DE102014015073B4 - Method for updating and / or expanding a map data set in a limited environment - Google Patents

Abstract

Method for updating and / or expanding a map data set of a limited environment (2) traveled by mobile devices (3) with environment sensors (16), which is a parking environment, a current map data record being sent to a mobile device (2) operated within the environment (2). 3) is transmitted and is used within the device (3) to localize the mobile device (3) within the environment (2), characterized in that at least one property of at least one feature in the environment not yet or differently contained in the map data set (2) Descriptive feature data are determined from sensor data from the environment sensors (16) of the mobile devices (3) and stored in the copy of the map data set stored in the mobile devices (3), after which the map data records expanded in this way are sent to a central server device assigned to the environment (18) transmitted and there, taking into account the new en and / or changed properties are merged into an updated map data set, the properties comprising a class and / or position of the feature obtained as a result of a classification and / or, in particular as a result of the classification, at least one purpose of the feature, the classification and / or the determination of the purpose from the sensor data takes place using an auxiliary database transmitted from the server device (18) to the mobile device (3), which allows the mapping of sensor data to at least one class and / or purpose of the feature.

Description

The invention relates to a method for updating and / or expanding a map data record of a limited environment traveled by mobile devices with environment sensors, which is a parking environment, a current map data record being transmitted to a mobile device operated within the environment and within the device for localizing the mobile device is used within the environment.

Mobile devices, for example motor vehicles, which have a system for automatically guiding the mobile device, are already known in the prior art. Such systems, for example fully autonomous vehicle systems, completely take over the longitudinal and lateral guidance of the mobile device in order to guide it to a destination. In the example of a motor vehicle, it has been proposed for limited surroundings, in particular parking areas such as parking garages, to guide a motor vehicle fully autonomously to a parking space within the parking area or the like. In order to be able to realize this, it must be possible within the mobile device to determine the position relative to the destination and to select a trajectory that can be driven by the mobile device and that leads to the destination.

Localization methods or localization algorithms for mobile devices in limited environments are already known in the prior art. Common approaches are, for example, the use of SLAM algorithms (Simultaneous Localization and Mapping) and / or MCL algorithms (Monte Carlo Localization). Usually, a position hypothesis is determined in that data from sensors of the mobile device, in particular also from environmental sensors, are evaluated, sensor data from at least some of these sensors being merged. Other input data that can be included in the determination of a position hypothesis are ego data of the motor vehicle, for example when its own movement is to be taken into account starting from a known point. If different types of localization are used, position hypotheses of the individual procedures can be combined to form an ultimately determined position of the mobile device.

As an example, reference is made to the book with regard to localization: Thrun, Sebastian; Burgard, tungsten; Fox, Dieter: Probabilistic Robotics (Intelligent Robotics and Autonomous Agents), The MIT Press - ISBN 9780262201629.

In order to support mobile devices with autonomous navigation within the limited environment, it is known to provide the mobile devices with map data sets which are intended to allow orientation and navigation within the limited environment. In this context, it is known, for example, to derive map data records of the limited environment from construction plans, to measure the limited environment and the like. In this context, semantic data, in particular data for the purpose of defining characteristics of the environment, are added manually. Such semantic data can include, for example, the identification of features of the environment as walls, doors, exits, signs, parking markings and the like.

A high level of manual effort is required to create such map data records. In particular with regard to the semantic data mentioned, which describe the content of meaning of features of the environment, it is initially true that these can change over time after, for example, signs can be removed or slipped and the like. This results in a high level of manual tracking effort. Furthermore, such manually created map data sets contain potential inaccuracies, especially when deriving map data from construction plans. For the fully automatic guidance of mobile devices, in particular motor vehicles, as well as other driver assistance functions, however, highly accurate and highly reliable map material is required. A need for manual map creation would make the introduction of such functions difficult due to high costs.

The post-published DE 10 2013 011 824 A1 discloses a method for determining and updating an occupancy map in a parking area, sensor data from at least one environmental sensor of at least one motor vehicle located in the parking area being used to determine or update the occupancy map via the communication device. In addition to the simple information as to whether a parking space is occupied or free, useful additional attributes for a parking space can be contained or derived, whereby adjacent parking spaces or other additional information can also be present. When determining the position of the motor vehicle, the occupancy map can be used at least partially; It is also conceivable that the raw sensor data are used to update a copy of the occupancy card stored in the motor vehicle, with at least changed portions of the occupancy card being transmitted as sensor data to the central computing device.

