KR20220084910A - Method of estimating the location of a moving object using vector map - Google Patents

Abstract

A method for estimating a map matching position of a moving object using a vector map is disclosed.
The method includes generating a vector map including a map representative vector in at least one part of an object based on a map point cloud of map data, and based on a surrounding point cloud obtained from a surrounding environment of a moving object generating detection feature information including a detection representative vector in at least one part of an object, and using the similarity between the map representative vector and the detection representative vector to be the same in the vector map and the detection feature information and estimating the position of the moving object by matching the vector map with the detection feature information based on the searched part.

Description

Method of estimating the location of a moving object using vector map}

The present disclosure relates to a map matching location estimation method, and more particularly, to a map matching location estimation method of a moving object using a vector map.

LiDAR (Light Detection And Ranging) is a technology that acquires distance information by measuring the time of a laser beam that is reflected off an object and returns using a high-power pulsed laser. and unmanned devices, etc.

Recently, as it is used as a core technology for 3D reverse engineering, laser scanners and 3D imaging cameras for autonomous driving and unmanned vehicles, its utility and importance are gradually increasing.

In particular, in a moving object such as an autonomous vehicle, a moving path selection strategy must be different depending on which path and which position it is in order to move to a destination. However, the conventional GPS technology has a problem in that the maximum error occurs up to 10 meters, and thus the location cannot be accurately determined in units of paths of moving objects, for example, units of driving lanes of vehicles.

Accordingly, in a moving object, such as an autonomous vehicle, in which it is important to accurately recognize the driving environment, research is being conducted on a driving environment recognition technology that combines map data and sensing information for observing surrounding objects of the moving object.

On the other hand, the easiest way to calculate location recognition is to substitute a moving object in a wide area to estimate the location of the moving object as the location of the moving object, which shows the highest matching rate between the surrounding map data acquired in real time from the moving object and the pre-stored map data. , since both map data, for example, point cloud data of both maps, must be prepared at all locations, there is a disadvantage in that the amount of computation is excessively increased.

Accordingly, in recent years, a particle filter technology for measuring a position at a corresponding position by randomly sampling without substituting data of a point cloud of all points has been used.

However, the particle filter technology also has a problem in that the accuracy and the amount of calculation are significantly different depending on the number of particles that are position information.

The technical problem to be achieved by the present disclosure is to reduce the amount of computation in matching between surrounding map data obtained from a moving object and pre-stored map data, while remarkably reducing the capacity of map data while securing the accuracy of estimating the location of a moving object. An object of the present invention is to provide a method for estimating a map matching position of a moving object using a vector map.

Objects of the present disclosure are not limited to the objects mentioned above, and other objects not mentioned will be clearly understood by those skilled in the art from the following description.

According to an aspect of the present disclosure for achieving the above technical problem, in a method for estimating a map matching position of a moving object using a vector map, a vector including a map representative vector in at least one part of an object based on a map point cloud of map data Generating a map (vector map); Generating detection feature information including a detection representative vector in at least one portion of an object based on a surrounding point cloud obtained in real time from a surrounding environment of a moving object; searching for at least one part estimated to be the same in the vector map and the detection feature information using the similarity between the representative vector and the detection representative vector, and the vector map and the detection based on the searched part and estimating the position of the moving object by matching the feature information.

In another embodiment, the part may be at least one of a face and an edge of the object.

In another embodiment, the generating of the vector map may include determining whether it is a single surface including pointers based on a vector between a predetermined point in the map point cloud and adjacent points; Repeating the process of determining the enlarged surface by determining whether it is one enlarged surface based on the distance between the determined surface and the adjacent points, and based on the distance and angle between the plurality of enlarged surfaces determined in the repeated process The method may include determining a final surface and a final edge, and generating a vector map including the map representative vectors for each of the final surface and the final edge.

In another embodiment, the step of determining whether it is a single surface comprises the steps of calculating an eigenvalue using a covariance matrix between the predetermined point and the adjacent points, and a minimum If the eigenvalue is less than the first threshold value, determining that it is one surface including the predetermined point and the adjacent point.

In another embodiment, the determining of the final surface and the final edge may include: an angle between the plurality of enlarged surfaces and a distance between a geometric centroid of each of the enlarged surfaces and adjacent enlarged surfaces is greater than a second threshold. In a small case, it is determined that the enlarged faces merge into one face, so that the final face and the final edge can be determined.

In another embodiment, the generating of the detection feature information may include: determining whether it is one surface including pointers based on a vector between a predetermined point in the neighboring point cloud and adjacent points; Repeating the process of determining the enlarged surface by determining whether it is one enlarged surface based on the distance between the determined surface and the adjacent points, and based on the distance and angle between the plurality of enlarged surfaces determined in the repeated process to determine a final surface for detection and a final edge for detection, and generating detection feature information including the detection representative vectors for each of the final surface for detection and the final edge for detection.

