阅读说明：本技术 对通过超声波识别的对象的高度进行分类的方法和装置 (Method and device for classifying the height of an object identified by ultrasound ) 是由 K·拜纽什卡 于 2021-05-13 设计创作，主要内容包括：本发明涉及一种用于对通过超声波在周围环境中识别的对象的高度进行分类的方法。其中,连续地发送超声波脉冲并且再接收超声回波。针对每个接收到的超声回波,将该超声回波配属给地图对象,或者当接收到的超声回波不能够配属给现有地图对象时,创建新地图对象,在创建新地图对象时,以起始值初始化高度指标。在另一方法步骤中,从接收到的超声回波和/或先前配属给地图对象的超声回波中提取至少一个属性。将高度指标改变一个点值。当高度指标的值超过预给定的极限值时,将由地图对象代表的对象分级为高的、不能驶过的对象。本发明还涉及一种用于对通过超声波在周围环境中识别的对象的高度进行分类的装置,所述装置设置为用于实施所述方法。(The invention relates to a method for classifying the height of an object identified in the surroundings by means of ultrasound. In which ultrasonic pulses are continuously transmitted and ultrasonic echoes are re-received. For each received ultrasound echo, the ultrasound echo is assigned to a map object, or a new map object is created if the received ultrasound echo cannot be assigned to an existing map object, and the height indicator is initialized with a starting value when the new map object is created. In a further method step, at least one property is extracted from the received ultrasound echoes and/or the ultrasound echoes previously associated with the map object. The altitude indicator is changed by one point value. When the value of the height indicator exceeds a predetermined limit value, the objects represented by the map objects are classified as high objects which cannot be driven through. The invention also relates to a device for classifying the height of an object identified in the surroundings by means of ultrasound, said device being designed to carry out the method.)

1. Method for classifying the height of an object (20) identified in the surroundings by means of ultrasound waves, wherein ultrasound pulses (30) are continuously transmitted and ultrasound echoes (32) reflected by the object (20) in the surroundings are received again, and wherein for each received ultrasound echo (32) at least the following steps are traversed:

-attaching the received ultrasound echoes (32) to map objects, wherein map objects represent objects (20) in the surroundings, or creating new map objects when the received ultrasound echoes (32) cannot be attached to existing map objects, wherein the height indicator attached to a map object is initialized with a starting value when creating the new map object,

-extracting at least one attribute from the received ultrasound echoes (32) and/or ultrasound echoes (32) previously assigned to the map object, wherein the at least one attribute gives an indication of the height of an object (20) represented by the map object,

-changing the height indicator by a point value, wherein the point value is recalled from an attachment table in dependence on the extracted at least one property and the distance of the object (20), and

-ranking the object (20) represented by the map object as a high, non-drivable object when the value of the height indicator exceeds a predefined limit value (42).

2. Method according to claim 1, characterized in that the assignment table is created from ultrasound echoes (32) received from objects (20) with a known true height classification, wherein a value of the at least one property is considered to be a value indicating a high object when it occurs more in the case of a high object than in the case of a low object, and conversely is considered to be a value indicating a low object when it occurs more in the case of a low object than in the case of a high object.

3. Method according to claim 2, characterized in that the entry of the attached table for the respective value of the at least one property for a determined distance has a positive point value when the determined value of the at least one property for the distance is a value indicating a high object and vice versa a negative point value when the determined value of the at least one property for the distance is a value indicating a low object.

4. Method according to claim 2 or 3, characterized in that the size of the point value of the assignment table for determining the distance from the determined value for the at least one property is determined from the quotient of the number of received ultrasonic echoes (32) for which the property value correctly indicates the height of the object (20) and the total number of received ultrasonic echoes (32) for this value.

5. Method according to claim 4, characterized in that a weighting function is taken into account when dimensioning the point values of the attachment table.

6. Method according to one of claims 1 to 5, characterized in that the assignment table is determined beforehand.

7. Method according to one of claims 1 to 6, characterized in that the assignment table is updated if the actual height classification of an object (20) has been determined using additional criteria and/or other measurement methods, wherein the ultrasound echoes (32) assigned to the map objects representing this object (20) are taken into account when updating using the now known actual object classification.

