KR20220151829A - Real-time precision road map update method and Service-linked real-time precision road map update system using the same - Google Patents

Abstract

A real-time precision road map update method includes the steps of: collecting, by a data collection device installed in each vehicle, road driving video data, road driving trajectory including latitude, longitude and altitude, acceleration information, angular velocity information, vehicle posture information including roll, pitch, and yaw and vehicle control data including light intensity information; receiving, by a server, the vehicle control data of each vehicle and matching the data by time series and space; calculating, by the server, quality data of the road driving video data in consideration of the road driving trajectory, the acceleration information, the angular velocity information, the vehicle posture information, and the light intensity information; and selecting, by the server, road driving video data for object recognition in consideration of the quality data, internal record data including vehicle statistics for each day and time of the week, and external reference data including road traffic and weather data. The present invention can reduce a server load due to redundant data analysis.

Description

Real-time precision road map update method and Service-linked real-time precision road map update system using the same}

The present invention relates to a road map updating method, and more particularly, to a real-time precision road map updating method and a service-linked real-time precision road map updating system using the same.

Recently, in preparation for the commercialization of self-driving vehicles, a vehicle equipped with a mobile mapping system (MMS: Mobile Mapping System) equipped with multiple sensors for producing precise maps with an error of 20 cm or less (hereinafter referred to as MMS vehicles) has been introduced to provide precise maps. is being put into production.

However, the existing precision map creation method using a mobile mapping system (MMS) vehicle cannot use many vehicles because it requires the use of expensive equipment, and requires considerable manual work in the process of processing the acquired data, which takes a long time. There is a problem that real-time performance is poor because a lot of cost must be invested.

To solve this problem, Korean Patent Publication No. 10-2020-0002217 “Apparatus and Method for Generating and Updating Precise Maps” discloses road linear information (e.g., lanes, road markings, traffic signs) from as many vehicles as possible running on the road. , traffic lights, etc.) is collected and transmitted to the server in real time, so that the server creates a map.

However, Korean Patent Publication No. 10-2020-0002217 extracts road linear information from a sensor collection device mounted on a vehicle and sends only that information to the server. There are disadvantages that are difficult to install on the vehicle.

KR 10-2020-0002217 A

The present invention has been proposed to solve the above technical problems, and after a server collects vehicle control data from a data collection device such as a black box installed in each vehicle, redundant meaningless data is removed through road data mining. To provide a real-time precision road map update method that can calculate road change points and update the map with precision, and a service-linked real-time precision road map update system using the same.

According to an embodiment of the present invention for solving the above problem, road driving video data, road driving trajectory including latitude, longitude and altitude, acceleration information, angular velocity information, Collecting vehicle attitude information including roll, pitch, and yaw, and vehicle control data including light quantity information, and receiving the vehicle control data of each vehicle from the server in time series and space Step of matching data for each type, step of calculating quality data of road driving video data in consideration of road driving trajectory, acceleration information, angular velocity information, vehicle posture information, and light amount information in server, step of calculating quality data of road driving video data in server and selecting road driving video data for object recognition in consideration of external reference data including road traffic and weather data and internal record data including statistics of roadside vehicles by day and time zone Real-time precision road map update including step of selecting A method is provided.

In addition, the present invention includes the steps of extracting an object on the road based on road driving video data for object recognition in a server and calculating a road change point by comparing it with a map with an existing precision, and using the calculated road change point to accurately It is characterized in that it further comprises the step of updating the map.

In addition, according to another embodiment of the present invention, road driving video data installed in each vehicle, a road driving trajectory including latitude, longitude and altitude, acceleration information, angular velocity information, roll, pitch ( A plurality of data collection devices that collect vehicle posture information including pitch, yaw, and vehicle control data including light quantity information, and time series receiving the vehicle control data of each vehicle from the plurality of data collection devices and matching the data by space, calculating the quality data of the road driving video data by considering the road driving trajectory, acceleration information, angular velocity information, vehicle posture information, and light amount information, and quality data, by day of the week and by time of day. A service-linked real-time precision road map updating system is provided, including a server that selects road driving video data for object recognition in consideration of internal record data including shoulder vehicle statistics and external reference data including road traffic and weather data. .

In addition, the server included in the present invention extracts an object on the road based on road driving video data for object recognition, compares it with the existing precision map, calculates road change points, and maps the precision road through the calculated road change points. It is characterized by updating.

