determining the walkable area based solely on images, there is a problem in that while ground information can be determined, terrain information cannot be properly analyzed, the accuracy of generating 3D information based on images is not high, and the degree of ground roughness cannot be properly identified, so the accuracy of determining the walkable area is also not high.

[0006] Accordingly, there has been an increased need for a technology that determines the walkable area by integratively considering geometry information and image information expressed as point cloud data. The problem to be solved

[0007] The objective of the present invention is to provide an AI-based walking robot that determines a walking area based on image and geometry information, and a control method thereof. means of solving the problem

[0008] A walking robot for determining an AI-based walking area according to one embodiment of the present invention for achieving such an objective includes a sensor unit for acquiring point cloud data, an image capturing unit for acquiring an image, and a processor for determining a walking area based on the point cloud data acquired through the sensor unit and the image acquired through the image capturing unit.

[0009] Here, the processor can distinguish between ground and non-ground based on the acquired image and reflect this in the acquired point cloud data, and determine a walkable area by calculating at least one of the slope, roughness, and flatness of the ground based on the point cloud data.

[0010] In addition, the processor may calculate the slope in each area based on the normal vector value in each area of ​​the ground, and determine that an area in which the calculated slope is less than a preset angle is a walkable area.

[0011] In addition, the processor may calculate the height difference for each area of ​​the ground and determine that an area where the calculated height difference is less than a preset value is a walkable area.

[0012] In addition, the processor can extract a plane equation for each area of ​​the ground and calculate a standard deviation for each area, and determine that an area where the calculated standard deviation falls within a preset range is a walkable area.

[0013] In addition, the processor can calculate the actual walking path of the walking robot based on the acquired image, and label the area corresponding to the calculated walking path as a walking area.

[0014] In addition, the processor can set the size of each area of ​​the ground and adjust the accuracy of the walkable area calculation by adjusting the size of each area.

[0015] Meanwhile, a control method for a walking robot that determines an AI-based walking area according to one embodiment of the present invention includes the steps of obtaining point cloud data through a sensor unit, obtaining an image through an image capturing unit, and determining a walking area based on the obtained point cloud data and image.

[0016] Here, the step of determining the walkable area may distinguish between the ground and non-ground based on the acquired image and reflect this in the acquired point cloud data, and determine the walkable area by calculating at least one of the slope, roughness, and flatness of the ground based on the point cloud data.

[0017] Additionally, the step of determining the walking area may include calculating the slope in each area based on the normal vector value in each area of ​​the ground, determining an area where the calculated slope is less than a preset angle as a walking area, calculating the height difference for each area of ​​the ground, determining an area where the calculated height difference is less than a preset value as a walking area, extracting a plane equation for each area of ​​the ground to calculate the standard deviation for each area, determining an area where the calculated standard deviation is within a preset range as a walking area, calculating the actual walking path of the walking robot based on the acquired image, and labeling an area corresponding to the calculated walking path as a walking area.

[0018] In addition, the step of determining the walkable area may set the size of each area of ​​the ground and adjust the accuracy of the walkable area calculation by adjusting the size of each area.

[0019] Meanwhile, a computer-readable recording medium according to one embodiment of the present invention may include a program for executing on a computer a control method for a walking robot that determines an AI-based walking area. Effects of the invention

[0020] According to various embodiments of the present invention as described above, the accuracy of determining the walking area is increased, thereby enabling rapid determination of the walking area of ​​the walking robot and enhancing autonomous walking capabilities. Furthermore, such a walking robot can replace human labor, leading to increased productivity and enhanced user convenience. Brief explanation of the drawing

[0021] FIG. 1 is a diagram illustrating the configuration of a walking robot that determines an AI-based walking area according to one embodiment of the present invention. FIG. 2 is a drawing for explaining the overall process of determining a walkable area according to one embodiment of the present invention. FIGS. 3 to 5 are drawings for explaining the process of determining a walkable area based on point cloud data according to an embodiment of the present invention. FIG. 6 is a diagram illustrating the process of determining an image-based walkable area according to an embodiment of the present invention. FIG. 7 is a flowchart illustrating a control method for a walking robot that determines an AI-based walking area according to an embodiment of the present invention. FIG. 8 is a diagram illustrating the specific configuration of a walking robot that determines an AI-based walking area as illustrated in FIG. 1. FIG. 9 is a drawing relating to a software module stored in a storage unit according to an embodiment of the present invention. Specific details for implementing the invention

