TW202020406A - Dynamic map classification device and method capable of reducing downloading time for map information - Google Patents

Abstract

The invention discloses a dynamic map classification device and a method thereof. The dynamic map classification device is arranged in a vehicle and comprises at least one automatic driving assistance system, a wireless communication interface, a storage device, a satellite positioning module, and a processor. The wireless communication interface is wirelessly connected to a cloud server. A high-resolution map (HD MAP) and three-dimensional point cloud map information are stored in the cloud server, therefore the satellite positioning module acquires the position coordinates of the vehicle on an electronic map. The storage stores at least one of the road curvature and at least one intersection feature of the road environment in a predetermined travel path of the vehicle. According to the automatic driving assistance degree of at least one road curvature and the intersection feature and the automatic driving assistance system, the processor classifies the map information to be downloaded, thereby reduces downloading time for map information.

Description

Device and method for categorizing dynamic graph resources

The invention relates to a classification technology, and in particular to a dynamic graph resource classification device and method.

For self-driving cars, driving safety is very important. Traditional digital navigation maps can no longer meet the needs of self-driving cars. Therefore, we must rely on high-resolution electronic maps to obtain road environment maps in the travel path and locate ourselves. For the location of the vehicle, if you want to ensure safe driving, a three-dimensional point cloud map is also necessary.

The function of the high-resolution electronic map is to provide road information to self-driving cars, such as lane lines, signs, road curvature, etc., to provide roadside characteristics of self-driving cars to estimate vehicle positions. However, the disadvantage of high-resolution maps is that they must be matched with other positioning technologies to accurately estimate the vehicle position. For example, three-dimensional point cloud map resources need to be synchronized with simultaneous localization and mapping (SLAM, Simultaneous localization and mapping) technology to locate the vehicle position at the time, so as to achieve simultaneous positioning and map construction. However, although high-resolution electronic maps and 3D point cloud maps can provide rich and accurate road environment information, the relative amount of data is huge, and the current 4G network transmission speed is still limited and unstable, so it requires more cost Long download time, but for self-driving cars, a little time delay may cause the system to misjudge and cause extremely serious vehicle accidents.

Therefore, in view of the above-mentioned problems, the present invention proposes a dynamic image data classification device and method to solve the problems caused by the conventional knowledge.

The main purpose of the present invention is to provide a dynamic image data classification device and method, because the data volume of high-resolution maps and 3D point cloud map information is huge, and the storage space of the dynamic image data classification device installed on the vehicle cannot cover all the image data , So you need to put the data to the cloud server to provide self-driving download. Therefore, according to at least one of the road curvature and intersection characteristics of the road environment and the degree of automatic driving assistance of the automatic driving assistance system, the map information to be downloaded is classified and a download request is submitted to the cloud based on the vehicle's position coordinates and speed information Server to download area map information including road environment. According to this, for the simpler road environment, the download amount of area map information can be reduced. For more complex road environments, area map information with higher data volume can be downloaded in advance.

To achieve the above object, the present invention provides a dynamic image data classification device, which is installed in a vehicle and includes at least an automatic driving assistance system, a wireless communication interface, a storage, a satellite positioning module, and a processor. The satellite positioning module obtains the position coordinates of the vehicle on an electronic map. The processor is electrically connected to the satellite positioning module. The processor plans a predetermined driving path on the electronic map according to the driving destination and position coordinates of the vehicle. The vehicle travels on the predetermined driving path. The road environment of the predetermined driving path includes at least one road curvature At least one of the features of at least one intersection. The automatic driving assistance system is electrically connected to the processor. The wireless communication interface is electrically connected to the processor, and wirelessly connects to a cloud server through a wireless network. The cloud server stores high-resolution map (HD Map) and three-dimensional point cloud map information. The storage is electrically connected to the processor, and the storage stores an electronic map and at least one of at least one road curvature and at least one intersection feature. Before the processor uses the satellite positioning module to find that the vehicle reaches the road environment including at least one of the at least one road curvature and the at least one intersection feature, the processor according to at least one of the at least one road curvature and the at least one intersection feature, and The degree of automated driving assistance of the automatic driving assistance system is to classify the high-resolution map and the 3D point cloud map information through wireless communication, and to find the area map information corresponding to the road environment from the high-resolution map or the 3D point cloud map information for download .