In the real-time precision road map updating method and service-linked real-time precision road map updating system using the same according to an embodiment of the present invention, after a server collects vehicle control data from a data collection device such as a black box installed in each vehicle, road data mining After removing redundant meaningless data through , it is possible to calculate road change points and update the map with precision.

In particular, by connecting inexpensive data collection devices such as black boxes (vehicle video recording devices) installed in each vehicle with other services, users are secured, and redundant data is analyzed after removing redundant and meaningless data through road data mining. Server load due to analysis can be reduced.

1 is a conceptual diagram of a service-linked real-time precision road map updating system according to an embodiment of the present invention.
2 is a block diagram of a service-linked real-time precision road map updating system according to an embodiment of the present invention.
3 is a flowchart of a real-time precision road map updating method processed in a service-linked real-time precision road map updating system;

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings in order to describe in detail enough for those skilled in the art to easily implement the technical idea of the present invention.

1 is a conceptual diagram of a service-linked real-time precision road map update system 1 according to an embodiment of the present invention, and FIG. 2 is a configuration diagram of a service-linked real-time precision road map update system 1 according to an embodiment of the present invention. 3 is a flowchart of a real-time precision road map update method processed in the service-linked real-time precision road map update system 1.

The service-linked real-time precision road map update system 1 according to the present embodiment includes only a simple configuration for clearly explaining the technical idea to be proposed.

Referring to FIGS. 1 to 3 , the service-linked real-time precision road map updating system 1 includes a plurality of data collection devices 100 and a server 200 .

The main operations of the service-linked real-time precision road map updating system 1 configured as described above are as follows.

The data collection device 100 may be defined as an information collection device installed in each vehicle, road driving video data, road driving trajectory including latitude, longitude and altitude, acceleration information, angular velocity information, roll ( It collects vehicle attitude information including roll, pitch, and yaw, vehicle control data including light quantity information, and satellite location information.

The data collection device 100 is preferably a device having a function of collecting road driving trajectory and satellite location information including at least road driving video data, latitude, longitude and altitude, and is defined as a vehicle driving recording device. A black box may also be used.

That is, the data collection device 100 installed in each vehicle essentially collects road driving video data, road driving trajectory and satellite location information including latitude, longitude and altitude, and other information (acceleration information, angular velocity information and , vehicle attitude information including roll, pitch, and yaw, and vehicle control data including light quantity information) may be selectively collected according to the type of data collection device.

The server 200 receives the vehicle control data of each vehicle from the plurality of data collection devices 100 and matches the data by time series and by space.

In addition, the server 200 calculates quality data of the road driving video data in consideration of the road driving trajectory, acceleration information, angular velocity information, vehicle posture information, and light amount information.

In addition, the server 200 selects road driving video data for object recognition in consideration of quality data, internal record data including statistics of vehicles on the shoulder by day and time, and external reference data including road traffic and weather data. .

In addition, the server 200 extracts an object on the road based on the road driving video data for object recognition, calculates a road change point by comparing it with an existing precision map, and then calculates a map with precision through the calculated road change point. can be renewed

In addition, the server 200 may provide a predetermined reward to the vehicle owner (information provider) of the data collection device 100 that provided the road change point.

That is, as described above, the process of storing the vehicle control data of the vehicle control data collection device 100 in the server 200 through wireless communication;

A process of classifying a plurality of vehicle control data by space and matching the data in time series;

The process of calculating the quality of the video by combining the position of the video with the Z-axis vibration, brightness, vehicle speed, etc.;

A process of selecting an image for object recognition, including the calculated video quality, existing collected data such as statistics of vehicles parked on the shoulder by day and time, and external reference data such as road traffic volume and weather;

A process of extracting an object on the road through the selected image and comparing it with a map with existing precision to calculate a road change point;

A process of updating the map with precision through the calculated road change points and a process of providing a reward to the change point provider may proceed.

Referring to the flowchart of the real-time precision road map update method processed in the service-linked real-time precision road map update system of FIG. 3,

The real-time precision road map updating method includes road driving video data, road driving trajectory including latitude, longitude, and altitude, acceleration information, angular velocity information, and roll A step of collecting vehicle attitude information including , pitch and yaw, and vehicle control data including light quantity information is performed.

Next, a step of receiving the vehicle control data of each vehicle from the server 200 and matching the data by time series and space is performed.

Next, a step of calculating quality data of the road driving video data in consideration of the road driving trajectory, acceleration information, angular velocity information, vehicle posture information, and light amount information in the server 200 is performed.