[0022] The present invention will be described in more detail below with reference to the drawings. Furthermore, in describing the present invention, specific descriptions of related known functions or configurations are omitted if it is determined that such detailed descriptions would unnecessarily obscure the essence of the invention. Additionally, the terms described below are defined considering their functions in the present invention, and these may vary depending on the intentions or relationships of the user or operator. Therefore, their definitions should be based on the content throughout this specification.

[0023] FIG. 1 is a diagram illustrating the configuration of a walking robot that determines an AI-based walking area according to one embodiment of the present invention.

[0024] Referring to FIG. 1, a walking robot (100) that determines an AI-based walking area according to one embodiment of the present invention includes a sensor unit (110), an image capturing unit (120), and a processor (130).

[0025] Here, the walking robot (100) may include any type of walking robot or electronic device capable of scanning or sensing the surrounding environment, ground, terrain, etc., and in this specification, a quadruped walking robot is mainly used as an example.

[0026] In particular, a walking robot (100) according to one embodiment of the present invention includes a sensor unit (110) capable of scanning or sensing the surrounding environment, ground, terrain, etc., and the sensor unit (110) may include various sensors such as, for example, radar, lidar, depth camera, RGB-D, ultrasonic, sound sensor, infrared camera, acoustic sensor, etc.

[0027] The sensor unit (110) can obtain point cloud data through the various sensors described above, where the point cloud data is data collected mainly by LiDAR sensors or RGB-D sensors, and by sending light or a signal to an object and recording the time it returns to calculate distance information per light or signal and generating it as a single point or point, the point cloud data or point cloud refers to a set of the aforementioned points or points spread out in a three-dimensional space. Since the point cloud data or point cloud is a known technology, a detailed explanation will be omitted.

[0028] Also, the video capturing unit (120) can acquire images and can be implemented mainly as a camera, motion camera, 3D camera, etc.

[0029] Additionally, the processor (130) can determine a walkable area based on point cloud data obtained through the sensor unit (110) and an image obtained through the image capturing unit (120).

[0030] Specifically, the processor (130) can distinguish between ground and non-ground based on the acquired image and reflect this in the acquired point cloud data, and calculate at least one of the slope, roughness, and flatness of the ground based on the point cloud data to determine the walkable area.

[0031] FIG. 2 is a drawing for explaining the overall process of determining a walkable area according to one embodiment of the present invention.

[0032] Referring to FIG. 2, the processor (130) can determine walking capability based on an image (220) obtained through an image capturing unit (120) equipped in a walking robot (100) that determines an AI-based walking area. For example, it can determine that a road indicated in the image is walkingable, while parts such as mountains, water, or building walls are not walkingable. That is, the processor (130) can distinguish between a surface area such as a road and a non-surface area that is not a road among the images obtained through the image capturing unit (120).

[0033] In addition, the processor (130) can distinguish between ground and non-ground areas by matching and reflecting information regarding the distinguished ground and non-ground areas in the point cloud data (210) obtained through the sensor unit (110), and determine the walkable area by calculating at least one of the slope, roughness, and flatness of the ground based on the point cloud data belonging to the ground, thereby making it possible to determine the walkable area more quickly and accurately.

[0034] Specifically, the processor (130) can match the location information of the point cloud data and the image, respectively, in order to match the point cloud data and the image, and can superimpose the point cloud data and the image at the same location onto each other.

[0035] Meanwhile, the process of calculating at least one of the slope, roughness, and flatness of the ground based on point cloud data belonging to the ground is as follows.

[0036] FIGS. 3 to 5 are drawings for explaining the process of determining a walkable area based on point cloud data according to an embodiment of the present invention.

