KR20230072060A - High precision map data processing and conversion method for autonomous driving - Google Patents

Abstract

A high-precision map data processing and conversion method for autonomous driving according to one embodiment of the present invention may be executed in a high-precision map data processing and conversion device for autonomous driving, and may comprise: a step of, when precision road map data is received from the national territorial information platform, filtering particular vertices and the height value of the corresponding vertices from one link in the precision road map data before generating filtering data; a step of using a spline interpolation method to process the coordinates of the vertices and the height values of the vertices in the filtering data before fragmenting the intervals between the current vertex and the next vertex; a step of, when a plurality of vertices are generated after the interval between the current vertex and the next vertex is fragmented, the distance between a first vertex and a second vertex may be calculated according to the coordinate interval between the first vertex and the second vertex among a plurality of vertices to store the same as the current distance; and a step of outputting each of the coordinate value of the starting vertex and the coordinate value of the end vertex corresponding to the previous length depending on whether the current distance is a predetermined critical distance or lower. Therefore, data may be processed by using a high-precision map.

Description

High-precision map data processing and conversion method for autonomous driving {HIGH PRECISION MAP DATA PROCESSING AND CONVERSION METHOD FOR AUTONOMOUS DRIVING}

The present invention relates to a method for processing and converting high-precision map data for autonomous driving, and more specifically, autonomous driving that processes the data by using a high-precision map of road data necessary for autonomous driving by utilizing a high-precision map provided by the National Geospatial Information Platform. It relates to a high-precision map data processing and conversion method for

Currently, the system for producing high definition electronic maps is MMS (Mobile Mapping) equipped with equipment such as DGPS (Differential GPS) receiver, multiple cameras, IMU (inertial measuring unit), and LiDAR (Light Detection and Ranging). System) is configured to scan roads and facilities around roads.

Worldwide, these devices are increasing map accuracy from tens of meters to 50 cm.

Meanwhile, an autonomous vehicle refers to a vehicle that drives by itself without driver's intervention, recognizes the surrounding environment and driving situation, and controls driving of the vehicle according to the recognition result to drive itself to a preset destination.

Recently, research and development related to such self-driving vehicles have been actively conducted. Thanks to this, the traffic accident rate of autonomous vehicles has decreased, and traffic efficiency and passenger convenience have improved. Accordingly, autonomous vehicles are attracting attention as a next-generation means of transportation.

However, self-driving vehicles inevitably require precise electronic maps including key road information in preparation for malfunctions of configured devices or malfunctions due to obstacles.

For this reason, existing electronic maps with errors of several tens of meters cannot be used due to their low accuracy. Therefore, in order to secure an accuracy of less than 50 cm positioning error, high-precision field positioning is performed using a high-precision map production system equipped with MMS equipment.

According to Patent Registration No. 10-2272143 (3D High-precision Road Mapping System for Autonomous Driving), it relates to "3D High-precision Road Mapping System for Autonomous Driving, and more specifically, voice recognition and roadside shooting. It is about a 3D high-precision map production system for autonomous driving that has been improved to automatically generate road alignments when producing high-precision electronic maps for self-driving vehicles using field survey method using image information and positioning information." has been disclosed.

In detail, “GPS is increasing the map accuracy from several tens of meters to 50 cm through these equipments worldwide.” Accuracy is dramatically improved." is disclosed in the background art.

In addition, the precision road map provided by the National Territory Information Platform is “the lane necessary for autonomous driving (regulation line, road boundary line, stop line, lane center line), road facilities (median strip, tunnel, bridge, underpass), marking facilities (traffic safety Signs, road markings, traffic lights) information is produced in 3D.” It was defined as an electronic map.

Although the National Geospatial Information Platform provides map data with the precision required for autonomous driving, there is a disadvantage of not being more detailed. Less detailed is the wide spacing between points in a link.

Here, a link means one lane on a road. If the distance between points in the link is wide, road data is mainly used for self-driving vehicles without cameras, and road sensor (Radar, LiDAR, IMU, GNSS) data is used. The details of the data are important.