EP 0 921 509 A2 relates to a system and method for updating, expanding or refining a geographic database at Use of feedback. In this case, sensor data are to be recorded from motor vehicles on roads in the geographic region, which are then compared with data of a physical feature in the database, with a significance then being determined for the comparison and a change in the geographic database taking place based on the significance, with mainly targeted at road networks.

WO 2008/009 965 A1 relates to the generation of a map of features, the position of the features being determined from images of a vehicle driving past. The card relates in particular to an intersection, a parking area or an individual parking lot.

US 2010/0 021 012 A1 discloses the generation of an open area map, with an end user generating images of a walk-in environment which are used to generate the map.

DE 10 2012 202 186 A1 describes a method for providing information about the surroundings, a map of the surroundings being supplemented with detailed information. So the card value should be increased.

US 2014/0 278 096 A1 relates to concepts for defining travel paths in parking areas. Parking routes are derived from paths actually driven.

US 2013/0 257 632 A1 discloses a system and a method for estimating the availability of parking on the street, for which purpose a parking probability generator is used. The probability is determined based on user densities.

The invention is therefore based on the object of specifying a possibility for determining highly precise map data sets describing the environment with regard to as many properties as possible for navigation in limited environments.

To achieve this object, the features of claim 1 are provided according to the invention in a method of the type mentioned at the beginning.

Mobile devices, in particular motor vehicles, exist in various equipment variants. It is therefore often the case that a single mobile device can actually only record certain properties of features of the environment in which the mobile device is operated. If, however, different mobile devices are considered, there are often different configurations of environment sensors which allow the most varied properties of features, in particular objects, to be measured in an environment. The present invention proposes using this treasure trove of data, which is already accumulated during the operation of the mobile device, on the limited environment in which the mobile device is operated, in particular completely automatically or partially automatically, with regard to the map data set by using the information from a large number of Mobile facilities that are operated within the limited environment are brought together and used for constant updating and, if necessary, expansion of the map data set.

The method according to the invention relates to the detection of properties of features via environmental sensors, and therefore not only to the recording of a traveled route in order to create a road network structure composed of route sections from several traveled routes. Rather, it is about the most accurate possible mapping of a limited environment, in particular a building, for example a parking garage, in which there is an increased use also for semantic information on individual features, in particular for their purpose. This clearly goes beyond approaches that only want to differentiate between drivable and non-drivable areas, because semantic metadata relating to the meaning of features of the environment are also collected and added to the map data set, which is advantageous not only in an at least partially, preferably fully automatic Management of the mobile devices can be used within the limited environment, but can also be used for assistance functions within the mobile devices or other devices, which will be discussed in more detail below.

The present invention therefore allows maps of a limited environment, in particular a parking garage, to be created and / or updated in a fully automated and precise manner, with in particular different environment sensors being used in different mobile devices, so that in particular an added value for mobile devices with less equipment at environmental sensors due to additional information in the map data set.

It should be noted that it is entirely within the scope of the present invention to use the term “feature” for segments of an occupancy card that is part of the card data record. For example, map data sets are known in which individual segments, for example defined by a grid, in addition to "occupied / free" information in the sense of "drivable / not drivable" further metadata, for example in a tree structure, are assigned. Of course, the term “feature” can also designate objects or parts of objects that are present within the limited environment.

It should also be pointed out that the expansion of the map data set can also be understood as adding a map of the limited surroundings created by the motor vehicle itself, if the motor vehicle is designed accordingly, for example by means of SLAM algorithms. Within the server device, for example a back-end server, it is then possible to combine the corresponding card components by means of card matching algorithms, which are based in particular on certain features / objects. This also applies with regard to the early phases of the creation of map data sets when different areas of the limited environment are recorded by different mobile devices and are to be merged.

The invention thus describes a possibility for the construction and exchange of map data sets of a limited environment in order to accumulate the measurements of a large number of mobile devices using a backend solution so that errors can be leveled out as well as always current map data sets of a large fleet of mobile devices, in particular To make available to motor vehicles and, if necessary, to other devices. Environment sensors are used by mobile devices to determine or update properties of features of the environment for the first time, the map data sets are merged in a central server device of the limited environment and lead to the updating of the map data set, which can be distributed to other mobile devices / devices , so that not only a localization within the limited environment is made possible, but also distances / routes to target locations can be determined, which can expediently also be selected and / or determined on the basis of semantic feature data that describe the meaning of a feature of the environment.

It should be pointed out at this point that the map data set can optionally be interpreted as a further environmental sensor for mobile devices.