In another embodiment, the moving object includes a lidar sensor that acquires the surrounding point cloud, an image sensor that acquires an image of the surrounding environment, positioning sensors that acquire location information of the moving object, and a posture of the moving object. Before the step of mounting multiple sensors including an inertial sensor (IMU), and determining whether it is a single surface including pointers based on a vector between a predetermined point in the surrounding point cloud and adjacent points, the step of determining, generating odometry information of the moving object by using the IDA sensor, the geometric information between the image sensor, and the positioning sensor, and based on the posture information of the moving object obtained from the inertial sensor, The method may further include correcting distortion generated in the data.

In another embodiment, the step of searching for at least one part estimated to be the same in the vector map and the detection feature information may include: determining a current position as an initial value, and based on the initial value, the vector map and the vector map related to representative vectors having a predetermined similarity between the map representative vector of the vector map and the detection representative vector of the detection feature information It may include searching for the faces of the detection feature information, and estimating the found faces as the same faces of the same object.

In another embodiment, before determining whether it is one plane including pointers based on a vector between a predetermined point in the neighboring point cloud and the neighboring points, downsampling the neighboring point cloud to a predetermined density It may further include the step of

In another embodiment, the estimating of the position of the moving object may include: estimating the position of the moving object by matching object information of the vector map with object information of the detection feature information based on the searched part; The pose value of the moving object by sampling points constituting the surface of the object in the vector map, and an optimization process using the distance between the sampled points and the surface of the object in the detection feature information , and correcting the estimated position based on the calculated pose value.

The features briefly summarized above with respect to the present disclosure are merely exemplary aspects of the detailed description of the present disclosure that follows, and do not limit the scope of the present disclosure.

According to the present disclosure, it is possible to reduce the amount of computation in matching between the surrounding map data obtained from the moving object and the pre-stored map data, and at the same time, it is possible to secure the accuracy of estimating the location of the moving object while remarkably reducing the capacity of the map data.

In addition, the effect derived through the configuration that can be understood by those skilled in the art through the present specification is not excluded. The effects obtainable in the present disclosure are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those of ordinary skill in the art to which the present disclosure belongs from the description below. will be.

1 is a block diagram of a map matching location estimating apparatus for implementing a map matching location estimation method of a moving object using a vector map according to an embodiment of the present disclosure.
2 is a flowchart illustrating a method for estimating a map matching position of a moving object using a vector map according to an embodiment of the present disclosure.
3 is a flowchart illustrating a process of generating a vector map.
4 is a diagram illustrating an example of determining whether it is one surface constituting an object based on vector information between mutually adjacent points.
5 is a diagram illustrating an example of determining whether an enlarged surface constituting an object by extending from one determined surface to adjacent points.
6 is a flowchart illustrating a process of generating detection feature information.
7 is a diagram illustrating an example in which a moving object acquires a surrounding point cloud of a surrounding environment.
FIG. 8 is a diagram illustrating an example of determining whether a surface constitutes an object based on vector information between mutually adjacent points in a process of generating detection feature information.
9 is a diagram illustrating an example of searching and matching a part of the same object between detection feature information and a vector map.

Hereinafter, preferred embodiments of the present disclosure will be described in detail with reference to the accompanying drawings and the content to be described later. However, the present disclosure is not limited to the embodiments described herein and may be embodied in other forms. Rather, the embodiments introduced herein are provided so that the disclosed subject matter may be thorough and complete, and that the spirit of the present disclosure may be sufficiently conveyed to those skilled in the art. Like reference numerals refer to like elements throughout. Meanwhile, the terminology used in this specification is for describing the embodiments and is not intended to limit the present disclosure. In this specification, the singular form also includes the plural form unless otherwise specified in the phrase. As used herein, “comprises” and/or “comprising” means that the stated component, step, operation and/or element is the presence of one or more other components, steps, operations and/or elements. or addition is not excluded.

In addition, "unit" to "module" generally refer to logically separable parts such as software (computer program) and hardware. Therefore, the module in this embodiment refers not only to the module in a computer and a program, but also the module in a hardware configuration. Therefore, the present embodiment provides a computer program for making them function as modules (a program for causing the computer to execute each step, a program for making the computer function as each means, a program for making the computer to realize each function) ), systems and methods are also explained. However, for convenience of explanation, "save", "save", and phrases equivalent to these are used, but when the embodiment is a computer program, these words are stored in a storage device or stored in a storage device It means to control as In addition, "part" to "module: may correspond to functions one-to-one. However, in implementation, one module may be configured as one program, multiple modules may be configured as one program, and conversely, one module may be configured as multiple programs. In addition, a plurality of modules may be executed by one computer, and one module may be executed by a plurality of computers by a computer in a distributed or parallel environment.In addition, one module may contain other modules In addition, hereinafter, "connection" is used not only for physical connection but also for logical connection (data exchange, instruction, reference relationship between data, etc.). It is said that it is determined before the existing processing, and not only before the processing according to the present embodiment is started, but also after the processing according to the present embodiment is started, and before the target processing, depending on the situation or state at that time, or It is used including the meaning of what is decided according to the situation and status up to that point.

