Electronic device, and data generation method for road object recognition thereof

Abstract

An electronic device according to one embodiment disclosed in the present document may comprise: a memory for storing point cloud data acquired on a road; and a processor functionally connected to the memory, wherein the processor can extract a driving line and floor surface data from the point cloud data, form split points having, individually, a first interval on the driving line, calculate unit areas designated on the basis of each of the split points, and generate, on the basis of artificial intelligence, recognition data for recognizing a designated road object from the floor surface data in each of the unit areas by using values of the floor surface data in each of the unit areas.

Description

Electronic devices and their methods for generating data for road object recognition

[0001] Various embodiments disclosed in this document relate to road object recognition technology.

[0002] Autonomous vehicles can perceive their surroundings and movements with little or no human intervention or input. To achieve this, autonomous vehicles are equipped with sensors capable of detecting their surroundings, such as radar, computer vision, GPS, odometry, and inertial measurement units.

[0003] To safely navigate roads designed for conventional vehicles (manned vehicles), autonomous vehicles utilize technology that recognizes road rules. For example, lane recognition technology, one of the most fundamental aspects of road rules, is essential for autonomous vehicles to safely navigate roads alongside manned vehicles. Conventional lane recognition in autonomous vehicles relies on images captured by cameras, but this method can be less accurate in situations with limited ambient light.

[0004] Autonomous vehicles can more accurately recognize road lanes through a mobile mapping system (MMS). A mobile mapping system can be a technology that collects and maps geographic information and the surrounding environment using GPS, LiDAR, cameras, and an inertial measurement unit (IMU). Specifically, a mobile mapping system can track the precise location coordinates of an autonomous vehicle using high-precision maps and GPS sensors, and generate 3D point cloud data of the surrounding environment using LiDAR data. Furthermore, a mobile mapping system can combine camera data with LiDAR data to perform more precise mapping, and an inertial measurement unit can detect changes in the autonomous vehicle's attitude to further improve data accuracy.

[0005] Autonomous vehicles use a portion of 3D point cloud data to recognize lanes. However, when autonomous vehicles use point cloud data to recognize lanes, gaps in the dotted lines within the lane can cause errors in lane recognition. This can lead to the vehicle drifting out of its lane, potentially leading to an accident.

[0006] Various embodiments disclosed in this document can provide an electronic device capable of generating data for artificial intelligence-based road object recognition from point cloud data and a data generation method for road object recognition.

[0007] An electronic device according to an embodiment disclosed in the present document comprises a memory for storing point cloud data acquired on a road; and a processor functionally connected to the memory, wherein the processor extracts driving line and floor data from the point cloud data, forms segmentation points each having a first interval on the driving line, calculates designated unit areas based on each segmentation point, and generates recognition data for recognizing a designated road object from the floor data within each of the unit areas based on artificial intelligence by using the value of the floor data within each of the unit areas.

[0008] In addition, a method for generating data for road object recognition by at least one processor according to an embodiment disclosed in the present document may include an operation of acquiring point cloud data acquired on a road; an operation of extracting a driving line and floor data from the point cloud data; an operation of calculating designated unit areas based on dividing points of a first interval on the driving line; and an operation of generating recognition data for recognizing a designated object based on artificial intelligence from the floor data for each unit area using a value of the floor data corresponding to each unit area.

[0009] According to the various embodiments disclosed in this document, data for AI-based road object recognition can be generated from point cloud data. Furthermore, various benefits, directly or indirectly identified through this document, can be provided.

[0010] FIG. 1 illustrates a flowchart of a method for generating data for lane recognition based on point cloud data according to one embodiment.

[0011] Figure 2 is an example diagram of point cloud data according to one embodiment.

[0012] Figure 3 illustrates an example of point cloud floor surface data according to one embodiment.

[0013] Figure 4 shows a schematic diagram of an electronic device according to one embodiment.

[0014] Figure 5 shows an example of a floor surface extraction process according to one embodiment.

[0015] Figure 6 shows an example of a unit area calculation process according to one embodiment.

[0016] FIG. 7 is a diagram for explaining the generation of PCL floor surface data by unit area according to one embodiment.

[0017] FIG. 8 is a diagram for explaining the generation of learning data related to PCL floor surface data by unit area according to one embodiment.

[0018] Figure 9 is a flowchart of a learning method of an artificial intelligence model according to one embodiment.