The above LiDAR is a relative positioning method that measures the distance to a target point based on the absolute location of all facilities around the road. Most of the facilities around the middle road can be expressed as a point cloud, and the measurable range is generally within a relative distance of 70 m, and the accuracy is less than 2 cm.

In addition, the IMU continuously measures the relative distance such as height and curvature based on the absolute position through calculation considering speed and acceleration, etc. play an auxiliary role in

Otherwise, a phenomenon that is not smooth occurs when changing lanes or driving on a curve turning left or right during autonomous driving.

For example, if the distance between the first point and the next point widens, the displacement value between the points increases. Therefore, in order to meet the time to reach the next point as much as the displacement value increases, an emergency lane line when changing lanes This means that there is a greater likelihood of change.

Therefore, if the distance between points is narrow, the time for the self-driving vehicle to go from the first point to the next point is shortened as much as the displacement is reduced, enabling smoother driving.

Patent Registration No. 10-2272143 (2021.06.28) Registered Patent No. 10-1871441 (2018.06.20)

An object of the present invention is to provide a high-precision map data processing and conversion method for autonomous driving that processes the data using a high-precision map of road data necessary for autonomous driving by utilizing a high-precision map provided by the National Geospatial Information Platform.

The objects of the present invention are not limited to the above-mentioned objects, and other objects and advantages of the present invention not mentioned above can be understood by the following description and will be more clearly understood by the examples of the present invention. It will also be readily apparent that the objects and advantages of the present invention may be realized by means of the instrumentalities and combinations indicated in the claims.

High-precision map data processing and conversion method for autonomous driving executed in a high-precision map data processing and conversion device for autonomous driving to achieve this purpose is, when precision map data is received from the National Geospatial Information Platform, among the high-precision map data Generating filtering data after filtering only specific vertices and height values of the corresponding vertices in one link, and calculating the coordinates of the vertices and the height values of the vertices among the filtered data using spline interpolation. Subdividing the interval between the current point and the next point by processing, when a plurality of points are generated by subdividing the current point and the next point, the coordinate interval between the first and second points among the plurality of points calculating the distance between the first and second points according to the distance and storing the distance as the current distance, and the coordinate value of the starting point and the ending point corresponding to the previous length according to whether the current distance is less than or equal to a predetermined threshold distance, respectively. It includes the step of outputting.

According to the present invention as described above, there is an advantage in that road data necessary for autonomous driving can be processed using the high-precision map by utilizing the high-precision map provided by the national land information platform.

1 is a block diagram illustrating a system for processing and converting high-precision map data for autonomous driving according to an embodiment of the present invention.
2 is a flowchart illustrating an embodiment of a method for processing and converting high-precision map data for autonomous driving according to the present invention.
3 to 6 are exemplary diagrams for explaining a process of processing and converting high-precision map data for autonomous driving according to an embodiment of the present invention.

The above objects, features and advantages will be described later in detail with reference to the accompanying drawings, and accordingly, those skilled in the art to which the present invention belongs will be able to easily implement the technical spirit of the present invention. In describing the present invention, if it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted. Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to indicate the same or similar components.

Among the terms used in this specification, “link” means one lane on a road.

1 is a block diagram illustrating a system for processing and converting high-precision map data for autonomous driving according to an embodiment of the present invention.

Referring to FIG. 1 , an apparatus 100 for processing and converting high-precision map data for autonomous driving includes a data input unit 110 , a data processing unit 120 and a data comparison unit 130 .

The data input unit 110 receives map data with high precision from the national land information platform. At this time, the precision road map data is a shp file, which is data that provides information necessary for precise roads and traffic among the geographic information system, and defines spatial data that provides spatial location and object properties. It consists of attribute data (Attribute) that

In one embodiment, the precision map data is a unique identification number of the speed information of the road (Max_Speed), a unique identification number of the starting point of the road (FromNodeID), a unique identification number of the end point of the road (ToNodeID), and a unique identification of the distance of the road. Number (Length), unique identification number (L_Link_ID) of the left lane on the road, and unique identification number (R_Link_ID) of the left lane on the road are stored in vector form.