In addition, a system or device means a plurality of computers, hardware, devices, etc. connected by communication means such as a network (including a one-to-one correspondence) and configured by one computer, hardware, device, etc. cases to be realized are included. "Device" and "system" are used as mutually synonymous terms. Of course, "system" does not include things that are nothing more than social "organizations" (social systems) that are artificial decisions.

In addition, for each processing by each unit or each module, or when a plurality of processing is performed within each unit or module, target information is read and inputted from the storage device for each processing, and after performing the processing, the processing result is displayed writing to the memory device. Therefore, the description of read input from the storage device before processing and writing to the storage device after processing may be omitted in some cases. Note that, as the storage device here, a hard disk, random access memory (RAM), an external storage medium, a storage device via a communication line, a register in a CPU (Central Processing Unit), and the like may be included.

Hereinafter, an apparatus for estimating a map matching position of a moving object using a vector map (hereinafter, referred to as a 'position estimating apparatus') according to an embodiment of the present disclosure will be described with reference to FIG. 1 .

The position estimating apparatus 100 is an apparatus for estimating the location of a moving object by matching sensor data of the surrounding environment acquired in real time through a sensor mounted on the moving object and map data pre-stored in the moving object in the course of moving the moving object. Although the present disclosure exemplifies that map data and a vector map are embedded in the location estimation apparatus 100, as another example, map data and a vector map may be transmitted and used from an external device, and in another example As such, map data may be transmitted from an external device, and a vector map may be processed from the transmitted map data and stored in the location estimation apparatus 100 .

Meanwhile, the moving object may be a fully or semi-autonomous driving vehicle, an autonomous driving robot, or the like.

1 is a block diagram of an apparatus for implementing a map matching position estimation method of a moving object using a vector map according to an embodiment of the present disclosure.

The position estimating apparatus 100 includes a multi-sensor 110 that acquires sensing data of an object from a surrounding environment in the process of moving a moving object, indexing, calibration, and matching of data of sensors constituting the multi-sensor 110. A conversion unit 120 that performs fusion, processing, and generation of odometry information, a generation unit 130 that generates a vector map and detection feature information based on map data and data of multiple sensors, and a map data unit (140), a matching unit 150 for matching the vector map and detection feature information to each other, and a position estimating unit 160 for estimating the location of the moving object using the matched maps.

The multi-sensor 110 includes a positioning sensor 112 that acquires location information of a moving object, a lidar sensor 114 that acquires a surrounding point cloud from the surrounding environment of the moving object, and an image sensor 116 that acquires an image of the surrounding environment. ) and an inertial sensor (Inertial Measurement Unit (IMU) 118) for detecting the posture of the moving object.

The positioning sensor 112 may include a position acquiring device that acquires the moving position of the mobile body through a global positioning system (GPS).

The lidar sensor 114 may be a sensor that is mounted on a moving object and acquires observation data for surveying by acquiring two-dimensional or three-dimensional geographic data related to an object around it, for example, terrain and feature-related data. For example, the lidar sensor 114 may be a laser or ultrasonic sensor, and in the case of a laser sensor, it may be a LiDAR (Light Detection and Ranging; LiDAR) sensor. Such a lidar sensor scans a laser on an object to acquire data, detects a parallax and energy change of electromagnetic waves that are reflected from the object and return, and calculates a distance to the object and a reflection intensity.

The image sensor 116 is a sensor that is mounted on a moving object and acquires observation data for image by photographing surrounding objects, such as terrain and features, as images, and may be a camera for surveying or non-surveying, a stereo camera, It is not limited thereto.

Although the present disclosure exemplifies that the inertial sensor is an IMU, the posture of the vehicle may be acquired using an inertial navigation system (INS).

Although not shown in FIG. 1 , the multi-sensor 110 may acquire a change value (movement distance, rotation angle, etc.) of a position of a moving object by using an odometry sensor (not shown). For example, the odometry sensor may calculate the position change value of the vehicle by measuring the number of wheel rotations while the moving object is traveling. Accordingly, the location estimation apparatus 100 may calculate relative location data of two points (nodes). In another example, even though the odometry sensor is omitted, the relative position data of two points is based on the image data of the image sensor 116 , the position information of the positioning sensor 112 , and data of the surrounding point cloud of the lidar sensor 114 . Based on this, relative position data between points moved by the moving object may be calculated.

