DE102013010335A1 - Method and device for selecting an object for navigation instructions - Google Patents

Abstract

The invention relates to a method and a device (1) for selecting an object for navigation instructions, comprising a digital road map (3) and a navigation device (2), the digital road map (3) having objects (O1, O2, O3) which can be selected by the navigation device (2) and output for navigation instructions, the objects (O1, O2, O3) being assigned features whose characteristic values are stored, at least one characteristic being dependent on supervision, the navigation device (2) being designed in such a way that this determines a driving maneuver as a function of a calculated route and an actual position of the navigation device (2) and selects objects (O1, O2, O3) located in the vicinity from the digital road map (3), with the selected objects (O1, O2 , O3) a first parameter vector (P1) is set up with the characteristic values, at least one second parameter vector (P2) f jedes is formed for each object (O1, O2, O3) which includes at least the actual position of the navigation device (2) and the position of the object (O1, O2, O3), the first parameter vector (P1) and the second parameter vector (P2) of an object (O1, O2, O3) is linked, the link in each case providing a relevance (RO1, RO2, RO3) for the object (O1, O2, O3), the relevance (RO1, RO2, RO3) of the Objects (O1, O2, O3) are compared with one another, the object (O1, O2, O3) with the greatest relevance (RO1, RO2, RO3) is selected and the navigation device (2) sends a navigation instruction including the object (O1, O2, O3) is generated.

Description

The invention relates to a method and a device for selecting an object for navigation instructions.

Such a method is for example from the DE 10 2004 009 278 A1 in which the route guidance information is output to the user by means of an output device, wherein the environment is recorded by means of an image capture device, the recorded image is decomposed into individual components by means of an image recognition method and the route guidance information to be output is formed and supplemented by at least one recognized component of the image , Preferably, a stereo image is recorded and evaluated, ie the stereo image is decomposed into individual objects. Object data is determined and assigned to the individual objects, for example, the object is a house. In addition, the objects are examined for further properties. Criteria of the further investigation are, for example, size, color, recognizability and / or relevance. The examined objects are thereby to differentiate into striking and less striking, static or mobile objects. Static objects are preferred in the selection as landmarks, since a change in their position and recognizability in the next few moments is not expected. It is also proposed to additionally store object data of static objects in a database in order to retrieve them at a later time.

The invention is based on the technical problem of providing a method and a device for selecting an object for navigation instructions, which on average selects the suitable objects with less technical outlay.

The solution of the technical problem results from the objects with the features of claims 1 and 9. Further advantageous embodiments of the invention will become apparent from the dependent claims.

The method for selecting an object for navigation instructions is carried out by means of a digital road map and a navigation device, the digital road map having objects that can be selected by the navigation device and output for navigation instructions. In this case, the term digital road map is to be understood as a general term for a suitable database which has street guides and objects. The navigation device comprises means for detecting an actual position (eg a GPS receiver), a computing unit for calculating a route and for selecting the objects and an input and output unit for inputting a destination and for outputting navigation instructions. The output of the navigation instructions can be done, for example, visually and / or acoustically. The navigation device can serve to navigate a motor vehicle or a person, for example by means of a portable device. Parts of the navigation device and / or the digital road map can be outsourced, d. H. be off-board. Thus, for example, the arithmetic unit may be outsourced, so that the calculations are carried out, for example off-board in a central computer, which then transmits the results to the on-board parts of the navigation device. Accordingly, it can also be provided, for example, that the digital road map is centrally managed and only relevant parts are transmitted to the navigation device.

The objects of the digital road map are assigned characteristics whose characteristic values are stored. The objects are, for example, houses or other static objects, such as bridges.

The characteristics are, for example, physical dimensions (height, width, etc.), color, lighting, façade structure, name, object class (eg house or bridge), permanence, name recognition (possibly depending on user preferences). It should be noted that not all features must have a feature expression for each object. The feature expression is, for example, in the feature color z. B. the color "yellow". One possible feature characteristic for the feature facade structure is z. B. "glazed". At least one characteristic is subject to supervision. For example, the color of a building may vary depending on the view, for example, because the walls of a house have different colors or the color impression varies depending on the viewing direction. Furthermore, depending on the supervision, a characteristic expression can be visible or not visible.

Depending on a calculated route and an actual position of the navigation device, a driving maneuver is determined, for example, that must be bent left at an intersection. Furthermore, objects located in the environment are selected, such as the objects that are in the area of the intersection.