[0037] Referring to FIG. 3, the processor (130) calculates the slope in each area based on the normal vector value in each area of ​​the ground, and can determine that an area where the calculated slope is less than a preset angle is a walkable area.

[0038] Specifically, the processor (130) can divide the ground into multiple regions and set a normal vector for each of the multiple regions, and can calculate the slope in each region according to the set normal vector value.

[0039] For example, the processor (130) can set a first normal vector (310) for a first area (311) of the ground and can calculate the slope in the first area (311) based on the set first normal vector (310) value. Specifically, the first normal vector (310) value can be calculated based on the length or coordinate values ​​on the x-axis, y-axis, and z-axis in each area, and the slope of the first area (311) is calculated based on the calculated first normal vector (310) value and the length or coordinate values ​​on the x-axis, y-axis, and z-axis.

[0040] Likewise, the processor (130) can set a second normal vector (320) for a second area (321) of the ground and can calculate the slope in the second area (321) based on the set second normal vector (320) value. Specifically, the second normal vector (320) value can be calculated based on the length or coordinate value on the x-axis, y-axis, and z-axis in each area, and the slope of the second area (321) is calculated based on the calculated second normal vector (320) value and the length or coordinate value on the x-axis, y-axis, and z-axis.

[0041] And, the processor (130) can determine that an area where the calculated slope is less than a preset angle is a walkable area, as the slope is relatively small and walking is possible.

[0042] Additionally, the processor (130) may determine that an area where the calculated slope is greater than a preset angle is an area where walking is impossible because the slope is relatively large.

[0043] For example, the preset angle for determining the standard of the slope calculated based on a quadruped walking robot may be 25 degrees, and the processor (130) may determine that the calculated slope is 25 degrees or more as an unwalkable area.

[0044] In addition, if the value of A is expressed as a value between 0 and 1 through the following mathematical formula 1, it can be determined as a walkable area.

[0045]

[0046] Here, the slope value represents the angle of inclination, and the A value represents the slope. Accordingly, an area where the slope is expressed as a value between 0 and 1 can be determined as a walkable area.

[0047] That is, the processor (130) can calculate the slope through the normal vector value and the slope angle, and if the calculated slope is expressed as a value between 0 and 1, it can be determined as a walkable area.

[0048] Of course, in the case of a bipedal robot, the aforementioned preset angle becomes smaller, and as the number of supports or legs for walking decreases, the sensitivity to the inclination is increased, and based on this, it becomes possible to determine whether a walking area is possible.

[0049] Meanwhile, referring to FIG. 4, the processor (130) calculates the height difference for each area of ​​the ground, and the area where the calculated height difference is less than a preset value can be determined as a walkable area.

[0050] Specifically, the processor (130) can divide the ground into multiple areas and calculate a height value for each of the multiple areas to calculate a height difference based on the highest and lowest height values, and can generate a slope map through the height difference calculated in this way.

[0051] And, the processor (130) can determine that an area where the calculated height difference is less than a preset value is a walkable area, considering that walking is possible.

[0052] Referring to FIG. 4, the processor (130) generates a map (430), i.e., a slope map, which contains the difference in height between a map (410) that displays the highest height value for each area of ​​the ground and a map (420) that displays the lowest height value for each area of ​​the ground, and can determine that an area calculated to be less than a preset value on the generated slope map is a walkable area.

[0053] For example, a preset value for determining the standard for the height difference calculated based on a quadruped walking robot may be 0.45, and the processor (130) may determine that the calculated height difference is 0.45 or greater and that it is an area where walking is impossible.

[0054] In addition, if the value of B is expressed as a value between 0 and 1 through the following mathematical formula 2, it can be determined as a walkable area.

[0055]

[0056] Here, step represents the difference in height, and B represents the flatness. Accordingly, an area where the flatness is expressed as a value between 0 and 1 can be determined as a walkable area.

[0057] That is, the processor (130) can calculate the flatness through the difference in height, and if the calculated flatness is expressed as a value between 0 and 1, it can be determined as a walkable area.