8. The method according to claim 7, characterized in that the behavior when approaching an object (20) is taken into account as an additional criterion for determining the true height classification of the object (20).

9. The method according to any one of claims 1 to 8, wherein the at least one attribute is selected from the group consisting of: the amplitude of the ultrasound echo (32), the significance of the ultrasound echo (32), the dependency on a further ultrasound echo (32), the time interval to the dependent ultrasound echo (32), the number of multiple reflections, the type of ultrasound echo (32), the pulse duration of the initially transmitted ultrasound pulse (30), and a combination of a plurality of these properties.

10. Device (10) for classifying the height of an object (20) identified in the surroundings by means of ultrasound waves, wherein the device (10) comprises at least one ultrasound sensor (12) for transmitting ultrasound pulses (30) and for receiving ultrasound echoes (32) reflected on the object (20) and comprises a controller (14), characterized in that the device (10) is provided for carrying out the method according to one of claims 1 to 9.

Technical Field

The invention relates to a method for classifying the height of an object identified in the surroundings by means of ultrasound, wherein ultrasound pulses are transmitted continuously and ultrasound echoes reflected by objects in the surroundings are received again. Another aspect of the invention relates to a device for classifying the height of an object identified in the surroundings by means of ultrasound, which device is provided for carrying out the method.

Background

Modern vehicles are equipped with a large number of driver assistance systems which assist the driver of the vehicle when performing different driving manoeuvres. Driver assistance systems are also known which warn the driver of hazards in the surroundings. In order to fulfill its function, driver assistance systems require precise data about the surroundings of the vehicle and in particular about objects located in the surroundings of the vehicle.

Ultrasound-based object localization methods are generally used, in which two or more ultrasound sensors are used. In this case, the ultrasonic sensors each emit an ultrasonic pulse and receive ultrasonic echoes reflected by objects in the surroundings. The distance between the reflecting object and the corresponding sensor is determined from the propagation time of the ultrasonic pulse until the reception of the respective ultrasonic echo and the known speed of sound. If the object is located in the field of view of more than one ultrasonic sensor, i.e. if the distance to the object can be determined by more than one ultrasonic sensor, the exact position of the reflecting object relative to the sensor or relative to the vehicle can also be determined by means of a least squares algorithm.

As the field of view and the sensitivity of sensors become larger and larger, objects on the ground, such as kerbs, sleepers or manhole covers, can increasingly also be recognized. In this case, it is important for the correct functioning of the driver assistance system to be able to distinguish between objects which are relevant to a collision (for example pillars, walls or traffic signs) and objects which are irrelevant to a collision and can be driven past (for example kerbs, sleepers or manhole covers).

DE 102015104940 a1 describes a method for providing height information of objects in the surrounding area of a motor vehicle. It is proposed that an object is assigned a first height level "high", a second height level "low" and a third height level "non-divisible object". In addition, a probability of correctness is assigned to the classification. Preferably, when the object is classified with a first probability in one height level and with a second probability in another height level, it is decided according to the two probabilities into which height level the object is finally classified.

DE 112013004908T 5 describes an object sensing device which is arranged on a mobile body and which transmits and receives waves, for example in the form of ultrasonic waves, again. A spacing relative to the reflecting object is calculated based on the received waves and a height determining variable, which may be a weighted average, is determined.

DE 102017202964 a1 describes a method for providing ultrasonic signal information, in which echo signals are received and homologies of the individual echo signals are determined on the basis of their propagation time and/or signal similarity with respect to the initially transmitted ultrasonic signal. The echo signals of the same category are combined into echo groups. In addition, it can be provided that the following objects are classified on the basis of the properties characterizing the echo group: the echo signal has been reflected on the object. For example, it is possible to distinguish between planar obstacles (e.g. vehicles) and structured obstacles (e.g. bushes).

A disadvantage of the known methods for classifying objects is that the usual criteria for distinguishing between high collision-relevant objects and low objects that can be driven over generally only work reliably over short distances of less than 1.5m to 2 m. Therefore, there is a need for a method for classifying the height of an object, which enables reliable classification, in particular even at large distances. It is also desirable that the method is adaptable to various environmental conditions.