[0019] In connection with the description of the drawings, the same or similar reference numerals may be used for identical or similar components.

[0020] FIG. 1 illustrates a flowchart of a method for generating data for lane recognition based on point cloud data according to one embodiment.

[0021] Referring to FIG. 1, in operation 110, an electronic device (e.g., 400 of FIG. 4) can extract driving line and floor data from point cloud data collected on a road. For example, when point cloud data is collected on a road, it is collected by dividing it into left and right viewpoints along the driving path of the vehicle. Accordingly, the electronic device (400) can calculate the driving line (10) by converting the left and right viewpoints collected at the time of collection into points and then converting them into lines. The floor data may be point cloud data corresponding to the floor among the point cloud data. For example, the electronic device (400) can extract the floor data by deleting data above a specified height from the ground from the entire point cloud data.

[0022] In operation 120, the electronic device (400) may calculate a designated unit area based on dividing points having a first interval on the driving line. For example, the electronic device (400) may divide the driving line into first intervals and calculate the width of second intervals to the left and right of each dividing point. In addition, the electronic device (400) may calculate a unit area related to each dividing point by connecting the width of the previous dividing point and the width of the next dividing point based on each dividing point.

[0023] In operation 130, the electronic device (400) may generate recognition data for learning or inferring an artificial intelligence model for a designated object on the road using the floor data values ​​within each unit area among the point cloud floor data. For example, the electronic device (400) may rotate the floor data within the unit area in a designated direction and project the signal intensity value of the rotated floor data onto a static two-dimensional array. In addition, the electronic device (400) may convert the static two-dimensional array into image data and generate the converted image data as data for recognizing a designated road object (recognition data). Here, the electronic device (400) may check reference coordinate information (coordinate values ​​corresponding to the corners of the unit area) of the floor data within the unit area and rotation angle information related to rotation in the designated direction. The electronic device (400) may store restoration data including the reference coordinate information and the rotation angle information in relation to the recognition data. In one embodiment, the recognition data may be used for learning and inferring an artificial intelligence model. The above restoration data can be used, for example, to confirm the location coordinates of a road object recognized according to the inference of an artificial intelligence model.

[0024] In this way, the electronic device (400) according to one embodiment can generate recognition data for an artificial intelligence model for recognizing a designated road object (e.g., lane, road rule marking) on ​​a floor surface based on point cloud data collected on the road.

[0025] Figure 2 is an example diagram of point cloud data according to one embodiment.

[0026] Referring to FIG. 2, point cloud (PCL) data may be a set cloud of multiple points spread across a three-dimensional space collected from a plurality of sensors. The plurality of sensors may include a Lidar sensor, an RGB-D sensor, etc. The plurality of sensors send light / signals to an object and record the return time, calculate distance information for each light / signal, and generate a single point corresponding to the calculated distance information. In the case of FIG. 2, point cloud data illustrates data (las file format) collected from a Lidar sensor equipped in an electronic device (e.g., MMS). Due to the characteristics of the las file format, the point cloud data is collected by dividing it into left and right viewpoints along the driving path of the vehicle when collected on the road. Therefore, the electronic device (400) can calculate a driving line (20) by gathering the left and right viewpoints into points and converting them into lines.

[0027] As illustrated in FIG. 2, point cloud data may include information related to not only road lanes and road surface markings, but also trees, structures, tunnels, and other objects surrounding the road. Therefore, the electronic device (400) may extract and utilize ground data, including road lanes and road surface markings, from the point cloud data for faster computation. This will be described below with reference to FIG. 3.

[0028] Figure 3 illustrates an example of point cloud floor surface data according to one embodiment.

[0029] Referring to FIG. 3, it can be seen that the point cloud floor data only includes information such as driving lines (20), lanes, and road markings, excluding information related to trees, structures, tunnels, etc. around the road. The extraction of point cloud floor data by the electronic device (400) will be described later.

[0030] Fig. 4 illustrates a configuration diagram of an electronic device according to one embodiment. Fig. 5 is a diagram for explaining unit area extraction according to one embodiment, and Fig. 6 is a diagram for explaining static two-dimensional array generation according to one embodiment.

