Method and system for optimizing detected parking spaces - Patents.com

Abstract

The present invention relates to a method for improved utilization of a parking space detected by a parking assistance system (1), comprising: a) detecting (S10) a potentially available parking space by initially detecting at least one sub-area of ​​the parking space using a sensor system (2) of the vehicle (F); b) determining an initial parking area contour (4) and an initial parking area orientation (α) for the detected potentially available parking space and determining a quality criterion for the parking space (S11); c) varying the parking area orientation (α) by rotating the initial parking area contour (4) about a vertical axis, whereby a parking area contour with a changed rotational position is generated (S12); d) calculating a revised quality metric based on the changed rotational position of the parking area contour (S13); e) checking whether the quality standards relating to the parking spaces are improving (S14); f) repeating steps c) to e) iteratively until a termination criterion is reached (S15); g) specifying (S16) a final parking target area (5) with a final parking area contour and a parking area orientation based on the identified best quality criterion; h) performing a parking maneuver (S17) so that the vehicle (F) is parked in the final parking target area (5); The present invention relates to a method comprising the steps of:

Description

[Technical field]

[0001] The present invention relates to a method and system for improved utilization of a parking space detected by a parking assistance system and parked at least partly autonomously. [Background technology]

[0002] Parking assistance systems are basically known. To detect parking spaces with a (partially) automatic parking assistance system, a parking lot or possible parking spaces are scanned as the car drives by. Due to the limited range of the sensors and / or the speed of the vehicle (e.g. up to 50 km / h), the exact parking situation is often not detected accurately. In particular, in many cases, only the edges of the parking lot adjacent to the vehicle are detected, so that further surrounding objects that border, such as further parked vehicles and borders or lines drawn on the ground, are detected only with insufficient accuracy.

[0003] The initial detection of the parking lot may for example not be exactly parallel or perpendicular to the optimal parking position of the vehicle in the parking lot, or the vehicle may not ideally scan the parking lot at a right angle, so that during the initial determination of the parking area contour by the parking assistance system it is not clear how the parking area contour should actually be oriented correctly to obtain an optimal parking result. This is particularly problematic when no imaging sensor, such as a camera, is used, which allows the interpretation of information provided by other sensors (for example based on lines drawn on the ground).

[0004] In addition, adjacent objects, such as vehicles that are also not optimally parked, may be at an angle to the actual parking space. If the parking area contour to be parked is now calculated along the orientation of the vehicle itself, the vehicle will be parked by the parking assistance system without being well positioned between the bordering objects, i.e. according to the usual parking position that a human driver would take in particular.

[0005] A problem with known parking assistance systems is that often the parking lot is not used optimally and the parking process may not be started or may be aborted because the parking lot is deemed not possible for parking, even if parking would be possible by changing the parking area orientation and / or the parking area size. Summary of the Invention [Problem to be solved by the invention]

[0006] Based on this, it is an object of the present invention to specify a method for improved utilization of parking spaces detected by a parking assistance system, which method allows for improved utilization of existing parking lots and thus allows parking the vehicle in a more natural manner modeled on the parking behavior of a human driver. [Means for solving the problem]

[0007] This object is achieved by a method having the features of independent claim 1. The dependent claims relate to preferred embodiments. Equivalent claim 14 relates to a parking assistance system for a vehicle and further equivalent claim 15 relates to a vehicle comprising such a parking assistance system.

[0008] According to a first aspect, the invention relates to a method for improved utilization of a parking space detected by a parking assistance system, the method comprising the steps of:

[0009] First, a potentially available parking space is detected by initially capturing at least one partial area of ​​the parking space with a sensor system of the vehicle. Any sensor system of the vehicle can be used as the sensor system, for example, ultrasonic sensors, radar sensors, cameras, LIDAR, etc. For example, during the initial capture, only a partial area of ​​the parking space adjacent to the vehicle is captured, but not the entire parking situation along the depth of the parking lot.

[0010] An initial parking area contour and an initial parking area orientation are then determined for the detected potentially available parking space. The initial parking area contour can be calculated, for example, from information available through an initial capture of at least one partial area of ​​the parking space. For example, the initial parking area contour can be a rectangular area, the size of which is derived at least in part from the initially captured information about the parking space, and preferably also in part from information about the vehicle itself, for example the length and / or width of the vehicle to be parked. An initial parking area orientation is understood to mean a rotational position of this parking area contour in the horizontal direction, which may point to a reference axis.

[0011] The parking area orientation is then modified by rotating the initial parking area contour or a parking area contour derived from the initial parking area contour about a vertical axis, thereby creating a parking area contour with a modified rotation position. This means that in a first iteration step, starting from an initial parking area orientation, the initial parking area contour is first rotated, and in subsequent iteration steps the process either restarts with this initial parking area orientation (e.g. if the first iteration step results in an invalid parking space) or a modified parking area contour or a modified parking area orientation is used to further improve the parking space.

[0012] After changing the rotational position of the parking area contour, a modified quality metric is calculated. Preferably, the quality metric has also already been defined or calculated for the initially captured parking space, i.e. the initial parking area contour and the associated initial parking area orientation. The quality metric may refer to a parameter of the parking space, such as the width of the parking area contour, the area of ​​the parking area contour, the alignment of the parking area contour with other objects, or a weighted combination of these parameters. The modified quality metric can be used to determine whether the usability of the parking space has changed due to the change in the rotational position of the parking area contour.