At this time, the center line of the road is given a unique identification number, LINK_ID (region code + serial number + direction identifier + lane), and the unique identification number (L_Link_ID) of the left lane of the road is "L" indicating the left side of the center line and It is created by combining the unique identification number "LINK_ID" of the center line, and the unique identification number (R_Link_ID) of the right lane of the road is created by combining "R" indicating the right side of the center line and "LINK_ID", the unique identification number of the center line.

The data input unit 110 filters only specific vertices and height values of the corresponding points in one link of the map data with precision from the data input unit 110, and then converts the filtered data to the data processing unit 120. ) is provided.

The data processing unit 120 receives filtered data extracted from the data input unit 110 by filtering only the points and the height values of the corresponding points among the precision map data and processes them.

In one embodiment, the data processing unit 120 processes the coordinates of points and height values of the points in the filtered data using the spline interpolation method, and subdivides and extracts the intervals between the points.

The above spline interpolation method is a method of obtaining a smooth function with a low-order polynomial by dividing the entire section into sub-sections. Spline interpolation provides a good approximation to the behavior of a locally rapidly changing function and utilizes low-order polynomials.

That is, when there are N points in the spline interpolation method, the data processing unit 120 divides between the N points into (n-1) subsections, each subinterval is i, and the spline function for each subinterval is S _i , the interpolation point in each subsection must be defined as in [Equation 1] below.

[Equation 1]

y _i = S _i (x _i ) (i=0, 1, ??, m)

i: subinterval divided between N points,

S _i (): subinterval spline function

Through the above process, the data processing unit 120 subdivides the interval between the current point and the next point to generate a plurality of points. Then, the data processing unit 120 applies the coordinates of a plurality of points to the Pythagorean theorem to calculate a straight line distance between points and stores it as the current distance, and between the points until the current distance becomes close to a predetermined threshold distance. The current distance can be calculated again by deleting the point in and calculating the straight line distance.

To this end, the data processing unit 120 stores the distance between the first and second points as the current distance according to the coordinate interval between the first and second points among the plurality of points.

Then, the data processor 120 checks whether the current distance is equal to or less than a predetermined threshold distance (eg, 30 cm).

If the current distance is less than a predetermined threshold distance (eg, 30 cm), the data processing unit 120 deletes the second point so that the distance between the coordinates of the first point and the coordinates of the third point is combined. The distance between the first point and the third point is calculated and stored as the current distance.

After that, the data processing unit 120 checks once again whether the current distance is equal to or less than a predetermined threshold distance (eg, 30 cm). As described above, the present invention deletes the points (ie, the third point, the fourth point ??) between the points while repeatedly performing the above process until the current distance is less than or equal to the predetermined threshold distance, thereby increasing the distance between the coordinates. are merged and the current distance is saved.

Meanwhile, if the current distance is greater than or equal to the predetermined threshold distance, the data processing unit 120 compares the previous distance and the current distance with the predetermined threshold distance to calculate a first difference value and a second difference value.

Then, the data processing unit 120 may determine that any one of the previous distance and the current distance is closer to the predetermined threshold distance by using the first difference value and the second difference value.

In an embodiment, when the first difference value is smaller than the second difference value, the data processing unit 120 may determine that the previous distance is closer to a predetermined threshold distance among the previous distance and the current distance.

As in the above embodiment, when the previous distance is determined to be closer to the predetermined threshold distance, the data processing unit 120 may output coordinate values of a starting point and an ending point corresponding to the previous distance.

In another embodiment, when the first difference value is greater than the second difference value, the data processing unit 120 may determine that the current distance is closer to a predetermined threshold distance among the previous distance and the current distance.

As in the above embodiment, the data processing unit 120 ends when it is determined that the current distance is closer to a predetermined threshold distance.

The data comparison unit 130 calculates the distance between the plurality of points received from the data processing unit 120 (ie, the coordinate value of the starting point and the coordinate value of the ending point corresponding to the previous distance), and the distance is at a predetermined threshold distance. If so, it is determined that data processing is normally completed.

That is, the data comparator 130 selects the first point from among the plurality of received points, calculates the coordinates of the second point and obtains the straight line distance using the Pythagorean theorem, and if the straight line distance corresponds to a predetermined critical distance, data processing is normally performed. can be considered complete.