The conversion unit 120 may converge by correcting geometric information between sensor data, such as data of a neighboring point cloud, image data, and location information of a moving object. When the surrounding point cloud obtained from the lidar sensor 114 is a color, the conversion unit 120 uses RGB data and image data of the surrounding point cloud to correct geometric information and data from the multi-sensor 110 . In the process of merging the data, depth data may be included in the fusion data. Detection feature information, which will be described later, may be generated by using RGB data and depth data included in the surrounding point cloud, in addition to the distance and reflection intensity included in the data of the surrounding point cloud.

The generator 130 may generate a vector map including a map representative vector in at least one portion of an object based on a map point cloud of map data. Also, the generator 130 may generate detection feature information including a detection representative vector in at least one portion of an object based on a surrounding point cloud obtained in real time from a surrounding environment of the moving object. Real-time localization of the moving object may be implemented by matching the detection feature information generated in real time with a pre-made vector map and using it for location estimation.

The vector map and detection feature information may additionally have other data including the representative vector by the generator 130 . That is, the object information included in the vector map and the detection feature information may be generated based on the point cloud, image data, and location information, and the shape and type of the object as well as the representative vector defining at least one part of the object , a name, and the like.

The map data unit 140 may store map information about the vicinity of the moving object. The map information may include spatial information previously generated according to a predetermined coordinate system based on sensor data previously acquired from the multi-sensor 110 and a vector map generated based on a map point cloud.

The predetermined coordinate system applied to the spatial information may be a map coordinate system, an absolute coordinate system, or a separate coordinate system, but is not limited thereto. Spatial information is data generated in the corresponding coordinates, etc., sensor (observation) data of the multi-sensor 110, internal geometry, external geometry, unique data, time information, acquisition environment information, probability information, point cloud data related information, image data It may include related information, object information, and map related information. The probability information may include an error probability for the position of the object based on observation data, time information, and a degree of error in the position of the object due to the erroneous measurement of each of the sensors 112 to 118 . The object information is data about an object existing in the vicinity that is already acquired before the moving object moves, and may include the type, name, shape, change history, etc. of the object. The object information may be object-related information derived by analyzing sensor data acquired from the multiple sensors 110 in the past.

In the present embodiment, although spatial information is described as including all of the above-described data or information, the present invention is not limited thereto, and may be set to include only some of the above-described data/information according to design specifications. In addition, the map information may be updated based on newly acquired sensor data.

The matching unit 150 may search for at least one portion that is estimated to be the same in the vector map and the detection feature information by using the similarity between the map representative vector and the detection representative vector.

The position estimator 160 may estimate the position of the moving object by matching the vector map with the detection feature information based on the portion searched for by the matching unit 150 . The position estimator 160 may estimate the position of the moving object by matching the object information of the vector map with the object information of the detection feature information based on the searched part. Each object information may be a map representative vector and a detection representative vector, and may be a shape, type, name, etc. of an object recorded in the representative vector and vector map and detection feature information.

In the above, the function of each module of the location estimation apparatus 100 has been schematically described. Hereinafter, with reference to FIGS. 1 to 9 , a map matching location estimation method of a moving object using a vector map according to an embodiment of the present disclosure; In addition, the function of each module of the location estimation apparatus 100 will be described in more detail.

2 is a flowchart illustrating a method for estimating a map matching position of a moving object using a vector map according to an embodiment of the present disclosure.

Referring to FIG. 2 , the generator 130 may generate a vector map including representative vectors for each face and edge of an object based on a map point cloud of map data ( S105 ).

Here, the part may be at least one of a face and an edge of the object, and in this embodiment, for convenience of description, it is described that the part includes all of the face and edge of the object. However, if the map representative vector of the vector map can be expressed, the part may be at least one of a face and an edge. An edge is a line located between adjacent faces.

The generation of the vector map will be described in detail with reference to FIG. 3 . 3 is a flowchart illustrating a process of generating a vector map.

First, the generator 130 may generate a spatial index in the map data (S205).

The spatial index may assign an index to a map point cloud of spatial information stored in the map data unit 140 for easy search of spatial information, matching with a vector map, and the like. For example, when spatial information is structured in a specific form, an index may be assigned according to a protocol of the structure. In addition, since spatial information is generated in advance according to a predetermined coordinate system based on sensor data previously acquired from the multi-sensor 110, the coordinate system associated with each spatial information, the channel of the map point cloud, the scanning interval, the shape, etc. An index may be assigned based on scan information or the like. In the above embodiment, it is exemplified that the generating unit 130 performs spatial indexing, but in another embodiment, after the map data unit 140 performs spatial indexing, the indexed map point cloud is generated by the generating unit 130 . may be provided in

Next, the generator 130 may determine whether it is a single surface including pointers based on a vector between a predetermined point 12 and adjacent points 12 in the map point cloud ( S210 ).

A process of determining whether it is a single surface will be described in more detail with reference to FIG. 4 . 4 is a diagram illustrating an example of determining whether it is one surface constituting an object based on vector information between mutually adjacent points.

