EP2830030B1 - Method for determining and updating an occupancy map in a parking area - Google Patents

Description

The invention relates to a method for ascertaining and updating a debit card in a parking area having a plurality of parking spaces, in particular a parking lot or a parking garage, wherein the occupancy card is stored in a central computing device having a communication device designed to communicate with motor vehicles in the parking area contains a data structure, in or out of at least one position and its occupancy state for each parking lot as an additional attribute or can be derived.

Parking areas, in particular so parking lot or multi-storey car parks, are locations that have a variety of parking lots of motor vehicles. Often, the use of this parking is subject to a charge. For drivers who are looking for a parking space within such a parking area, it is sometimes difficult, especially at a high occupancy, to find a parking space. To solve this problem, it has already been proposed in the prior art to use occupancy sensors for the parking spaces. This provides the information as to whether a parking space is free or occupied. This information can be entered in a badge map, so that a data structure exists from which it can be determined which parking lots are occupied and which are free. In this case, various types of parking sensors are known, for example the use of an ultrasonic sensor installed in a ceiling, which determines via the distance to the next object (motor vehicle or floor) whether a parking space is occupied. Also, measurements of distortion of the earth's magnetic field and the like are possible.

In the use of such parking-specific occupancy sensors, however, is to be expected enormous costs, since each parking lot of the parking area, especially parking garage, must be equipped individually. In addition to the simple occupancy information (free, occupied), these sensors provide no further information. Another expense arises after every occupancy sensor has to be serviced.

DE 10 2010 033 215 A1 discloses a method for assisting a parking operation in a parking garage. A parking system is provided with information about the parking garage at the entrance, from which control data for automatic navigation in the parking garage are generated. Furthermore, an occupancy state recognition system is described which is to receive information about the occupancy of parking spaces from the vehicles and to store it in a database.

EP 1 006 503 A2 concerns the parking space detection in a traffic network. A vehicle-side terminal is to transmit parking space information, representing a stationary phase of the vehicle and the vehicle location, to a control center, which generates information regarding free parking space.

DE 10 2011 086 268 A1 discloses a method for detecting a free parking space. A motor vehicle recognizes by means of a sensor, whether a parking space is free. A location of the free parking space is determined, after which the information about the free parking space and the location position are forwarded to an external receiver.

A system for detecting the parking position of vehicles in an area, especially in parking garages, is disclosed in DE 10 2008 006 356 A1 described. In this case, a parking ticket is provided with an RFID tag to determine the parking position of a motor vehicle and to rediscover the motor vehicle to the driver.

The invention is therefore based on the object, a method for determining and updating a badge map in a parking lot, in particular a parking garage, indicate which avoids the use of parking-specific occupancy sensors, can lead to a more extensive information containing occupancy map and in particular is easier and less expensive to implement.

To solve this problem, a method having the features of claim 1 is provided.

It should already be noted at this point that the sensor data does not necessarily have to be raw sensor data, but that different evaluation states can be present, that is to say the sensor data can at least partially be evaluated by the motor vehicle before it can be sent to the communication device of the Computing be transmitted. The parking area is preferably a parking garage.

The idea of the present invention is therefore to use the surroundings sensor system which is already present anyway on the motor vehicle in order to create, update or refine an occupancy map of the parking area. There are therefore standard in the vehicles already existing environmental sensors (often environment sensors) used to fulfill an additional function, namely to have constantly available a possible current occupancy map of the parking area. Expensive, additional occupancy sensors for each parking place is then no longer necessary. Modern motor vehicles already have a plurality of sensors that can provide useful information, wherein at least one imaging sensor, in particular a camera and / or a PMD sensor, and / or a distance-measuring sensor, in particular a radar sensor and / or a lidar sensor and / or an ultrasonic sensor can be used. Particularly useful as environmental sensors of the motor vehicle also prove in the prior art already proposed "top-view cameras", that is, wide-angle cameras, for example, use fish eyes and are arranged at the top of the body of the motor vehicle such that they , especially together with a front camera, can capture the entire environment of the motor vehicle.

