DE102017210830A1 - Create and use an edge development map - Google Patents

Abstract

A method of creating a lane boundary map of a lane includes steps of determining obstacle distances, each indicating the depth of an obstacle-free space to the side of the lane at an associated altitude; determining distance profiles respectively indicating the course of one of the obstacle distances along the lane at the respective height; mapping the distance gradients into a reference altitude; and providing the depicted distance gradients.

Description

The invention relates to a peripheral building map. In particular, the invention relates to the generation and use of a highly accurate edge development map for a motor vehicle.

A motor vehicle is equipped with a navigation system that allows it to determine its own position, for example by means of a receiver for a satellite navigation system, and to set in relation to positions of other objects, the information being stored in a map memory. The accuracy of the position determination for the motor vehicle is usually about one meter and information in a conventional card memory have a similar accuracy. If the motor vehicle is to be controlled partially or completely autonomously in the longitudinal and / or transverse direction, more precise information regarding the position of the motor vehicle and objects in its surroundings is required.

One object of the present invention is to provide an improved technique for generating and using a peripheral building map, in particular in connection with a partially or completely autonomously controllable motor vehicle. The invention solves this problem by means of the subjects of the independent claims. Subclaims give preferred embodiments again.

According to a first aspect, a method of constructing an edge development map of a lane comprises steps of determining obstacle distances each indicating the depth of an obstacle-free space at the side of the lane at an associated height; determining distance profiles which respectively indicate the course of one of the obstacle distances along the lane in the respective height of the assigned obstacle distances; mapping the distance gradients into a reference altitude; and providing the depicted distance gradients.

This makes it possible to bring information about objects that can be detected on the side of a lane, in a universally usable form that can be exploited by different applications on different vehicles. The information about the edge development can be reduced so that their memory consumption is acceptable, at the same time detailed or accurate enough to support demanding positioning or control tasks of the motor vehicle. Once provided distance gradients can be easily adjusted based on new information, without requiring the original original data. Access to the distance profiles shown can be done quickly. The distance gradients can be easily related to the course of the lane, so that fast access is possible. The generated distance plots produced can be provided by means of a generic interface.

By mapping to the reference altitude, characteristic properties of the distance gradients can be preserved, while at the same time the data can be processed more easily. A required storage space can be reduced. The reference altitude is preferably fixedly predetermined and is preferably within a height range covered by the obstacle distances. The reference altitude may coincide with the predetermined altitude of one of the obstacle distances. The reference altitude is typically related to a vertical altitude above the lane and may be zero in a further preferred embodiment.

It is preferred that mutually identical sections of the depicted distance courses are reduced to one section. Sections on which several distances are superimposed can be stored in a space-saving way. It can be determined how far apart the sections may be to be considered identical. Such a distance may be, for example, a few centimeters or millimeters.

It is further preferred that the depicted distance courses are stored in vector form. On the one hand, this can further reduce the memory requirement for the distance profiles depicted; on the other hand, the accuracy can be easily adapted or selected as a function of a surrounding system. Potentially, the achievable accuracy is significantly higher than with a storage that relates to a predetermined grid of the environment.

It is of particular advantage that the obstacle distances can come from different data sources. It is also possible to merge different data sources with each other. The accuracy or reliability of the generated data can thus be increased. Existing data collections can be used in an improved way, even retroactively, in order to improve existing distance courses or to make them more accurate. For example, cartographic data of objects along the lane can be used as well as current scans that are created on board the motor vehicle.

The distance plots shown may be provided together with an indication of an underlying data source of the obstacle distances. As a result, it is possible to improve using a peripheral building map Application to assess the quality, timeliness or intention of the data source on the basis of which the obstacle distances were determined. Among several mapped distance histories, one can be selected that goes back to a predetermined data source.

The mapped distance plots may also be provided along with an indication of the relative location of points along the lane. The edge development map can thus be better managed. In particular, access to distance profiles can be made easier or faster. The depicted distance profiles can be better related to positions that are also defined with respect to points along the lane, such as the positions of visible objects along the lane.