[0031] Referring to FIG. 4, an electronic device (400) according to one embodiment may include a memory (410) and a processor (450). In one embodiment, the electronic device (400) may omit some components or include additional components. In addition, some of the components of the electronic device (400) may be combined to form a single entity, but may perform the same functions of the corresponding components prior to combination. In one embodiment, the electronic device (400) may be, for example, a mobile mapping system or an autonomous vehicle system including a mobile mapping system.

[0032] The memory (410) may include various types of volatile memory or non-volatile memory. For example, the memory (410) may include read-only memory (ROM) and random access memory (RAM). In one embodiment, the memory (410) may be located inside or outside the processor (450), and the memory (410) may be connected to the processor (450) via various known means. The memory (410) may store various data used by at least one component (e.g., the processor (450)) of the electronic device (400). The data may include, for example, input data or output data for software and commands related thereto. For example, the memory (410) may store at least one instruction and data for generating data for training (or inference) of an artificial intelligence model for lane recognition based on cloud point data. The memory (410) may store cloud point data.

[0033] The processor (450) can control at least one other component (e.g., hardware or software component) of the electronic device (400) and perform various data processing or calculations. The processor (450) can include, for example, at least one of a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, an application processor, an application specific integrated circuit (ASIC), and a field programmable gate array (FPGA), and can have a plurality of cores. According to one embodiment, the processor (450) can extract driving line and floor data from point cloud data, calculate designated unit areas based on segmentation points having a first interval on the driving line, and generate input data of an artificial intelligence model for extracting lanes from each of the unit areas using the values ​​of the floor data within each of the unit areas.

[0034] According to one embodiment, the processor (450) may include a filter unit (451), a calculation unit (453), and a conversion unit (455). The filter unit (451), the calculation unit (453), and the conversion unit (455) may be hardware modules or software modules included in the processor (450) or executed by the processor (450). However, for the convenience of explanation, in the following document, the operations of each of the filter unit (451), the calculation unit (453), and the conversion unit (455) are described based on the processor (450).

[0035] According to one embodiment, the processor (450) (filter unit (451)) can filter point cloud data with a road floor filter to extract cloud data related to the road floor.

[0036] According to one embodiment, the processor (450) (filter unit (451)) can remove information unnecessary for floor extraction within a predetermined distance from the point cloud data. The point cloud data may include information related to not only lanes or road surface guidance marks (e.g., direction guidance marks, yield marks, direction strings) on the road, but also trees, structures, tunnels, etc. around the road. For example, the processor (450) (filter unit (451)) can perform a drive near filter that extracts only point cloud data within a predetermined distance based on the driving line (10). In addition, the processor (450) can leave only point cloud data necessary for floor extraction through the drive near filter. Additionally, the point cloud data can remove information unnecessary for floor extraction, for example, trees, structures, etc. around the road, through the drive near filter. The predetermined distance may be, for example, a distance necessary for leaving only information related to road rule marks on the road based on the driving line (20) in the point cloud data, and may change depending on the width of the road. there is.

[0037] The processor (450) (filter unit (451)) can generate a minimum z-grid table from point cloud data that includes only the information necessary for floor extraction, and extract the floor point cloud using the initial z-grid table. For example, the processor (450) (filter unit (451)) can generate a point (dividing point) of a first interval on the driving line in the driving direction using only the x-coordinate value and the y-coordinate value of the point cloud data (ignoring the z-coordinate value of the point cloud data). The processor (450) can generate a grid corresponding to the driving interval point using the x-coordinate and the y-coordinate. The processor (450) can check the minimum z-coordinate value of the point cloud data included in each cell of the generated grid, and set the checked minimum z-coordinate value as the z-coordinate value of each cell. However, when there is a road slope (uphill or downhill slope), the height between each cell can be set unevenly. To prevent this, the processor (450) can further based on the actual road slope of the precision map data. The height of the cell can be gradually compensated for. For example, if there is an increase or decrease in the z-coordinate value between cells located forward and backward along the driving direction (or, if there is a road gradient), the processor (450) can interpolate the z-coordinate value for the contact point between cells located forward and backward. The cells divided into grids containing the interpolated z-coordinate values ​​can be a road surface filter (minimum z-grid table).

[0038] Additionally, the processor (450) (filter unit (451)) can extract floor data by extracting points located under two points included in each cell (minimum z-grid table) from the point cloud data. For example, the processor (450) can extract points located under points included in each cell by determining whether they are located under two points included in each cell using a counter clockwise (CCW) algorithm. The point cloud data including the extracted points can be floor data.