As shown in FIG. 4 , the process of determining whether it is a single surface is a predetermined point 12 and adjacent points 12 distributed in the object 10 identified through the map point cloud by the generator 130 . ) can be used to calculate an eigenvalue using a covariance matrix between As illustrated in the two boxes of FIG. 4 , the covariance matrix may be calculated by deriving a vector between a predetermined point and each adjacent point and using the vectors. Eigenvalues may be calculated as values of vectors of the covariance matrix.

Next, when the minimum eigenvalue is less than the first threshold value, the generator 130 may determine that the surface is one surface including the predetermined point 12 and the adjacent point 12 .

4 as an example, if it is determined that the eigenvalue calculated from the covariance matrix based on the vectors between the points in the upper box is smaller than the first threshold value, the dotted rectangle surrounding the points may be determined as one face. . Alternatively, if it is determined that the eigenvalue in the lower box is greater than the first threshold value, it may be determined that the dotted rectangle corresponding to the lower box is not a single face.

Next, the generator 130 may repeat the process of determining the enlarged surface by determining whether the previously determined one surface is one enlarged surface based on the distance between the previously determined surface and adjacent points ( S215 ).

The process of determining the enlarged surface will be described with reference to FIG. 5 . 5 is a diagram illustrating an example of determining whether an enlarged surface constituting an object by extending from one determined surface to adjacent points.

The generator 130 may determine whether it is one enlarged surface based on the distance between the one surface 14 and the adjacent points 12 shown in the upper box of FIG. 4 . When the distance is smaller than the magnification threshold, the generator 130 may determine a rectangular plane corresponding to reference numeral 16 surrounding one face 14 and the adjacent points 12 as the magnification plane 16 . This process is extended to all points of the object 10, so that the magnification surface 16 may be expanded to include all points that satisfy a predetermined condition such as a relative difference from the magnification threshold. When the enlarged surface 16 is maximally expanded through the repetition of step S215 , a surface equation based on points included in the enlarged surface 16 may be obtained.

In addition, the generating unit 130 repeats the process of steps S210 and S215 for points not included in the one side 14 and the enlarged side 16 during the implementation of steps S210 and S215 to create the object 10 . All the constituent faces can be estimated.

Next, the generator 130 may determine the final surface and the final edge based on the distance and angle between the plurality of enlarged surfaces determined by repeating steps S210 and S215 for all the points 12 of the object 10 . (S220).

Specifically, the determination of the final face and the final edge is determined that, when the angle between the plurality of enlarged faces and the distance between the geometric centroid of each of the enlarged faces and the adjacent enlarged faces are less than a second threshold, the enlarged faces are one By determining to merge into a face, a final face and a final edge can be determined. When the angle and distance are greater than the second threshold, the enlarged faces are determined as final distinct faces in the object 10, and if the separate faces are adjacent to each other, the line between the separate faces can be determined as the final edge. . In addition, the surface determined as the final surface can be obtained by using the points included in the surface equation.

Subsequently, the generator 130 may generate a vector map including the final surface and the map representative vector for each edge ( S225 ).

The map representative vectors V_P1 to V_P6 and V_E1 to V_E6 may be assigned to each final surface and final edge of each object 20m, 22m, and 28m, as in the picture map indicated by the solid line in FIG. 9 . V_P1 to V_P6 and V_E1 to V_E6 are exemplified as map representative vectors for the final face and the final edge, respectively. The map representative vector may be calculated using a surface equation for each final surface and/or a vector between points selected by a predetermined number. Also, the map representative vector for the final edge may be set as a vector of points constituting the edge, or the like.

After generating the vector map in this way, it is possible to generate detection feature information by returning to FIG. 2 . This embodiment does not clearly illustrate that the map point cloud is a colored point. In another embodiment, when the map point cloud is a color, a vector map may be generated using the intensity of RGB data of the map point cloud. In addition, when image data is further included in spatial information, depth data may be included in the fusion data during the correction of geometric information between the map point cloud and the image data and the fusion of data of the multi-sensor 110 . A vector map may be generated by using RGB data and depth data included in the map point cloud, in addition to the distance and reflection intensity included in the data of the map point cloud.

Referring back to FIG. 2 , the generator 130 may generate, in real time, detection feature information including a detection representative vector in at least one part of an object based on a surrounding point cloud obtained in real time from the surrounding environment of the moving object. There is (S110).

Here, the part may be at least one of a face and an edge of an object, and in this embodiment, for convenience of description, it is described that the part includes all of the face and an edge of the object. However, if the detection representative vector of the detection feature information can be expressed to correspond to the map representative vector, the part may be at least one of a surface and an edge. An edge is a line located between adjacent faces.

The detection feature information may be generated by the generator 130 in a similar manner to the vector map, and the generation of the detection feature information will be described in detail with reference to FIG. 6 . 6 is a flowchart illustrating a process of generating detection feature information.