Disclosure of Invention

A method for classifying the height of an object recognized in the surroundings by means of ultrasound is proposed. In the method, ultrasonic pulses are continuously transmitted and ultrasonic echoes reflected by objects in the surroundings are received again. Furthermore, it is proposed that for each ultrasound echo a received ultrasound echo is assigned to a map object, wherein the map object represents an object in the surroundings, or that a new map object is created if the received ultrasound echo cannot be assigned to an existing map object, wherein, when the new map object is created, the height indicator assigned to the map object is initialized with a start value. In a further method step, at least one property is extracted from the received ultrasound echoes and/or the ultrasound echoes previously associated with the map object, wherein the at least one property gives an indication of the height of the object represented by the map object. Furthermore, the height indicator is changed by a point value, wherein the point value is called from an attachment table (look-up table LUT) depending on the extracted at least one property and the distance of the object. In this case, the distance of the object represented by the map object is preferably determined from the propagation time of the ultrasound echo assigned to the map object. Subsequently, when the value of the height indicator exceeds a predefined limit value, the objects represented by the map objects are classified as high, non-drivable objects.

The limit value is preferably dependent on the distance and can be fixedly predefined or dynamically selected. Preferably, the distance-related limit value is fixedly predefined.

The method is used, for example, in connection with an environment sensor of a vehicle, wherein the vehicle preferably comprises at least two ultrasonic sensors, with which ultrasonic pulses can be transmitted and ultrasonic echoes can be received in each case.

Preferably, for each received ultrasound echo, at least the following steps are carried out: assigning to a map object, extracting attributes, changing a height indicator, and ranking, wherein the steps can be performed in this order. However, other sequences are also conceivable. For example, the properties of the ultrasound echo can already be extracted before the assignment to the map object.

An ultrasonic pulse is transmitted by the ultrasonic sensor and an ultrasonic echo of the ultrasonic pulse is received again. The ultrasonic echo can be received by the ultrasonic sensor that transmitted the initial ultrasonic pulse. In this case, the received ultrasonic echo is referred to as a direct echo. If an ultrasonic echo is received by another ultrasonic sensor, it is called a cross echo.

The time elapsed between the transmission of an ultrasonic pulse and the reception of an ultrasonic echo and the known speed of sound in air make it possible to determine the distance between the corresponding ultrasonic sensor and the object which is reflecting. If ultrasonic echoes are received by a plurality of ultrasonic sensors, in addition to the distance, the relative position of the reflecting object relative to the sensors can be determined using a least squares algorithm. Such methods are known in principle to the person skilled in the art and are described, for example, in DE 102017202964 a 1.

The received ultrasound echoes are each assigned to a map object. The map object represents the actual object in the surroundings and can be stored, for example, in a memory of the controller. The map object is associated with at least one position of the object, which is specified, for example, with respect to the ultrasonic sensor or with respect to a vehicle which uses the ultrasonic sensor to monitor the surroundings. In addition, the map object can be assigned further characteristics, which in each case represent specific characteristics of the represented object. In particular, a height indicator is associated with the map object, which height indicator is used to classify the height of the object.

In order to assign the received ultrasound echoes to map objects, it can be checked, for example, whether the determined position of the reflecting object lies in the vicinity of the position of an existing map object. If this is the case, the ultrasound echo is assigned to the map object. Otherwise, a new map object is created and the ultrasound echo is assigned to the new map object.

From the associated ultrasound echo and the ultrasound echo which may have been previously associated with the map object, an attribute is extracted which gives an indication of the height of the object.

Preferably, the at least one attribute is selected from: the amplitude of the ultrasonic echo, the significance of the ultrasonic echo, the dependency on further ultrasonic echoes, the time interval with dependent ultrasonic echoes, the number of multiple reflections, the type of ultrasonic echo, the pulse duration of the initially transmitted ultrasonic pulse and a combination of a plurality of these properties. In particular, it can be provided that a plurality of different combinations of these properties are used.

