DE102016105022A1 - Method for detecting at least one object in an environment of a motor vehicle by an indirect measurement with sensors, control device, driver assistance system and motor vehicle - Google Patents

Abstract

The invention relates to a method for detecting at least one object (3) in an environment (4) of a motor vehicle (1), in which a first sensor (6) is driven to emit a sensor signal, sensor data are received by a second sensor (6) , which describe the sensor signal reflected by the at least one object (3), from the sensor data a blurred feature (xU) is determined as object feature (xP, xL, xU) for describing the at least one object (3), the fuzzy feature ( xU) describes a distance between the at least one object (3) and a position (xS1) of the first sensor (6) and / or a position (xS1) of the second sensor (6), the fuzzy feature (xU) being an ellipse ( xE), wherein focal points (xF1, xF1) of the ellipse (xU) are determined on the basis of the position (xS1) of the first sensor (6) and the position (xS2) of the second sensor (6), a length of a large semiaxis ( a) the ellipse (xE) based on a Lau Time of the sensor signal is determined and a length of a small half-axis (b) of the ellipse (xE) based on the focal points (xF1, xF1) and the length of the semi-major axis (a) is determined.

Description

The present invention relates to a method for detecting at least one object in an environment of a motor vehicle, in which a first sensor is driven to emit a sensor signal, sensor data is received from a second sensor, which describe the sensor signal reflected by the at least one object Sensor data is a blurred feature as an object feature for describing the at least one object is determined, wherein the fuzzy feature describes a distance between the at least one object and a position of the first sensor and / or a position of the second sensor. In addition, the invention relates to a control device and a driver assistance system for a motor vehicle. Finally, the invention relates to a motor vehicle.

Different methods are known from the prior art with which objects in the environment of a motor vehicle can be detected. For example, the objects can be detected with the aid of corresponding sensors, for example ultrasound sensors, radar sensors, laser sensors or cameras. Furthermore, it is known to use the sensor signals of the sensors to determine a distance between the motor vehicle and the object or a relative position between the motor vehicle and the object.

In particular in connection with ultrasonic sensors, it is known that an object can be determined by means of an indirect measurement. For this purpose, a first ultrasonic sensor emits an ultrasonic signal. This ultrasonic signal is then reflected at the object and received by a second ultrasonic sensor. Based on the transit time between the emission of the ultrasound signal from the first sensor and the reception of the ultrasound signal reflected by the object with the second sensor, a distance of the object to the first sensor and / or the second sensor can then be determined. For example, it can be simplified to assume that the object is located on a circle, wherein the center of the circle is arranged centrally between the two sensors or the positions of the two sensors. An angular information regarding the object can not be derived from such an indirect measurement.

Moreover, it is known in the art to fuse the results of multiple or different sensors to more reliably detect objects. For this purpose, it can be provided, for example, to determine from the sensor data provided by the respective sensor, in each case an object feature which describes the object. Such an object feature may be, for example, a point feature or a line feature that describes the object or a part thereof. For the object feature also a spatial uncertainty or blur can be determined. The object features can also be entered in a digital environment map, which describes the environment of the motor vehicle. To merge the measurements of the various sensors, the object features can be merged.

It is an object of the present invention to provide a solution, as objects in an environment of a motor vehicle, in particular in an indirect measurement with sensors of the motor vehicle can be detected easier and more reliable.

This object is achieved by a method by a control device, by a driver assistance system and by a motor vehicle with the features according to the respective independent claims. Advantageous developments of the present invention are the subject of the dependent claims.

In one embodiment of a method for detecting at least one object in an environment of a motor vehicle, a first sensor for emitting a sensor signal is preferably activated. Furthermore, sensor data are received, in particular from a second sensor, which describe the sensor signal reflected by the at least one object. Furthermore, it is preferably provided that from the sensor data, a blurred feature is determined as an object feature for describing the at least one object, wherein the fuzzy feature describes a distance between the at least one object and a position of the first sensor and / or a position of the second sensor , In particular, the fuzzy feature is described as an ellipse, wherein focal points of the ellipse are determined based on the position of the first sensor and the second sensor. Furthermore, a large semiaxis of the ellipse is preferably determined on the basis of a transit time of the sensor signal, and a length of a small semiaxis of the ellipse is determined in particular on the basis of the focal points and the length of the large semiaxis.

An inventive method is used to detect at least one object in an environment of a motor vehicle. In the method, a first sensor for emitting a sensor signal is driven. In addition, sensor data are received from a second sensor, which describe the sensor signal reflected by the at least one object. From the sensor data, a blurred feature is determined as an object feature for describing the at least one object, wherein the fuzzy feature describes a distance between the at least one object and a position of the first sensor and / or a position of the second sensor. In this case, the fuzzy feature is described as an ellipse, wherein focal points of the ellipse are determined by the position of the first sensor and the second sensor, a length of a large half-axis of the ellipse is determined based on a transit time of the sensor signal and a length of a small half-axis of the ellipse based on the Focal points and the length of the semi-major axis is determined.