As already indicated, it is particularly advantageous if mobile devices with different environmental sensors are used. By using a variety of equipment with regard to the environment sensor system, a wide variety of properties of features can be recorded as feature data and added to the map data set, so that the map data set forms the most complete picture possible of the limited surroundings that would not be available to individual mobile devices . Because the information from a large number of mobile devices is collected centrally and bundled in the current map data set, mobile devices can compensate for insufficient equipment with regard to their environment sensors using the information in the map data set that was collected by other mobile devices.

The map data set is preferably transmitted to the respective mobile device when entering the surroundings. If the environment is, for example, a parking garage, it can be useful to provide the current map data set in an entrance area, for example at a barrier, so that it can be used, for example in the context of localization, and updated accordingly / can be added. It is also conceivable that the current map data set is transmitted from the server device as soon as a communication connection can be established between the server device and the mobile device. It is of course also possible that, in particular on request, a further updated map data set can also be transmitted to mobile devices at later times, in particular if they are moving within the limited environment for a longer period of time. In the example of a multi-storey car park or other parking area, it can also be useful to transmit the map data record in its current form when the mobile device is restarted after a long period of time.

Accordingly, it can be provided that an expanded map data set is transmitted to the server device with each expansion and / or when the environment is left. On the one hand, it is conceivable that whenever new feature data are added to the map data record for updating and / or supplementation, a transmission to the server device also takes place. If the data traffic is to be restricted, it may be more expedient to carry out the retransmission to the server device less frequently, for example when the mobile device is switched off if it is a parking area and / or when the environment is left. It is of course also possible to carry out a cyclical transmission of the extended map data set to the server device.

According to the invention it is provided that the properties a class and / or position of the feature obtained as a result of a classification and / or, in particular as a result of the classification, at least one purpose of the Include feature. With regard to the mapping of the surroundings carried out here, it is of course expedient if the properties include a position of the feature. A particularly advantageous embodiment is obtained when the properties include a class obtained as a result of a classification. In particular, systems that are designed for autonomous guidance of the mobile device, but also a large number of other systems that can be provided in mobile devices, in particular of vehicle systems, use different classifications of objects / features of the environment. For example, classifiers can be used that are applied to the combined sensor data of the environment sensors in order to identify features, in particular objects, in the environment of the mobile device and to further treat them accordingly. For example, objects in the vicinity can be classified as further mobile devices, columns, markings, movable or immovable objects and the like. Such a classification is therefore a special example for the determination of a semantic property, because a class contains information on the meaning of the feature, which can go well beyond the pure degree of blocking for trajectories of the mobile device.

It is therefore provided in a variant of the invention that the properties, in particular as a result of a classification, include at least one purpose of the feature, which is then stored accordingly in the feature data. In particular, it can be a purpose that goes beyond the operation of the mobile device. The purpose of a wall in a parking garage is based on a load-bearing effect and structuring of the same; The purpose of a sign is to orient people within the limited environment. Especially for applications that go beyond the fully autonomous or semi-autonomous operation of the mobile device, for example in the case of mobile devices, which are also contemplated within the scope of the invention, use such additional properties of features contained in the map data set for further functions which can then be implemented appropriately and precisely.

With regard to the classification and / or intended purpose as a property, information signs within the surroundings and / or lane markings and / or parking lot markings and / or aids, in particular fire extinguishers, can be identified particularly advantageously as features by determining a class and / or purpose. In this way, for example, functions are made possible that can lead people to the next exit and / or the next fire extinguisher and the like. In particular, mobile devices with higher-quality sensors, for example image processing, can recognize properties that describe the purpose / meaning of features, for example standardized information signs pointing to an exit or the like. Of course, such additional semantic feature data can also be used in the context of the operation of the mobile device, for example if the mobile device is a motor vehicle and it is relevant whether parking spaces are designated for certain groups of people or the like, for example in the case of Women parking spaces.

The invention provides that the classification and / or the determination of the intended purpose from the sensor data takes place using an auxiliary database that is present in the mobile device or transmitted from the server device to the mobile device, which the mapping of sensor data to at least one class and / or Purpose of the feature allowed. Thus, in a preferred embodiment, the tool is given to the mobile devices by the server device itself in order to be able to determine the desired semantic feature data. Ultimately, the auxiliary database thus describes the appearance and / or other recognition properties of possible semantic feature data with regard to the recognizability by environmental sensors. The auxiliary database can therefore contain, for example, a sub-database on exit and emergency exit signs and the like.