The significance of an ultrasound echo is a value which describes the similarity between the ultrasound echo and the associated initially transmitted ultrasound wave pulse or with the expected ultrasound echo. The significance is in particular the value determined by means of an optimization filter or a matched filter. The significance describes, in particular, the signal similarity of the ultrasound echo to the ultrasound pulse or the signal similarity of the ultrasound echo to the expected ultrasound echo.

The two received echoes can for example be considered as belonging to one another when the propagation time difference between the propagation times of two temporally successive ultrasonic echoes is smaller than a determined time value. In other words, this means that two individual ultrasound echoes can be evaluated as belonging to the same category when they are received within a predefined time interval. This usually involves multiple reflections, in which case a single ultrasound pulse is reflected multiple times by an object and thus multiple ultrasound echoes are received from the object.

In terms of the type of ultrasound echo, a distinction is made, for example, between a direct echo, which is received by the ultrasound sensor that initially sent the ultrasound pulse, and a cross echo, in which case the ultrasound echo is received by another ultrasound sensor.

A plurality of such properties can be combined to form a composite property, if necessary using a predetermined weighting factor.

In the method, a distance to the ultrasonic sensor or to the vehicle is assigned to the map object, and the distance is determined as a function of a determined signal propagation time of the assigned ultrasonic echo. The distance changes over time, so that it can be provided here that only the last assigned ultrasound echo or a predetermined number of previously assigned ultrasound echoes in addition to the last assigned ultrasound echo are taken into account.

For the classification of objects, a subdivision into high, collision-related objects and low, objects that can be driven through is proposed. Tall, impact-related objects are, for example, pillars, walls or traffic signs. Low, drivable objects are, for example, kerbs, sleepers or manhole covers. For the classification, the corresponding map object is assigned a height indicator, which is initialized to a starting value, for example 0, when the map object is created.

For each object received and assigned to a map object, the height indicator is changed by a specific point value, which is obtained from an assignment table (LUT). For example, the assignment table is stored in a memory of the controller. In this case, the height indicator increases when the associated table gives a positive point value, and decreases when the associated table gives a negative point value. In this case, the point value of the entry of the assignment table is dependent, in particular, on the determined object distance and on the extracted at least one property. If only unique attributes are used, the attachment table is two-dimensional. The number of dimensions of the assignment table increases correspondingly with the number of attributes used. For example, the attachment table can associate three attributes along with a distance with a saved point value. Here, the three attributes are preferably selected from: the length of the transmitted ultrasonic pulse, the type of ultrasonic echo and the high degree of significance. The attribute "high significance" is formed by a combination of attributes, wherein the dependency on further ultrasound echoes, the time interval between dependent ultrasound echoes and the number of multiple reflections are preferably included in the high significance. By forming such additional compound attributes, the number of dimensions in the attachment table can be reduced.

Preferably, the method is implemented in connection with sensing of the surroundings of the vehicle. In this case, it is preferably provided that the method is carried out only while the vehicle is moving, so that the height indicator is not increased accordingly when the vehicle is just stationary.

In the case of continuous variables, for example distances, it is preferable to perform stepping (abstufang) with a predetermined resolution. For example, a resolution of 200mm can be predefined for distances, such that the assignment table comprises one entry per 200mm, wherein a first entry is used for distances greater than or equal to 0mm and less than 200mm, a second entry is used for distances greater than or equal to 200mm and less than 400mm, and so on.

After the change of the height indicator, the height indicator is compared with a predefined limit value, which is preferably predefined as a function of the distance. If the height indicator is above the limit value, the object is classified as a tall, non-drivable object. If the height indicator has a value below a limit value, this either involves low objects that can be driven through, or reliable classification is not yet possible.

Preferably, the assignment table is created from ultrasound echoes received from objects with a known true height classification, wherein values of the at least one property are considered to indicate values of high objects when they occur more in the case of high objects than in the case of low objects, and conversely, values of the at least one property are considered to indicate values of low objects when they occur more in the case of low objects than in the case of high objects. The at least one attribute is preferably a composite attribute, such as the high significance value explained above. The attributes can have discrete values, i.e. only certain values, or can be continuous, the assignment table then containing entries each representing a certain value range at a predetermined resolution.