With the help of the method, one or more objects in the environment of the motor vehicle should be detected or detected. The method can be carried out, for example, with a control device or an electronic control unit of the motor vehicle. With the control device, a first sensor of the motor vehicle can be controlled, so that the first sensor emits a sensor signal as a result of the control. This emitted sensor signal then hits the at least one object and is reflected by this. The reflected sensor signal or the echo of the sensor signal is then received by the second sensor. The first and / or the second sensor may be, for example, ultrasonic sensors. When the second sensor has received the reflected ultrasonic signal, it provides sensor data which is transmitted to the controller. These sensor data may include, for example, a plurality of sensor values or measured values which describe the reflected sensor signal. With the aid of the control device, a transit time can now be determined on the basis of the transit time between the emission of the sensor signal with the first sensor and the reception of the reflected sensor signal with the second sensor. On the basis of this transit time can now be determined how far the object from the first sensor and / or the second sensor is removed. Furthermore, it is provided that a blurred feature or a blurred object feature is determined by means of the control device, which serves to describe the object.

Such a blurred feature represents an object feature for describing the object. The fuzzy feature results, for example, from the fact that based on the sensor data or the transit time of the sensor signal only distance information relating to the object can be derived. No angle information regarding the object can be derived from the sensor data. Thus, the relative position between the sensors and the object can not be determined accurately. The fuzzy feature can be entered, for example, in a digital environment map, which describes the environment of the motor vehicle. In this case, the fuzzy feature can be entered with a spatial uncertainty or a blur in the map. In this environment map also other object features can be entered. These object features may have been determined, for example, from measurements with other sensors of the motor vehicle. It can also be provided that these further object features were determined with the first and / or the second sensor.

According to the invention, it is now provided that the fuzzy feature is described as an ellipse, wherein focal points of the ellipse are determined based on the position of the first sensor and the second sensor. The positions of the first sensor and the second sensor are known. These may for example be stored in a memory of the controller. The ellipse is now determined so that the respective positions of the first sensor and the second sensor are assigned to the two focal points of the ellipse. Furthermore, the length of the semi-major axis of the ellipse is determined on the basis of the transit time of the sensor signal. Here, it is assumed that the object from which the sensor signal was reflected is on the ellipse. The transit time of the sensor signal describes the path from a first focus to the point on the ellipse describing the object plus the path from that point to a second focus. The distance that can be assigned to the term thus corresponds to twice the large semiaxis or the length of the major axis of the ellipse. From the position of the foci and the length of the major half-axis, the length of the semi-minor axis can be derived. The ellipse can thus be determined in a simple manner on the basis of the known positions of the sensors and the transit time, which is determined from the sensor signal. Characterized in that the fuzzy feature is described in the indirect measurement with the sensors as an ellipse and not as a circle, a more accurate information about the position of the object can be provided. Overall, this allows a more accurate and reliable detection of the at least one object.

Preferably, an eccentricity of the ellipse is determined on the basis of a distance of the focal points, and the length of the semi-minor axis is determined from the length of the major semiaxis and the eccentricity. Since the respective positions or installation positions of the two sensors are known, the distance between the first sensor and the second sensor can also be determined. This distance can also be referred to as eccentricity of the ellipse. From this eccentricity and the length of the semi-major axis of the ellipse then the length of the semi-minor axis of the ellipse can be determined. The length of the semi-minor axis of the ellipse is the square root of the length of the semimajor axis squared minus the eccentricity squared. Thus, the length of the semi-minor axis of the ellipse, which describes the fuzzy feature, can be determined easily and within a short computing time.

In a further embodiment, based on sensor data of a further measurement of the first and / or the second sensor, a further object feature is determined, which describes the at least one object, and the fuzzy feature is merged with the object feature. Such an object feature can be, for example, a point or a point feature which describes a position of the object or a part thereof in space. The object feature may also be a line or a line feature that describes a boundary of an object or a part thereof. Such object features are now to be merged with the fuzzy feature or ellipse to further characterize the object in the environment. It may be the case that the object feature is already entered in the area map. Following this, the fuzzy feature can then be determined. Now it is possible to check first whether the object feature and the fuzzy feature describe the same object in the environment. It can therefore be checked whether the fuzzy characteristic can be assigned to the object feature. If an assignment is made, it can be checked whether the fuzzy feature can be merged with the object feature. In this merging, which can also be referred to as merging, based on the object feature, a merged or updated object feature can be determined. This merged object feature is determined based on the fuzzy object feature. This allows the object to be detected more reliably.