[0013] It is then checked whether the quality metric for the parking space is improving, which can be done in particular by comparing the current quality metric value with an initial quality metric value or with a quality metric value from a previous iteration step.

[0014] The above steps of changing the parking area orientation, calculating a revised quality metric and checking the improvement of the quality metric are then repeated iteratively until a stop criterion is reached. The stop criterion can be, for example, that the quality metric no longer improves significantly, i.e. that the maximum or essentially maximum value of the quality metric is reached. Alternatively, a threshold value can be specified as the stop criterion, so that the iterations end if the quality metric exceeds the threshold value.

[0015] A final parking target area having a final parking area contour and a parking area orientation is then determined based on the determined best quality criterion.

[0016] Finally, the parking process is performed so that the vehicle is parked in the final parking target area.

[0017] The technical advantage of the method is that, in particular, advantageous parking in a parking space is achieved by the parking assistance system regardless of how the parking space was initially scanned, i.e. even after passing the parking space at an angle with the car, it is possible to park in the parking space in a manner adapted to human driving behavior, which additionally makes it possible to park in parking spaces that would be rejected by conventional parking assistance systems.

[0018] According to an exemplary embodiment, the parking area orientation is changed by rotation in a clockwise and / or counterclockwise direction. For example, initially, an attempt can be made to improve the use of the parking area by rotation in a first rotation direction. For example, this first rotation direction can be fixed or determined based on available information about the parking situation or on information about the surroundings provided by a sensor system of the vehicle. The information can be, for example, the orientation of adjacent vehicles, the orientation of parking space signs, and / or the orientation of other objects.

[0019] According to an exemplary embodiment, the parking area contour is first rotated iteratively in each case in a first rotation direction and in a second rotation direction after a stop criterion is reached, which can for example be that no improvement in the quality criterion has been achieved in the last iteration step.

[0020] According to an exemplary embodiment, the first rotation direction is determined based on the parking situation derived during initial capture of at least one partial area of ​​the parking space. This means that it is already possible to start with an advantageous rotation direction, where an improvement in the quality metric is expected. For example, the orientation of adjacent vehicles, the orientation of the parking space sign, or other objects can be used to determine the first rotation direction.

[0021] According to an exemplary embodiment, the step size of the rotation of the parking area contour is changed. In particular, if an improvement of the quality criterion is achieved in an iteration step, the rotation can be continued in the same direction with a larger step size. However, if an improvement of the quality criterion is not achieved in an iteration step or if an iteration step results in an invalid parking space, the step size is preferably reduced in the next iteration step. This means that an optimal or essentially optimal use of the parking space can be achieved.

[0022] Preferably, after an iteration step has been performed in which the quality criterion has improved, i.e. the usability of the parking space has improved, an iteration step is performed in which the rotational position is not changed, which means that it is possible to check whether the quality criterion has changed due to a rotation of the parking area contour or due to the detection of new objects that are newly detected due to the movement of the vehicle into the parking lot.

[0023] According to an exemplary embodiment, after the iteration step, it is checked whether the parking area contour meets the defined size criteria, which means that the parking space can be classified as valid, i.e. a size that allows parking is available.

[0024] According to an exemplary embodiment, the size criterion indicates a lower and / or upper limit for the length and width of the parking area contour. This size criterion makes it possible to check whether the vehicle to be parked can be parked in the parking area anyway, since the parking area has a specified minimum size at the current rotational position of the parking area contour. Defining an upper limit for the length and width of the parking area contour may mean that any large area will not be considered as a parking space and will not be parked there.

[0025] According to an exemplary embodiment, in at least some iteration steps after the rotation of the parking area contour, it is checked whether the length and / or width of the parking area contour can be enlarged without collisions, in which case the length and / or width of the parking area contour is enlarged in such a way that the result is an enlarged parking area contour that does not overlap with surrounding objects.

[0026] According to an exemplary embodiment, the parking area contour Expansion of is first subjected to overlap with at least one surrounding object, and preferably with all adjacent surrounding objects. It is carried out so that , then The expanded parking area contour is Continue moving the camera until there is no overlap with the detected surrounding objects. Shrink with Small do. This allows the parking area contour to be extended utilizing the maximum parking area coverage.

[0027] According to an exemplary embodiment, the step of changing the parking area orientation by rotating the parking area contour around the vertical axis is only performed when the parking process has already been started by the parking assistance system and the vehicle has already been partially moved into the parking lot. This is advantageous when sensor systems are used that can only capture the surrounding area of ​​the vehicle in a very narrow environment, such as up to 5 m. Starting the optimization of the parking area after the vehicle has been partially moved into the parking lot means that more information about the parking situation is already available than when the parking area contour and its rotational position were initially determined. This means that the parking area contour and its rotational position can be adapted in an improved manner to the parking situation, which can only be detected when the vehicle is moved into the parking lot.

[0028] If a sensor system is used to detect the parking situation and this sensor system is already able to capture a parking space with sufficient depth before the start of the parking process, and only if the parking process has not yet been started by the parking assistance system and the vehicle has not yet been moved into the parking space, the step of changing the parking area orientation by rotating the parking area contour around a vertical axis can already be carried out, which means that the parking process can already be started with an at least partially optimized parking area orientation.