The assignment table is the result of a statistical evaluation in the form of a model that can be easily used. In this case, it is possible to pre-incorporate a large number of ultrasound echoes into the assignment table, wherein the complexity of the assignment table and thus the complexity for carrying out the method are not increased even in the case of large data volumes used for this purpose.

Preferably, the entry of the attachment table for the respective value of the distance for the at least one property has a positive point value when the determined value of the at least one property for the determined distance is a value indicating a high object and conversely has a negative point value when the determined value of the at least one property for the distance is a value indicating a low object. If the determined value of the at least one property for the determined distance cannot be concluded, the corresponding entry of the assignment table is preferably set to "0".

Preferably, the size of the point values of the assignment table for determining the value of the distance determined for the at least one property is determined from the quotient of the number of received ultrasound echoes for which the property value correctly indicates the height of the object and the total number of received ultrasound echoes for that value. The quotient therefore indicates in which case the respective value of the attribute in the assignment table indicates the real classification.

Preferably, the weighting function is taken into account when determining the size of the point values of the attachment table. For example, a step function is used as the weighting function, wherein for the case where the quotient of the number of received ultrasonic echoes, those for which the attribute value correctly indicates the height of the object, and the total number of received ultrasonic echoes for that value is 0.95 or more, the point value is selected to be of size 5; selecting a point value size of 1 when the quotient is in a range of greater than or equal to 0.85 and less than 0.95; in other cases, the point value size is selected to be 0. In this way, values that correctly indicate a true classification with a high probability are taken into account intensively and values that indicate a true classification with only a low probability are taken into account only or only marginally.

Preferably, the assignment table is determined beforehand.

Preferably, the assignment table is updated if the actual height classification of the object has been determined using additional criteria and/or other measurement methods, wherein the ultrasound echoes assigned to the map objects representing the object are taken into account using the actual object classification known at the time of the update. In practice, the received ultrasound echoes can be incorporated in this way into the statistical data for creating the assignment table.

Preferably, the behavior when approaching an object is considered as an additional criterion for determining the true height classification of the object. For example, the following are used: when approaching a low, drivable object, this object can no longer be sensed by the ultrasonic sensor at less than the minimum distance, whereas a high, collision-related object remains visible to the ultrasonic sensor when approaching and can continue to be sensed. The minimum distance is for example 800 mm. If the object appears to disappear at less than the minimum distance, it must be a low, drivable object, and conversely, if the object continues to be sensed by the ultrasonic sensor, it must be a high, collision-related object. Alternatively or additionally, for example, changes in the number of received multiple reflections of the object and in the case of multiple reflections the distance to be determined can also be taken into account as criteria.

Another aspect of the present invention is to provide an apparatus for classifying a height of an object recognized in a surrounding environment by ultrasonic waves. The apparatus comprises at least one ultrasonic sensor for transmitting ultrasonic pulses and for receiving ultrasonic echoes reflected on the object and a controller.

The apparatus is constructed and/or arranged to carry out the method described herein. Correspondingly, the features described in the context of the method apply correspondingly to the device, and conversely the features described in the context of the system apply correspondingly to the method.

Preferably, the device comprises two or more ultrasonic sensors, the fields of view of which at least partially overlap. Here, the field of view indicates the following regions: within the region, the ultrasonic sensor is capable of detecting the object by receiving ultrasonic echoes.

Preferably, the device is part of a driver assistance system which is provided for assisting a driver of the vehicle and/or for noticing the driver of a hazard in the surroundings of the vehicle when carrying out a driving maneuver.

THE ADVANTAGES OF THE PRESENT INVENTION

The method according to the invention makes it possible to use statistical models, which are obtained by evaluating large data volumes in an easy manner and with low complexity, for reliably identifying tall, non-drivable objects using an assignment table. In this case, high objects that cannot be driven through can already be reliably recognized at distances in the range of 3 to 4 meters, whereas the typical classification methods can only achieve reliable recognition at distances close to approximately 1.5 to 2 meters.

In an advantageous embodiment, the method can also be adapted to the conditions occurring in practice by updating the assignment table after recognition of the real object classification using further criteria or sensors. This makes it possible, for example, for conditions not previously sensed in laboratory tests to also be included in the statistical model used and thus in the assignment table. Thus, the method is learning-capable.