Next, the server 200 selects road driving video data for object recognition in consideration of quality data, internal record data including vehicle statistics for each day and time of the week, and external reference data including road traffic and weather data steps are in progress.

Next, the server 200 extracts an object on the road based on the road driving video data for object recognition, compares it with the existing precision map, calculates a road change point, and maps the road with precision using the calculated road change point. The step of updating is in progress. In this case, a step of processing a control system for selective purchase of lane-by-lane data may be further performed.

Finally, a step of providing a reward to the changepoint provider may proceed while updating the map with precision.

For reference, the following vehicle control data (sensing data) collected by the data collection device 100 of each vehicle is managed by matching the data by time series and space.

(1) Road driving video data

(2) Road driving trajectory (latitude, longitude, altitude)

(3) Acceleration and angular velocity (X, Y, Z [vehicle shaking])

(4) Vehicle posture (Roll, Pitch, Yaw)

(5) Light quantity (LUX)

(6) Satellite location information

In addition, examples of external reference data are as follows.

(1) Weather data (sunny, rain, fog, snow, etc.) collected in real time from the Korea Meteorological Administration server

(2) Road traffic collected from the Ministry of Land, Infrastructure and Transport (road traffic information, https://www.its.go.kr/)

In addition, examples of internal recording data (existing collected data) are as follows.

(1) Road condition data by day/time (number of vehicles on the shoulder of the road)

(2) Standard node link data

In addition, the process of calculating the road driving image for object recognition in the server 200 is performed as follows.

(1) The sensing data (vehicle control data) matched in time series is divided into link units using standard node links.

(2) After gathering the sensing data divided by link unit and determining the priority by considering at least one or more of the following items, a road driving image for object recognition is calculated.

- Whether it belongs to a certain average light intensity range (excluding night data)

- Weather Priority (Sunny > Cloudy > Rain > Fog > Snow, except when rain/fog/snow exceeds a certain range)

- Considering the number of vehicles on the shoulder of the road (if the number of vehicles on the shoulder is large, multiple data fusions)

- Consider road traffic (multiple data fusion in case of heavy traffic)

- Considering the straight road trajectory [road driving trajectory, acceleration and angular velocity, vehicle posture data] (distortion due to left and right turns is excluded as much as possible)

- Considering the degree of maintaining a constant speed (exclude distortion due to speed change as much as possible)

Here, after giving absolute scores to each item, the final priority can be given by summing them up. At this time, quality data of the road driving video data is calculated in consideration of the road driving trajectory, acceleration information, angular velocity information, vehicle attitude information, and light amount information, and the calculated quality data and shoulder vehicle statistics by day and time are calculated. It is also possible to select road driving video data for object recognition after priorities are determined in consideration of the included internal record data and external reference data including road traffic and weather data.

On the other hand, the server 200 extracts an object on the road based on road driving video data for object recognition, compares it with the existing precision map, calculates a road change point, and updates the map with precision over a predetermined period when updating the map. Areas that have not been identified can be designated as areas of interest.

Since the server 200 is interlocked with each vehicle's navigation system, when the vehicle's navigation system sets the optimal route for a road driving destination, it can provide information on the region of interest so that it can present a corrected route to pass through the region of interest. have.

In addition, when the road driving video data is a 360-degree circular image, the server 200 divides the circular image to eliminate partial distortion and process the partial image to generate a flat image from the circular image. That is, the unit of the divided circular image (circular division unit image) may be divided in consideration of resolution, arc, pixel, image distortion, and the like.

A method of generating a flat image based on a circular segmentation unit image is as follows.

First, a pixel structure of a circular division unit image may be grasped. Reference focus pixels for the circular segmentation unit image of which the pixel structure is determined may be extracted. A reference focus pixel may be a pixel that does not require expansion or contraction. The pixel structure may be divided into pixels to be reduced and pixels to be expanded based on the reference focus pixel.

A correction process per pixel may be performed to eliminate partial distortion for generating a plane division unit image. For example, reduction may be performed by determining a reduction area for an area expanded based on a reference focus pixel, and expansion may be performed through interpolation after determining an expansion area for the reduced area. A plane segmentation unit image may be extracted through additional image processing in consideration of resolution, arc, pixel, image distortion, and the like. Through this method, a plurality of circular division unit images may be generated as a plurality of planar division unit images.