[0039] According to one embodiment, the processor (450) (filter unit (451)) can extract floor data by deleting data above a specified height from the ground from the entire point cloud data.

[0040] According to one embodiment, the processor (450) (the output unit (453)) can output a unit area (polygon area) using the driving line of point cloud data. For example, the processor (450) can divide the driving line (50) into a first interval (e.g., 20 m) in the driving direction. The processor (450) can identify the positions of a first point and a second point spaced apart from each of the division points of the driving line (50) by a second interval (e.g., 15 m) to the left and right, respectively.

[0041] Referring to FIG. 5, the processor (450) may calculate a polygon area (unit driving area) (e.g., 40 m*30 m) formed by connecting the first point (e.g., L1) and the second point (e.g., R1) of the previous division point (e.g., S1) and the first point (e.g., L3) and the second point (e.g., R3) of the subsequent division point (e.g., S3) based on each division point (e.g., S2) from the second division point (e.g., S2) of the driving line. The calculated polygon area includes a driving line corresponding to twice the second interval, but may overlap a portion (e.g., half) of the polygon area before and after the driving line.

[0042] According to one embodiment, the processor (450) (conversion unit (455)) can extract PCL floor data (or PCL floor data within each polygon area) corresponding to each polygon area from point cloud floor data (hereinafter, 'PCL floor data').

[0043] For example, the processor (450) (conversion unit (455)) can rotate the PCL floor data in each polygon area in a specified direction (e.g., north direction). The processor (450) can insert the signal intensity value of the rotated PCL floor data for each unit area into a static two-dimensional array. The static two-dimensional array can be, for example, a two-dimensional array having a size of 600 * 800. Here, the processor (450) can check restoration data including rotation angle information and reference coordinate information for each unit area. The reference coordinate information can be, for example, a coordinate value of one corner of each polygon area. The restoration information can be information for restoring the PCL floor data before rotation from the image data.

[0044] In one embodiment, if the number of PCL floor data is greater than the number of values ​​in the static two-dimensional array, the processor (450) may accumulate (e.g., average) multiple values ​​of the floor data and insert (or project) them into the static two-dimensional array.

[0045] For example, referring to FIG. 6, if the PCL bottom surface data is 4*6 and the static two-dimensional array is 2*3, the processor (450) can input the average value of the 2*2 PCL bottom surface data into the static two-dimensional array.

[0046] Additionally, the processor (450) can convert a two-dimensional array value for each unit area into image data and store the image data in memory (410) in relation to restoration information. The image data may be, for example, at least one of PNG format, bitmap table data, or bitmap file data.

[0047] In this way, the electronic device (400) according to one embodiment can generate data for learning / inference of an artificial intelligence model for recognizing lanes based on point cloud data collected on the road.

[0048] Figure 7 shows an example of a floor surface extraction process according to one embodiment.

[0049] Referring to FIG. 7, the filter unit (451) can retrieve original cloud point data from the memory (410), as in the image (710).

[0050] The filter unit (451) can divide the original cloud point data into a certain unit, such as image (720) and image (730), and extract a driving line for the certain unit of point cloud data.

[0051] The filter unit (451) can extract PCL floor data by filtering a certain unit of point cloud data using a floor filter, such as in image (740) and image (750). In image (740) and image (750), it can be confirmed that the PCL floor data includes road lanes.

[0052] Figure 8 illustrates an example of a unit area calculation process according to one embodiment. Figure 8 illustrates an example in which the first interval is set to 20 m and the second interval is set to 15 m.

[0053] Referring to FIG. 8, as in image (810), the output unit (453) can check the driving line of the point cloud data.

[0054] As shown in image (820), the calculation unit (453) can divide the driving line into 20m units in the driving direction. The calculation unit (453) can generate a width (or range) with 15m intervals on the left and right sides at each division point. For example, the calculation unit (453) can calculate the first and second points with 15m intervals on the left and right sides at each division point. The calculation unit (453) can calculate a polygon area (unit area) with 40m intervals and 30m width of the driving line by connecting the first and second points of the previous division point and the first and second points of the next division point based on each division point.

[0055] FIG. 9 is a diagram for explaining the generation of PCL floor surface data by unit area according to one embodiment.

[0056] Referring to FIG. 9, the conversion unit (455) can relate the PCL floor surface data and the unit area to each other, and extract the PCL floor surface data corresponding to each unit area based on each division point from the second division point in the driving direction from the starting point (hereinafter, referred to as “PLC floor surface data by unit area”).