[0029] According to an exemplary embodiment, the vehicle surroundings are captured during the vehicle movement after an initial capture of at least one partial area of ​​the parking space, and updated information on the vehicle surroundings is generated. The calculation of the corrected quality criterion and / or the check of whether the parking area contour satisfies the defined size criterion is performed on the basis of this updated information. This allows the orientation and / or the size of the parking area contour to be continuously adapted to the newly determined surrounding situation.

[0030] According to an exemplary embodiment, the step of performing the parking process to place the vehicle in the final parking target area includes maneuvering the vehicle, for example, if the final parking target area cannot be occupied in the correct position during the parking operation, i.e., when driving directly into the parking lot, rather, multiple parking operations may need to be performed so that the vehicle occupies the final parking target area.

[0031] According to an exemplary embodiment, the information on the parking area contour, the parking area orientation and the associated quality criteria is at least partially stored in the iteration steps. In other words, in each iteration step, a backup of the information on the parking area contour, the parking area orientation and the associated quality criteria is generated. Only the backup of the information from the last iteration step or the backup of the information from several previous iteration steps can be stored. As a result, in a subsequent iteration step, if it is determined that this iteration step did not achieve a positive change of the quality criteria or that a valid parking space was not obtained, the backup of the previous iteration step can be used.

[0032] According to a further aspect, the invention relates to a parking assistance system for a vehicle for improved utilization of a detected parking space, comprising at least one sensor system for capturing the vehicle surroundings and a computer unit. a) detecting a potentially available parking space by initially capturing at least one sub-area of ​​the parking space using a sensor system of the vehicle; b) determining an initial parking area contour and an initial parking area orientation for the detected potentially available parking space and determining a quality criterion for the parking space; c) changing the parking area orientation by rotating the parking area contour about a vertical axis; d) calculating a modified quality metric based on the modified rotational position of the parking area contour; e) checking whether quality standards regarding parking spaces are improving; f) iteratively repeating steps c) to e) until a stopping criterion is reached; g) determining a final parking target area with a final parking area contour and a parking area orientation based on the determined best quality criterion; h) performing a parking procedure so that the vehicle is parked in the final parking target area; The method is configured to:

[0033] According to a final aspect, the invention relates to a vehicle including a parking assistance system.

[0034] For the purposes of the present invention, the terms "approximately", "substantially" or "about" refer to deviations of ±10%, preferably ±5%, from the respective exact value and / or deviations that are not significant for functionality.

[0035] Further developments, advantages and possible applications of the invention can also be seen from the following description of exemplary embodiments and from the drawings, in which all the features described and / or illustrated in the figures, either on their own or in any desired combination, are essentially subject matter of the invention, regardless of the combination of claims or their subsequent references, the subject matter of which also becomes part of this description.

[0036] The invention will now be described in more detail with the aid of exemplary embodiments and with reference to the drawings. [Brief description of the drawings]

[0037] [Figure 1] 1 shows a schematic diagram of a vehicle with a parking assistance system as an example. [Diagram 2] A schematic diagram of a parking situation is shown as an example, where a vehicle passes a parking lot, which is captured by the vehicle's sensor system and an initial parking area contour is determined. [Diagram 3] 1 shows, as an example, a schematic diagram of a parking process in which a vehicle enters a parking lot and rotates the initial parking area contour to allow improved utilization of the parking lot. [Figure 4]1 shows an example schematic diagram of a parking situation with surrounding obstacles where parking in the parking lot is possible by rotating the parking area contour. [Diagram 5] A schematic diagram of a parking situation with surrounding obstructions is shown as an example, where improved usability of the parking lot is achieved by both rotating the parking area contour and by changing the size of the parking area contour. [Figure 6] 1 shows, by way of example, a flow chart of an exemplary embodiment of a method for improved utilization of parking spaces. [Figure 7] A diagram illustrating the steps of a method for improved utilization of parking spaces is shown as an example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0038] 1 shows, by way of example, a vehicle F with a parking assistance system 1. The parking assistance system 1 comprises a sensor system 2 with a number of individual sensors distributed around the vehicle F. The sensor system 2 may, for example, capture only a partial area of ​​the surrounding area of ​​the vehicle F or may be designed to capture the entire surrounding area of ​​the vehicle F (360° capture). The sensor system may comprise any sensor type, such as ultrasonic sensors, one or more radar sensors, one or more cameras and / or one or more LIDAR sensors.

[0039] The sensor system 2 is coupled to a computer unit 3 having at least one processor and at least one memory unit, which computer unit 3 is designed to execute the method sequences disclosed in this document and thus to achieve an improved utilization of the available parking space.

[0040] 2 shows by way of example a parallel parking situation detected by the parking assistance system 1 of a vehicle F. In the illustrated Cartesian coordinate system, the vehicle passes laterally through a parking space in the positive x-direction, as represented by the arrow. In this case, the sensor system 2 of the parking assistance system 1 detects a parking space 6, which in the illustrated exemplary embodiment is formed between two objects U1, U2, as indicated by the dashed semicircle.

[0041] As shown in FIG. 2, the longitudinal direction of the parking lot does not run perpendicular to the direction of travel of the vehicle F, but at an angle to it, i.e. it is neither a parallel nor a parallel parking situation, but a tiered parking situation that is between these two orientations.

[0042] As shown by the dashed semicircle, the sensor system 2 is often not only unable to capture the entire parking situation that actually exists, but can only detect the opening of the parking lot facing the vehicle F, so that the actual orientation of the parking area of ​​the parking lot or its actual size is not available when starting the parking process.