For the selected objects, a first parameter vector with the characteristic values is respectively set up or the already created first parameter vector is retrieved. Furthermore, a second parameter vector is formed for each selected object, which comprises at least the actual position of the navigation device and the position of the object. The actual position of the navigation device is thereby the position of the user. For example, if parts of the navigation device outsourced, the actual position of the navigation device is the position of the parts of the navigation device that remain in the immediate vicinity of the user, such as the output unit. For applications in motor vehicles, for example, the actual position of the navigation device is equal to the actual position of the vehicle. It should also be noted that the method can be performed several times along the trajectory of the driving maneuver. It is also possible to carry out the method with a future actual position of the navigation device. For example, the arithmetic unit knows a desired transition (trajectory) and can thus estimate, for example, the actual position in the next few seconds. The second parameter vector preferably also includes the driving maneuver in addition to the actual position or the future actual position of the navigation device and the actual position of the objects.

The respectively first parameter vector of an object is in each case linked to the second parameter vector of the object, the link in each case providing a relevance for the object. The relevance is preferably a scalar quantity, for example a numerical value. The combination of the two parameter vectors can be done in different ways. For example, the link is via characteristics or maps. It should be noted that not all features have to be considered in the link. For example, the characteristic "Name" can be ignored during the linkage.

The relevances of the objects determined by the combination of the parameter vectors are compared with each other and the object with the greatest relevance is selected. The navigation device then generates a navigation instruction including the object. The navigation hint does not have to be restricted to an object. So it is conceivable to use situational two objects. In this case, for example, the objects with the largest and the second largest relevance would then be selected. The navigation instruction including the object is carried out such that at least one feature expression of the first parameter vector is called.

The advantage of the method according to the invention is that no camera is needed. Rather, the linkage determines what the user will see or sees. Therefore, the method can be implemented with lower hardware costs for many users. Another advantage is that the linkage allows a better adaptation to changing conditions. Thus, for example, in the prior art, an object is detected and stored when this is detected for the first time. However, the prior art does not take into account that the stored object is unsuitable for navigational indications, for example in the case of a different approach to an intersection, possibly due to a lack of visibility.

In one embodiment, the relevances of the feature values of the selected object are determined, and the navigation device generates the navigation hint including the object and at least the feature expression with the greatest relevance. For example, in one situation the characteristic color and in another situation the feature lighting or physical size can be used. The relevance of the characteristic values includes the visibility or perception of the feature expression, which can also be user-dependent and / or time-dependent, for example. Further preferably, the feature expression of the feature "object class" is preferably used, this being preferably combined with a feature expression of another feature (eg color or physical dimensions). Further preferably, the feature expression of the object class is combined with the feature expression of the feature that has the highest relevance.

In a further embodiment, therefore, at least one feature expression is time-dependent. Thus, for example, it can be taken into account that a feature expression can be perceived differently depending on the day and / or season. For example, the color of a house may not be perceived in the dark. The time can be included via a parameter vector (eg the second parameter vector) or directly in the link. In addition to the characteristic color, for example, the characteristic values for the facade structure and / or lighting can also be time-dependent.

In a further embodiment, a third parameter vector is set up, which includes environmental conditions, which are preferably a position of the sun and / or a weather state and / or a vegetation state. For each object, the first, second and third parameter vector are linked to determine the relevance. In this case, it can be taken into account, for example, that depending on the position of the sun, a feature characteristic of the first parameter vector is only poorly or not at all perceptible. The same applies to the state of vegetation. For example, in summer, an object can be heavily hidden by vegetation. Furthermore, for example, color perception may change during heavy cloud cover. The consideration can be stepped or continuous. For example, it may be provided that, depending on an environmental condition Characteristic expression is classified as perceptible or imperceptible or that the perceptibility of the feature expression is expressed as a continuous function.

In a further embodiment, a fourth parameter vector is set up, which comprises preferences for objects, the four parameter vectors or the first, second and fourth parameter vectors being linked for each object. It can be considered, for example, that a user is color blind or red-green-blind. In this case, the color feature is set to zero, so to speak. Another possible preference may be, for example, that architecturally interesting buildings are preferred as objects. But also preferred locations with eye-catching billboards can preferably be selected as objects. Also preferences of the manufacturer can be considered.

In a further embodiment, the navigation device is associated with at least one camera, wherein it is checked on the basis of the images of the camera, as the selected object is recognizable. This makes it possible to adjust the link continuously. It can be provided that these data are made available to a central office, which then adapts the link accordingly to other users. The images of the camera thus only verify the result of the arithmetic unit.

With regard to the device according to the invention can be fully taken to the statements on the method claims reference.

The invention will be explained in more detail below with reference to a preferred embodiment. The figures show:

1 a schematic block diagram of a device for selecting an object for navigation instructions and

2 a schematic representation of an exemplary navigation situation.