It has already been explained that, in a particularly expedient embodiment, the properties of the features are determined within the framework of a function that evaluates the sensor data of the environmental sensors, so that little or no additional effort is required to contribute to the function of updating the map data set. For example, it can therefore be provided that the determination of the properties takes place at least partially in the context of a localization of the own mobile device, in particular in the context of the implementation of a SLAM method. In the context of SLAM methods, but also with other localization methods, the surroundings of the mobile device are often mapped anyway, possibly also with a view to determining the position of the mobile device in the map data set also considered for this purpose. As a result, while a localization function is being carried out, a large number of feature data are created which describe the properties of features in the environment of the mobile device, which can therefore also be used to update and / or supplement the map data set.

The server device can provide for an assignment of new and / or changed properties to a feature by means of a map comparison based on predefined features on the server device. Ultimately, then, the server device collects the expanded map data sets and reconciles them with one another. Such methods are already known in principle in the prior art under the catchphrase “map matching” and can, for example, be based on specific, prominent features. Of course, the map comparison can also be possible in a simplified manner if the mobile devices can be reliably localized themselves and thus the feature data are in any case in a clear relation to the other features contained in the map data set. Different, basically known configurations are conceivable here.

In general, within the scope of the self-localization of the mobile devices, it is expediently conceivable that clearly identifiable markers are used within the environment to localize the mobile device. Localization methods have already been proposed that use unique markings, for example two-dimensional barcodes, which can be recognized by the environmental sensors so that a unique position can be determined, including any sensor data from the environmental sensors and thus also the feature data in relation to a unique coordinate system of the limited environment can be set.

A device-to-infrastructure communication, for example in the case of a motor vehicle, a motor vehicle-to-infrastructure communication (Car2x communication), can be used to transmit the map data records and, if necessary, further information. The use of WLAN connections is particularly useful here. Concrete communication links with which mobile devices can communicate in a limited environment with a server device assigned to them have already been proposed and can be used within the scope of the present invention.

At least one camera and / or at least one ultrasonic sensor and / or at least one radar sensor and / or at least one lidar sensor can particularly expediently be used as the environment sensors. Cameras can be assigned to image processing devices. Line detection in the sensor data of cameras can, for example, make properties of floor markings, for example parking lot boundaries, identifiable; In the context of image processing, standardized symbols, for example emergency exit signs, can also be recognized in order to obtain semantic feature data. Radar sensors make it possible to detect metallic objects and their position, for example other mobile devices, in particular motor vehicles. By jointly evaluating sensor data from radar sensors and the camera, for example, the position of a further mobile device can be compared with recognized lane / parking lot markings in order to determine the occupancy of a parking space, in particular a parking lot, and the like. Ultrasonic sensors make it possible, for example, to detect a free space around one's own mobile device, so that drivable areas can be detected on the assumption that the mobile device starts on a drivable route. In the case of a parking area, parking spaces or parking spaces can be measured in terms of their depth. Lidar sensors make it possible, for example, to measure object contours precisely, particularly with regard to building features and the like.

A scan matching method is preferably used to assign the sensor data to a position in the map data record. Scan matching methods are already known in the prior art and ultimately map the known position and orientation of the environmental sensor relative to a localized position of the mobile device in the map data set.

A further development of the invention provides that a path covered by the own mobile device is also stored in the map data set with a property indicating whether it can be driven on. In addition to the semantic feature data at the center of the method according to the invention, properties of features, in particular areas or segments, of the environment can also be derived from other systems of the mobile device. If a mobile device covers a certain route, which can be tracked from odometry data of the mobile device, the drivability of this covered route follows. Paths covered can, however, also be appropriately evaluated by the server device. For example, it can be provided that the server device identifies areas of the environment that have been bypassed, in particular road damage and / or danger spots, by comparing routes traveled by several mobile devices. Thus, if several mobile devices repeatedly drive around places that have been marked as free in the map data set, this can be an indication of an occupancy of these places and / or damage in the roadway, which is also important additional information in the map data set and when planning trajectories subsequent journeys further mobile facilities can be considered. Further measures can of course also be derived from such information, for example the infrastructure facility can output information to the responsible persons if there are indications of damage within the limited environment, for example also together with further feature data.

If the method according to the invention relates to a parking area, it can also be used to determine current occupancy states of parking spaces from the expanded map data sets. The occupancy status of parking spaces is information that can be changed dynamically and quickly, which can therefore be easily detected and noticed through changes in the expanded map data sets. In addition to the constant improvement of the map, in particular by adding semantic feature data, current occupancy information for parking spaces can therefore also be derived from the expanded map data records in the server device.