The present invention also provides a method for classifying dynamic map data. First, a vehicle is driven on a predetermined driving path by using a vehicle's position coordinates and driving destination on an electronic map. The road environment of the predetermined driving path includes at least At least one of a road curvature and at least one intersection feature. Next, at least one of at least one road curvature and at least one intersection feature is stored. Before the vehicle reaches a road environment including at least one of at least one road curvature and at least one intersection feature, according to at least one of at least one road curvature and at least one intersection feature, and at least one automatic driving assistance provided in the vehicle The degree of automated driving assistance of the system, classify the high-resolution map and 3D point cloud map information stored in the cloud server, and find the area map information corresponding to the road environment from the high-resolution map or 3D point cloud map information for download .

In order to make your reviewer have a better understanding and understanding of the structural features and achieved effects of the present invention, I would like to use the preferred embodiment drawings and detailed descriptions, the explanations are as follows:

The embodiments of the present invention will be further explained in the following with the related drawings. As much as possible, in the drawings and the description, the same reference numerals represent the same or similar components. In the drawings, the shape and thickness may be exaggerated for simplicity and convenience. It can be understood that elements that are not specifically shown in the drawings or described in the specification are in a form known to those of ordinary skill in the technical field. Those of ordinary skill in the art may make various changes and modifications according to the content of the present invention.

Please refer to Figure 1, Figure 2, Figure 3 and Figure 4 below, where Figure 2 is a schematic diagram of a straight road of the present invention, Figure 3 is a schematic diagram of a curved road of the present invention, and Figure 4 is a junction of the present invention Schematic. The following describes the dynamic image data classification device 10 of the present invention, which is installed in a vehicle. The dynamic image classification device 10 includes at least one automatic driving assistance system 12, a wireless communication interface 14, a storage 16, a satellite positioning module 18, and a processor 20, where the number of automatic driving assistance systems 12 is taken as an example The automatic driving assistance system 12 may be, for example, an automatic lane switching system (Lane Changing System, LCS), a lane keeping system (Lane Keeping System, LKS), an automatic emergency braking system (Autonomous Emergency Braking, AEB), and a lane following system (Lane Following System, LFS) or Active Cruise Control (Adaptive Cruise Control, ACC), etc. The satellite positioning module 18 obtains the position coordinates of the vehicle on an electronic map. The processor 20 is electrically connected to the satellite positioning module 18, and the processor 20 plans a predetermined travel path on the electronic map according to the travel destination and position coordinates of the vehicle. The vehicle travels on the predetermined travel path. The road environment 22 of the predetermined travel path includes At least one of at least one road curvature and at least one intersection feature. The automatic driving assistance system 12 is electrically connected to the processor 20. The wireless communication interface 14 is electrically connected to the processor 20 and wirelessly connects to a cloud server 24 through a wireless network. The cloud server 24 stores high resolution map (HD Map) and three-dimensional point cloud map information. Because the amount of data in high-resolution maps and three-dimensional point cloud maps is huge, the storage space of the dynamic map resource classification device cannot cover all the map resources, so the data needs to be placed on the cloud server 24 to provide self-driving download. The storage 16 is electrically connected to the processor 20. The storage 16 stores at least one of at least one road curvature and at least one intersection feature and an electronic map, wherein at least one of the road curvature and intersection feature can be used by the processor 20 wirelessly The communication interface 14 and the cloud server 24 are downloaded from the high-resolution map, but the invention is not limited to this. The electronic map stored in the storage 16 may also include at least one of at least one road curvature and at least one intersection feature. Before the processor 20 uses the satellite positioning module 18 to find that the vehicle reaches the road environment 22 that includes at least one of at least one road curvature and at least one intersection feature, the processor 20 uses at least one of the at least one road curvature and at least one intersection feature One, and the degree of automated driving assistance of the automatic driving assistance system 12, classifies the high-resolution map and the three-dimensional point cloud map information through the wireless communication interface 14, and finds the corresponding road environment 22 from the high-resolution map or the three-dimensional point cloud map information Area map information for download.