[0057] FIG. 10 is a diagram for explaining the generation of learning data related to PCL floor surface data by unit area according to one embodiment.

[0058] Referring to FIG. 10, in operation 1010, the conversion unit (455) can check the reference coordinate information of the PCL floor surface data for each unit area. For example, the conversion unit (455) can check the coordinate value of one corner of the PCL floor surface data for each unit area as the reference coordinate information of the PCL floor surface data for each unit area.

[0059] In operation 1020, the conversion unit (455) can rotate the PCL floor surface data for each unit area toward the north direction. For example, the conversion unit (455) can rotate the PCL floor surface data for each unit area so that the reference coordinate of the PCL floor surface data for each unit area is located at the lower left. The conversion unit (455) can check the rotation angle information of the PCL floor surface data for each unit area.

[0060] In operation 1030, the conversion unit (455) converts the PCL floor surface data for each rotated unit area into image data, and can relate the image data to restoration data. The restoration data can include reference coordinate information and rotation angle information.

[0061] In operation 1040, the conversion unit (455) can store related image data and restoration data in the memory (410). The restoration data

[0062] Figure 11 is a flowchart of a learning method of an artificial intelligence model according to one embodiment.

[0063] Referring to FIG. 11, in operation 1110, when the electronic device (400) acquires point cloud data, in operation 1120, it can generate image data for learning of an artificial intelligence model for lane recognition as described above.

[0064] In operation 1130, as the user creates a learning mask (correct answer) for the artificial intelligence model, the electronic device (400) can obtain it.

[0065] In operation 1140, an artificial intelligence model can learn a learning mask and learning image data. The artificial intelligence model can be included in an electronic device (400) or an external electronic device.

[0066] In operation 1150, the electronic device (400) can recognize a lane by acquiring image data for each unit area from PCL data collected on a driving road using a learned artificial intelligence model. The electronic device (400) can confirm the coordinate values ​​of the lane recognized from the image data for each unit area using restoration data.

[0067] Figure 12 illustrates an example of road rule markings on a road surface according to one embodiment.

[0068] Referring to FIG. 12, an electronic device (400) according to one embodiment may generate recognition data related to various road rule markings. The road rule markings may include at least one of a road linearity and a road surface shape. The road linearity may include, for example, at least one of a center line, a dotted line, a safety zone marking, a stop line, and a road curb. The road surface shape may include, for example, at least one of a turn sign, a speed suggestion sign, a crosswalk sign, and other instruction signs.

[0069] According to various embodiments, the electronic device (400) may generate recognition data for recognizing road facilities. In this case, the electronic device (400) may filter data on the placement of road facilities from point cloud data and generate recognition data using point cloud data related to the area where the road facilities are placed. The road facilities may include, for example, at least one of a traffic light, a crosswalk post, and a traffic sign.

[0070] The various embodiments of this document and the terminology used herein are not intended to limit the technical features described in this document to specific embodiments, but should be understood to include various modifications, equivalents, or substitutes of the embodiments. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the items, unless the context clearly indicates otherwise. In this document, phrases such as "A or B," "at least one of A and B," "at least one of A or B," "A, B, or C," "at least one of A, B, and C," and "at least one of A, B, or C" can each include any one of the items listed together in the corresponding phrase, or all possible combinations thereof. Terms such as "first," "second," or "first" or "second" may be used simply to distinguish the corresponding component from other corresponding components, and do not limit the corresponding components in any other respect (e.g., importance or order). When a component (e.g., a first component) is referred to as being “coupled” or “connected” to another component (e.g., a second component), with or without the terms “functionally” or “communicatively,” it means that the component can be connected to the other component directly (e.g., wired), wirelessly, or through a third component.

[0071] The term "module" as used herein may include a unit implemented in hardware, software, or firmware, and may be used interchangeably with terms such as logic, logic block, component, or circuit. A module may be an integral component, or a minimum unit or portion of such a component that performs one or more functions. For example, according to one embodiment, a module may be implemented in the form of an application-specific integrated circuit (ASIC).