For a portion of the depicted distance courses along the lane, an indication of an object can be provided which limits the obstacle-free space of one of the distance courses in the depth direction. The depth direction preferably extends perpendicular to the direction of travel of the motor vehicle in a direction parallel to a ground, in particular a lane, or parallel to the normal zero area. In other words, it can be indicated in the depicted distance profiles which object delimits the distance profile at a predetermined location along the lane. The evaluation or use of the distance profiles shown can be further improved.

Before determining a distance course, the surface of an object can be determined which limits the obstacle-free space in the depth direction. For example, a measurement error, measurement noise or a seasonal deviation from comparison values can be better interpreted. If, for example, the object is a tree or a shrub, a greater deviation must be expected in the case of repeated sampling than, for example, in the case of a building such as a house.

The surface between certain obstacle distances can be interpolated. The scanned surface can be closed by post-processing. Vegetation like a shrub can be less realistic, but it can be modeled better and therefore more useful.

It is further preferable that noise of the obstacle distances is reduced. In particular, a low-pass filter can be used for this purpose. It is also possible to use a plurality of noise filters at different intermediate or final results of the method. The comparability of the achievable mapped distance profiles with, for example, current measurements can thereby be increased.

Generally, it is preferred that portions of the depicted distance courses be organized in tiles, with an indication of another map information of the same section being provided for each tile. In particular, the tiles may be formed in a manner used in the Navigation Data Standard (NDS). In this case, the depicted distance courses can be stored on a dedicated map level, in particular level 13, and preferably referenced with objects or elements of other levels.

An apparatus for establishing an edge development map of a lane comprises an interface for receiving obstacle distances, each of which indicates the depth of an obstacle-free space at the side of the lane at an assigned height; a storage device and a processing device. In this case, the processing device is set up to determine distance profiles on the basis of the determined obstacle distances, each describing the course of an obstacle distance along the traffic lane at a predetermined height. Furthermore, the processing device is set up to map the determined distance courses to a reference height and to provide the distance profiles shown.

The device can be used for example on board a motor vehicle. The processing device may in particular comprise a programmable microcomputer or microcontroller. The device and in particular the processing device can be configured to carry out the method described above in whole or in part. For this purpose, the method may be in the form of a computer program product with program code means. Features and benefits of the method can be transferred to the device and vice versa.

According to a second aspect, a method for determining the position of a motor vehicle along a lane comprises steps of determining obstacle distances, each indicating the depth of an obstacle-free space to the side of the lane at an associated height; determining distance profiles respectively indicating the course of one of the obstacle distances along the lane at the respective height; mapping the distance gradients into a reference altitude; comparing the mapped distance plots with predetermined mapped distance plots associated with an indication of the location of at least one point along the lane; and determining the position of the motor vehicle based on the at least one point.

The determination of the distance profiles shown can be carried out as in the above-mentioned method, wherein the obstacle distances are preferably currently scanned from the motor vehicle. The distance profiles shown can then be compared with the distance profiles shown, which can be stored in an edge development map according to the method described above. Determining the position of the motor vehicle with respect to the distance plots depicted is a preferred application of the results of the method described above. The position of the motor vehicle along the lane can be determined quickly and with very high accuracy. As a result, for example, a method for controlling the motor vehicle in the longitudinal and / or transverse direction on the basis of the highly accurately determined position of the motor vehicle can function better.

A device for determining a position of a motor vehicle along a lane comprises a scanning device for determining obstacle distances, each of which indicates the depth of an obstacle-free space at the side of the lane at an assigned height; a data source for an edge building map; and a processing device. The processing device is set up to determine at least one distance profile based on the determined obstacle distances, which indicates the course of an obstacle distance along the lane at a predetermined height; to map the distance course to a predetermined height; compare the distance profile shown with a predetermined distance profile from the data source, wherein the predetermined distance profile is associated with an indication of the location of at least one point along the lane; and determine the position of the motor vehicle based on the at least one point.