In addition, the processor 20 is electrically connected to an Inertial Measurement Unit (IMU) 26 provided in the vehicle. The processor 20 uses the wireless communication interface 14 to obtain the network speed and the file size of the area map information, and uses inertia The measurement unit 26 obtains the speed of the vehicle, and uses the satellite positioning module 18 to obtain the latitude and longitude of the road environment 22 to determine the download area map according to the vehicle's position coordinates, the above-mentioned latitude and longitude, the speed of the vehicle, the network speed and the file size of the area map information The point in time of the information. For example, if the file size of the area map information is 50M bytes (MB), the speed of the vehicle is 45 kilometers per hour (Km/hr), and the network speed is 20M bits per second (bps), then the processor 20 According to the speed, network speed, and file size of the area map information, a time point for downloading the area map information is determined. At this time point, a download request is sent to the cloud server 24 through the wireless communication interface 14 to download the area map information. According to the above conditions, it can be seen that it takes 2.5 seconds to download the area map information, and the vehicle travels 12.5 meters per second. Therefore, the vehicle should be at least 31.25 kilometers away from the next road environment 22 that contains at least one of the road curvature and intersection characteristics. The download of the ruler will start to download the required area map. Therefore, for the more complicated road environment 22, the area map information with higher data volume can be downloaded in advance. For a simpler road environment 22, the download amount of area map information can also be reduced.

The degree of automatic driving assistance of the automatic driving assistance system 12 includes low, medium and high. The American Association of Automatic Machinery Engineering (SAE) classifies autonomous driving levels as No Automation Level 0, Driver Assistance Level 1, Partial Automation Level 2, Conditional Automation Level 3 , High Automation (High Automation) Level 4 and Full Automation (Full Automation) Level 5. Corresponding to this classification, the low of the present invention belongs to Driving Assistance Level 1, the middle belongs to partial automation Level 2 or conditional automation Level 3, and the high belongs to highly automated Level 4 or fully automated Level 5. In general, when the level of automated driving assistance is low, information on lane lines and road curvature is required. The relatively low level will provide more information for the vehicle to use. When there is a danger, the automatic driving assistance system 12 provides the driver with sufficient reaction time. High indicates that the vehicle needs to have a high-precision graphics system for the automatic driving assistance system 12 to drive.

Please refer to Fig. 1, Fig. 2, Fig. 3, Fig. 4 and Fig. 5, the following introduces the dynamic picture classification method of the dynamic picture classification device of the present invention. First, as shown in step S10, the processor 20 uses the position coordinates of the vehicle and the travel destination to plan a predetermined travel path on the electronic map, and the vehicle travels on the predetermined travel path. Next, as shown in step S12, the processor 20 uses the wireless communication interface 14 and the cloud server 24 to download at least one of at least one road curvature and at least one intersection feature from the high-resolution map to store at least one road curvature and at least one At least one of the intersection features is stored in the storage 16; in addition, one of the road curvature and the intersection features can be downloaded from the cloud server 24, and can also be built into the electronic map on the vehicle. Finally, as shown in step S14, the processor 20 uses the satellite positioning module 18 to find that the vehicle arrives at the road environment 22 that includes at least one of at least one road curvature and at least one intersection feature before the vehicle At least one of the features of at least one intersection, and the degree of automated driving assistance of the automatic driving assistance system 12, classify the high-resolution map and the three-dimensional point cloud map information through the wireless communication interface 14, and select from the high-resolution map or the three-dimensional point cloud map Information Search for area map information corresponding to the road environment 22 for download. The area map information has different levels. The higher-level area map information has a larger amount of data and needs to be downloaded early. The lower-level area map information has a smaller amount of data and can be downloaded when it is closer to the intersection feature. However, the area map information is bound to be downloaded before the vehicle reaches the road environment 22.

After step S14, step S16 can also be selectively performed. In step S16, the processor 20 uses the wireless communication interface 14 to obtain the network speed and the file size of the area map information, the inertial measurement unit 26 to obtain the vehicle speed, and the satellite positioning module 18 to obtain the latitude and longitude of the road environment 22, The time of downloading the area map information is determined based on the vehicle's position coordinates, the above-mentioned latitude and longitude, the speed of the vehicle, the network speed, and the area map information file size.

表一The following describes the classification of high-resolution maps and 3D point cloud map information. This classification is shown in Table 1 and Figure 6.