[0058] Of course, in the case of a bipedal robot, the aforementioned preset value becomes smaller, so as the number of supports or legs for walking decreases, the sensitivity to height differences increases, and based on this, it becomes possible to determine whether a walking area is possible.

[0059] Referring to FIG. 5, the processor (130) extracts a plane equation for each area of ​​the ground and calculates a standard deviation for each area, and can determine that an area where the calculated standard deviation is within a preset range is a walkable area.

[0060] Specifically, the processor (130) can divide the surface into multiple regions, extract a plane equation for each of the multiple regions, and calculate a standard deviation based on the plane equation extracted for each region.

[0061] For example, the processor (130) can extract a first plane equation for a first area (510) of the ground and calculate a first standard deviation based on the extracted first plane equation.

[0062] Likewise, the processor (130) can extract a second plane equation for a second area (520) of the ground and calculate a second standard deviation based on the extracted second plane equation.

[0063] And, the processor (130) can determine that the first area (510) is a walkable area by considering that the first standard deviation falls within a preset range, and determine that the second area (520) is a non-walkable area by considering that the second standard deviation falls outside the preset range.

[0064] For example, the preset range for determining the standard deviation calculated based on a quadruped walking robot can be 0.3 cm, and the processor (130) can determine that the calculated standard deviation is 0.3 cm or more as an unwalkable area.

[0065] In addition, if the value of C is expressed as a value between 0 and 1 through the following mathematical formula 3, it can be determined as a walkable area.

[0066]

[0067] Here, roughness refers to the standard deviation based on the plane equation calculated for each region, and the value of C represents the roughness. Accordingly, regions where the roughness is expressed as a value between 0 and 1 can be determined as walkable regions.

[0068] That is, the processor (130) can calculate the roughness through the standard deviation, and if the calculated roughness is expressed as a value between 0 and 1, it can be determined as a walkable area.

[0069] Of course, in the case of a bipedal robot, the aforementioned preset range becomes smaller, and as the number of supports or legs for walking decreases, the sensitivity to roughness is increased, and based on this, it becomes possible to determine whether a walking area is possible.

[0070] As described above, the processor (130) can determine a walkable area based on point cloud data, i.e., geometry information, obtained through the sensor unit (110).

[0071] Meanwhile, the processor (130) calculates the walking path of the actual walking robot (100) based on the image obtained through the image capturing unit (120), and the area corresponding to the calculated walking path can be labeled as a walking area.

[0072] FIG. 6 is a diagram illustrating the process of determining an image-based walkable area according to an embodiment of the present invention.

[0073] Referring to FIG. 6, the processor (130) can calculate a walking path using footprint tracking information (610, 620) of the walking robot (100), and can label the calculated footprint tracking information (610, 620) as an area where walking is possible, and can determine that walking is impossible in areas other than the footprint tracking information (610, 620).

[0074] That is, the processor (130) tracks footprint data captured while the walking robot (100) actually walks, and by referring to the tracked information, namely the location information of the footprints, determines that the location is an area where walking is possible, and the area where walking is possible and the area where walking is not possible can be displayed in different colors.

[0075] Meanwhile, the processor (130) can label the fire-affected area (630) and the non-fire-affected area (631) as described above, by marking the fire-affected area (630) in red as an area where walking is impossible (631), and marking the non-fire-affected area (631) in green as an area where walking is possible (641).

[0076] And, the processor (130) may also project the labeled areas, distinguished as walkable areas and non-walkable areas, onto the point cloud data to aggregate them.

[0077] Meanwhile, the processor (130) can set the size of each area of ​​the ground and adjust the accuracy of the walkable area calculation by adjusting the size of each area.

[0078] Specifically, as the processor (130) sets the size of each area of ​​the ground to be smaller, the number of times the slope, roughness, and flatness of each area as described above are calculated for the ground increases, which means that the number of samples for calculating data for the ground increases, and thus the accuracy of determining the walkable area for the ground increases.