[0043] Therefore, when starting a parking process, often only information is available that indicates that the detected parking space is a parallel or perpendicular parking space. Therefore, in the parking situation according to Fig. 2, it is first assumed that the parking space has an initial parking area contour 4, i.e. that when the parking space is initially detected, its longitudinal axis runs perpendicular to the driving direction of the vehicle, as shown by way of example in Fig. 2 with a dashed rectangle.

[0044] If vehicle 4 were to perform the parking maneuver according to the original plan, vehicle F would not stop at a tiered parking position similar to the orientation of the surrounding objects U1, U2 after completing the parking maneuver, but would instead stop, for example, with the vehicle's longitudinal axis perpendicular to the original driving direction.

[0045] 3 shows diagrammatically the optimization steps used to enable improved utilization of an initially detected parking space, where the surroundings of the vehicle F are captured at least occasionally or continuously and the size and / or orientation of the initial parking area contour 4 is modified based on the detected surroundings such that an improved utilization of the existing parking space occurs based on the detected surroundings of the vehicle F.

[0046] Starting from the initial parking area contour 4 and the initial parking area orientation α, a quality metric is first calculated which indicates a measure of the quality of the parking space. The quality metric may for example refer to the length of the parking space opening, through which the vehicle enters the parking space, or the total area of ​​the detected parking area.

[0047] Subsequently, as shown in Fig. 3 with a curved arrow, the parking area orientation α is changed, in particular by rotating the initial parking area contour 4 around a vertical axis (around the z-axis in the depicted coordinate system), which results in a modified parking area contour 4' with a modified parking area orientation α', as shown in Fig. 3. The direction of rotation can be determined randomly or based on information regarding the orientation of the parking space derived from detected information about the surroundings.

[0048] After changing the parking area orientation, a modified quality metric is calculated, which may differ from the original quality metric due to the rotation of the parking area contour, i.e. the quality metric may be improved or worse.

[0049] It is then checked whether the quality criterion has improved or deteriorated. If so, the method continues on the basis of this modified parking area contour 4'. If no improvement in the quality criterion has been found, the result of the last iteration step is discarded and a backup of the last successful iteration step, in this example the initial parking area contour 4, is loaded in order to achieve an improved utilization of the parking area on this basis. For example, in a subsequent iteration step the magnitude of the rotation angle by which the parking area contour is rotated can be reduced and / or the direction of rotation can be changed, i.e. an attempt is made to improve the utilization of the parking area by rotating the parking area contour in the opposite direction.

[0050] Preferably, after an iteration step has been performed in which the quality metric has improved, i.e. the usability of the parking space has improved, an iteration step is performed in which the rotational position is not changed. This means that it can be ascertained whether the quality metric has changed due to a rotation of the parking area contour or due to the detection of a new object that is newly detected due to the movement of the vehicle F into the parking lot. If the quality metric has changed without rotating the parking area contour, it can be concluded that a change in the quality metric has been found due to a changed and detected surrounding situation.

[0051] For example, Fig. 4 shows a surrounding scenario similar to Fig. 3, with the difference that a further surrounding object U3, not previously detected by the sensor system 2, is detected at the edge of the parking lot. Now, if the surroundings are intermittently or continuously captured during the movement of the vehicle F, this further surrounding object U3 can also be detected, and it can be determined in an iterative step, without changing the rotational position, whether the further surrounding object U3 affects the quality criterion. In other words, it can therefore be determined in an iterative step, without changing the rotational position, whether the newly detected surrounding object has caused a deterioration of the quality criterion and thus a changed parking situation, which requires a new adaptation of the parking area contour.

[0052] As can be seen in FIG. 4, a further rotation of the parking area contour makes it possible to determine a final parking target area 5 which allows the vehicle F to be parked in the parking area.

[0053] After changing the parking area orientation and / or the parking area contour, a check step can be carried out to determine whether a valid parking area is obtained with the changed parking area orientation and / or the changed parking area contour. In this case, it is possible to check whether the parking area contour does not collide towards surrounding objects and also whether the length of the parking area contour has a specified minimum length and the width of the parking area contour has a specified minimum width. The minimum length and the minimum width are preferably vehicle-specific variables. In addition, when checking for a valid parking space, it is preferably also checked whether the parking area contour does not exceed a specified maximum length or a specified maximum width. This can mean that free areas with excessive length or width are not recognized as a parking area and a parking process is not performed in these free areas.

[0054] In addition, a check step can be performed to check whether one side of the parking area contour is too long. This may be the case, for example, if a parking lot that is actually too wide is scanned at an angle, so that the initial parking area contour is rotated / diagonal to the actual parking lot. The initial parking area contour is then valid since its opening is shortened by the diagonal position. If the optimization were to rotate the parking area contour to the correct position, the opening would expand and would no longer fit the maximum allowed width. The parking space would then be discarded. This step prevents the detection of parking spaces that are too wide.

[0055] In addition to changing the parking area orientation, the size of the parking area contour can also be changed, in particular the length and width of the parking area contour can be expanded, at least temporarily, in particular after each change of the parking area orientation α, in such a way that the parking area contour fills the parking lot without collisions and preferably while keeping a specified safety distance from surrounding objects U1, U2, U3.

[0056] FIG. 5 shows, by way of example, such a change in size of the parking area contour.