First, the moving object may acquire a surrounding cloud point from the surrounding environment (S305).

As illustrated in FIG. 7 , the movable body 18 may acquire the surrounding environment while moving while mounting the multi-sensor 110 . 7 is a diagram illustrating an example in which a moving object acquires a surrounding point cloud of a surrounding environment. The moving object 18 may use the lidar sensor 114 of the multi-sensor 110 to obtain a surrounding cloud point from the objects 20 and 22 in the surrounding environment. Of course, the positioning sensor 112 , the image sensor 116 , and the inertial sensor 118 included in the multi-sensor 110 may acquire location information of the moving object, an image of the surrounding environment, and the posture of the moving object. In addition, if the multi-sensor 110 further includes an odometry sensor, it is possible to obtain a change value (movement distance, rotation angle, etc.) of the position of the moving object.

Next, the conversion unit 120 may generate odometry information using the multi-sensor 110 of the moving object 18 and may correct data of the surrounding point cloud ( S310 ).

Specifically, the conversion unit 120 is configured to convert between points moved by the moving object based on the image data of the image sensor 116 , location information of the positioning sensor 112 , and data of the surrounding point cloud of the lidar sensor 114 . It is possible to generate odometry information representing relative position data. Although the present embodiment describes an example of generating odometry information even if the odometry sensor is omitted, if the odometry sensor is provided, the converter 120 converts the odometry information based on the data from the odometry sensor. information can be created.

Also, data correction of the surrounding point cloud may be performed by correcting data distortion of the surrounding point cloud with reference to the posture of the moving object obtained from the inertial sensor 118 . The data of the surrounding point cloud may include pointers expressed in objects, the shape of each object according to the distance and reflection intensity between the moving object and each object, the relative position between objects, probability information of points, location information of each object, etc. have. In addition, the conversion unit 120 may converge by correcting geometric information between sensor data, such as data of a neighboring point cloud, image data, and location information of a moving object. Accordingly, the geometrically transformed data of the surrounding point cloud may be corrected with reference to the posture of the moving object.

In addition, when the surrounding point cloud obtained from the lidar sensor 114 is a color, the conversion unit 120 uses RGB data and image data of the surrounding point cloud to correct geometric information and the multi-sensor 110 . ), depth data may be included in the fusion data. The detection feature information may be generated using not only pointers, distances to objects, and reflection intensity included in data of the surrounding point cloud, but also RGB data and depth data included in the surrounding point cloud.

Subsequently, the conversion unit 120 may generate an index in the data of the neighboring point cloud (S315).

For easy search of the neighboring point cloud, matching with map data and/or a vector map, etc., an index may be assigned to the neighboring point cloud based on a lidar channel and scanning information, and the like.

Next, the converter 120 may downsample the corrected data of the neighboring point cloud to a predetermined density in order to reduce the calculation of the detection feature information ( S320 ).

The transform unit 120 may downsample data of the neighboring point cloud, for example, points to a predetermined density in a similar range regardless of the distance to the object.

Subsequently, the generator 130 may generate detection feature information including a detection representative vector based on the corrected surrounding point cloud ( S325 ).

The detection feature information may be generated in a manner similar to the vector map described with reference to FIG. 3 , and this will be described.

First, similar to step S210 , the generator 130 may determine whether it is a single surface including pointers based on a vector between a predetermined point in the neighboring point cloud and adjacent points.

A process of determining whether it is a single surface will be described in more detail with reference to FIG. 8 . FIG. 8 is a diagram illustrating an example of determining whether a surface constitutes an object based on vector information between mutually adjacent points in a process of generating detection feature information.

The points 24 may be selected from points of the remaining channels except for the uppermost and lowermost channels in the LiDAR channels scanned into the object. In addition, the points 24 may be selected as points located in two adjacent channels 26a and 26c at each point 24 and a channel 26b to which each point 24 belongs, as shown in FIG. 8 . have.

In the process of determining whether it is a single surface, similarly to FIG. 4 , an eigenvalue may be calculated using a covariance matrix between the points 24 selected by the generator 130 . The covariance matrix may be calculated by deriving a vector between a predetermined point and each adjacent point illustrated in FIG. 8 and using the vectors. Eigenvalues may be calculated as values of vectors of the covariance matrix.

Next, when the minimum eigenvalue is less than the third threshold value, the generator 130 may determine that it is one surface including a predetermined point and an adjacent point. A detailed description related thereto is omitted because it is similar to step S210.

Next, the generator 130 may repeat the process of determining the enlarged surface by determining whether it is an enlarged surface based on the distance between the previously determined one surface and adjacent points, similar to the step S215 . The process of determining the enlarged surface is similar to the description with reference to FIG. 5 and thus will be omitted.

When the enlarged surface 16 is maximally expanded through the repetition of step S215 , a surface equation based on points included in the enlarged surface 16 may be obtained.