A particularly advantageous embodiment of the present invention provides that feature data relating to the same property of a feature from various mobile devices are statistically evaluated by the server device, in particular using outlier detection. Since limited environments are mostly used by many mobile devices, especially in the case of motor vehicles in a parking area, there is a large number of extended map data sets, which of course also make it possible to carry out a statistical evaluation, in particular to exclude individual incorrect measurements of individual mobile devices and to improve the overall quality of the transmitted feature data, since ultimately several different measurements are made. In one embodiment, error values estimated by the mobile device can also be added to the feature data, which ultimately also reflect a statement about the quality of the individual measurements and, for example, in a weighting of feature data from different mobile devices that relate to the same feature of the environment, precipitation can find. In this way, the server device also levels out errors in particular through the statistical analysis.

As has already been explained, at least one property of at least one feature contained in the map data set is expediently evaluated by a function of the mobile device in addition to the localization, in particular by a driver assistance function. Thus, for example, an occupant can already be shown in the mobile device itself where the next exit from the restricted area is, exactly as is possible, for example, in the case of parking garages for pay machines or the like. This is where the semantic data that assign meaning to features of the map data set proves extremely useful. Supporting functions can also be used when selecting the destination, regardless of whether this is done automatically or manually by a driver, for example if it is known that the driver is a man and a parking space is reserved for women and the like.

In a particularly advantageous embodiment of the present invention, however, it is provided that the map data set is transmitted to a mobile terminal of a user, in particular a mobile phone, and is evaluated there as part of a software application, in particular a navigation application. If the map data set is also available on a mobile device that is carried by a user of the mobile device, he can, for example after getting out of a motor vehicle, on the way back to his mobile device or the like, additional assistance functions that are provided by the mobile Device are provided, use. For example, on the basis of the map data set, the person can be navigated to the next possible exit, guided to an aid, for example a fire extinguisher, and the like. The map data set, which has been improved by the present invention in particular with regard to the semantic feature data, can therefore be used profitably beyond the mobile devices.

Appropriately, for the initial localization of one's own mobile device in the map data set, as has already been indicated, a predefined feature, in particular located at an entrance to the surroundings, can be detected. An example of such detectable features is a barrier system, which is often found at the entrance to a parking garage as a limited environment. All mobile devices that want to penetrate the limited environment pass this point, so that a particularly simple determination of a starting position for the localization within the limited environment is made possible.

It can also be provided that a GPS position determination is taken into account to identify the surroundings and / or the initial position in the map data set. The mobile devices can therefore have a GPS sensor which allows a corresponding absolute position determination in a geodetic coordinate system. On the basis of this geodetic position it can be determined in which limited Environment the motor vehicle is just driving in, whereby it is also expedient to take such a geodetic position into account during the initial positioning in the map data set. The identification of the limited environment ultimately also identifies the server device with which the mobile device is to make contact.

As already indicated, the method according to the invention can be applied particularly advantageously to motor vehicles, so that a motor vehicle can be used as the mobile device. The limited environment can in particular be a parking environment, preferably a parking garage, or some other limited building in which the motor vehicle is to navigate.

If the devices participating in the method according to the invention are combined, the result is a system for updating and / or expanding a map data set of a limited environment traveled by mobile devices with environment sensors, which includes a server device that is assigned to the limited environment. The server device is designed to transmit current map data sets to mobile devices which are also part of the system and which are operated in the limited environment and to receive map data sets from them that have been expanded, that is to say updated with feature data. The server device, which is stationary and which is an infrastructure device, is also designed to merge the extended or updated map data records obtained in order to determine a new, current map data record. The mobile devices, in particular motor vehicles, in the limited environment are in turn provided with environment sensors and designed to do so, in particular in the case of characteristics of the environment not contained in the map data set and / or properties of the environment that are changed compared to the map data record with current characteristic data from the sensor data of the environment sensors were determined to add, which are generally taken into account after the return of the map data set to the server device there as part of the updating of the map data set. The statements relating to the method according to the invention naturally apply analogously to the system described here and the corresponding partial devices, in particular the server device and the mobile devices.

It should also be pointed out that it can also be useful to take into account, in addition to the mobile devices, other devices within the limited environment for map updates, for example the aforementioned mobile devices, which often also have environmental sensors, in particular cameras, thus themselves within the Can locate map data set and determine properties of features of the environment, which can be passed on to the server device just as useful.

• 1 eine Parkumgebung mit mobilen Einrichtungen und einer Servereinrichtung,
• 2 einen Ablaufplan eines Ausführungsbeispiels des erfindungsgemäßen Verfahrens, und
• 3 eine Fahrsituation eines Kraftfahrzeugs in der begrenzten Umgebung.