In a further embodiment, the further object feature and the fuzzy feature are converted into a common state space, in the state space an innovation function is defined which describes a similarity between the object feature and the fuzzy feature and the merged object feature is determined from the object feature based on the innovation function. The further object feature and the fuzzy feature can be converted or transformed into a common state space. This state space can represent, for example, a common coordinate system for the further object feature and the fuzzy feature. In particular, the state space may be different from the environment map. In the state space, a state of the further object feature and a state of the fuzzy feature may be described. In this state space, the innovation function, which may represent a similarity measure for determining a similarity between the further object feature and the fuzzy feature, is also determined. In particular, the innovation function may describe how similar the state of the further object feature is to the state of the fuzzy feature. With the aid of the innovation function, the respective states of the further object feature and the fuzzy feature can be represented in such a way that only the states and / or dimensions are taken into account, which are also of interest for determining the similarity. For example, the innovation function may describe a distance between the further object feature and the fuzzy feature or a difference in the orientation of the further object feature and the fuzzy feature. On the basis of the innovation function, it can first be determined whether the fuzzy feature can be assigned to the further object feature. If the assignment has taken place, the innovation function can also be used to derive the merged object feature from the further object feature. Thus, the innovation function can be determined with little computational effort and yet reliable.

In one embodiment, to determine the merged object feature from the further object feature, a linear mapping rule is determined that includes the innovation function and an amplification factor for the innovation function, where the amplification factor depends on spatial uncertainty of the further object feature and spatial uncertainty of the fuzzy feature. In order to determine the merged object feature, the linear mapping rule is determined. This mapping rule can also be referred to as a filter. In principle, the mapping rule can be defined in the manner of a Kalman filter. In order to determine the merged object feature, the mapping rule can be added to the further object feature. The mapping rule may include the innovation function and the amplification factor with which the innovation function is multiplied. This gain factor may also be referred to as Kalman gain. Of the Amplification factor can be determined as a function of the respective spatial uncertainty of the further object feature and the fuzzy feature. The spatial uncertainty can be determined on the basis of the respective expectations or expected values of the states or the characteristics. The spatial uncertainty can also be determined with the aid of the covariance or the covariance matrix of the respective states or features. Thus, the merged object feature can be determined taking into account the spatial uncertainty.

Furthermore, it is advantageous if the amplification factor is determined such that a spatial uncertainty of the merged object feature is minimal. In order to determine the merged object feature as reliably as possible, the amplification factor can be determined such that the spatial uncertainty of the merged object feature is minimized. The amplification factor can thus be determined so that the covariance or the covariance matrix of the merged object feature is minimal. The covariance matrix can be minimized, for example, by minimizing the eigenvalues of the covariance matrix. Also, unit vectors of the covariance matrix can be determined and minimized. Thus, the gain can be reliably determined.

It can also be provided that a limit is determined on the basis of the innovation function and the fuzzy feature is disregarded when determining the merged object feature if the innovation function exceeds the limit. In order to decide whether the fuzzy feature and the object feature should be merged, the innovation function can be used. In particular, the expectation of the innovation function and / or the covariance of the innovation function can be determined. From this a boundary, which is also called Mahalanobis distance, can be derived. If the innovation function is smaller than this limit, the merged object feature can be determined using this innovation function. In the event that the innovation function is greater than this limit, the fuzzy feature can not be taken into account.

Preferably, a point feature or a line feature is determined as the further object feature. Such a point feature can be determined, for example, by detecting the object with the first sensor and / or the second sensor at a plurality of temporally successive time points or in a plurality of temporally successive measurements. For example, by means of triangulation, the relative position between the first and / or the second sensor can be determined to the object. The further object feature may also be a line or a line feature which describes a boundary of an object or a part thereof. Such a line feature is obtained, for example, when the motor vehicle is moved past the object and continuous measurements are made during the passing. The respective measured values or sensor data, which describe the distance to the object, can then be combined into a line feature. It can also be provided that a plurality of measured values is determined and a line is laid through the measured values in order to obtain the line feature.

Furthermore, it is advantageous if, for determining the innovation function, a measurement model is determined which describes a distance between the position of the first sensor or the position of the second sensor on the one hand and the point feature or the line feature on the other hand. The measuring model can be determined, for example, in a measuring room. If the additional object feature is a point feature, the measurement model can describe the distance between one of the sensors and the point feature. If the further object feature is the line feature, the measurement model can describe the distance between one of the sensors and the line feature. For this purpose, for example, a perpendicular to the line feature can be determined, which runs through the position of the first sensor or the position of the second sensor. The measurement model is determined in particular in a local coordinate system of the ellipse. This allows the simple determination of the respective measurement model.

If the further object feature is the point feature, for determining the measurement model, in particular the measurement ellipse is determined in such a way that the point feature lies on the measurement ellipse. Thus, a measurement ellipse for determining the measurement model is determined. This measuring ellipse is determined so that the point feature lies on the measuring ellipse. Thus, it is checked how a blurred feature would look like an ellipse when measuring the point feature as the object. The measurement model describes the distance between one of the sensor positions or focal points to the point feature. This distance corresponds to the large semiaxis of the scale ellipse. This allows a simple determination of the measurement model in the event that the object feature is the point feature.