In addition, the generating unit 130 repeats the above steps for points that are not included in one plane and an enlarged plane among the aforementioned steps similar to steps S210 and S215 to estimate all the planes constituting the object. can

Next, the generator 130 determines the final surface for detection and the final edge for detection based on the distances and angles between the plurality of enlarged surfaces determined by repeating the above steps for all points of the object, similarly to the step S220. can

Determination of the final surface for detection and the final edge for detection is substantially the same as in step S220, and thus a detailed description thereof will be omitted.

Accordingly, for the surface determined as the final surface for detection, a surface equation may be obtained using points included in the corresponding surface.

Subsequently, the generator 130 may generate detection feature information including a detection representative vector for each final surface for detection and a final edge for detection.

Detection representative vectors (V_SP1 to V_SP6, V_SE1 to V_SE6) may be assigned to each of the final detection surface and detection edge of each object 20s, 22s, and 28s, as in the detection feature information indicated by the dotted line in FIG. 9 . . V_SP1 to V_SP6 and V_SE1 to V_SE6 are exemplified as detection representative vectors for the last face for detection and the last edge for detection, respectively. The detection representative vector may be calculated using a surface equation for each final surface for detection and/or a vector between points selected with a predetermined number or the like. In addition, the detection representative vector for the final edge for detection may be set as a vector of points constituting the edge, or the like.

After generating the detection feature information in this way, it is possible to return to FIG. 2 to search for the same estimated portion in the vector map and the detection feature information.

Referring back to FIG. 2 , the matching unit 150 may search for a part of the vector map located at the minimum distance from the initial value of the detection feature information ( S115 ).

Specifically, the matching unit 150 sets the current position of the moving object estimated based on the position information of the moving object by the positioning sensor 112, the posture information of the moving object by the inertial sensor 118 and odometry information as an initial value. can decide Next, the matching unit 150 searches for surfaces of a vector map and an exit vector map related to representative vectors having a predetermined similarity between the map representative vector of the vector map and the detection representative vector of the detection feature information based on the initial value. can Accordingly, the matching unit 150 may estimate the searched faces as the same faces of the same object.

That is, the matching unit 150 may search for at least one portion that is estimated to be the same in the vector map and the detection feature information by using the similarity between the map representative vector and the detection representative vector.

The part may be a face and/or an edge, and although an example of searching for a face has been described in the above embodiment, an edge or a combination of face/edge may be searched for in another embodiment.

Next, the position estimator 160 may match the vector map and the detection feature information based on the searched part to estimate the position of the moving object in real time (S120).

Specifically, as shown in FIG. 9 , the position estimator 160 may estimate the position of the moving object by matching the object information of the vector map with the object information of the detection feature information based on the searched part. 9 is a diagram illustrating an example of searching and matching a part of the same object between detection feature information and a vector map.

The object information may include a map representative vector and a detection representative vector defining at least one part of the object, as well as the shape, type, name, and the like of the object. In this embodiment, in order to reduce the computational burden during position estimation, position estimation using a representative vector is taken as an example, but is not limited thereto. An example of estimating the location is not excluded.

As in the example of FIG. 9 , the location estimator 160 detects the map representative vectors (V_P1 to V_P6, V_E1 to V_E6 distributed in each object 20m, 22m, and 28m) of the vector map based on the searched part. By matching the detection representative vectors of feature information (V_SP1 to V_SP6 and V_SE1 to V_SE6 distributed in each object 20s, 22s, and 28s), the position of the moving object 18 can be estimated in real time.

Meanwhile, the location estimator 160 may additionally perform a process of sampling points constituting the surface of the object in the vector map. Thereafter, for accuracy of localization, the location estimator 160 performs an optimization process using the distance between points sampled from the vector map and the surface of the object in the detection feature information to determine the pose of the moving object. A value may be calculated, and the estimated position may be corrected based on the calculated pose value.

According to the present disclosure, it is possible to reduce the amount of computation in matching between the surrounding map data obtained from the moving object and the pre-stored map data, and at the same time, it is possible to secure the accuracy of estimating the location of the moving object while remarkably reducing the capacity of the map data.

The components constituting the apparatus shown in FIG. 1 or the steps according to the embodiments shown in FIGS. 2, 3 and 6 may be recorded in a computer-readable recording medium in the form of a program realizing the function. Here, the computer-readable recording medium refers to a computer-readable recording medium in which information such as data or a program is stored by electrical, magnetic, optical, mechanical, or chemical action. Among these recording media, for example, portable storage, flexible disk, magneto-optical disk, CD-ROM, CD-R/W, DVD, DAT, memory card, etc. may be used as the recording medium detachable from the computer. In addition, as a recording medium fixed to a location estimation device and a computer, there are a solid state disk (SSD), a hard disk, a ROM, and the like.