In this context, but also generally in the case of imaging sensors, it is expedient if image data of the imaging sensor are at least partially evaluated in the motor vehicle and / or in the computing device by an image processing algorithm. If the imaging sensor delivers two-dimensional image data, it can nevertheless be used to determine three-dimensional sensor data describing the environment using a suitable image processing algorithm. For example, two-dimensional images taken of different positions of the motor vehicle between which the movement of the motor vehicle is known may also be considered whole motion vector fields. Such image processing algorithms are known, for example, under the keyword "structure from motion". In most cases, motor vehicles also have control units that are already suitable for carrying out such image processing algorithms, which in the present case can also suitably prepare the raw sensor data in order then to transmit the evaluated sensor data to the computing device.

Such a determined and up-to-date occupancy map can serve as a basis for further navigation in the case of manual control of the motor vehicle, for automated allocation of a parking space and / or for automated guidance of the motor vehicle for path planning based thereon. Accordingly, it can be provided that the occupancy card is used for allocating a parking space to a motor vehicle to be parked and / or for route planning within the parking area by a motor vehicle and / or for determining at least one navigation instruction for a motor vehicle. Especially in this context, the advantages of the present invention are particularly clear, because in addition to the simple information whether a parking space is occupied or free, additional useful additional attributes can be included or derivable to a parking lot, the navigation, especially the fully automatic guidance of the motor vehicle (piloted driving) significantly simplify. Already Today, motor vehicles are known, in which an automatic parking is made possible without any intervention by the driver. It may then be expedient to provide for the additional attributes of at least one parking space adjacent to the target parking space to be taken into account when planning a parking procedure of a motor vehicle to be carried out, in particular automatically. Additional attributes may therefore be relevant for a piloted parking operation, for example, whether a larger vehicle protrudes on a neighboring parking space and the like. Damage can be avoided here. When overall piloted, so fully automatic driving a motor vehicle within the parking area, especially parking garage, of course, further, related to all passed parking lot attributes and included in the occupancy map additional information can be considered, for example, "wild", ie outside a parking lot, parked vehicles and / or objects.

The use of the sensor data of the environmental sensors of motor vehicles traveling in the parking area thus makes it possible to determine useful additional attributes. The invention provides that at least one position of a motor vehicle parked in the parking space relative to the parking space and / or at least one dimension of the parked motor vehicle and / or at least one identification feature of the parked motor vehicle are used as additional attributes. In addition to the simple statements "occupied" or "free" so the occupancy information (additional attributes) are enriched to each parking lot to concrete vehicle contours, color and appearance of the occupying vehicle and the exact position and dimension of the currently parked motor vehicle within the parking lot. As already stated, such additional attributes are for the most part extremely useful for safe rail planning in the fully automated operation of motor vehicles in the parking area after, for example, objects that project beyond their parking lot and / or parked outside a parking space are also adequately taken into account. It is already pointed out at this time that whenever a fully automatic (piloted) method of the motor vehicle is to take place in the parking area, the occupancy map expediently also contains information about the architecture may include the parking area itself, in particular the location of immovable components, such as walls, so that the occupancy map can be used as a whole as a basis for fully automatic control of the motor vehicle and the previous path planning.

It is also useful, as additional attributes save more relevant for navigation information about parked vehicles in the occupancy map, for example, the color of the motor vehicle, the model and the manufacturer of the motor vehicle, if detectable in the sensor data or in any case from a sending, eingeparkten or parked motor vehicle as Egodaten sent, for example, a license plate of the motor vehicle and the like. This proves to be particularly useful when it comes to finding a motor vehicle. It can therefore be provided that the occupancy card is used for identifying and locating a particular parked motor vehicle. For this purpose, only the desired motor vehicle descriptive identification data must be compared with corresponding, contained in the occupancy map or derivable from this additional attributes. Corresponding inputs and searches can take place, for example, via suitable terminals in the parking area, for example terminals already used as automatic pay stations, whereby information stored on a parking card can also be taken into account if necessary.