Further advantages and details of the present invention emerge from the exemplary embodiments described below and with reference to the drawing. Show:

• 1 a parking area with mobile facilities and a server facility,
• 2 a flow chart of an embodiment of the method according to the invention, and
• 3 a driving situation of a motor vehicle in the limited environment.

1 shows a system in which the method according to the invention can be applied. A section is shown 1 from a limited environment 2 , here a park environment in which as mobile facilities 3 Motor vehicles 4th be moved. The park environment 2 , which in this case is a parking garage, is characterized by a large number of different features. walls 5 delimit structurally drivable and accessible areas, parking lot markings 6th show where you can park, further lane and parking lot markings 7th , 8th , 9 should notify drivers of the motor vehicles 4th give; Information signs 20th on the walls 5 indicate, for example, exits and emergency exits, aids such as fire extinguishers 10 , are on walls 5 hung up the walls 5 are through doors 11 interrupted, pillars 12 carry the ceiling of the corresponding floor of the parking garage, on another as a pay machine 13 trained tools, payments can be made and the like. The lane markings are shown here, for example 7th the directions in which you can continue driving; the lane marking 8th indicates a zebra crossing and the parking lot marker 9 indicates a parking space for women. Motor vehicles 14th that are not currently in operation occupy individual parking spaces.

Here are the motor vehicles 4th , 14th especially designed to be within the parking area 2 semi-automatically or fully automatically, so that they have a corresponding fully autonomous vehicle system, which, for example, by a control unit 15th is realized. Environmental sensors 16 of motor vehicles 4th , 14th capture the environment, the motor vehicles 4th in the present case, a different set of environmental sensors 16 have. Via a motor vehicle-to-x communication device 17 (Car2x communication device), the motor vehicles 4th with each other and / or with Infrastructure facilities exchange communication information.

Such infrastructure facilities also include that of the park area 2 associated server facility 18th that the motor vehicles 4th , 14th Information about the parking area 2 provides, in particular a current map data set of the parking area 2 . Approaching a motor vehicle 4th , 14th the park environment 2 , for example, based on position information from a GPS sensor or due to the fact that the communication range of the server device 18th occurs, it is concluded that the server device 18th for the park area currently being driven on 2 responsible is. Passed a motor vehicle 4th , 14th an entrance to the park area 2 , in 1 for the sake of clarity not shown in more detail, in the present case specifically a barrier device, the current map data set is transmitted via a communication link 19th from the server facility 18th to the corresponding motor vehicle 4th , 14th transfer. This current map data record is also used to send an auxiliary database to the corresponding motor vehicle 4th , 14th transmitted to the motor vehicle 4th , 14th when recognizing features of the environment 2 supported, for example by information signs 20th , Lane and / or parking lot markings 6 - 9 and the same.

The map data set is used to locate the motor vehicle in operation 4th within the park area 2 to improve, which can be improved in particular using clearly detectable markers, for example two-dimensional bar codes. Through the environmental sensors 16 recorded sensor data can be merged and evaluated in order to identify the surroundings of the motor vehicle in operation 4th evaluate them as part of the localization function and bring them into line with the map data set; However, it should be noted that other functions in the operated motor vehicle 4th Sensor data from the environmental sensors 16 can evaluate and can access map data of the map data set.

In the park area 2 however, the method according to the invention is also carried out, that is to say, new or changed feature data on features of the parking garage that are generated by the environmental sensors 16 of the motor vehicles operated 4th are recorded are added to the map data set, which is consequently expanded and forms an expanded map data set. The extended map data records of the vehicles in operation 4th are in the server facility 18th collected and merged there into a new, updated map data set. The retransmission of the extended map data set to the server device can thereby 18th with parking an operated motor vehicle 4th in a parking lot and / or when leaving the parking area 2 take place, but also at other times, for example cyclically.

An embodiment of the method according to the invention is intended with regard to 2 are now shown in more detail. There drives in one step S1 a motor vehicle 4th in the park area 2 and receives over the communication link 19th the current map data set is transmitted. While the motor vehicle 4th through the park environment 2 moved, are in the context of functions in the motor vehicle, in particular a localization function, but also other functions, sensor data of the environment sensors 16 evaluated, also with regard to properties of various features of the park environment 2 . In the present case, a SLAM localization can in particular be carried out as a localization function. In other words arise in the step S2 to a variety of features of the park environment 2 Characteristic data describing properties of these characteristics, with the properties that are created by the respective motor vehicle 4th can be determined by its equipment with environmental sensors 16 depends, but in particular also on the quality of these environmental sensors 16 , for example an image processing device which is assigned to a camera. As environmental sensors 16 can in motor vehicles 4th overall, for example, in addition to cameras, radar sensors, lidar sensors, ultrasonic sensors and the like are also used. When determining properties of features of the park environment 2 the already mentioned auxiliary data set is also taken into account.