The device can be arranged on board a motor vehicle. The processing device is preferably configured to carry out the method of the second aspect described above in whole or in part. Advantages or features of the method can be related to the device. For the embodiment of the device or the processing device, the degrees of freedom mentioned above with respect to the first aspect are possible.

According to a further aspect, a motor vehicle comprises one of the above-mentioned devices. It is also possible to provide a device which fulfills the features of both devices or is set up to carry out both of the methods described above in whole or in part.

• 1 ein Kraftfahrzeug auf einer Fahrspur,
• 2 ein Ablaufdiagramm eines Verfahrens zum Erzeugen einer Randbebauungskarte;
• 3 ein übergeordnetes Diagramm zur Erzeugung einer Randbebauungskarte;
• 4 beispielhafte Darstellungen von Informationen während der Erstellung von Daten für eine Randbebauungskarte; und
• 5 ein Ablaufdiagramm eines Verfahrens zum Positionieren eines Kraftfahrzeugs auf der Basis von Daten einer Randbebauungskarte

darstellt.The invention will now be described in more detail with reference to the accompanying drawings, in which:

• 1 a motor vehicle on a lane,
• 2 a flowchart of a method for generating an edge development map;
• 3 a parent diagram for generating an edge development map;
• 4 exemplary representations of information during creation of data for an edge building map; and
• 5 a flowchart of a method for positioning a motor vehicle based on data of an edge development map

represents.

1 shows a motor vehicle 100 on an exemplary lane 105 , 1A shows a view of the motor vehicle 100 from behind and 1B from above. Here are some details in 1B not repeated. In the following, both representations will be described together.

The lane 105 extends here by way of example between limitations 110 , which may be formed, for example, as lane markings, mechanical limitations such as a curb or theoretical limitations such as a trajectory. For easier reference, the lane 105 on a baseline 115 be reduced, for example, in the middle between the boundaries 110 extends. It is assumed that the motor vehicle 100 along the baseline 115 emotional.

On one side of the motor vehicle 100 - in the representation of 1 right, in another embodiment additionally or alternatively also left - can be one or more objects 120 located, the obstacle-free space to the side of the vehicle 100 or the side of the lane 105 limit. The space at the side of the lane 105 is up to a surface of the objects 120 unobstructed. This distance can be at different heights 125 over the lane 105 are determined, each with an associated object or obstacle distance 130 results.

Along the lane 105 form the obstacle distances 130 in different heights 125 each a distance course 135 who in 1B is visible. The distance course 135 usually corresponds to the course of the smallest distance in the associated height 125 in the lateral direction between the lane 105 and an object 120 consists. Lying multiple objects 120 from the lane 105 seen from behind, so can a distance 135 that results in different sections 140 on different objects 120 is related.

It is suggested, distance courses 135 different heights 125 to a common height above the roadway 105 to relate or depict. The shown distances 135 can then form an edge building map or add edge building information to an existing map. The shown distances 135 can be characteristic of a body or a section 140 the lane 105 be. By comparing currently mapped distance profiles with previously stored distance profiles, a position along the lane can be obtained 105 be determined with high accuracy.

On board the motor vehicle 100 is preferably a device 150 arranged in different embodiments for generating edge development information or for determining the position of the motor vehicle 100 may be used on the basis of sampled and predetermined edge building information. The device 150 preferably comprises a processing device 155 , a storage device 160 a scanning device 165 for determining obstacle distances 130 , and preferably a positioning device 170 , Optionally, a communication interface 175 for in particular wireless communication with a device outside the motor vehicle 100 be provided. Next optional is a scanner 180 provided a relative position or speed along the lane 105 scan. The scanning device 180 For example, a hodometer may comprise, for example, the speed sensors of wheels of the motor vehicle 100 can determine a covered distance or a driven speed.