In the method according to the invention, the use of different data structures is conceivable. For example, it is possible for the data structure to be sorted mainly according to the parking spaces and their then explicitly stored additional attributes. In cases in which the occupancy maps can also be used in the fully automatic operation of motor vehicles, however, other data structures are available, from which the additional attributes can be derived at least partially, but in some cases can also be explicitly stored. For example, it is conceivable that the data structure comprises a grid structure in which subareas associated with grid lines are assigned subarea attributes. Objects can then, for example, be linked via subordinate occupancy information be identified and the like. Of course, however, other data structures are conceivable, as they are already known for example from environmental models in motor vehicles. In particular, a CAD structure also proves advantageous. Usable data structures for occupancy maps are already widely known in the prior art and will not be discussed in more detail here.

If a grid structure is used as the data structure, it is also possible to define parking spaces within the grid, for example by entering the center vector of the parking marker at one end of the parking marker into the corresponding subarea or the respective grid point, indicating the direction of the vector in which direction the marker extends, the length of the vector describes the length of the marker. Of course, other possibilities are conceivable. Incidentally, similar ways of entering parking spaces as certain vectors are also occupancy maps based on a CAD model. Also, such CAD-occupancy maps are useful in the process of the invention.

It should again be noted at this point that the present invention relates to a centrally managed occupancy card. The computing device acts as a data collection point and, with its calculations for updating the occupancy map, ensures that there is always a current occupancy map which matches the incoming sensor data of different motor vehicles. This can then be distributed via the communication device in turn to the motor vehicles or they can send special requests, for example for a free parking space and / or navigation instructions there.

It is expedient for the sensor data from the motor vehicle to be transmitted at least partially wirelessly to the computing device-side communication device via a communication device on the motor vehicle side. In this case, the data transmission can expediently take place via a WLAN network. Various variants of motor vehicle-to-X communication (Car2X communication) can be used in the context of the present invention, with corresponding standardized communication networks, in which motor vehicles can participate, are already known in the prior art.

An expedient embodiment of the present invention further provides that a parked or parked motor vehicle also transmits ego data about the motor vehicle itself to the computing device, which are used in the determination or updating of the occupancy map, in particular with regard to a parked parking lot. In motor vehicles themselves, a large amount of information about one's own motor vehicle is usually present, for example its dimensions, its type, special extensions and the like. If it is already known, in any case, that a motor vehicle has parked in a certain parking space, it may be useful, as described above, to store as far as possible information about the parked motor vehicle into the map which is obtained firsthand when the ego data is transmitted can.

A development of the invention provides that in the evaluation of the sensor data for determining or updating the occupancy map, a position of the motor vehicle relative to the features described by the sensor data is taken into account. This position can be determined in the motor vehicle and transmitted with the sensor data, but it is also possible to assign the sensor data to a position of the motor vehicle determined, for example, via a position determination system, which can be done by the computing device. If the position of the motor vehicle to which the sensor data refer is first known in the parking area, it is possible in a simple manner to draw conclusions therefrom for the creation or updating of the occupancy map. To determine the position of the motor vehicle a variety of ways are conceivable. The position should also be comprehensively understood as the orientation of the motor vehicle.

Thus, it can be provided that the position of the motor vehicle is provided by an evaluation of sensor data, in particular of sensor data of an imaging sensor, and / or position determination data of a positioning system installed at least partially in the parking area and / or by a communication network, the part of which forms the communication device Communication data and / or from the route of the motor vehicle from a known position descriptive position data is determined. In this case, some variants are already known in the prior art, by which the position of an object in a certain area can be determined, which can also be used in the context of the present invention. In this case, different positioning measures are often combined, which may also be useful in the context of the present invention. The aim is to determine the position of the motor vehicle as accurately as possible, in particular with a tolerance of less than 10 cm, in particular less than 5 cm. Some concrete possibilities are described in more detail below.

For example, it may be provided that the position determination system comprises lidar sensors which are arranged at tire height but below the body height of motor vehicles and which detect the tires of motor vehicles. Such a concept has recently been proposed under the heading "Outside-In-Lidar Localization", including the post-published German patent application DE 10 2012 016 800 the applicant is referenced. In this case lidar sensors are used, which detect as objects in their detection range the tires of motor vehicles, which can be assigned to certain motor vehicles due to their special correlation. Is now, for example, from a barrier forth, possibly additionally in consideration of odometry, known to which motor vehicle an arrangement of four particular tires that have been detected, is assigned, can be highly accurately determine its position.