The determination of properties of the features in the present case includes the determination of semantic feature data that go beyond the mere existence or position of features and describe the meaning of the features. This is done in the control unit 15th of the motor vehicle 4th Classifiers are used that allow the assignment of features to a class, for example information signs 20th , Parking lot markings 6th , 9 , Doors 11 , parked motor vehicles 14th and the same. This also provides information about the purpose of the characteristics as a property, as an information sign 20th serves as a reference, for example to an exit, and the like, whereby the properties can of course go beyond the mere naming of the class and the purpose information transported thereby. To information signs 20th can be determined, for example, whether the information sign 20th to a normal exit, an emergency exit, a pay machine, a fire extinguisher 10 or the like. Such properties cannot by everyone in the park area 2 powered motor vehicles 4th can be determined as they are equipped with environmental sensors 16 is different; however, by adding them to the map data set, motor vehicles can 4th , 14th obtained from the sensor data of their environmental sensors 16 cannot derive these properties, use the information to operate functions of their vehicle systems.

In one step S3 it is now checked whether the feature data correspond to a change and / or addition to the map data set. The mapping of features through environmental sensors 16 to features in the map data set is due to the localization carried out and the known installation position and alignment of the environmental sensors 16 easily possible. It should be noted that the step S3 can in principle also be omitted if all by the motor vehicle operated 4th determined feature data are added to the map data set, which, however, is ideally avoided due to the high volume of data by using in step S3 the expansion criteria are evaluated with regard to the change or the addition of new properties.

In one step S4 becomes that in the operated motor vehicle 4th Saved copy of the map data set expanded to include the new / changed feature data.

In one step S5 the extended map record is sent to the server facility 18th retransmitted.

The steps S1 - S8 according to 2 are of course used by everyone within the park area 2 powered motor vehicles 4th executed.

In one step S6 are on the part of the server device 18th the expanded map data sets are brought into congruence with one another so that they can then be statistically evaluated together in order to determine an updated, new, generally applicable map data set. If the underlying map data set is already very precise, this can also be used for the localization of the motor vehicles in operation 4th be accepted, so that the "map matching" to be carried out can be carried out easily; the map data set is still under construction and will be completely new areas of the parking area 2 (ideally through several operated vehicles 4th ), special features can be used to make the comparison between the extended map data records. Such procedures are basically already known.

A statistical evaluation of the new feature data that has been added then takes place with regard to whether the generally applicable map data set needs to be updated. In particular, an outlier detection can be carried out in order to detect measurement errors of individual operated motor vehicles 4th if possible not to be included in the map data set. In this evaluation step S7 is created using new information from within the parking area 2 powered motor vehicles 4th consequently an updated, possibly supplemented map data set for other vehicles operated in the parking garage 4th can be used there for better orientation and better use of vehicle functions.

Regarding the step S7 it should be noted that a further development of the procedure provides in step S4 also within the park area 2 distance covered by the respective motor vehicle 4th add to the map data set, since conclusions about properties can be drawn from this by means of statistical evaluation, for example when a actually passable point in the parking area 2 from all motor vehicles 4th is bypassed, damage to the roadway or the like can be inferred.

As already mentioned, the map data set can not only be used in the context of self-localization within the parking area 2 powered motor vehicle 4th can also be used by other vehicle functions. For example, it can be inside a motor vehicle 4th , 14th navigation information to the next exit is output if there are signs 20th are known in the map data set, but it is also possible, for example, to point out unsuitable parking spaces or to take this into account when choosing the destination, ie a parking space. An example of a driving situation of this kind is shown 3 looking into the front passenger area of a motor vehicle 4th shows in a driving situation, being through the windshield 21st Parking lot markings 6th and the marking indicating a women's car park 9 can be recognized. Even if a camera is built into the motor vehicle itself 22nd or the assigned image processing device is not able to mark the parking lot 9 to be recognized, this is contained in the map data set, which ultimately functions as a type of additional environment sensor, so that a male driver, for example, has a display device 23 in the motor vehicle 4th can be informed that the car park is a women's car park.

Similar refinements are also conceivable if certain aids are sought, for example pay machines 13 , Fire extinguisher 10 and the same.