The device 150 is preferably set up in the manner described above to a common height above the lane 105 shown distances 135 due to samples by means of the scanner 165 provide. The scanning device 165 may in particular comprise a radar, LiDAR or optical sensor. In one embodiment, the mapped spacing curves 135 in the storage device 160 stored. In another embodiment, they are included in the memory device 160 filed or via the communication interface 175 received predetermined mapped distance profiles compared to the position of the motor vehicle 100 along the lane 105 to determine. It should be noted that the provision of mapped distance courses also outside of the motor vehicle 100 can take place, wherein instead of the scanning results of the scanning device 165 already collected and possibly pre-processed databases on obstacle distances 130 be used.

2 shows a flowchart of a first method 200 for generating a peripheral building map. The procedure 200 can on the device 150 on board the motor vehicle 100 be executed.

In one step 205 become obstacle distances 130 between the lane 105 and one or more objects 120 that are on a predetermined side next to the lane 105 are at a predetermined position determined. The obstacle distances 130 can in one step 210 be subjected to a noise filter to allow easier processing. In one step 215 For example, a height filter may be applied to only such obstacle distances 130 to be taken into account between a predetermined minimum and a predetermined maximum height 125 regarding the lane 105 are located. For example, a height range between 30 cm and 1.5 m above the lane 105 be taken into account.

The obstacle distances 130 define surfaces of the objects 120 , For better processing, the surfaces between the individual scans in one step 220 be interpolated or closed. Single measurements that have no apparent relationship to an object 120 can be discarded. Another noise filter can be in one step 225 be applied. Similar objects 120 can in one step 230 be united with each other. For example, individual trees can be combined to form an avenue, individual guard rails to form a continuous guardrail, or individual houses to form a façade.

Be different scanners 165 used, for example, an optical camera and an ultrasonic sensor, so the existing data in terms of their sampling rates in one step 235 be adapted to each other. In one step 240 can distance gradients 135 be determined, taking a distance course 135 each the course of an obstacle distance 130 along the lane 105 at a predetermined height 125 indicates. The heights 125 are preferred with respect to the traffic lane 105 Are defined. Is the lane 105 not even, so can be done a geometric adjustment. In one step 245 become the previously determined distances 135 to a common height 125 displayed. This can in particular by a projection along the height direction, that is generally perpendicular to a plane in which the lane 105 extends, take place. The common height is preferably zero, so it is at the same height or in the same plane as the lane 105 ,

Optionally be in one step 250 individual sections 140 the pictured Distance courses 135 sporadically. That means identical sections 140 two or more distance courses 135 be shown on only one. In one step 255 can a section 140 or a combination of obstacle distances 130 at one point along the lane 105 be assigned a point whose position relative to the lane 105 , in particular in the longitudinal direction, is known. The individual distances 135 can thereby each at positions along the lane 105 be coupled.

Optionally, in one step 260 an indication of a data source 165 . 175 a section 140 or a point along the lane 105 be assigned. In a similar way, in one step 265 an indication of an object 120 be assigned.

The information provided and any references to other information are preferred in one step 270 provided. These can be the distance gradients 135 in particular in vectorial form, preferred with respect to the points along the lane 105 , be specified. In one embodiment, the provided data is set in relation to other map data relating to the lane 105 already known. In particular, it is proposed to use a data format defined in the Navigation Data Standard.

3 a parent diagram 300 for generating a marginal development map 305 , In general, the method described above 200 be used, not only on current from a moving motor vehicle 100 scanned obstacle distances 130 Alternatively, work on other data sources. Exemplary are here as the first data source 310 Data from a road DNA, as a second data source 315 Obstacle Map data and third data source 320 called three-dimensional data of a occupancy grid (3D Occupancy Grid). Alternative or additional data sources are also possible.