In another variant of the present invention can be provided that for the evaluation of the sensor data in a given description of the parking area, in particular the occupancy card, a particle filter and / or a Kalman filter is used for position determination. In this approach, therefore, the sensor data itself is used to derive the position from them by ultimately working with previously known information of the vehicle environment (the given description) to perform vehicle localization. The more accurately the environment of the motor vehicle is described a priori, the better it is possible with such methods to localize itself therein. It is particularly expedient if the occupancy card itself is used as the a priori data, that is to say the predetermined description of the parking area.

Particle filters, often referred to as Monte Carlo localization or sequential Monte Carlo method, are already known from the field of robotics for solving the localization problem. In the solution by means of particle filters, the pose of the motor vehicle is represented by a particle cloud. Each particle represents a possible position of the motor vehicle. Via the particle filter, each particle, ie each position represented thereby, is checked for plausibility. The probability of plausible particles can then be increased, for example, the probability of less plausible particles is reduced. If particles fall below a certain probability threshold, they are discarded.

Specifically, within the scope of the present invention, it may be provided that a data set describing at least the position of immovable, detectable components, in particular walls, of the parking area is determined from the sensor data and compared with hypothesis data records determined for different positions in the parking area by simulation of sensor data , wherein the position of the motor vehicle is determined by comparing the data set with the hypothesis data sets. If the badge card is used, of course, in addition to immovable components such as the walls of a parking garage, other objects, such as parked cars, add. In the process, a simulation of sensor data, in particular based on the occupancy map, is ultimately used here. This is particularly useful, because the a priori data thus correspond to a same or a similar environmental sensor as that which is used in the motor vehicle. Incidentally, it is also conceivable that the given description itself contains sensor data recorded with a sensor. This leads to an improvement of the localization quality, because similar data are compared and do not have to be adjusted lossy via a model.

In this context, it may also be expedient if the position is slowly refined to reduce the computing power, so that, for example, after a comparison, all hypothesis data records having a plausibility threshold that are lower than a plausibility threshold and having a certain plausibility value for matching the positions are discarded and their Position in a predefined area around the positions of the non-discarded hypothesis data sets for the determination of new hypothesis data sets are awarded for at least one further comparison.

Such a method for determining the position based on the perception of sensor data of the motor vehicle itself can also be improved by using additional markers visible to the environmental sensors, at least in sections of the parking area which differ little at different positions, for example driveways bordered by two walls become.

As already stated for the case of using the particulate filter and / or Kalman filter, it is most convenient, then, when using the sensor data itself to locate the motor vehicle, to provide comparative data that has been acquired with a same or similar sensor or simulated accordingly , in particular based on the occupancy card. Thus, in general terms, it can be provided that the position of the motor vehicle is determined at least partially by comparing the sensor data with reference data of the parking area, in particular by comparing the sensor data with reference data recorded with the same type of environmental sensor and / or a same environmental sensor and / or correspondingly, in particular based on the occupancy card, simulated data. Thus, an increase in the quality of localization can be achieved.

Furthermore, as was also indicated in the example, it is generally expedient for the occupancy map to be used at least partially when determining the position of the motor vehicle, in particular as a predetermined description of the parking area.

Of course, the position data can also be at least partially determined and transmitted in the motor vehicle itself, with suitable data, in particular sensor data, also being able to be determined in the motor vehicle itself and transmitted by it to determine the position.

As already mentioned, it can be particularly expedient if different positioning methods or localization methods are used in combination, for example for refinement purposes and / or plausibility purposes. For example, it may be fundamentally expedient to compare sensor data recorded with an imaging environment sensor of the motor vehicle with the environment, as described, for example, by the occupancy map. Appropriately here is the use of oriented in different directions cameras of the motor vehicle with wide-angle lenses as environmental sensors. Incidentally, the aforementioned additional attributes describing the appearance of a parked motor vehicle prove to be useful in this connection. In addition, the use of odometry, which means that it is concluded on the basis of the path traveled by the motor vehicle from a known location to a current location, is expedient. This may be used, for example, as a starting point for defining a range of start values for a particulate filter and the like. Obviously, various possibilities are conceivable.