If the further object feature is the line feature, in particular a measuring ellipse is determined for determining the measuring model such that the line feature runs tangentially to the ellipse. Basically, to determine the measurement model, the measurement ellipse could be determined such that it intersects with the line feature. But this results in two points of intersection, whereby no clear solution can be determined. Therefore, the Messellipse is determined so that it has only a single intersection with the line feature. This is the case when the line feature is tangent to the measurement ellipse. Thus, the measurement model for the case that the object feature is the line feature, can be determined easily.

The innovation function is preferably determined on the basis of a difference between the length of the large semiaxis of the ellipse and the measurement model. As already explained, the measurement model describes the distance between one of the focal points and the point feature or the line feature. From this distance, the length of the major half-axis of the Messellipse can be determined. This can be compared with the length of the semi-major axis of the ellipse, which was determined as a blurred feature. Thus it can be determined how the ellipse differs as a blurred feature or the measured ellipse from the Messellipse. Thus, the similarity between the fuzzy feature or the ellipse and the object feature, the point feature or the line feature, can be determined in a simple manner.

A control device according to the invention for a driver assistance system of a motor vehicle is designed to carry out a method according to the invention. The controller may include a computer, a microprocessor, a digital signal processor, or the like. Preferably, the control device is formed by an electronic control unit of the motor vehicle.

An inventive driver assistance system for a motor vehicle comprises a control device according to the invention as well as a first sensor and a second sensor, which are in particular ultrasonic sensors. The control device is connected to the first and the second sensor for data transmission. The first sensor serves as a transmitter for transmitting the sensor signal. The second sensor serves as a receiver for receiving the sensor signal reflected by the at least one object. It may also be provided that the driver assistance system has further sensors, which may be an ultrasound sensor, a camera, a radar sensor, a lidar sensor, a laser scanner or the like.

A motor vehicle according to the invention comprises a driver assistance system according to the invention. The motor vehicle is designed in particular as a passenger car.

The preferred embodiments presented with reference to the method according to the invention and their advantages apply correspondingly to the control device according to the invention, the driver assistance system according to the invention and the motor vehicle according to the invention.

Further features of the invention will become apparent from the claims, the figures and the description of the figures. The features and feature combinations mentioned above in the description, as well as the features and combinations of features mentioned below in the description of the figures and / or shown alone in the figures, can be used not only in the respectively specified combination but also in other combinations or in isolation, without the frame to leave the invention. Thus, embodiments of the invention are to be regarded as encompassed and disclosed, which are not explicitly shown and explained in the figures, however, emerge and can be produced by separated combinations of features from the embodiments explained. Embodiments and combinations of features are also to be regarded as disclosed, which thus do not have all the features of an originally formulated independent claim. Moreover, embodiments and combinations of features, in particular by the embodiments set out above, are to be regarded as disclosed, which go beyond or deviate from the combinations of features set out in the back references of the claims.

The invention will now be described with reference to preferred embodiments and with reference to the accompanying drawings.

Showing:

1 a motor vehicle according to an embodiment of the invention, which comprises a driver assistance system with a plurality of sensors for detecting an object;

2 a circle, which is used to describe the position of the object in a direct measurement of the sensors;

3 an ellipse, which is used to describe the position of the object in an indirect measurement of the sensors;

4 the ellipse off 3 in a further embodiment;

5 a schematic representation of the ellipse and a point feature, wherein the point feature and the ellipse to be merged; and

6 a schematic representation of the ellipse and a line feature, wherein the line feature and the ellipse to be merged.

In the figures, identical and functionally identical elements are provided with the same reference numerals.

1 shows a motor vehicle 1 according to an embodiment of the present invention in a plan view. The car 1 is designed here as a passenger car. The car 1 includes a driver assistance system 2 with which at least one object 3 in an environment 4 of the motor vehicle 1 can be detected.

The driver assistance system 2 comprises a control device 5 , for example, by an electronic control unit (ECU - Electronic Control Unit) of the motor vehicle 1 can be formed. In addition, the driver assistance system includes 2 a plurality of sensors 6 . 7 , With the sensors 6 . 7 These may be, for example, ultrasonic sensors, cameras, radar sensors, lidar sensors, laser scanners or the like. In the present embodiment, the driver assistance system comprises 2 twelve ultrasonic sensors 6 of which six in a front area 8th and six in a tail area 9 of the motor vehicle 1 are arranged. The ultrasonic sensors 6 For example, at the bumpers of the motor vehicle 1 be arranged. With the respective ultrasonic sensors 6 an ultrasonic signal can be emitted and that of the object 3 reflected ultrasonic signal are received again. It is also possible to carry out an indirect measurement in which one of the ultrasonic sensors 6 emits the ultrasonic signal and an adjacent ultrasonic sensor 6 that of the object 3 reflected ultrasonic signal is received.