The device 1 for selecting an object for navigation instructions comprises a navigation device 2 , a digital road map 3 , Sensors 4 - 6 as well as an air interface 7 for data exchange with a central office 8th with an air interface 9 ,

The navigation device 2 includes a computing unit 10 , Means for detecting an actual position 11 and an input and output unit 12 ,

The arithmetic unit 10 calculated at least on the basis of a destination input a route, wherein the input and output unit 12 Navigation instructions are issued. It should be noted that the route can also be calculated externally and the calculated route then the computing unit 10 can be transmitted.

In the digital road map 3 In addition to the streets or street guides and objects are stored. These objects are assigned characteristics whose characteristic values are also stored, with at least one feature expression being supervisory-dependent. The features are for example physical dimensions, color, lighting, facade structure, name, object class and / or permanence. For example, depending on the supervision of the color of an object may be different or have a different facade structure. Also, for example, a lighting can be perceived differently, depending on the supervision, so from which direction an object is considered. More preferably, at least one feature expression is time-dependent. For example, the lighting is usually off during the day or difficult to perceive, whereas this evening can be very noticeable. The feature values of the objects are used as a first parameter vector P ₁ in the digital road map 3 stored.

In particular, in places where there are direction changes, a navigation instruction is output by the navigation device. First, the arithmetic unit determines 10 which objects are in the vicinity of the location of the change of direction. Subsequently, for each selected object, a second parameter vector P _{2 is} set up, which is at least the actual position of the object and the actual position of the device 1 or the navigation device 2 (supplied by the means for detecting an actual position 11 ) as well as the driving maneuver to be carried out. From the actual positions of object and device, a plan view of the object can be determined, wherein a future trajectory is known due to the driving maneuver to be performed. For example, it is also possible to take into account a change in supervision since, for example, the vehicle has to be classified on the left. The second parameter vector P ₂ is thus situationally by the arithmetic unit 10 established.

Optionally, a third parameter vector P _{3 is} set up, which includes environmental conditions, such as, for example, the position of the sun and / or a weather state and / or a vegetation state. The environmental conditions can be sensory, for example, by means of the sensors 4 - 6 be captured or even by the central office 8th be transmitted. The position of the sun, for example, can also be calculated from the time of day, taking into account the date. Due to the knowledge of geographic Alignment of the vehicle can thus be calculated, the direction of irradiation of the sun on the vehicle or the object. Furthermore, optionally, a fourth parameter vector P _{4 can be} stored, in which personal preferences of a user are stored, for example by means of the input and output unit 12 entered. The third parameter vector P ₃ is thus set up situationally, whereas the fourth parameter vector P ₄ in the digital roadmap 3 can be stored.

In a next step, at least the first and the second parameter vector P ₁ , P _{2 of} an object are linked together. As a result of the link results in a relevance of the object. Then at least the object with the greatest relevance is selected. The arithmetic unit 10 then generates a navigation hint, including the selected object with at least one feature expression. Such a navigation hint could for example be: "Turn left behind the large, yellow house." Due to the fact that characteristic values are also time-dependent, the result of the linkage is also time-dependent, ie with the same driving situation, the same object does not always have to be selected the appropriate object is always selected.

The navigation device 2 can a camera 13 be assigned, which is formed for example as a stereo camera. By means of the camera 13 captured images can be verified whether the selected object is actually clearly visible. If there are deviations, ie the selected object is not clearly visible in the pictures of the camera 13 so can the head office 8th be informed. As a result, depending on the cause of the deviation, the database of the digital road map can be 3 and / or the linking of the parameter vectors P ₁ -P ₄ are updated for the users, whereby this update also benefit users without their own camera.

In the 2 an exemplary situation is shown having three objects O1-O3 at an intersection. The vehicle F should make the dashed driving maneuver ("turn right"). The dashed driving maneuver can also be called a trajectory or transition.

Taking into account the parameter vectors P ₁ -P _4, the relevance R _{Oi of} the individual objects Oi with i = 1, 2, 3 can now be calculated for each position of this transition, this being explained as an example for the position P.

At this position P, the object O1 should have poor visibility. The poor visibility is given, for example, due to the facade structure of the building in connection with the current position of the sun (strong sunlight on glass building). Therefore, the relevance R _O1 for the object O1 is particularly low and R _O1 <R _O2 , R _O3 .