In addition, even though it has been described above that all components constituting the embodiment of the present disclosure operate by being combined into one, the present disclosure is not necessarily limited to this embodiment. That is, within the scope of the object of the present disclosure, all of the components may operate by selectively combining one or more. In addition, all of the components may be implemented as one independent hardware, but some or all of the components are selectively combined to form a program module that performs some or all of the functions combined in one or a plurality of hardware. It may be implemented as a computer program having

Although the present disclosure has been described in detail through representative embodiments above, those of ordinary skill in the art to which the present disclosure pertains can make various modifications to the above-described embodiments without departing from the scope of the present disclosure. will understand Therefore, the scope of the present disclosure should not be limited to the described embodiments and should be defined by all changes or modifications derived from the claims and equivalent concepts as well as the claims to be described later.

Claims (10)

generating a vector map including a map representative vector in at least one portion of an object based on the map point cloud of the map data;
generating detection feature information including a detection representative vector in at least one portion of an object based on a surrounding point cloud obtained in real time from a surrounding environment of the moving object;
using the similarity between the map representative vector and the detection representative vector to search for at least one portion that is estimated to be the same in the vector map and the detection feature information; and
and estimating the position of the moving object by matching the vector map with the detection feature information based on the searched part.
The method of claim 1,
A method for estimating a map matching position of a moving object using a vector map in which the part is at least one of a face and an edge of the object.
The method of claim 1,
The step of generating the vector map comprises:
determining whether it is one surface including pointers based on a vector between a predetermined point in the map point cloud and adjacent points;
repeating the process of determining an enlarged surface by determining whether it is one enlarged surface based on the distance between the determined surface and adjacent points;
determining a final surface and a final edge based on the distance and angle between the plurality of enlarged surfaces determined in the repeated process; and
and generating a vector map including the map representative vectors for each of the final surface and the final edge.
4. The method of claim 3,
The step of determining whether it is a single surface is,
calculating an eigenvalue using a covariance matrix between the predetermined point and the adjacent points; and
and determining that it is a single surface including the predetermined point and the adjacent point when the minimum eigenvalue is less than a first threshold value.
4. The method of claim 3,
The determining of the final surface and the final edge may include: When an angle between the plurality of enlarged surfaces and a distance between a geometric centroid of each of the enlarged surfaces and adjacent enlarged surfaces are less than a second threshold, the enlargement A map matching position estimation method of a moving object using a vector map for determining the final surface and the final edge by determining that the surfaces are merged into one surface.
The method of claim 1,
The step of generating the detection feature information includes:
determining whether it is one surface including pointers based on a vector between a predetermined point in the neighboring point cloud and the neighboring points;
repeating the process of determining an enlarged surface by determining whether it is one enlarged surface based on the distance between the determined surface and adjacent points;
determining a final surface for detection and a final edge for detection based on the distances and angles between the plurality of enlarged surfaces determined in the repeated process;
and generating detection feature information including the detection representative vectors for each final surface for detection and the final edge for detection.
7. The method of claim 6,
The moving object includes a lidar sensor that acquires the surrounding point cloud, an image sensor that acquires an image of the surrounding environment, positioning sensors that acquire location information of the moving object, and an inertial sensor (IMU) that detects the posture of the moving object. It is equipped with multiple sensors including
Before the step of determining whether it is a single surface including pointers based on a vector between a predetermined point in the neighboring point cloud and adjacent points,
generating odometry information of the moving object by using geometric information between the lidar sensor, the image sensor, and the positioning sensor; and
The method of estimating a map matching position of a moving object using a vector map further comprising correcting distortion generated in the data of the surrounding point cloud based on the posture information of the moving object obtained from the inertial sensor.
8. The method of claim 7,
The step of searching for at least one part estimated to be the same in the vector map and the detection feature information,
determining, as an initial value, a current position of the moving object estimated based on the position information, the posture information, and the odometry information of the moving object; and
Based on the initial value, the vector map and faces of the detection feature information related to representative vectors having a predetermined similarity between the map representative vector of the vector map and the detection representative vector of the detection feature information are searched, and the search is performed. A method for estimating a map matching position of a moving object using a vector map, comprising estimating the surfaces of the same object as identical surfaces.
7. The method of claim 6,
Before the step of determining whether it is a single surface including pointers based on a vector between a predetermined point in the neighboring point cloud and adjacent points,
The method of estimating a map matching position of a moving object using a vector map further comprising the step of downsampling the neighboring point cloud to a predetermined density.
The method of claim 1,
The step of estimating the position of the moving object includes:
estimating the position of the moving object by matching the object information of the vector map with the object information of the detection feature information based on the searched part;
sampling points constituting the surface of the object in the vector map; and
A pose value of the moving object is calculated by an optimization process using the distance between the sampled points and the surface of the object in the detection feature information, and the estimated position is corrected based on the calculated pose value A method for estimating a map matching position of a moving object using a vector map, comprising the step of:

Images