As already mentioned, a preferred embodiment of the invention provides that the raw sensor data are at least partially evaluated by the motor vehicle itself and the evaluated sensor data is transmitted to the computing device. This can take place in arbitrarily far-reaching configurations, so that it is also conceivable in particular for the raw sensor data to be used to update a copy of the occupancy card stored in the motor vehicle, wherein at least changed portions of the occupancy card are transmitted to the central processing unit as sensor data. Ultimately, the computing device then receives proposals for modification, which must be entered, if appropriate, according to a temporal priority, only in the central version of the occupancy card, whereupon the updated central version of the occupancy card can be redistributed to the motor vehicles. However, this is less preferred since the motor vehicles should in principle always have access to the most recent version of the occupancy map.

It is expedient, however, if the raw sensor data for updating objects in the environment of the motor vehicle descriptive environment map are evaluated, wherein at least a portion of the data contained in the environment map, in particular parked motor vehicles as objects descriptive data, or the entire environment map as sensor data to the Computing device to be transmitted. Such environment maps, often referred to as environment models, are already widely known in the art and are used in the motor vehicle as part of a sensor and information fusion to provide, for example, driver assistance systems that require environmental information to provide a standardized data collection from which the required environmental information is retrieved can. If such an environment map (such an environment model) kept up-to-date anyway in the motor vehicle, it can be usefully transmitted as sensor data to the central computing device. With the known position of the motor vehicle, it is then easily possible to convert the corresponding described objects and their attributes for the occupancy map and enter there.

It should also be noted at this point that (for reasons of energy consumption alone) it is assumed that the sensor data are supplied by currently operated motor vehicles, not by parked, out of service motor vehicles, the latter also being conceivable in principle, for example, when They are electric or hybrid motor vehicles, which are connected to a charging port, which would enable the energization of the environmental sensors.

Overall, the present invention thus describes a system that is formed by the parking area with the central computing device and within the parking area, in particular powered vehicles, the motor vehicles have environmental sensors whose optionally already at least partially evaluated sensor data via a communication link, in particular part of Communication network forms can be transmitted to a communication device of the computing device. The computing device is designed to evaluate sensor data received from motor vehicles in order to determine or update a badge map designed as described with regard to the inventive method. Of course, the central computing device also fulfills certain functions in certain embodiments, for example the allocation of parking spaces to motor vehicles, the availability of the occupancy map to motor vehicles, the determination of navigation instructions for motor vehicles, the finding of parked motor vehicles and the like. Of course, all statements regarding the method according to the invention also apply to such a system.

Fig. 1

ein Parkareal, in dem das erfindungsgemäße Verfahren durchgeführt wird und

Fig. 2

eine Prinzipskizze eines am Verfahren teilnehmenden Kraftfahrzeugs.

Further advantages and details of the present invention will become apparent from the embodiments described below and with reference to the drawings. Showing:

Fig. 1

a parking area in which the inventive method is carried out and

Fig. 2

a schematic diagram of a participant in the process vehicle.

Fig. 1 shows a schematic diagram of a parking area 1, which in the present case is designed as a parking garage, which has several parking levels with parking lots 2, which in the present case for the sake of clarity only in a section 3 is shown. These parking spaces 2 are separated from each other by parking markings 4 and provide space for parked motor vehicles 5. Within the parking area 1 also move motor vehicles 6, which are to be parked or leave the park area 1 after parking.