2 is a flowchart illustrating an embodiment of a method for processing and converting high-precision map data for autonomous driving according to the present invention.

Referring to FIG. 2, when the high-precision map data processing and conversion device 100 for autonomous driving receives precision map data from the land information platform, a specific point (Vertices) in one link of the high-precision map data ) and filtering only the height value of the corresponding point, and then filtering data is generated (step S210).

The apparatus 100 for processing and converting high-precision map data for autonomous driving subdivides the interval between the current point and the next point by processing the coordinates of points and height values of the points in the filtering data using the spline interpolation method.

When the high-precision map data processing and conversion apparatus 100 for autonomous driving is subdivided between the current point and the next point to generate a plurality of points, according to the coordinate interval between the first and second points among the plurality of points. The distance between the first and second points is calculated and stored as the current distance (step S220).

The apparatus 100 for processing and converting high-precision map data for autonomous driving deletes a second point among a plurality of points when the current distance is equal to or less than a predetermined threshold distance (step S230) to coordinates of the first point to the coordinates of the third point. After the intervals between the points are combined, the distance between the first to third points is calculated and stored as the current distance (step S240). At this time, as the distance between the first and third points is stored as the current distance, the distance between the first and second points is changed to the previous distance.

After that, the apparatus 100 for processing and converting high-precision map data for autonomous driving checks once again whether the current distance is equal to or less than a predetermined threshold distance (eg, 30 cm). As described above, the present invention deletes the points (ie, the third point, the fourth point ??) between the points while repeatedly performing the above process until the current distance is less than or equal to the predetermined threshold distance, thereby increasing the distance between the coordinates. are merged and the current distance is saved.

At this time, the high-precision map data processing and conversion apparatus 100 for autonomous driving, if the current distance is greater than or equal to a predetermined threshold distance (step S230), compares the previous distance and the current distance, and the current distance is close to the predetermined threshold distance (step S230). Step S250) ends, and if the previous distance is close to the predetermined threshold distance, each of the coordinate values of the starting point and the ending point corresponding to the previous distance is output (step S260).

In one embodiment of step S260, the high-precision map data processing and conversion apparatus 100 for autonomous driving compares each of the previous distance and the current distance with the predetermined threshold distance when the current distance is greater than or equal to the predetermined threshold distance, A first difference value and a second difference value are calculated.

Thereafter, the apparatus 100 for processing and converting high-precision map data for autonomous driving adds any one of the previous distance and the current distance to the predetermined threshold distance by using the first difference value and the second difference value. It can be judged as a close distance.

In one embodiment, the apparatus 100 for processing and converting high-precision map data for autonomous driving, when the first difference value is smaller than the second difference value, the previous distance among the previous distance and the current distance is greater than a predetermined threshold distance. It can be judged as a close distance.

As in the above embodiment, when the apparatus 100 for processing and converting high-precision map data for autonomous driving determines that the previous distance is closer to a predetermined threshold distance, the coordinate values of the starting point and the coordinates of the ending point corresponding to the previous distance Each value can be output.

In another embodiment, in the apparatus 100 for processing and converting high-precision map data for autonomous driving, when the first difference value is greater than the second difference value, the current distance among the previous distance and the current distance is greater than a predetermined threshold distance. It can be judged as a close distance.

As in the above embodiment, the apparatus 100 for processing and converting high-precision map data for autonomous driving ends when it is determined that the current distance is closer to a predetermined threshold distance.

3 to 6 are exemplary diagrams for explaining a process of processing and converting high-precision map data for autonomous driving according to an embodiment of the present invention.

3 to 6, the apparatus 100 for processing and converting high-precision map data for autonomous driving receives map data with high precision from the land information platform. At this time, the precision road map data is a shp file as shown in FIG. 3, which is data that provides information necessary for precise roads and traffic among the geographic information system, spatial data that provides spatial location, and objects It is composed of attribute data (Attribute) that defines the attribute of

The apparatus 100 for processing and converting high-precision map data for autonomous driving generates filtering data after filtering only specific vertices and height values of the corresponding vertices in one link of map data with precision. . At this time, the apparatus 100 for processing and converting high-precision map data for autonomous driving may filter necessary data from a link as shown in FIG. 4 using a QGIS3 program. Figure 4 (a) shows a part of the list of points (Vertices) in one link (Link), Fig. 4 (b) shows a part of the list of heights (Height) in one link (Link).