In addition, those attributes of the ultrasound echo that allow for a high degree of classification can be easily identified and used.

Drawings

Embodiments of the invention are explained in more detail with reference to the figures and the following description.

The figures show:

figure 1 shows a vehicle with a device for classifying the height of objects recognized in the surroundings by means of ultrasound,

figure 2 is an exemplary excerpt from the attachment table,

figure 3a shows the trend of the height indicator for a low object when approaching,

figure 3b shows the trend of the height indicator for a tall object when approaching,

FIG. 4a the number of received echoes of an object that is low in proximity, an

Figure 4b number of received echoes of a high object at approach.

Detailed Description

In the following description of the embodiments of the invention, identical or similar elements are denoted by identical reference numerals, wherein a repeated description of these elements in individual cases is omitted. The figures only schematically show the content of the present aspect.

Fig. 1 shows a vehicle 1, which comprises a device 10, by means of which an object 20 in the surroundings of the vehicle 1 can be identified. In addition, the device 10 is provided for classifying the height of the object 20 into a high, collision-relevant object and a low, drivable-through object.

To this end, the device 10 comprises an ultrasonic sensor 12, wherein, in the example of fig. 1, a first ultrasonic sensor 121 and a second ultrasonic sensor 122 are arranged at the front of the vehicle 1. In addition, the device 10 includes a controller 14 that operates the ultrasonic sensor 12.

An ultrasound pulse 30 is transmitted by the ultrasound sensor 12 and an ultrasound echo 32 reflected at the object 20 is received again. In the situation shown in fig. 1, for this purpose the first ultrasonic sensor 121 transmits an ultrasonic pulse 30, which is reflected by the object 20. Here, the first ultrasonic sensor 121 receives the direct echo 321 and the second ultrasonic sensor 122 receives the cross echo 322. The distance from the ultrasonic sensor 12 to the object 20 can be determined from the propagation time from the transmission of an ultrasonic pulse 30 to the reception of the corresponding ultrasonic echo 32. Using these distances and the known arrangement of the ultrasonic sensors 12 relative to one another, the relative position of the object 20 relative to the vehicle 1 can also be determined by the least squares method.

In order to classify the height of the object 20 identified by the received ultrasound echoes 32, it is provided that, in a first step, for each ultrasound echo 32 a map object is assigned to this received ultrasound echo 32, wherein this map object represents the object 20. Such a map does not yet exist at the first reception of the ultrasound echo 32, so that the map object is newly created.

The map object is assigned a height indicator H, which is initialized with a starting value when the map object is created, see fig. 3a and 3 b. For example, a value of 0 is used as the start value.

The received ultrasound echoes 32 are further analyzed, wherein at least one property is extracted from the received ultrasound echoes 32 and/or the ultrasound echoes 32 previously assigned to the map object. The at least one property is selected in such a way that it gives an indication of the height of the object 20 represented by the map object.

Furthermore, the height indicator H is changed by a point value, which is called from an assignment table (LUT, lookup table) depending on the extracted at least one property and the distance of the object 20. In this case, the distance of the object 20 represented by the map object is preferably determined from the propagation time of the ultrasound echo 32 assigned to the map object.

Next, the object 20 represented by the map object is ranked as a high collision-relevant object if the value of the height indicator H exceeds a predefined distance-related limit value 42, see fig. 3a and 3 b. If the limit value 42 is not exceeded, then no reliable classification can be carried out or a low, drivable object is involved. In this case, the method is preferably carried out continuously, so that, in particular during the approach of the vehicle 1 to the object 20, the ultrasonic pulses 30 are continuously transmitted, the ultrasonic echoes 32 are received, and the height indicator H is updated.

Fig. 2 shows an exemplary excerpt from the assignment table, in which the distance of the object 20 from the vehicle 1 is plotted in steps of 200mm in the horizontal direction and the discrete value a of the composite property is plotted in the vertical direction. The composite property is formed, for example, by a combination of properties, wherein the dependency on the further ultrasound echo 32, the time interval of the dependent ultrasound echo 32 and the number of multiple reflections are preferably included in the composite property. By forming such a composite attribute, the number of dimensions in the attachment table can be reduced. In the example of fig. 2, the illustrated excerpt of the assignment table has two dimensions.