In order to ensure the smoothest possible operation in the parking area 1, the motor vehicles 5, 6 or their drivers are provided by a central computer 7 information from which, for example, free parking 2 can be inferred as destination parking, navigation instructions to parking 2 or can be generated from the parking area 1 and / or even fully automatic motor vehicles 5, 6 even automatic parking planning is possible. For this purpose, the computing device 7 has a memory device 8 in which a busy map of the parking area 1 is stored. This occupancy map can be designed in various ways, but contains at least the positions of the parking spaces 2 and their occupancy state, which can be stored directly as an additional attribute or can be derived from the data of the occupancy map. However, the occupancy card preferably also contains further information, on the one hand, the basic architecture of the parking area 1, in particular the immovable components such as walls 9, as well as additional additional attributes to the parking spaces 2 and additional data from which further additional attributes to the parking spaces 2 can be derived.

The occupancy map can have, for example, a CAD-like grid structure as a data structure into which various objects can be entered. Then, architectural data and the like are also preferably included. However, other possibilities are also conceivable, for example a data structure organized according to the parking spaces 2, in which, for example a record associated with a parking lot 2 may be realized as a container containing the additional attributes. Of course, further containers may be provided in such a data structure, for example, fixed components of the parking area 1 or parked outside parking 2 objects.

In the entire parking area 1, a WLAN network is available as communication network, the central computing device 7 having a communication device 10 designed for communication in the WLAN network, so that a communication link 11 to motor vehicles 5, 6 can be established. Via the communication link 11, the badge map and information derived therefrom can be made available to the motor vehicles 5, 6, for example a destination car park with an associated navigation instruction or without such when automatic planning of the vehicle 5, 6 is to be carried out automatically Preferably, the occupancy card is made available to the motor vehicle 5, 6 as a whole. If the occupancy map, which, as explained, is also preferred, also contains architectural data for parking area 1, a path planning can be carried out on this basis, especially if variable objects, thus mainly parked motor vehicles 5, are also detected.

In the present case, however, the communication link 11 is also used to transmit from environmental sensors of the motor vehicles 6 in operation and optionally at least partially evaluated sensor data to the computing device 7, which uses the sensor data to create the occupancy map and keep current, thus to update , Overall, so in the in Fig. 1 In the parking area 1 system formed the inventive method can be performed.

The use of the sensor data of the motor vehicle 6 not only allows to do without an additional parking-specific occupancy sensors, but it can also be a significantly improved occupancy map are created, after the additional attribute "free" or "occupied" for a parking lot 2 outgoing information, in particular additional attributes, can be stored. In the present case, the position of a motor vehicle 5 parked in a parking space 2 relative to the parking space 2 and the dimensions of the parked motor vehicle 5 are also considered as additional attributes, and also identification features of the parked motor vehicles 5, for example their color, model, type and the like.

Such additional attributes are made possible by the high information content as standard in motor vehicles 5, 6 built environment sensors. It should be noted that motor vehicles 5, 6 can also provide data relating to themselves (ego data), which can be incorporated into the update of the occupancy map as soon as the motor vehicle 5, 6 has been parked and thus is a parked motor vehicle 5, its properties as an additional attribute the corresponding parking 2 can be assigned.

Fig. 2 shows a schematic diagram of a co-operating in the inventive motor vehicle 12, which may correspond to one of the motor vehicles 5, 6. This has a plurality of environmental sensors, of which radar sensors 13 and ultrasonic sensors 14 are shown as distance sensors in the present case. As an imaging environment sensors, the motor vehicle 12 has a forwardly directed, for example, behind the windshield camera 15 and three so-called top-view cameras 16, two of which are directed to the sides, one to the rear. The cameras 16 are all provided with a wide-angle lens in order to be able to record the widest possible area of the surroundings of the motor vehicle 12. The raw sensor data of the various environmental sensors 13 - 16 are already at least partially evaluated in at least one control unit 17 of the motor vehicle 12 in order to obtain at least partially evaluated sensor data, which can be transmitted to the communication device 10 of the computing device 7 via a communication device 18 on the motor vehicle side.

Various evaluation steps can already take place in the motor vehicle 12, in particular those which are already carried out in motor vehicles today anyway. A first example for this are image processing algorithms, in particular those which can evaluate two-dimensional image data of the imaging environment sensors, in this case the cameras 15, 16, in order to obtain three-dimensional information. In this case, for example, two-dimensional images taken at different positions of the motor vehicle 12 can be viewed, but the movement between these positions is known ("structure from movement"). It should be noted, however, that of course, 3D cameras can be used, such as PMD sensors. Other image processing algorithms are also conceivable, for example those which detect and classify objects in images and the like.