The different data sources 130 . 310 to 320 can by means of the method described above 200 edited and preferably fused together. The result is an edge development map 305 For example, in a central database 325 can be stored. If necessary can by a motor vehicle 100 a request via the communication interface 175 to the central database 325 be placed on which the distance profiles shown 135 a predetermined section 140 can be transmitted back. The edge building map 305 may also be in whole or in part aboard the motor vehicle 100 in the storage device 160 be stored. This is the part of the motor vehicle 100 both online and offline operation possible. The first data source 310 (Road DNA) is known by TomTom. Parallel to a lane 105 Right and left vertical projection surfaces are assumed, which are divided into tiles of predetermined size. For each tile, the depth of the obstacle-free space becomes an object 120 determined and noted. Symbolic is in 3 shown a false color representation of the depth information.

The second data source 315 shows obstacles that can be provided by the map service HERE, for example. These obstacles may be based on a fusion of different sensor scans aboard a motor vehicle 100 be created.

The exemplary third data source 320 can also be provided by the map service HERE. For a three-dimensional strip to the right and left of the lane 105 are simple geometric objects (cuboid) defined, all objects 120 include the lateral free space to the lane 105 limit.

The edge building map 305 is in the representation of 3 in perspective, the individual distances 135 each represented as lines.

4 shows exemplary representations of information during the creation of data for an edge building map 305 , 4A shows a false color representation of obstacle distances 130 in a view transverse to the direction of travel of the motor vehicle 100 from 1 , The darker an area is displayed, the greater the assigned obstacle distance 130 , This representation coincides, at least in part, with that of the first data source 310 from 3 ,

4B shows the same information after applying a noise filter. 4C shows a plan view of the mapped to a common height distances 135 , The further a distance course 135 in the presentation of 4C above, the greater is its obstacle distance 130 to the lane 105 ,

5 shows a flowchart of another method 500 , The procedure 500 can be used to determine a position of a motor vehicle 100 on the basis of an edge development map 305 be used. The procedure 500 is especially for draining on a device 150 on board a motor vehicle 100 from 1 set up.

In one step 505 becomes a lateral environment of the motor vehicle 100 sampled and in one step 510 becomes at least one distance course based on the samples 135 certainly. These steps are preferred by corresponding parts of the method described above 200 realized.

In parallel, in one step 515 an approximate position of the motor vehicle 100 be determined, for example by means of the positioning 170 , In the area of the approximate position becomes in one step 520 from the storage device 160 or by means of the communication interface 175 from the central database 325 at least a predetermined distance profile 135 certainly. This predetermined distance profile 135 gets in one step 525 with the locally determined distance course 135 of the step 510 compared. The predetermined distance course 135 At least one point is assigned to the lane 105 and preferably a geographical position is assigned. Based on the comparison and this point can in one step 530 the position of the motor vehicle 100 be determined. Due to the accuracy of the determined distance gradients 135 and its preferred vectorial storage, the position can be determined with high accuracy. It should be noted that the determination is not limited to only one side of the vehicle, but can be performed on any side of the vehicle or on both sides of the vehicle.

LIST OF REFERENCE NUMBERS

100

motor vehicle

105

lane

110

limit

115

baseline

120

object

125

height

130

obstacle distance

135

Distance course

140

section

150

contraption

155

processing device

160

storage device

165

scanning

170

positioning

175

Communication Interface

180

scanning

200

method

205

Determine obstacle distances

210

noise filter

215

height filters

220

object surface

225

Noise filter on object

230

Association of similar objects

235

Adjust sampling rates

240

Determine distance gradients

245

Depict distance gradients at common height

250

seperate

255

Map point along lane

260

Map data source

265

Assign object

270

Provide

300

diagram

305

Peripheral development map

310

first data source (Road DNA)

315

second data source (Obstacle Map)

320

third data source (3D Occupancy Grid)

325

central database

500

method

505

Scanning environment

510

Determine distance course

515

Determine approximate position

520

Determining predetermined distance profile

525

to compare

530

Determine position

Claims (15)