[0079] Additionally, as the processor (130) sets the size of each area of ​​the ground larger, the number of times the slope, roughness, and flatness of each area as described above are calculated for that ground decreases, which means that the number of samples for calculating data for that ground decreases, and thus the accuracy of determining the walkable area for that ground decreases.

[0080] Therefore, the smaller the size of each area of ​​the ground is set by the processor (13), the higher the accuracy of calculating the walkable area, and the larger the size of each area of ​​the ground is set, the lower the accuracy of calculating the walkable area.

[0081] In addition, as described above, as the number of walking supports, i.e., legs, of the walking robot (100) decreases, the sensitivity to the slope, roughness, and flatness of the ground increases, so the walking area can be determined by lowering the preset angle, preset value, and preset range for determining this.

[0082] FIG. 7 is a flowchart illustrating a control method for a walking robot that determines an AI-based walking area according to an embodiment of the present invention.

[0083] Referring to FIG. 7, a control method for a walking robot that determines an AI-based walking area according to an embodiment of the present invention includes the steps of obtaining point cloud data through a sensor unit (S710), obtaining an image through an image capturing unit (S720), and determining a walking area based on the obtained point cloud data and image (S730).

[0084] Here, the step of determining the walkable area (S730) can determine the walkable area by distinguishing between the ground and non-ground based on the acquired image and reflecting this in the acquired point cloud data, and by calculating at least one of the slope, roughness, and flatness of the ground based on the point cloud data.

[0085] Additionally, the step of determining a walkable area (S730) calculates the slope in each area based on the normal vector value in each area of ​​the ground, determines an area where the calculated slope is less than a preset angle as a walkable area, calculates the height difference for each area of ​​the ground, determines an area where the calculated height difference is less than a preset value as a walkable area, calculates the walking path of the actual walking robot based on the acquired image, and can label an area corresponding to the calculated walking path as a walkable area.

[0086] In addition, the step of determining the walkable area (S730) sets the size of each area of ​​the ground and can adjust the accuracy of the walkable area calculation by adjusting the size of each area.

[0087] Meanwhile, as described above, a computer-readable recording medium may be provided that records a program for executing on a computer a control method for a walking robot that determines an AI-based walking area according to one embodiment of the present invention.

[0088] FIG. 8 is a diagram illustrating the specific configuration of a walking robot that determines an AI-based walking area as illustrated in FIG. 1.

[0089] Referring to FIG. 8, a walking robot (100) that determines an AI-based walking area includes a sensor unit (110), an image capturing unit (120), a processor (130), and a storage unit (140).

[0090] The processor (1130) generally controls the operation of the walking robot (100) that determines the AI-based walking area. Specifically, the processor (130) includes RAM (131), ROM (132), main CPU (133), graphics processing unit (134), first to n interfaces (135-1 to 135-n), and a bus (136).

[0091] RAM (131), ROM (132), main CPU (133), graphics processing unit (134), first to n interfaces (135-1 to 135-n), etc. can be connected to each other via a bus (136).

[0092] The first to n interfaces (135-1 to 135-n) are connected to the various components described above. One of the interfaces may be a network interface connected to an external device through a network.

[0093] The main CPU (133) accesses the storage unit (140) and performs booting using the O / S stored in the storage unit (140). Then, it performs various operations using various programs, content, data, etc. stored in the storage unit (140).

[0094] In particular, the main CPU (133) can determine the walking area based on point cloud data obtained through the sensor unit and images obtained through the image capturing unit.

[0095] A set of instructions for booting the system is stored in the ROM (132). When a turn-on command is input and power is supplied, the main CPU (133) copies the O / S stored in the storage unit (140) to the RAM (131) according to the instructions stored in the ROM (132), and executes the O / S to boot the system. When booting is complete, the main CPU (133) copies various application programs stored in the storage unit (140) to the RAM (131), and executes the application programs copied to the RAM (131) to perform various operations.