[0057] The dotted line shows as an example the parking area contour 4' obtained due to a change in the rotational position of the parking area contour in an iterative step of the method. After this change in the rotational position it is checked whether the parking area contour can be enlarged. In this case the parking area contour is expanded, for example by enlarging the length and width of the parking area contour. This results in a modified parking area contour which does not collide with the surrounding objects U1, U2, U3 but which uses the parking area in an improved way.

[0058] For example, the length and width of the parking area contour can be expanded such that the extent of the parking area contour into the depth of the parking area (e.g., into the negative y direction according to the exemplary embodiment of FIG. 5) is extended to the extent that it reaches the maximum allowed extent in this direction. The maximum allowed extent can be a vehicle-specific value. In a perpendicular parking situation, the longitudinal direction of the parking area contour is preferably expanded up to a vehicle-specific maximum value, and in a parallel parking situation, the transverse direction of the parking area contour is expanded.

[0059] The parking space is then expanded along further spatial directions perpendicular to the spatial direction in which the parking space was previously expanded, for example in the positive and negative x direction, in each case to the extent that an overlap occurs with the surrounding objects U1, U2, U3, in particular with all adjacent surrounding objects U1, U2, U3 located within the defined area. In particular, the expansion can be performed such that an overlap occurs with a defined distance.

[0060] In other words, after the rotation of the parking area contour, an attempt is made to expand the area of ​​the parking area contour in each case, in particular so that the preferably rectangular parking area contour is expanded within its longitudinal and / or lateral extents until an overlap with the surrounding objects occurs, such an expansion of the parking area contour is indicated by the outward arrows in Figure 5.

[0061] Preferably, the parking area contour is also extended in a direction where there are no surrounding objects, but there are other lines or contours that define the boundaries of the parking area. Thus, for example, in a parking lot that is open to the road, with the opening facing the positive y-direction in the illustrated exemplary embodiment, there may be a sign indicating the boundaries of the parking area relative to the road. This sign can be captured using a suitable sensor system, such as a camera, and can be taken into account when extending the parking area contour. This may mean that the final parking target area, and thus the vehicle in the final parked state, is positioned such that the parking target area or the parked vehicle is located in a manner that is aligned with the adjacent surrounding objects U1, U2 on the left and / or right side or on the lines that define the boundaries of the parking space.

[0062] After the parking area contour is expanded, all surrounding objects that overlap with the expanded parking area contour are captured. For example, the objects can be stored as a series of 2D polygons.

[0063] The overlapping surrounding objects are successively processed in a list to capture the vector along which the parking area contour must be reduced in each case so that there is no overlap with the respective surrounding object. This vector is represented in FIG. 5 by an inward arrow. If there are several surrounding objects that overlap in one direction, the vector with the smallest magnitude is used first, i.e. the parking space is reduced along this vector. It should be ensured that the reduction vector does not point away from the vehicle F. This means that a part of the object is located between the vehicle F and the parking space. Therefore, in the illustrated exemplary embodiment, only those reduction vectors that do not have a negative y component are allowed.

[0064] After the parking area contour has been reduced by the smallest of the calculated vectors, the list of remaining overlapping surrounding objects is scanned again and the necessary reduction vectors are calculated again, until the parking area contour is completely free of overlaps with all objects.

[0065] In order to also allow for the expansion of the parking area contour when this is possible due to the movement of the vehicle into the parking lot rather than due to a rotation of the parking area contour, the length and width of the parking area contour are also preferably expanded in the aforementioned iterative steps which do not rotate the parking area contour.

[0066] The quality criterion of the modified parking area contour is preferably determined each time the parking area contour is rotated and thereafter the size of the parking area contour is changed or only the size of the parking area contour is changed.

[0067] FIG. 6 shows a flow chart illustrating an exemplary implementation of the method described above.

[0068] At the start of the method, several variables are initialized in step S20. In particular, a variable J in which the quality metric value from the previous iteration step is stored. old , and a variable J in which the quality metric from the current iteration step is stored. new is initialized with the value 0. In addition, the step size of the rotation angle Δα is initialized with the initial value Δα0. In addition, the variable "mode_current", indicating the current mode, is set to "no rotation". The other modes are "rotation_left" and "rotation_right". In addition, the variable "mode_old", indicating the mode in the previous iteration step, is set to "rotation_left". Finally, the variable "last_rotation_successful", indicating whether the rotation of the parking area contour, also called "parking space" below and in Fig. 6, was successful in the previous iteration step, is set to "false".

[0069] Then the variable "mode_old", i.e. which mode was used in the previous iteration step, is checked in step S21. In the first iteration step, initialization sets the variable to "rotation_left", i.e. step S23 is performed, then the variable "mode_current" is set to "no rotation", i.e. do not rotate the parking area contour.

[0070] When checking the variable "mode_old", if it is determined in step S21 that the variable "mode_old" has the value "no_rotation", i.e. no rotation was performed in the previous iteration step, step S22 first checks whether the change in the parking area orientation in the penultimate iteration step was successful, i.e. whether the change in the parking area orientation in the penultimate iteration step resulted in an improvement of the quality metric.

[0071] In this case, the rotation direction from the penultimate iteration step is kept, otherwise the rotation direction is changed by setting the variable "mode_current" to the appropriate value.