Table I

When the degree of automated driving assistance is low, and before the vehicle reaches the road environment 22 that includes the intersection feature as shown in Figure 4, the area map information includes the second level area map information M2 of the high-resolution map, that is, the intersection feature, where the intersection feature has Traffic signs, zebra crossings or stop lines. When the degree of automated driving assistance is low, and before the vehicle reaches the road environment 22 that contains a road curvature less than a predetermined value but does not include intersection features, the area map information includes the first-level area map information M1 of the high-resolution map, that is, the lane Attributes, which include road curvature, lane lines, speed limits, and number of lanes, where intersection features have traffic signs, zebra crossings, or stop lines. This road curvature is a straight road as shown in Figure 2. When the degree of automated driving assistance is low, and before the vehicle reaches the road environment 22 that includes a road curvature greater than a preset value but does not include intersection features, the area map information includes the first-level area map information M1 of the high-resolution map, which is the lane attribute , Which includes road curvature, lane line, speed limit and lane number. As shown in Figure 3, the curvature of this road is a curved road. The above-mentioned lane attributes may include at least one of lane line, lane number, speed limit and lane curvature, but the invention is not limited thereto. For example, when the automatic driving assistance system 12 is an automatic lane switching system, the lane attributes include lane line, road curvature, lane number, and speed limit. When the automatic driving assistance system 12 is a lane maintenance system, the lane attributes include lane line, road curvature, and speed limit. When the automatic driving assistance system 12 is a lane following system, the lane attributes include lane line, road curvature, lane number and speed limit. When the automatic driving assistance system 12 is an active distance adjustment cruise system, the lane attributes include lane line, lane curvature, and speed limit. When the automatic driving assistance system 12 is an automatic emergency braking system, the lane attribute includes a speed limit. When the degree of automated driving assistance is medium, and before the vehicle arrives at the road environment 22 that contains the intersection feature as shown in Figure 4, the area map information includes the fourth-level area map information M4 of the high-resolution map, that is, the lane attributes, intersection features and dynamics Attributes, where intersection features have traffic signs, zebra crossings, or stop lines; lane attributes include road curvature, lane lines, speed limits, and number of lanes; dynamic attributes include weather, traffic speed, accidents, congestion, emergency ambulance, road construction, and scattered objects , Potholes and signs are abnormal. When the automated driving assistance level is medium, and before the vehicle reaches the road environment 22 that contains a road curvature less than a predetermined value but does not include intersection features, the area map information includes the third-level area map information M3 of the high-resolution map, that is, the lane Attributes and dynamic attributes, where lane attributes include road curvature, lane line, speed limit, and number of lanes; dynamic attributes include weather, traffic speed, accidents, congestion, emergency ambulance, road construction, scattered objects, pits, and signs anomalies. This road curvature is a straight road as shown in Figure 2. When the degree of automated driving assistance is medium, and before the vehicle reaches the road environment 22 that includes a road curvature greater than a preset value but does not include intersection features, the area map information includes the third-level area map information M3 of the high-resolution map, that is, the lane attributes And dynamic attributes, where lane attributes include road curvature, lane line, speed limit, and number of lanes; dynamic attributes include weather, traffic speed, accidents, congestion, emergency ambulance, road construction, scattered objects, pits, and signs anomalies. As shown in Figure 3, the curvature of this road is a curved road.

When the degree of automated driving assistance is high, and before the vehicle reaches the road environment 22 that includes the intersection features as shown in Figure 4, the area map information is the sixth level area map information M6, that is, lane attributes, intersection features, dynamic attributes, and 3D points Cloud map information, combined with simultaneous localization and mapping (SLAM) technology to locate vehicles, where intersection features have traffic signs, zebra crossings, or stop lines; lane attributes include road curvature, lane lines, speed limits, and lanes number. When the degree of automated driving assistance is high, and before the vehicle reaches the road environment 22 that includes a road curvature less than a predetermined value but does not include intersection features, the area map information includes the fifth level area map information M5 of the high-resolution map, that is, the lane Attributes, dynamic attributes, and 3D point cloud map information, where lane attributes include road curvature, lane lines, speed limits, and number of lanes; dynamic attributes include weather, traffic speed, and accidents, etc.). This road curvature is a straight road as shown in Figure 2. When the degree of automated driving assistance is high, and before the vehicle reaches the road environment 22 that includes a road curvature greater than a preset value but does not include intersection features, the area map information includes the fifth level area map information M5 of the high-resolution map, that is, the lane attributes , Dynamic attributes and 3D point cloud map information. This road curvature is a curved road as shown in Figure 3, where the lane attributes include road curvature, lane line, speed limit, number of lanes; dynamic attributes include weather, traffic speed, accidents, congestion, emergency ambulance, road construction, scattered objects , Potholes and signs are abnormal.

In summary, the present invention classifies high-resolution maps and three-dimensional point cloud map information according to at least one of the road curvature and intersection characteristics of the road environment and the degree of automatic driving assistance of the automatic driving assistance system, and corresponds to different conditions according to different conditions. The level of regional map information can further reduce the download time and download volume of regional map information.