[0096] The graphics processing unit (134) generates a screen containing various objects such as icons, images, and text using a calculation unit (not shown) and a rendering unit (not shown). The calculation unit (not shown) calculates attribute values ​​such as coordinate values, shape, size, and color for each object to be displayed according to the layout of the screen based on a received control command. The rendering unit (not shown) generates a screen of various layouts containing objects based on the attribute values ​​calculated by the calculation unit (not shown).

[0097] In particular, the graphics processing unit (134) can implement objects generated by the main CPU (133) into a GUI (Graphic User Interface), icon, user interface screen, etc.

[0098] Meanwhile, the operation of the above-described processor (130) can be performed by a program stored in the storage unit (140).

[0099] The storage unit (140) stores various data, such as an O / S (Operating System) software module for driving a walking robot (100) that determines an AI-based walking area, and various multimedia content.

[0100] In particular, the storage unit (140) may include a software module for determining a walkable area based on point cloud data obtained through the sensor unit and an image obtained through the image capturing unit.

[0101] FIG. 9 is a drawing relating to a software module stored in a storage unit according to an embodiment of the present invention.

[0102] Referring to FIG. 9, the storage unit (140) may store programs such as a slope calculation module (141), a roughness calculation module (142), a flatness calculation module (143), a walkable area labeling module (144), and a walkable area calculation accuracy adjustment module (145).

[0103] Meanwhile, the operation of the processor (130) described above can be performed by a program stored in the storage unit (140). Below, the detailed operation of the processor (130) using the program stored in the storage unit (140) will be explained in detail.

[0104] Specifically, the slope calculation module (141) calculates the slope in each area based on the normal vector value in each area of ​​the ground, and the area where the calculated slope is less than a preset angle can be determined as a walkable area.

[0105] Additionally, the roughness calculation module (142) calculates the height difference for each area of ​​the ground, and the area where the calculated height difference is less than a preset value can be determined as a walkable area.

[0106] In addition, the flatness calculation module (143) can calculate the standard deviation for each area of ​​the ground by extracting a plane equation for each area, and the area where the calculated standard deviation falls within a preset range can be determined as a walkable area.

[0107] Additionally, the walking area labeling module (144) can calculate the walking path of the actual walking robot based on the acquired image, and the area corresponding to the calculated walking path can be labeled as a walking area.

[0108] Also, the walking area calculation accuracy adjustment module (145) can set the size of each area of ​​the ground and adjust the walking area calculation accuracy by adjusting the size of each area.

[0109] Meanwhile, a non-transitory computer-readable medium storing a program that sequentially performs the control method according to the present invention may be provided.

[0110] A non-transient readable medium refers to a medium that stores data semi-permanently and can be read by a device, rather than a medium that stores data for a short period of time, such as a register, cache, or memory. Specifically, the various applications or programs described above may be stored and provided on non-transient readable media such as CDs, DVDs, hard disks, Blu-ray discs, USBs, memory cards, and ROMs.

[0111] In addition, although the bus is not depicted in the aforementioned block diagram illustrating a walking robot that determines an AI-based walking area, a processor such as a CPU or a microprocessor that determines the walking area based on point cloud data acquired through a sensor unit and an image acquired through an image capturing unit may be further included.

[0112] Furthermore, although preferred embodiments of the present invention have been illustrated and described above, the present invention is not limited to the specific embodiments described above. It is understood that various modifications can be made by those skilled in the art without departing from the essence of the invention as claimed in the claims, and such modifications should not be understood individually from the technical spirit or perspective of the present invention. Explanation of the symbols

[0113] 100: AI-based walking robot determining walking range 110: Sensor unit 120: Video recording unit 130: Processor