[0072] After determining the mode in the current iteration step, the value of the variable "mode_current" is then checked in step S24, after which, depending on the value of the variable "mode_current", a rotation to the left, i.e. in the counterclockwise direction, a rotation to the right, i.e. in the clockwise direction, or no rotation is performed. If the variable "mode_current" has the value "rotation_left", then in step S25, first all information required to restore the current optimization state of the parking area, in particular the current parking area orientation and the current parking area contour, for example its length and width or its local position, is stored. This information virtually forms a backup of the optimization state of the parking area, which originates from a previous iteration step, on the basis of which the optimization state of the parking area from the previous iteration step can be restored if it turns out that the current iteration step has not led to an improvement of the quality criterion or the method stopping criterion has been reached. After the backup has been created, the parking area contour is rotated to the left, i.e. in the counterclockwise direction, by an angle Δα.

[0073] In particular, in the other cases where the variable "mode_current" has the value "rotation_right", step S26 again first stores all information necessary to restore the current optimized state of the parking area, in particular the current parking area orientation and the current parking area contour, for example its length and width or its local position. This information virtually forms a backup of the optimized state of the parking area resulting from a previous iteration step, on the basis of which the optimized state of the parking area from the previous iteration step can be restored if it turns out that the current iteration step has not led to an improvement of the quality criterion or the stopping criterion of the method has been reached. After the backup has been created, the parking area contour is rotated to the right, i.e. in the clockwise direction, by an angle Δα.

[0074] After rotation of the parking area contour and associated modification of the parking area orientation α, an attempt is made to expand the parking area contour in step S27 so that the parking lot is filled in the best possible way but without collisions with surrounding objects.

[0075] This step S27 is also performed immediately, i.e. without rotating the parking area contour, if the variable "mode_current" has the value "no_rotation", i.e. an attempt is made to directly enlarge the parking area contour without changing the parking area orientation α.

[0076] After carrying out step S27 of expanding the parking area contour, step S28 checks whether the parking space is valid, i.e. whether it has a given minimum length and width. Preferably, as already mentioned above, it is also checked whether the parking space does not exceed a certain maximum length and / or width.

[0077] If the check in step S28 reveals that the parking space is valid, step S29 checks whether a rotation has been performed previously, i.e. whether the value of the variable "mode_current" has the value "no_rotation" or whether the newly calculated quality criterion J new is the old quality metric J from the previous iteration step. old Check if it is greater than.

[0078] If one of these criteria is met, the newly determined parking area contour and / or parking area orientation is adopted in step 30 and the newly determined quality criterion J new is the quality standard J old is stored as

[0079] If the variable "mode_current" does not have the value "no_rotation", i.e. the parking area orientation has just been changed, the variable Δα defining the step size of the rotation angle is preferably increased and the value of the variable "last_rotation_successful" is set to true.

[0080] The process then returns to step S21 and a new iterative step is performed.

[0081] A check in step S29 determines whether the variable "mode_current" does not have the value "no_rotation" and whether the newly calculated quality metric J new is the old quality metric J from the previous iteration step. old If it turns out that the rotation angle step size is not greater than 0, step S34 is performed. The optimized state of the parking area from the previous iteration step is restored (backed up) so that the next iteration step can be resumed using this information or these values ​​of the variables. It is then checked whether the variable "mode_current" has the value "no_rotation". In this case, the variable Δα defining the step size of the rotation angle is preferably decreased and the value of the variable "last_rotation_successful" is set to false.

[0082] The process then returns to step S21 again and a new iterative step is performed.

[0083] If the check in step S28 reveals that there is no valid parking space, step S31 checks whether the variable "mode_current" has the value "no_rotation". In this case, i.e. if an invalid parking space occurs in an iteration step in which the parking area orientation α has not been changed, this parking space is discarded in step S33 and thus the parking process of the parking assistance system is aborted.

[0084] If it is determined in step S31 that the variable "mode_current" does not have the value "no_rotation", i.e. the parking area orientation α has just been changed, then in step S32 it is additionally checked whether one side of the parking area contour is too long, i.e. whether a threshold value has been exceeded. This may be the case if a parking lot that is actually too wide is scanned at an angle, so that the initial parking area contour is rotated / diagonal to the actual parking lot. The initial parking area contour is then valid since its opening is shortened by the diagonal position. If the optimization rotates the parking area contour to the correct position, the opening will be enlarged and will no longer fit the maximum permitted width. The parking space is then discarded in step S28 in order to avoid detecting a parking space that is too wide. By checking in step S31 whether one side is too long, it is determined whether the parking space should be completely discarded or whether the backup from the previous step can be further used.

[0085] If the check in step S34 reveals that one side is too long, the parking space is discarded again in step S33 and the parking process of the parking assistance system is therefore aborted. Otherwise, i.e. if the parking area orientation α has been changed in an iteration step and none of the sides of the parking area contour are too long, step S34 is executed again. The optimized state of the parking area from the previous iteration step is restored (backed up) so that the next iteration step can be resumed using this information or these values ​​of the variables. Then it is checked whether the variable "mode_current" has the value "no_rotation". In this case, the variable Δα defining the step size of the rotation angle is preferably decreased and the value of the variable "last_rotation_successful" is set to false.

[0086] Once step S34 has been performed, the process returns to step S21 and a new iterative step is performed.

[0087] FIG. 7 shows a diagram illustrating the method steps of the proposed method.