In addition, the driver assistance system includes 2 a camera 7 by means of which the object 3 can be detected. The camera 7 can be followed by an evaluation unit, by means of which the images of the camera 7 can be evaluated. The object 3 can be detected, for example, with a corresponding object detection algorithm. The ultrasonic sensors 6 and the camera 7 are each for data transmission with the control device 5 connected. Corresponding data lines are not shown here for the sake of clarity. With the sensors 6 . 7 In each case measurements can be carried out in which sensor data are generated. These sensor data can then be from the respective sensors 6 . 7 to the controller 5 be transmitted.

The control device 5 can then use the sensor data provided by one of the sensors 6 . 7 has an object feature x _P , x _L , x _U determine which object 3 describes. It can also be provided that the sensors 6 . 7 determine the object features themselves and then to the controller 5 transfer. The determination of the object features x _P , x _L , x _U can take place in a so-called feature extraction plane. Here, measured values of the sensor data can be combined or combined. The object features x _P , x _L , x _U , for example, in a digital environment map showing the environment 4 of the motor vehicle 1 describes be entered. The object features x _P , x _L , x _U may be point features x _P , line features x _L or fuzzy features x _U.

These object features x _P , x _L , x _U can be updated in the environment map. For this new object features x _P , x _L , x _U can be determined. If an object feature x _P , x _L , x _{U is} already present in the environment map and a new object feature x _P , x _L , x _{U is} determined, it can first be checked whether the new object feature x _P , x _L , x _U corresponds to the in the environment map existing object feature x _P , x _L , x _U can be assigned. If the object features x _P , x _L , x _U are assigned to one another, the object features x _P , x _L , x _U can be combined. If no assignment takes place, the new object feature x _P , x _L , x _{U can be included} in the environment map.

The object features x _P , x _L , x _U can be obtained from the sensor data of the ultrasonic sensors 6 be determined. Based on the sensor data of the ultrasonic sensors 6 can be a distance to the object 3 be determined. For this purpose, sends an ultrasonic sensor 6 an ultrasonic signal and receives that from the object 3 reflected ultrasound signal again. Based on the transit time between the emission of the ultrasonic signal and the receiving of the object 3 reflected ultrasonic signal 3 then the distance can be determined. If the driver assistance system 2 - as in the example of 1 shown - several ultrasonic sensors 6 can, with the ultrasonic sensors 6 direct and indirect measurements are carried out.

A direct measurement of the ultrasonic sensor 6 is schematically in 2 shown. In a direct measurement, the ultrasonic sensor sends 6 the ultrasonic signal and the sensor signal and also receives that from the object 3 reflected ultrasound signal or the echo of the ultrasonic signal again. From the duration of the ultrasonic signal can then be determined that the object 3 on a circle x _K or semicircle around the ultrasonic sensor 6 lies around. The radius of the circle x _K results from the half of the transit time of the ultrasonic signal. The arrows 10 describe the signal paths of the sensor signal or the ultrasonic signal.

An indirect measurement of the ultrasonic sensor 6 is schematic in 3 illustrated. Here, a first ultrasonic sensor 6 operate as a transmitter. This ultrasonic sensor 6 is at a position x _s1 and transmits the ultrasonic signal. A second or adjacent ultrasonic sensor 6 which is at a position x _s2 is operated as a receiver and receives that from the object 3 reflected ultrasonic signal from the first ultrasonic sensor 6 was sent out. This is the object 3 on an ellipse x _E or half ellipse, the positions x _s1 , x _{s2 of} the ultrasonic sensors 6 respective focal points x _F1 and x _{F2 of} the ellipse x _{E are} assigned.

As already explained, can be used as object characteristics _P x, x _L, x _U x _P point features or characteristics line x _L by a combination of sensor data in a plurality of measurements from the ultrasonic sensors 6 were generated. However, in some cases it is not possible to determine an object feature x _P , x _L , x _U with an explicit spatial position. This is the case, for example, if insufficient sensor data is available. In this case, a blurred feature x _{U is} provided as the object feature x _P , x _L , x _U. In a direct measurement, this is described by placing a point on the sensor axis with the measured distance from the sensor position. In an indirect measurement, the ellipse is approximated by a circle whose center x _{C is} the average of the two sensor positions x _s1 , x _s2 . This center point x _C is also called a virtual sensor position. The radius of the circle corresponds to the length of the semi-minor axis a of the ellipse. This can be described as follows:

Similarly, the fuzzy feature x _{U is} generated on the axis of the virtual sensor using the calculated radius. Even if it is not possible to use the sensor data to determine point features x _P or line features x _L , it may be provided to combine a point feature x _P or a line feature x _L with a fuzzy feature x _U.