[0072] Various embodiments of this document may be implemented as software (e.g., a program) including one or more commands stored in a storage medium (e.g., memory (410) of FIG. 4) (e.g., built-in memory or external memory) readable by a machine (e.g., an electronic device). For example, a processor (e.g., processor (450)) of a device (e.g., electronic device (400)) can call at least one command from among one or more commands stored from a storage medium and execute it. This enables the device to operate to perform at least one function according to the called at least one command. The one or more commands may include code generated by a compiler or code executable by an interpreter. A storage medium readable by the device may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' only means that the storage medium is a tangible device and does not include a signal (e.g., electromagnetic waves), and this term does not distinguish between cases where data is stored semi-permanently and cases where data is stored temporarily in the storage medium.

[0073] According to one embodiment, the method according to various embodiments disclosed in the present document may be provided as a computer program product. The computer program product may be traded between sellers and buyers as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or may be provided through an application store (e.g., Play Store). TM ) or directly between two user devices (e.g., smartphones), online distribution (e.g., downloading or uploading). In the case of online distribution, at least a portion of the computer program product may be at least temporarily stored or temporarily created in a machine-readable storage medium, such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0074] Components according to various embodiments of this document may be implemented in the form of software or hardware such as a digital signal processor (DSP), a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC), and may perform certain roles. The term "components" is not limited to software or hardware, and each component may be configured to be on an addressable storage medium or configured to play one or more processors. As examples, components may include components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, properties, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

[0075] According to various embodiments, each component (e.g., a module or a program) of the above-described components may include a single or multiple entities. According to various embodiments, one or more components or operations of the aforementioned components may be omitted, or one or more other components or operations may be added. Alternatively or additionally, a plurality of components (e.g., a module or a program) may be integrated into a single component. In such a case, the integrated component may perform one or more functions of each of the plurality of components identically or similarly to those performed by the corresponding component among the plurality of components prior to the integration. According to various embodiments, the operations performed by a module, program, or other component may be executed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be executed in a different order, omitted, or one or more other operations may be added.

Claims

1. In electronic devices, Memory for storing point cloud data acquired on the road; and A processor functionally connected to the memory, the processor comprising: Extract driving line and floor surface data from the above point cloud data, Forming dividing points each having a first interval on the above driving line, and calculating designated unit areas based on each dividing point, An electronic device that uses the values ​​of the floor surface data within each of the above unit areas to generate recognition data for recognizing a designated road object from the floor surface data within each of the above unit areas based on artificial intelligence.

2. In claim 1, the processor, An electronic device that extracts a first point and a second point at a second interval to the left and right of each of the above-mentioned dividing points, and calculates the respective unit areas related to each of the above-mentioned dividing points by connecting the first and second points of the previous dividing point and the first and second points of the subsequent dividing point based on each of the above-mentioned dividing points.

3. In claim 2, the processor, An electronic device that calculates each unit area from the second dividing point among the dividing points on the above driving line.

4. In claim 1, the processor, An electronic device that rotates the floor data corresponding to each of the above unit areas in a specified direction and generates recognition data corresponding to the unit area by projecting the signal intensity value of the rotated floor data onto a static two-dimensional array.

5. In claim 4, the processor, An electronic device that generates reference coordinate information of the floor surface data corresponding to each of the above unit areas and rotation angle information of the floor surface data corresponding to each of the above unit areas, and relates restoration data including the reference coordinate information and the rotation angle information to the recognition data.

6. In claim 5, the processor, An electronic device that generates the reference coordinate information by using the coordinate values ​​of one corner of the floor surface data corresponding to each of the above unit areas.

7. In claim 5, the processor, An electronic device that extracts the specified object from the floor surface data using an artificial intelligence model and determines the lane location coordinates corresponding to the floor surface data of each unit area based on the restoration data.

8. In claim 4, the processor, An electronic device that converts the two-dimensional array onto which the signal intensity values ​​are projected into image data and determines the image data as the recognition data.

9. In claim 1, the specified object is, An electronic device comprising at least one road rule object selected from the group consisting of a lane, a directional sign, a yield sign, and a direction string.

10. A method for generating data for road object recognition by at least one processor, An action to acquire point cloud data acquired on the road; An operation of extracting driving line and floor surface data from the above point cloud data; An operation of calculating designated unit areas based on the first interval division points on the above driving line; and An operation of generating recognition data for recognizing a designated object based on artificial intelligence from the floor surface data for each unit area using the values ​​of the floor surface data corresponding to each unit area. A method for generating data for road object recognition including:

Images