Claims

Claim 1 In a walking robot that determines an AI-based walking area, a sensor unit for acquiring point cloud data; an image capturing unit for acquiring images; and a processor that determines a walkable area based on point cloud data acquired through the sensor unit and an image acquired through the image capturing unit; wherein the processor distinguishes between ground and non-ground based on the acquired image and reflects this in the acquired point cloud data, matches the position information of the point cloud data and the acquired image respectively, and superimposes the point cloud data and the image at the same location and reflects them together; wherein the processor can set the size of each area of ​​the ground and adjusts the accuracy of calculating the walkable area by adjusting the size of each area; as the size of each area of ​​the ground is set smaller, the number of times the slope, roughness, and flatness of each area are calculated for the corresponding ground increases, and as the number of data calculation samples for the corresponding ground increases, the accuracy of determining the walkable area for the corresponding ground increases, and as the size of each area of ​​the ground is set larger, the number of times the slope, roughness, and flatness of each area are calculated for the corresponding ground decreases, and as the number of data calculation samples for the corresponding ground decreases, the corresponding ground AI-based walking robot for determining a walking area, which reduces the accuracy of determining the walking area. Claim 2 A walking robot that determines an AI-based walking area, wherein, in claim 1, the processor determines a walking area by calculating at least one of the slope, roughness, and flatness of the ground based on the point cloud data. Claim 3 A walking robot that determines an AI-based walking area, wherein, in paragraph 2, the processor calculates the slope in each area based on the normal vector value in each area of ​​the ground, and determines the area where the calculated slope is less than a preset angle as a walking area. Claim 4 A walking robot that determines an AI-based walking area, wherein, in paragraph 3, the processor calculates the height difference for each area of ​​the ground and determines the area where the calculated height difference is less than a preset value as a walking area. Claim 5 A walking robot that determines an AI-based walking area, wherein, in paragraph 4, the processor extracts a plane equation for each area of ​​the ground and calculates a standard deviation for each area, and determines an area where the calculated standard deviation falls within a preset range as a walking area. Claim 6 A walking robot that determines an AI-based walking area, wherein, in claim 5, the processor calculates the actual walking path of the walking robot based on the acquired image and labels the area corresponding to the calculated walking path as a walking area. Claim 7 delete Claim 8 A control method for a walking robot that determines an AI-based walking area, comprising: a step of acquiring point cloud data through a sensor unit; a step of acquiring an image through an image capturing unit; and a step of determining a walkable area based on the acquired point cloud data and image; wherein the step of determining the walkable area includes distinguishing between ground and non-ground based on the acquired image and reflecting this in the acquired point cloud data, matching the location information of the point cloud data and the acquired image respectively, and superimposing and reflecting the point cloud data and image at the same location; wherein the step of determining the walkable area includes setting the size of each area of ​​the ground and adjusting the accuracy of calculating the walkable area by adjusting the size of each area; wherein the smaller the size of each area of ​​the ground is set, the more times the slope, roughness, and flatness of each area are calculated for the corresponding ground, and as the number of data calculation samples for the corresponding ground increases, the accuracy of determining the walkable area for the corresponding ground increases, and as the size of each area of ​​the ground is set larger, the number of times the slope, roughness, and flatness of each area are calculated for the corresponding ground decreases, and as the number of data calculation samples for the corresponding ground decreases, the accuracy of determining the walkable area for the corresponding ground A control method for a walking robot that determines an AI-based walking area, wherein the accuracy of determining the walking area is reduced. Claim 9 A control method for a walking robot that determines an AI-based walking area, wherein the step of determining the walking area is to determine the walking area by calculating at least one of the slope, roughness, and flatness of the ground surface based on the point cloud data. Claim 10 In claim 9, the step of determining the walking area comprises calculating the slope in each area based on the normal vector value in each area of ​​the ground, determining an area where the calculated slope is less than a preset angle as a walking area, calculating the height difference for each area of ​​the ground, determining an area where the calculated height difference is less than a preset value as a walking area, extracting a plane equation for each area of ​​the ground to calculate the standard deviation for each area, determining an area where the calculated standard deviation is within a preset range as a walking area, calculating the actual walking path of the walking robot based on the acquired image, and labeling an area corresponding to the calculated walking path as a walking area, thereby controlling a walking robot based on AI determining a walking area. Claim 11 delete Claim 12 A computer-readable recording medium having a program for executing on a computer a control method for a walking robot that determines an AI-based walking area as described in any one of paragraphs 8 through 10.

Images