Point features x _P , line features x _L and fuzzy features x _U are calculated as Gaussian random variables x with the expectation E [x] = x ^ and the covariance matrix Cov [x] = E [(x - x ^) (x - x ^) ^T ] = P. Thus, a point feature x _P = (x _p , y _p ) ^{T is represented} by its two-dimensional coordinates in space. A line feature x _L = (x ₁ , y ₁ , α ₁ ) ^T is described by a base point or fulcrum in two-dimensional space and an angle α _{1 of} the line. In the present case, the length of the line is not stochastically modeled. Finally, a blurred feature x _{U is described} by a scalar random variable x _U = d _u , where d _{u is} its Euclidean distance from the actual or virtual sensor position x _S = (x _S , y _S ) ^T.

To merge the object features x _P , x _L , x _U , a mapping function in the form of a generic, linear estimator is used: x '/ 1 = x ₁ + Kh (x ₁ , x ₂ )

In this case, x ₁ and x _{2 can be} replaced by the object features x _P , x _L , x _U described above, that is to say a point feature x _P , a line feature x _P or a blurred feature x _U. For this purpose, in particular the collapsing of a point feature x _P with a fuzzy feature x _U or the Merging a line feature x _L with a fuzzy feature x _U described. The mapping rule combines two object features x _P , x _L , x _U to obtain further information about the already existing object feature x _P , x _L , x _U. With the aid of the mapping rule, from the first state x ₁ or the first object feature x _P , x _L , x _U, a combined object feature or a merged state can be obtained x '/ 1 be determined. For this purpose a stochastic innovation function h (x ₁ , x ₂ ) is used.

The innovation function h describes a similarity between the object features x _P , x _L , x _U or their states. The mapping rule or the innovation function h can also not be linear. In this case, the innovation function h can then be linearized. This can be done with the help of a Taylor series of the first order:

The factors H ₁ and H _{2 are} the Jacobi matrices. Furthermore, the mapping rule comprises a gain factor K or a gain matrix. The gain K is determined by the projection e ^T P '/ 1e where e may be any unit vector that can be optimally used for the estimator or mapping rule used in the linear form. The gain K can be described as follows: K = -P ₁ HT / 1 (H ₁ P ₁ HT / 1 + H ₂ P ₂ HT / 2) ^-1 .

In the present case it is considered that x ₁ and x ₂ are uncorrelated. This yields E [(x ₁ - x ^ ₁ ) (x ₂ - x ^ ₂ ) ^T ] = 0. Thus, the merged result and the updated covariance matrix result in:

In order to check whether the object features x _P , x _L , x _U or the states are to be assigned to one another, the so-called Mahalanobis distance is defined:

The Mahalanobis distance sets a limit. If this limit is undershot, an assignment of the states or object features x _P , x _L , x _{U is} plausible. In addition, a distance d, a difference in orientation or further geometric investigations can be used to check the plausibility.

To check whether the object x features, x _L, x _U are to be associated with each other and merged _P, is an appropriate innovation function h for the object features x _P, x x _U to determine _L. Subsequently, the models for merging of point features x _P with blurred features are x _L and the folding line of features x _L x _L with blurred features explained. Both models are determined as the difference between a measurement model a _P , a _L and a measured object distance du of the fuzzy feature x _U. For this purpose, the measurement model a _P , a _L can be transferred in a measurement space in order to be able to compare it with the current measurements.

When a point feature x _{P is to be} merged with a fuzzy feature x _U , the measurement model a _{P is} the Euclidean distance d _P between the point feature x _P and the sensor position x _S. The innovation function h results in:

When a line feature x _{L is to be} merged with a fuzzy feature x _U , the measurement model a _{L is} the absolute value of the distance d _L between the line feature x _L and the sensor position x _S. The innovation function h results in:

A graphic description of the two innovation functions h will be explained in more detail below.

As described _above , fuzzy features x _{U are described} as circular distance measurements with the distance d _u from a sensor position x _S.

This model is used regardless of the type of measurement that produces the fuzzy feature. In a direct measurement with an ultrasonic sensor 6 this model describes the information available regarding the signal path. For indirect measurements with the ultrasonic sensor 6 Such a model is only an approximation, the accuracy of which depends on how similar the ellipse is to a circle. In order to remedy this limitation, the description of a fuzzy feature x _U is first extended to also be able to store information from an indirect measurement. Furthermore, the innovation function h is determined in order to be able to combine point features x _P and line features x _L with fuzzy features x _U in the form of the elliptical distance measurements.

4 shows a schematic representation of an ellipse x _E as a blurred feature x _U. An ellipse x _E can be described mathematically in different ways. A common description is to consider an ellipse x _E as a set of points, the sum of the Euclidian distances to the two foci x _F1 = (x _f1 , y _f1 ) ^T and x _F2 = (x _f2 , y _f2 ) ^T the length of the large main axis 2a corresponds to:

The advantage of this description is that it is independent of any Cartesian coordinate system. In addition, the ellipse x _{E can be described} by the implicit ellipse equation using the major semiaxis a and the minor semiaxis b:

This equation is described in the coordinate system of the ellipse x _E , in which the x-axis is oriented at the large semiaxis a and the y-axis at the small semiaxis b. The ellipse x _E is inclined at an angle α. Another property of the ellipse x _E is the so-called eccentricity e: e = √ a² - b² ,

The eccentricity e describes the distance between each focal point x _F1 , x _F2 and the center x _C = (x _c , y _c ) ^{T of} the ellipse x _E. A linear eccentricity e is given by 0 ≤ e <a. For the case e = 0, that the eccentricity is 0, the ellipse x _{E would be} a circle. If e = a, the ellipse would degenerate x _E into a line.