After that, the high-precision map data processing and conversion apparatus 100 for autonomous driving uses the spline interpolation method as shown in FIG. extract As shown in FIG. 5, green dots represent data provided by the precision map, and red dots represent data generated by using spline interpolation between green dots.

In FIG. 5, when the distance between the green dots is wide, the displacement value between the green dots increases, so the displacement value increases to meet the time to reach the next point. will grow

However, as in the present invention, since a plurality of red dots are generated using the spline interpolation method between green dots, the distance between dots is narrowed, so the time required for the autonomous vehicle to go from the current point to the next point as much as the displacement is reduced. It has the advantage of being able to drive a little smoother because it is shorter.

When the high-precision map data processing and conversion apparatus 100 for autonomous driving is subdivided between the current point and the next point to generate a plurality of points, according to the coordinate interval between the first and second points among the plurality of points. The distance between the first and second points is calculated and stored as the current distance. Depending on whether the current distance is less than or equal to a predetermined threshold distance, as shown in FIG. 6 (a), the coordinate value of the starting point and the ending point corresponding to the current distance Each coordinate value can be output.

Although described by the limited embodiments and drawings, the present invention is not limited to the above embodiments, and those skilled in the art can make various modifications and variations from these descriptions. Therefore, the spirit of the present invention should be grasped only by the claims described below, and all equivalent or equivalent modifications thereof will be said to belong to the scope of the spirit of the present invention.

100: High-precision map data processing and conversion device for autonomous driving,
110: data input unit,
120: data processing unit,
130: data comparison unit

Claims (4)

In the high-precision map data processing and conversion method for autonomous driving executed in a high-precision map data processing and conversion device for autonomous driving,
When receiving precision map data from the land information platform, generating filtering data after filtering only specific vertices and height values of the corresponding points in one link of the precision map data;
subdividing a space between a current point and a next point by processing coordinates of points and their height values among the filtering data using a spline interpolation method;
When a plurality of points are generated by subdividing between the current point and the next point, the distance between the first and second points is calculated according to the coordinate interval between the first and second points among the plurality of points, store as distance; and
And outputting each of the coordinate values of the starting point and the ending point corresponding to the current distance according to whether the current distance is less than or equal to a predetermined threshold distance.
High-precision map data processing and conversion method for autonomous driving.
According to claim 1,
The step of outputting each of the coordinate values of the starting point and the ending point corresponding to the current distance according to whether the current distance is less than or equal to a predetermined threshold distance
If the current distance is less than or equal to the predetermined threshold distance, the second point is deleted from among the plurality of points so that the distances between the coordinates of the first point and the third point are combined, and then the distance between the first and third points is determined. Comprising the step of calculating the distance and storing it as the current distance
High-precision map data processing and conversion method for autonomous driving.
According to claim 2,
The step of outputting each of the coordinate values of the starting point and the ending point corresponding to the current distance according to whether the current distance is less than or equal to a predetermined threshold distance
calculating a first difference value and a second difference value by comparing each of the previous distance and the current distance with the predetermined threshold distance when the current distance is greater than or equal to the predetermined threshold distance;
determining that any one of the previous distance and the current distance is closer to the predetermined threshold distance by using the first difference value and the second difference value; and
When it is determined that the previous distance is closer to the predetermined threshold distance, outputting each of the coordinate values of the starting point and the ending point corresponding to the previous distance
High-precision map data processing and conversion method for autonomous driving.
According to claim 3,
Determining that any one of the current distance and the current distance is a distance closer to the predetermined threshold distance using the first difference value and the second difference value
determining that the previous distance and the current distance are closer to a predetermined threshold distance if the first difference value is less than the second difference value; and
And determining that the current distance among the previous distance and the current distance is closer to a predetermined threshold distance when the first difference value is greater than the second difference value.
High-precision map data processing and conversion method for autonomous driving.

Images