The above is only a preferred embodiment of the present invention and is not intended to limit the scope of the implementation of the present invention. Therefore, all changes and modifications based on the shape, structure, characteristics and spirit described in the patent application scope of the present invention are cited. , Should be included in the scope of patent application of the present invention.

10: Dynamic image data classification device 12: Automatic driving assistance system 14: Wireless communication interface 16: Storage 18: Satellite positioning module 20: Processor 22: Road environment 24: Cloud server 26: Inertial measurement unit

FIG. 1 is a block diagram of an embodiment of a dynamic image classification device of the present invention. Figure 2 is a schematic diagram of a straight road of the present invention. Figure 3 is a schematic diagram of a curved road of the present invention. Figure 4 is a schematic view of the intersection of the present invention. FIG. 5 is a flowchart of an embodiment of the dynamic picture information classification method of the present invention. FIG. 6 is a corresponding flowchart of Table 1 of the present invention.

10: Dynamic image data classification device

12: Autonomous driving assistance system

14: Wireless communication interface

16: memory

18: Satellite positioning module

20: processor

24: Cloud server

Claims (8)

A dynamic image data classification device is provided in a vehicle. The dynamic image data classification device includes: a satellite positioning module, which obtains the position coordinates of the vehicle on an electronic map; a processor, which is electrically connected to the satellite A positioning module, the processor plans a predetermined driving path on the electronic map according to the driving destination of the vehicle and the position coordinates, the vehicle travels on the predetermined driving path, and the road environment of the predetermined driving path includes at least one road At least one of curvature and at least one intersection feature; at least one automatic driving assistance system, electrically connected to the processor; a wireless communication interface, electrically connected to the processor, and wirelessly connected to a cloud server through a wireless network, The cloud server stores high resolution map (HD Map) and 3D point cloud map information; and a storage, electrically connected to the processor, the storage stores the electronic map and the at least one road curvature and the at least one At least one of the intersection features, before the processor uses the satellite positioning module to find that the vehicle reaches the road environment including at least one of the at least one road curvature and the at least one intersection feature, the processor determines At least one of the at least one road curvature and the at least one intersection feature, and the degree of automatic driving assistance of the automatic driving assistance system, classify the high-resolution map and the three-dimensional point cloud map information through the wireless communication interface, and Find the area map information corresponding to the road environment from the high-resolution map or the three-dimensional point cloud map information for download. The dynamic image data classification device according to claim 1, wherein at least one of the at least one road curvature and the at least one intersection feature is used by the processor from the high-resolution map using the wireless communication interface and the cloud server download. The dynamic image data classification device according to claim 1, wherein the electronic map stored in the storage includes at least one of the at least one road curvature and the at least one intersection feature. The dynamic image data classification device according to claim 2 or 3, wherein the processor is electrically connected to an inertial measurement unit (IMU) provided in the vehicle, and the processor uses the wireless communication interface to obtain Internet speed and the file size of the map information of the area, and use the inertial measurement unit to obtain the speed of the vehicle, and use the satellite positioning module to obtain the latitude and longitude of the road environment, according to the position coordinates, the latitude and longitude, the vehicle's The speed, the network speed and the file size of the area map information determine the time point of downloading the area map information. A dynamic image data classification method, which includes the following steps: using a vehicle's position coordinates and driving destination to plan the predetermined driving path on an electronic map, the vehicle driving on the predetermined driving path, the predetermined driving path The road environment includes at least one of at least one road curvature and at least one intersection feature; storing at least one of the at least one road curvature and the at least one intersection feature; and when the vehicle arrives includes the at least one road curvature and the at least one Before the road environment of at least one of the intersection features, according to at least one of the at least one road curvature and the at least one intersection feature, and the degree of automatic driving assistance of at least one automatic driving assistance system provided in the vehicle , Classify the high-resolution map and 3D point cloud map information stored in the cloud server, and find the area map information corresponding to the road environment from the high-resolution map or the 3D point cloud map information for download. The dynamic map classification method according to claim 5, wherein at least one of the at least one road curvature and the at least one intersection feature refers to being stored in the electronic map. The dynamic image data classification method according to claim 5, wherein at least one of the at least one road curvature and the at least one intersection feature refers to a download from the cloud server and storage. The dynamic image data classification method according to claim 6 or 7, wherein after classifying the high-resolution map and the three-dimensional point cloud map information stored in the cloud server, the vehicle speed and network speed are obtained , The file size of the area map information and the latitude and longitude of the road environment, according to the location coordinates, the latitude and longitude, the speed of the vehicle, the network speed and the file size of the area map information to download the area map information Time.

Images