The object O2 in this example does not correspond to the personal preference of the user or manufacturer, for example, it is an architecturally less interesting building. The object O3 is on the one hand well visible (eg particularly high), therefore R _O3 > R _O1 . In addition, the object O3 corresponds to the personal preference of the user, e.g. For example, the object O3 is an architecturally particularly interesting building and the user is interested in architecture. Therefore, R _O3 > R _O2 and the object O3 is selected for the navigation hint.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102004009278 A1 [0002]

Claims (10)

Method for selecting an object (O1, O2, O3) for navigation instructions by means of a digital road map ( 3 ) and a navigation device ( 2 ), the digital road map ( 3 ) Exhibits objects (O1, O2, O3) which are detected by the navigation device ( 2 ) are selectable and can be output for navigation instructions, characterized in that the objects (O1, O2, O3) are assigned characteristics whose characteristic values are stored, wherein at least one feature expression is supervisory-dependent, depending on a calculated route and an actual position of the navigation device ( 2 ) a driving maneuver is determined and in an environment located objects (O1, O2, O3) are selected, wherein for the selected objects (O1, O2, O3) in each case a first parameter vector (P ₁ ) is set up with the feature expressions, wherein at least one second parameter vector (P ₂ ) is formed for each object (O1, O2, O3) that at least the actual position of the navigation device ( 2 ) and the position of the object (O1, O2, O3), wherein in each case the first parameter vector (P ₁ ) and the second parameter vector (P ₂ ) of an object (O1, O2, O3) are linked, the link in each case having a relevance (R _O1 , R _O2 , R _O3 ) for the object (O1, O2, O3), the relevances (R _O1 , R _O2 , R _O3 ) of the objects (O1, O2, O3) are compared, the object ( O1, O2, O3) with the greatest relevance (R _O1 , R _O2 , R _O3 ) is selected by the navigation device ( 2 ) a navigation instruction including the object (O1, O2, O3) is generated. Method according to Claim 1, characterized in that relevances of the characteristic values of the selected object (O1, O2, O3) are determined and the navigation device ( 2 ) generates the navigation hint, including the object (O1, O2, O3) and at least the feature expression with the greatest relevance. A method according to claim 1 or 2, characterized in that at least one feature expression is time-dependent. Method according to one of the preceding claims, characterized in that a third parameter vector (P ₃ ) is set up, which comprises environmental conditions, wherein for each object (O1, O2, O3) the first, second and third parameters (P ₁ , P ₂ , P ₃ ) to determine the relevance (R _O1 , R _O2 , R _O3 ) are linked. A method according to claim 4, characterized in that the environmental conditions are a position of the sun and / or a weather condition and / or a vegetation state. Method according to one of the preceding claims, characterized in that a fourth parameter (P ₄ ) is set up, dr comprises preferences for objects (O1, O2, O3), wherein for each object (O1, O2, O3) the first, second, third and fourth parameter vector (P ₁ , P ₂ , P ₃ , P ₄ ) or the first, second and fourth parameter vectors (P ₁ , P ₂ , P ₄ ) for determining the relevance (R _O1 , R _O2 , R _O3 ) linked become. Method according to one of the preceding claims, characterized in that the navigation device ( 2 ) is assigned to a camera, is checked using the images of the camera, as the selected object (O1, O2, O3) is recognizable. Method according to one of claims 4 to 7, characterized in that at least one environmental condition is sensed. Contraption ( 1 ) for selecting a navigational information object comprising a digital road map ( 3 ) and a navigation device ( 2 ), the digital road map ( 3 ) Exhibits objects (O1, O2, O3) which are detected by the navigation device ( 2 ) and can be output for navigation instructions, characterized in that the objects (O1, O2, O3) are assigned characteristics whose characteristic values are stored, whereby at least one feature expression is supervisory-dependent, wherein the navigation device ( 2 ) is designed such that it depends on a calculated route and an actual position of the navigation device ( 2 ) determines a driving maneuver and located in an environment objects (O1, O2, O3) from the digital road map ( 3 ), wherein for the selected objects (O1, O2, O3) in each case a first parameter vector (P ₁ ) is set up with the characteristic values, wherein at least one second parameter vector (P ₂ ) for each object (O1, O2, O3) is formed that at least the actual position of the navigation device ( 2 ) and the position of the object (O1, O2, O3), wherein in each case the first parameter vector (P ₁ ) and the second parameter vector (P ₂ ) of an object (O1, O2, O3) is linked, the link in each case having a relevance (R _O1 , R _O2 , R _O3 ) for the object (O1, O2, O3), the relevances (R _O1 , R _O2 , R _O3 ) of the objects (O1, O2, O3) are compared, the object ( O1, O2, O3) with the greatest relevance (R _O1 , R _O2 , R _O3 ) is selected by the navigation device ( 2 ) a navigation instruction including the object (O1, O2, O3) is generated. Apparatus according to claim 9, characterized in that at least one feature expression is time-dependent.

Images