To obtain an extended description of a fuzzy feature x _{U, the} feature extraction layer requires all available information for the indirect measurement with the ultrasonic sensors 6 transferred to the digital environment map. For example, the positions x _{S of} the two ultrasonic sensors 6 and the information about the duration of the ultrasonic signal to be known. The length of the large semiaxis a can then be determined from the transit time: α = transit time distance / 2. From the distance of the ultrasonic sensors 6 the eccentricity e can be determined: e = sensor distance / 2. From this, the length of the small semiaxis b can then be determined: b = √ a² - e² ,

It is assumed that the distance between the ultrasonic sensors 6 or the sensor distance is completely known. This can, for example, in a memory of the control device 5 be deposited. The uncertainty of the measurement is reflected in the length of the large semiaxis a and the length of the small semiaxis b. Since the lengths of the two semiaxes a, b depend on each other and on the linear eccentricity e, the uncertainty of the ellipse x _{E is} completely reproduced by considering individual scalar random variables a or b for the semiaxes a, b. Since a ^ is measured directly with the ultrasonic sensors, a blurred feature x _{U is determined} stochastically by the length of the large semiaxis a. This results in x _U = a _u , where a _u is modeled as Gaussian size. In addition, either both sensor positions x _S, which are described by the focal points x _F1 and x _F2, or a compilation from the center x _C of the ellipse, the linear eccentricity e and the orientation α of the ellipse x _E required to the fuzzy feature to specify clearly.

To determine the assignment and the collapsing, two innovation functions h (x _P , x _U ) and h (x _L , x _U ) are to be determined, with which point features x _P and line features x _L with fuzzy features x _U can be combined. The innovation functions h are determined in such a way that they take into account a difference between a measurement model a (x) and the large semiaxis a of the measured ellipse x _E : h (x, x _U ) = h (x, a _u ) = a (x) - a _U.

Here, the state x may be replaced by the state of a point feature x _P or the state of a line feature x _L. The measurement model a _P (x _P ) can determine the length of the large semiaxis a of the ellipse x _E on which the point feature x _P would lie when used with the ultrasonic sensor 6 or more ultrasonic sensors 6 is measured. When a line feature x _{L is to be} merged with the fuzzy feature x _U , the measurement model a _L (x _L ) determines the length of the major semiaxis a of the ellipse x _E at which the line feature x _L would be tangent. It can be considered that the line feature x _L from a measurement with the same ultrasonic sensors 6 from whose measurement the fuzzy characteristic x _{U was} determined.

The determination of the measurement models for the point features x _P and line features x _{P will be} explained below. The determination of the measurement model a _P for merging the point feature x _P with the ellipse x _E is based on 5 which shows a schematic representation of the point feature x _P and the ellipse x _E. For a point feature x _P , the measurement model is derived from the following definition:

To determine the measurement model a _P , a measurement ellipse x ' _{E is} determined in the present case. This measuring ellipse x ' _E is determined such that the point feature x _P lies on the measuring ellipse x' _E. This definition must be adhered to for each point feature x _E on the measuring ellipse x ' _E. From this, the length of the large semiaxis a of the measuring ellipse x ' _E can be determined:

This applies to a> e. If the condition a = e is satisfied, the point feature x _P lies on the line connecting the foci x _F1 , x _F2 . In this case, no valid measurement can be derived from the point feature x _P. For all other point features x _P valid ellipses can be determined. The measurement model a _P results from the distance d between the first focus x _F1 and the point feature x _P.

This model is used for the innovation function h:

und

bestimmt werden:

After the linearization of the innovation function h, the Jacobi-Mrizen

and

be determined:

ist definiert für alle Punkte außer

Dieser kann aber nicht vorkommen wenn gilt, dass

The Jacobi matrix

is defined for all points except

This can not happen if it is true that

The determination of the measurement model a _L for merging the line feature x _L with the ellipse x _E is based on 6 which shows a schematic representation of the line feature x _L and the ellipse x _E. The measurement model a _L for a line feature x _L can be determined by cutting the line feature x _L with the measurement ellipse x ' _E. Since the solution must be unique, there must be only one point of intersection. This means that the line or the line feature x _{L has to run} tangentially to the measuring ellipse x ' _E. In a first step, the solution is derived from a line feature x _L existing in the local coordinate system of the ellipse x _E. Thereafter, the measurement model a _{L can be} extended so that line features x _{L can be used} in each Cartesian coordinate system.