A method (200) of creating a roadmap (305) of a lane (105), the method (200) comprising the steps of: - determining (205) obstacle distances (130) each indicative of the depth of an obstacle-free space at the side of the lane (105) at an associated height (125); - determining (240) distance profiles (135) which respectively indicate the course of one of the obstacle distances (130) along the lane (105) at the respective height (125); - mapping (245) the distance gradients (135) into a reference height (125); - Providing (270) the mapped distance courses (135). Method (200) according to Claim 1 , wherein mutually identical sections (140) of the illustrated distance courses (135) are reduced to a section (140). Method (200) according to Claim 1 or 2 , wherein the illustrated distance courses (135) are stored in vector form (270). Method (200) according to one of the preceding claims, wherein the obstacle distances (130) originate from different data sources (130, 310, 315, 320). Method (200) according to Claim 4 wherein the mapped distance traces (135) are provided (260) together with an indication of an underlying data source (130, 310, 315, 320) of the obstacle distances (130). The method (200) of any one of the preceding claims, wherein the mapped distance gradients (135) are provided (255) along with an indication of the relative location of points along the lane (105). Method (200) according to one of the preceding claims, wherein, for a section (140) of the illustrated distance courses (135) along the traffic lane (105), an indication of an object (120) is provided (265) representing the obstacle-free space of one of the distance courses (135) bounded in the depth direction. Method (200) according to one of the preceding claims, wherein prior to determining a distance profile (135), the surface of an object (120) which limits the obstacle-free space in the depth direction is determined (220). Method (200) according to Claim 8 wherein the surface is interpolated (220) between certain obstacle distances (130). Method (200) according to Claim 8 or 9 wherein noise of the obstacle distances (130) is reduced (210, 225). A method (200) according to any one of the preceding claims, wherein portions (140) of the mapped spacings (135) are organized in tiles, and for each tile an indication of other card information of the same portion (140) is provided. Apparatus (150) for creating a peripheral building map (305) of a lane (105), the apparatus (150) comprising: - an interface (165, 175) for receiving obstacle distances (130) each indicative of the depth of an obstacle-free space at the side of the lane (105) at an associated height (125); a storage device (160); and a processing device (155), which is set up, o to determine, based on the determined obstacle distances (130), distance courses (135) each describing the course of an obstacle distance (130) along the lane (105) at a predetermined height (125); o map the determined distance courses (135) to a reference height (125); and o to provide the illustrated distance gradients (135). A method (500) for determining the position of a motor vehicle (100) along a lane (105), the method (500) comprising the steps of: - determining (205) obstacle distances (130), each defining the depth of an obstacle-free space laterally of Indicate lane (105) at an associated altitude (125); - determining (240) distance profiles (135) which respectively indicate the course of one of the obstacle distances (130) along the lane (105) at the respective height (125); - mapping (245) the distance gradients (135) into a reference height (125); - Comparing (525) the mapped distance gradients (135) with predetermined, mapped distance gradients (135) to which an indication associated with the location of at least one point along the lane (105); and determining (530) the position of the motor vehicle (100) based on the at least one point. Device (150) for determining a position of a motor vehicle (100) along a lane (105), the device (150) comprising: - a scanning device (165) for determining obstacle distances (130), each indicating the depth of an obstacle-free space laterally of the lane (105) in an associated height (125); - a data source (130, 310, 315, 320) for an edge building map (305); and a processing device (155), which is set up, determining, on the basis of the determined obstacle distances (130), at least one distance course (135) which indicates the course of an obstacle distance (130) along the lane (105) at a predetermined height (125); o image the distance profile (135) into a predetermined height (125); o compare the distance profile (135) shown with a predetermined distance profile (135) from the data source (130, 310, 315, 320), wherein the predetermined distance profile (135) an indication of the location of at least one point along the lane (105) assigned; and o determine the position of the motor vehicle (100) based on the at least one point. Motor vehicle (100) with a device (150) according to Claim 12 or 14 ,

Images