[0088] First, a potentially available parking space is detected by initially capturing at least one sub-area of ​​the parking space using the vehicle's sensor system (S10).

[0089] Then, an initial parking area contour and an initial parking area orientation are determined for the detected potentially available parking spaces and a quality criterion for the parking spaces is determined (S11).

[0090] Afterwards, the parking area orientation is changed (S12) by rotating the initial parking area contour or a parking area contour derived from the initial parking area contour about a vertical axis, thereby generating a parking area contour with a changed rotational position.

[0091] After changing the parking area orientation a revised quality metric is calculated based on the changed rotational position of the parking area contour (S13).

[0092] It is then checked whether the quality metric for the parking spaces has improved, i.e. a revised quality metric indicates that an improvement in the parking spaces has occurred (S14).

[0093] Steps S12 to S14 are then repeated iteratively until a stopping criterion is reached (S15).

[0094] A final parking target area having a final parking area contour and a parking area orientation is then determined based on the highest quality criteria (S16).

[0095] Finally, the parking process is executed by the parking assistance system so that the vehicle is parked in the final parking target area (S17).

[0096] The invention has been described above using exemplary embodiments, it being understood that numerous variations and modifications are possible without departing from the scope of protection defined by the claims. This application relates to the invention described in the claims, but also includes the following as other aspects. 1. A method for improved utilization of a parking space detected by a parking assistance system (1), comprising: a) detecting (S10) a potentially available parking space by initially capturing at least one sub-area of ​​said parking space using a sensor system (2) of a vehicle (F); b) determining an initial parking area contour (4) and an initial parking area orientation (α) for the detected potentially available parking space and determining a quality criterion for the parking space (S11); c) modifying the parking area orientation (α) by rotating the initial parking area contour (4) or a parking area contour derived from the initial parking area contour (4) about a vertical axis (S12), resulting in a parking area contour with a modified rotation position; d) calculating a revised quality metric based on the changed rotational position of the parking area contour (S13); e) checking (S14) whether the quality metric for the parking space is improving; f) repeating steps c) to e) iteratively until a stopping criterion is reached (S15); g) determining (S16) a final parking target area (5) with a final parking area contour and a parking area orientation based on the determined best quality criterion; h) executing a parking process (S17) so that the vehicle (F) is parked in the final parking target area (5); A method comprising: 2. 2. The method according to claim 1, characterized in that the parking area orientation is changed by rotation in a clockwise and / or counterclockwise direction. 3. 3. The method according to claim 1 or 2, characterized in that the parking area contour is first rotated iteratively in each case in a first rotation direction and then in a second rotation direction after the stopping criterion has been reached. 4. 4. The method according to claim 3, characterized in that the first rotation direction is determined based on a parking situation derived during initial capture of at least one partial area of ​​the parking space. 5. 5. The method according to any one of claims 1 to 4, characterized in that a step width of the rotation of the parking area contour is varied. 6. 6. The method according to any one of claims 1 to 5, characterized in that after the iteration step it is checked whether the parking area contour meets a defined size criterion. 7. 7. The method according to claim 6, characterized in that the size criteria indicate lower and / or upper limits for the length and width of the parking area contour. 8. 8. The method according to any one of claims 1 to 7, characterized in that in at least some iteration steps after the rotation of the parking area contour it is checked whether the length and / or width of the parking area contour can be enlarged without collisions, and the length and / or width of the parking area contour is enlarged based on the check result so as to result in an enlarged parking area contour that does not overlap surrounding objects (U1, U2, U3). 9. The parking area contour Expansion of First, there is an overlap with at least one surrounding object (U1, U2, U3). It is carried out so that , then The enlarged parking area contour is Until there is no overlap with at least one surrounding object (U1, U2, U3) Shrink with Small Ruko 9. The method according to claim 8, wherein 10. 10. The method according to any one of claims 1 to 9, characterized in that the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed only when the parking process has already been started by the parking assistance system and has already partially moved the vehicle (F) into the parking lot, or the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed only when the parking process has not yet been started by the parking assistance system and the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed after the parking process has been started by the parking assistance system. 11. 11. The method according to any one of claims 1 to 10, characterized in that the vehicle surroundings are captured during the vehicle movement after an initial capture of at least a partial area of ​​the parking space, updated information about the vehicle surroundings is generated, and the calculation of the revised quality criterion and / or the check whether the parking area contour meets the defined size criterion is performed on the basis of the updated information. 12. 12. The method according to any one of claims 1 to 11, characterized in that the step of performing the parking process of placing the vehicle (F) in the final parking target area (5) includes maneuvering the vehicle (F). 13. 13. The method according to any one of claims 1 to 12, characterized in that the information on the parking area contour, the parking area orientation and the quality criterion is at least partially stored in the iteration step. 14. A parking assistance system for a vehicle for improved utilization of a detected parking space, comprising at least one sensor system (2) for capturing the surroundings of the vehicle and a computer unit (3), said computer unit (3) comprising: a) detecting a potentially available parking space by initially capturing at least one partial area of ​​the parking space with the sensor system (2) of the vehicle (F); b) determining an initial parking area contour (4) and an initial parking area orientation (α) for the detected potentially available parking space and determining a quality criterion for the parking space; c) modifying the parking area orientation (α) by rotating the parking area contour around a vertical axis; d) calculating a modified quality metric based on the modified rotational position of the parking area contour; e) checking whether the quality metric for the parking space is improving; f) iteratively repeating steps c) to e) until a stopping criterion is reached; g) determining a final parking target area with a final parking area contour and a parking area orientation based on the determined best quality criterion; h) executing a parking process so that the vehicle (F) is parked in the final parking target area (5); A parking assistance system configured to implement the above. 15. 15. A vehicle including the parking assistance system according to claim 14. [Explanation of symbols]