In the present case, an environment map is managed in the control unit 17 of the motor vehicle 12. In this objects are registered in the environment of the motor vehicle 12 and their properties, after several vehicle systems that are not shown here require such information. Appropriately, it is then provided in this embodiment that at least for the occupancy map relevant shares of the environment map are transmitted as sensor data to the computing device 7.

For the update of the occupancy map, in some data structures also directly the additional attributes, it is relevant to spatially assign the sensor data, which in the present case on the position of the sending motor vehicle 5, 6, 12 (which should also include its orientation). This means that when evaluating the sensor data for determining or updating the occupancy map, a position of the motor vehicle 5, 6, 12 relative to the features described by the sensor data is taken into account. It is inventively preferred if the position of the motor vehicle 5, 6, 12 is determined by an evaluation of at least a portion of the sensor data itself, in addition, the occupancy map as a predetermined description of the environment of the motor vehicle 5, 6, 12 is taken into account. Because the badge card contains, in particular then, even if immovable components, ie architectural data, of the parking area 1 are included, already a large amount of information that allows a comparison with sensor data or creating an expectation result for sensor data at certain positions. In particular, this is made possible by the more detailed, determined and updated by the method according to the invention occupancy map. For example, it is conceivable to use a particle filter to localize the location of the sensor data, as is already known in principle from applications of robotics. In addition to objects by distance measurements detecting environmental sensors also generally environmental imaging sensors can provide the sensor data that allows entry into such a comparison. In areas of the parking area 1, for example driveways or the like, where positions can be distinguished only with difficulty, additional markers can be used.

Of course, in addition or alternatively, other possibilities for determining the position are conceivable, for example, a positioning system of the parking area 1, which for the sake of clarity in Fig. 1 not shown. Such a position determination system may, for example, use lidar sensors whose detection range is at tire height, but below the body of motor vehicles 5, 6, 12, so that only the tires of the motor vehicles 5, 6, 12 are detected and further conclusions can be drawn therefrom. Incidentally, positioning systems of this kind may also use communication data, for example, when there are various receiving stations to enable coarse localization and the like. Also odometric approaches are, preferably complementary, conceivable in which the driving distance of the motor vehicle 5, 6, 12 from a known position descriptive position data are taken into account.

In summary, therefore, within the parking area 1 operated motor vehicles 5, 6, 12 sensor data, which were recorded with the environmental sensors of the motor vehicle 5, 6, 12, report via the communication link 11 to the computing device 7, so that in the sensor data contained information can be entered from the perspective of the current (known) motor vehicle position in the occupancy map.

It should also be noted at this point that another useful application of the map created or updated by means of the method according to the present invention is locating parked motor vehicles 5. For if characteristics of the parked motor vehicles 5 are stored in the occupancy map, in particular as additional attributes, these can be used to identify and locate motor vehicles 5. This can be done for example via terminals in the parking area 1, for example, those that also serve to pay a fee for parking.

Claims (17)