For a determined distance and for a determined value of the composite property, an entry of the attachment table has a positive point value when the corresponding value of the property for the distance is a value indicating a high object, whereas an entry of the attachment table has a negative point value when the corresponding value of the property for the distance is a value indicating a low object. If the determined value of the attribute for the determined distance cannot be concluded, the corresponding entry of the assignment table is preferably set to "0".

The size of the point values of the assignment table for determining the distance of the attribute values is determined by the quotient of the number of those ultrasonic echoes 32 received for which the attribute values correctly indicate the height of the object 20 and the total number of ultrasonic echoes 32 received for which the attribute values. The quotient thus indicates in which case the respective attribute value in the assignment table indicates the real classification.

If an ultrasound echo 32 is now received, for example in the situation shown in fig. 1, it is assigned to a map object corresponding to the object 20. Furthermore, the distance to the vehicle 1 is determined and the composite property is determined. If the distance is, for example, 2000mm and a value of 4 is determined as the value of the composite attribute, the height indicator H of the map object is increased by a value of 0.88. In a preferred embodiment, a weighting function is used, wherein high quotient values are taken into account intensively and low quotient values are not taken into account or are taken into account only less intensively when the height indicator H changes.

Fig. 3a shows a plurality of curves 40, which describe the changing course of the height indicator H of a low object when the vehicle 1 approaches the object 20. Fig. 3b shows a plurality of curves 44 in a similar manner, which show the changing course of the height indicator H of a tall object when the vehicle 1 approaches the object 20.

Here, the distance d of the object 20 from the vehicle 1 is plotted on the X axis and the value of the height index H is plotted on the Y axis, respectively.

As can be seen from the diagrams in fig. 3a and 3b, the height indicator H increases both when the vehicle 1 approaches a low, drivable-through object and when the vehicle approaches a high, collision-related object. However, in the case of an approaching low object, all curves 40 of fig. 3a remain below the distance-dependent limit value 42, whereas in the case of an approaching high object, as shown in fig. 3b, after a distance of less than about 2000mm to 3000mm, almost all of the curves 44 lie above the distance-dependent limit value 42. At distances of less than approximately 600mm, all of the course in the curve 44 lies above the limit value 42.

As can be seen from the diagrams in fig. 3a and 3b, a low, drivable object is not erroneously classified as a high, collision-relevant object. This method therefore does not lead in particular to false obstacle warnings. It can also be seen that, with large distances of approximately 3000mm, high objects can already be classified reliably as high objects in most cases, and only with a few observations can the classification be carried out upon further approach. It can also be seen that objects that were once ranked as "high and relevant to collision" are no longer reclassified in the process of approaching. The proposed method therefore enables in most cases the early and reliable identification of high collision-related objects.

Fig. 4a shows the number # of ultrasonic echoes 32 received from a low object in the proximity, while fig. 4b correspondingly shows the number # of echoes received from a high object when the vehicle 1 is in the proximity of the object 20. Here, the distance d between the vehicle 1 and the object 20 is plotted on the X axis, and the number # of received ultrasonic echoes 32 is plotted on the Y axis, respectively.

As can be seen from fig. 4a and 4b, below the minimum distance 46 (approximately 800mm in the example of fig. 4a and 4 b), low objects that can pass by are no longer detectable by the ultrasonic sensor 12 and therefore the number # of received ultrasonic echoes 32 does not increase further. In contrast, in the case of a tall, collision-related object, the number # of ultrasound echoes 32 continues to increase even at less than the minimum spacing 46. This characteristic can be used as a further criterion in order to determine the true height classification of the object 20. This real height classification can then be used to update the assignment table see fig. 2. For updating, the individual entries of the assignment table are recalculated, wherein the ultrasound echoes 32 assigned to the objects 20 with the now known real classification or to the respective map objects are taken into account.

The present invention is not limited to the embodiments described herein and the aspects emphasized therein. Rather, many modifications are possible within the scope of the measures conventional to a person skilled in the art within the scope given by the claims.