[0097] 1. Parking Assistance System 2 Sensor System 3 Computer Unit 4 Initial parking area contour 4' Modified parking area contour 5 Final parking target area 6. Parking F vehicle α Initial parking area direction α' Corrected parking area orientation U1, U2, U3 Surrounding objects

Claims

1. A method for improved utilization of a parking space detected by a parking assistance system (1), comprising: a) detecting (S10) a potentially available parking space by initially capturing at least one partial area of ​​said parking space using a sensor system (2) of a vehicle (F); b) determining an initial parking area contour (4) and an initial parking area orientation (α) for the detected potentially available parking space and determining a quality criterion for the parking space (S11); c) modifying the parking area orientation (α) by rotating the initial parking area contour (4) or a parking area contour derived from the initial parking area contour (4) about a vertical axis, resulting in a parking area contour with a modified rotation position (S12); d) calculating a revised quality metric based on the changed rotational position of the parking area contour (S13); e) checking (S14) whether the quality metric for the parking space is improving; f) repeating steps c) to e) iteratively until a stopping criterion is reached (S15); g) determining (S16) a final parking target area (5) with a final parking area contour and a parking area orientation based on the determined best quality criterion; h) executing a parking process (S17) so that the vehicle (F) is parked in the final parking target area (5); Including, in at least some iteration steps after the rotation of the parking area contour it is checked whether the length and / or width of said parking area contour can be enlarged without collisions, said length and / or width being enlarged based on said check result in such a way that an enlarged parking area contour is obtained which does not overlap with surrounding objects (U1, U2, U3); method.

2. 2. The method according to claim 1, characterized in that the parking area orientation is changed by a rotation in a clockwise and / or counterclockwise direction.

3. 3. Method according to claim 1 or 2, characterized in that the parking area contour is first rotated iteratively in each case in a first rotation direction and then in a second rotation direction after the stopping criterion has been reached.

4. 4. The method according to claim 3, characterized in that the first rotation direction is determined based on a parking situation derived during initial acquisition of at least one partial area of ​​the parking space.

5. 3. The method according to claim 1 or 2, characterized in that the step width of the rotation of the parking area contour is varied.

6. 3. The method according to claim 1, further comprising checking, after the iteration step, whether the parking area contour satisfies a defined size criterion.

7. 7. The method according to claim 6, characterized in that the size criteria indicate lower and / or upper limits for the length and width of the parking area contour.

8. 3. The method according to claim 1 or 2, characterized in that the parking area contour is first enlarged in such a way that an overlap with at least one surrounding object (U1, U2, U3) occurs and then the enlarged parking area contour is reduced until there is no overlap with the at least one surrounding object (U1, U2, U3).

9. 3. The method according to claim 1 or 2, characterized in that the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed only when the parking process has already been initiated by the parking assistance system and has already partially moved the vehicle (F) into the parking lot, or the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed only when the parking process has not yet been initiated by the parking assistance system and the step of changing the parking area orientation by rotating the parking area contour about a vertical axis is performed after the parking process has been initiated by the parking assistance system.

10. 3. The method according to claim 1, characterized in that the vehicle surroundings are captured during the vehicle movement after an initial capture of at least a partial area of ​​the parking space, and updated information about the vehicle surroundings is generated, and the calculation of the revised quality criterion and / or the check as to whether the parking area contour meets a defined size criterion is performed on the basis of the updated information.

11. 3. The method according to claim 1 or 2, characterized in that the step of performing the parking process of placing the vehicle (F) in the final parking target area (5) comprises maneuvering the vehicle (F).

12. 3. The method according to claim 1 or 2, characterized in that information regarding the parking area contour, the parking area orientation and the quality criterion is at least partially stored during the iteration steps.

13. A parking assistance system for a vehicle for improved utilization of a detected parking space, comprising at least one sensor system (2) for capturing the surroundings of the vehicle and a computer unit (3), said computer unit (3) comprising: a) detecting a potentially available parking space by initially capturing at least one partial area of ​​said parking space with said sensor system (2) of said vehicle (F); b) determining an initial parking area contour (4) and an initial parking area orientation (α) for said detected potentially available parking space and determining a quality criterion for said parking space; c) modifying the parking area orientation (α) by rotating the parking area contour around a vertical axis; d) calculating a modified quality metric based on the modified rotational position of the parking area contour; e) checking whether the quality metric for the parking space is improving; f) iteratively repeating steps c) to e) until a stopping criterion is reached; g) determining a final parking target area having a final parking area contour and a parking area orientation based on the determined best quality criterion; h) performing a parking process so that the vehicle (F) is parked in the final parking target area (5); configured to implement in at least some iteration steps after the rotation of the parking area contour it is checked whether the length and / or width of said parking area contour can be enlarged without collisions, said length and / or width being enlarged based on said check result in such a way that an enlarged parking area contour is obtained which does not overlap with surrounding objects (U1, U2, U3); Parking assistance system.

14. A vehicle including a parking assistance system according to claim 13.

Images