1. Method for determining and updating an occupancy map in a parking area (1) having a plurality of parking spaces (2), more particularly a ground-level car park or multistorey car park, wherein the occupancy map is stored in a central computing device (7) having a communication device (10) configured for communication with motor vehicles (5, 6, 12) located within the parking area (1) and contains a data structure, in or from which for each particular parking space (2), as additional attribute, at least its position and its occupancy state is contained or derivable, wherein, for the determining and updating of the occupancy map, sensor data received via the communication device (10) from at least one surroundings sensor of at least one motor vehicle (5, 6, 12) located within the parking area (1) are used,
   characterised in that,
   as additional attributes, also at least one position of a motor vehicle (5) parked on the parking space (2) relative to the parking space (2) and/or at least one dimension of the parked motor vehicle (5) and/or at least one identification feature of the parked motor vehicle (5) are used.
2. Method according to claim 1,
   characterised in that
   the sensor data from the motor vehicle (5, 6, 12) are transmitted at least in part wirelessly via a communication device (18) on the vehicle side to the communication device (10) on the computing device side.
3. Method according to claim 2,
   characterised in that
   the data transmission takes place via a WLAN network.
4. Method according to any one of the preceding claims,
   characterised in that
   a motor vehicle (5, 6, 12) that is being parked or is already parked also transmits data about the motor vehicle (5, 6, 12) itself to the computing device (7), said data being used in the determining or updating of the occupancy map, more particularly with regard to an occupied parking space (2).
5. Method according to any one of the preceding claims,
   characterised in that,
   for the evaluation of the sensor data for the determining or updating of the occupancy map, a position of the motor vehicle (5, 6, 12) relative to the features described by the sensor data is taken into account.
6. Method according to claim 5,
   characterised in that
   the position of the motor vehicle (5, 6, 12) is determined through an evaluation of sensor data, more particularly of sensor data of an imaging sensor, and/or from position determination data of a position determination system installed at least in part in the parking area (1) and/or from communication data supplied by a communication network, the part of which forms the communication device (10), and/or from position data describing the route of the motor vehicle (5, 6, 12) from a known position.
7. Method according to claim 6,
   characterised in that
   the position determination system comprises lidar sensors, which are arranged at tyre height, but below the body of motor vehicles (5, 6, 12), and detect the tyres of motor vehicles (5, 6, 12).
8. Method according to any one of claims 5 to 7,
   characterised in that,
   for the evaluation of the sensor data in respect of a given description of the parking area (1), more particularly of a given description by the occupancy map, a particle filter and/or a Kalman filter is used to determine the position.
9. Method according to any one of claims 5 to 8,
   characterised in that
   the position of the motor vehicle (5, 6, 12) is determined at least in part through comparison of the sensor data with reference data of the parking area (1), more particularly through comparison of the sensor data with reference data recorded with the same type of surroundings sensor and/or an equivalent surroundings sensor and/or from reference data derived from the occupancy map.
10. Method according to any one of claims 5 to 9,
    characterised in that,
    in the determination of the position of the motor vehicle (5, 6, 12), the occupancy map is used at least in part.
11. Method according to any one of the preceding claims,
    characterised in that
    the raw sensor data are evaluated at least in part by the motor vehicle (5, 6, 12) itself and the evaluated sensor data are transmitted to the computing device (7).
12. Method according to claim 11,
    characterised in that
    the raw sensor data are evaluated to update a map of the surroundings describing objects around the motor vehicle (5, 6, 12), wherein at least some of the data contained in the map of the surroundings, more particularly data describing parked motor vehicles (5) as objects, or the entire map of the surroundings are transmitted as sensor data to the computing device (7).
13. Method according to any one of the preceding claims,
    characterised in that
    at least an imaging sensor, more particularly a camera (15, 16) and/or a PMD sensor, and/or a distance-measuring sensor, more particularly a radar sensor (13) and/or a lidar sensor and/or an ultrasonic sensor (14), are used as surroundings sensor.
14. Method according to claim 13,
    characterised in that
    image data of the imaging sensor are evaluated at least in part in the motor vehicle (5, 6, 12) and/or in the computing device (7) by an image-processing algorithm, more particularly an image-processing algorithm obtaining three-dimensional information from two-dimensional image data.
15. Method according to any one of the preceding claims,
    characterised in that
    the occupancy map is used to allocate a parking space to a motor vehicle (5, 6, 12) that is to be parked and/or to plan the route to be taken by a motor vehicle (5, 6, 12) within the parking area (1) and/or to determine at least a navigational instruction for a motor vehicle (5, 6, 12).
16. Method according to claim 15,
    characterised in that,
    in the planning of a parking operation for a motor vehicle (5, 6, 12) that is to be carried out more particularly in an automated way, the additional attributes of at least one parking space adjacent to the target parking space are taken into account.
17. Method according to any one of the preceding claims,
    characterised in that
    the occupancy map is used to identify and to find a certain parked motor vehicle (5).

Images