That is, when the panoramic image is transmitted, the server 200 identifies objects or markers on the screen after generating a plurality of plane division unit images as described above, so that the probability of automatic identification is further increased.

The server 200 may feed back object information recognized in the image to the data collection device 100. Each object information includes an identification code pre-allocated for each object type and location information for each object's central area by time. do.

For example, when an object called “lane” is recognized in the image on the screen, an identification code pre-assigned to the “lane” and location information about the position (coordinates) of the center area of the “lane” by time are transmitted.

For reference, the identification code includes an object code and an additional code. The object code is a code given to the shape of “lane”, and the additional code is defined as coded additional data information such as color and type of “lane”. do.

Therefore, from the point of view of the data collection device 100, it is possible to receive the object information transmitted from the server 200 without directly processing the image processing work, and to grasp the object called “lane” and its moving position, so that autonomous driving and other lane changing functions , lane keeping function, etc.

In addition, the server 200 feeds back a portion of the captured image in which the image does not change to the data collection device 100, and the data collection device 100 automatically removes the portion in which the image does not change, thereby reducing its own storage capacity. can

In addition, after dividing the image into a plurality of regions, the server 200 may feed back information capable of changing the stored image frame for each region to the data collection device 100 in consideration of the movement speed of the object.

Meanwhile, the data collection device 100 may divide a preset capturing area into a plurality of sub-regions, and then independently preset sensing time, sensing objects, and sensing events for each area.

In the real-time precision road map updating method and service-linked real-time precision road map updating system using the same according to an embodiment of the present invention, after a server collects vehicle control data from a data collection device such as a black box installed in each vehicle, road data mining After removing redundant meaningless data through , it is possible to calculate road change points and update the map with precision.

In particular, by connecting inexpensive data collection devices such as black boxes (vehicle video recording devices) installed in each vehicle with other services, users are secured, and redundant data is analyzed after removing redundant and meaningless data through road data mining. Server load due to analysis can be reduced. For reference, data mining is defined as a process of systematically and automatically analyzing statistical rules or patterns in large-scale stored data to extract valuable information.

As such, those skilled in the art to which the present invention pertains will be able to understand that the present invention may be embodied in other specific forms without changing its technical spirit or essential features. Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting. The scope of the present invention is indicated by the following claims rather than the detailed description above, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

100: data collection device
200: server

Claims (7)

Collecting vehicle control data including at least one of road driving video data, road driving trajectory, acceleration information, angular velocity information, vehicle attitude information, and light amount information from a data collection device installed in each vehicle;
receiving the vehicle control data of each vehicle in a server and matching the data by time series and by space; and
calculating quality data of the road driving video data in consideration of the road driving trajectory, the acceleration information, the angular velocity information, the vehicle posture information, and the light amount information in the server;
Real-time precision road map update method comprising a.
According to claim 1,
In the server, selecting road driving video data for object recognition in consideration of the quality data, internal record data including statistics of vehicles on the road by day and time, and external reference data including road traffic and weather data; A real-time precision road map update method further comprising.
According to claim 1,
The road driving trajectory includes latitude, longitude, and altitude, and the vehicle attitude information includes roll, pitch, and yaw.
According to claim 2,
extracting an object on the road based on the road driving video data for object recognition in the server and calculating a road change point by comparing it with a map with an existing precision; and
A real-time precision road map updating method further comprising: updating a map with precision using the calculated road changing points.
According to claim 4,
The real-time precision road map updating method further comprising providing a reward to the change point provider.
Vehicles installed in each vehicle and including road driving video data, road driving trajectory including latitude, longitude and altitude, acceleration information, angular velocity information, roll, pitch, and yaw a plurality of data collection devices for collecting vehicle control data including posture information and light quantity information; and
The vehicle control data of each vehicle is received from the plurality of data collection devices, the data is matched by time series and space, and the road driving trajectory, the acceleration information, the angular velocity information, the vehicle attitude information, and the light quantity are matched. Calculate quality data of the road driving video data in consideration of information, and consider the quality data, internal record data including statistics of shoulder vehicles by day and time, and external reference data including road traffic and weather data A server that selects road driving video data for object recognition;
Service-linked real-time precision road map update system that includes.
According to claim 6,
The server,
Characterized in that an object on the road is extracted based on the road driving video data for object recognition, a road change point is calculated by comparing it with an existing precision map, and the map is updated with precision through the calculated road change point Service-linked real-time precision road map update system.

Images