Another advantageous development results when the map data set is not only within the parking area 2 operated motor vehicles 4th is made available, but also mobile devices 24 , compare 1 , particularly cell phones used by users of motor vehicles 4th , 14th be carried with you. A communication link can also be used here 25th , for example a WLAN connection, to the server device 18th build up, so that the current map data set by, for example, a software application used for navigation 26th (“App”) can be used to guide the user to an exit, an aid and the like, with corresponding instructions on a display 27 can be represented.

It should also be noted that in the mobile terminal 24 a localization function can be implemented on the basis of environmental data if corresponding environmental sensors, for example cameras, and / or motion sensors, for example acceleration sensors, are located within the mobile terminal 27 are available, although other localization options can of course also be used.

Finally, it should be pointed out that within the scope of the procedure according to the invention, the occupancy status of parking spaces can of course also be analyzed using the expanded map data sets.

Claims (18)

A method for updating and / or expanding a map data set of a limited environment (2) traveled by mobile devices (3) with environment sensors (16), which is a parking environment, a current map data record being sent to a mobile device (2) operated within the environment (2). 3) is transmitted and used within the device (3) to localize the mobile device (3) within the environment (2), characterized in that at least one property of at least one feature in the environment not yet or differently contained in the map data set (2) Descriptive feature data is determined from sensor data from the environment sensors (16) of the mobile devices (3) and stored in the copy of the map data set stored in the mobile devices (3), after which the map data records expanded in this way are sent to a central server device assigned to the environment (18) and there, taking into account the ne uen and / or changed properties are merged into an updated map data set, the properties including a class and / or position of the feature obtained as a result of a classification and / or, in particular as a result of the classification, at least one purpose of the feature, the classification and / or the determination of the purpose from the sensor data takes place using an auxiliary database transmitted from the server device (18) to the mobile device (3), which allows the mapping of sensor data to at least one class and / or purpose of the feature. Procedure according to Claim 1 , characterized in that mobile devices (3) with different environmental sensors (16) are used. Procedure according to Claim 1 or 2 , characterized in that the map data set is transmitted to the respective mobile device (3) when entering the environment (2) and an expanded map data record is transmitted to the server device (16) with each expansion and / or when leaving the environment (2) . Method according to one of the preceding claims, characterized in that information signs (20) within the surroundings (2) and / or lane markings (7, 8) and / or parking lot markings (6, 9) and / or aids, in particular fire extinguishers (10 ) can be identified by determining a class and / or purpose. Method according to one of the preceding claims, characterized in that the properties are determined at least partially in the context of a localization of the own mobile device (3), in particular in the context of the implementation of a SLAM method. Method according to one of the preceding claims, characterized in that an assignment of new and / or changed properties to a feature is carried out by a map comparison on the basis of predefined features on the part of the server device (18). Method according to one of the preceding claims, characterized in that clearly identifiable markers are used within the environment to localize the mobile device (3). Method according to one of the preceding claims, characterized in that the map data records are transmitted via a device-to-infrastructure communication, in particular a WLAN communication. The method according to any one of the preceding claims, characterized in that at least one camera (22) and / or at least one ultrasonic sensor and / or at least one radar sensor and / or at least one lidar sensor are used as environmental sensors (16) and / or for assigning the Sensor data for a position in the map data set, a scan matching method is used. Method according to one of the preceding claims, characterized in that a path covered by the own mobile device (3) is also stored in the map data set with a property indicating whether it can be driven on. Procedure according to Claim 10 characterized in that the server device (18) identifies areas of the environment (2) bypassed, in particular road damage and / or danger spots, by comparing paths traveled by several mobile devices (3). Method according to one of the preceding claims, characterized in that current occupancy states of parking spaces in the parking area are determined from the expanded map data sets. Method according to one of the preceding claims, characterized in that feature data relating to the same property of a feature from various mobile devices (3) are statistically evaluated by the server device (18), in particular using outlier detection. Method according to one of the preceding claims, characterized in that at least one property of at least one feature contained in the map data set is evaluated by a function of the mobile device (3) besides the localization, in particular by a driver assistance function. Method according to one of the preceding claims, characterized in that the map data set is transmitted to a mobile terminal device (24) of a user, in particular a mobile phone, and is evaluated there as part of a software application (26), in particular a navigation application. Method according to one of the preceding claims, characterized in that for an initial localization of the mobile device (3) in the map data set, a predefined feature, in particular located at an entrance to the surroundings (2), is detected. Method according to one of the preceding claims, characterized in that a GPS position determination is taken into account in order to identify the surroundings (2) and / or an initial position in the map data set. Method according to one of the preceding claims, characterized in that a motor vehicle (4, 14) is used as the mobile device (3) and / or the surroundings (2) is a parking garage.

Images