To determine the points of intersection, the line or line feature x _{L is described} in the explicit parameter form:

This is used in the implicit ellipse equation:

The equation is solved for a parameter t, because from this the intersection points can be derived. Since the equation is quadratic, it is put into the form of a general quadratic equation. After that is solved:

To get a unique solution, the discriminant must be 0: 0 = p ² - q.

This requirement can be solved for a by substituting p and q and replacing b ² = a ² - e ² with:

This solution for calculating the length of the major semiaxis a is valid for line features x _L , which are defined in the local coordinate system of the ellipse x _E. Therefore, the line feature x _L must be transformed into the local coordinate system of the ellipse x _E :

By using the transformations into the solution we get:

und ebenfalls eingesetzt. Dies führt zu:

Finally, the series vector is rewritten to:

and also used. This leads to:

It should be noted that a rotation of the tilted position of the line feature x _L completely disappears in the first term and that in the second term only one angle transformation is required. A rotation of the point feature x _P is not required since the first term is the square of the distance between the line feature x _L and the center x _{C of} the ellipse x _E and this is invariant. The solution obtained is valid only if a> e. If this condition does not hold, the line feature x _L intersects with the line segment between the two foci x _F1 , x _F2 . In this case, no correct ellipse x _E can be derived from the line feature x _L.

Geometrically speaking, the measurement model a _L is calculated by the projection of one of the focal points x _F1 , x _F2 on the line feature x _L and the calculation of its Euclidean distance d to the center x _{C of} the ellipse x _E determined. With this approach, it can be easily shown that the condition a = e is maintained if the line feature x _L lies at one or both focal points x _F1 , x _F2 .

The solution described above is used for the innovation function h, which is described as follows:

Furthermore, the Jacobi matrices are determined:

ist definiert für alle Fälle außer

In diesem Fall ist das Linienmerkmal x_L senkrecht zu der Hauptachse der Ellipse x_E und verläuft durch den Mittelpunkt x_C der Ellipse x_E. Dieser Fall wurde bereits durch die Vorgabe

ausgeschlossen.The Jacobi matrix

is defined for all cases except

In this case, the line feature x _{L is} perpendicular to the major axis of the ellipse x _E and passes through the center x _{C of} the ellipse x _E. This case was already by default

locked out.

Finally, it is shown that the determined innovation functions h for the ellipses x _{E is} a generalization of the already existing models for circles. For the special case of a circle, for example: e = 0 or x _F1 = x _F2 = x _C = x _S , the previously determined measurement models are mathematically the same:

It should be noted that the normal line of the existing measurement models for lines pointing in the opposite direction.

In the present case, an indirect measurement using ultrasonic sensors 6 explained. Such indirect measurement can also be done with other sensors of the motor vehicle 1 , For example, laser scanners, lidar sensors or radar sensors are performed. It can also be provided that the sensor data of further sensors of the motor vehicle 1 be used. For example, from the measurements of the camera 7 Object features are generated and entered in the map. That way the object can be 3 be reliably detected.

Claims (15)

Method for detecting at least one object ( 3 ) in an environment ( 4 ) of a motor vehicle ( 1 ), in which a first sensor ( 6 ) is driven to emit a sensor signal, from a second sensor ( 6 ) Sensor data is received, which the from the at least one object ( 3 ), from the sensor data, a blurred feature (x _U ) as an object feature (x _P , x _L , x _U ) for describing the at least one object ( 3 ), the fuzzy feature (x _U ) defining a distance between the at least one object ( 3 ) and a position (x _S1 ) of the first sensor ( 6 ) and / or a position (x _S1 ) of the second sensor ( 6 ) Describes, characterized in that the fuzzy feature _(U x) is described (as an ellipse x _E), the focal points (x _F1, x _F1) of the ellipse (x _U) (based on the position (x _S1) of the first sensor 6 ) and the position (x _S2 ) of the second sensor ( 6 ), a length of a semimajor axis (a) of the ellipse (x _E ) is determined on the basis of a transit time of the sensor signal, and a length of a semi-minor axis (b) of the ellipse (x _E ) on the basis of the focal points (x _F1 , x _F1 ) and the length of the major half-axis (a) is determined. Method according to claim 1, characterized in that an eccentricity (e) of the ellipse (x _E ) is determined on the basis of a spacing of the foci (x _F1 , x _F1 ) and the length of the semi-minor axis (b) is determined from the length of the semi-major axis ( a) and the eccentricity (e) is determined. A method according to claim 1 or 2, characterized in that based on sensor data of a further measurement of the first and / or the second sensor ( 6 ) a further object feature (x _P , x _L ) is determined which the at least one object ( 3 ), and the fuzzy feature (x _U ) is merged with the further object feature (x _P , x _L ). A method according to claim 3, characterized in that the further object feature (x _P , x _L ) and the fuzzy feature (x _U ) are converted into a common state space, in the state space an innovation function (h) is determined which a similarity between the fuzzy feature (x _U ) and the further object feature (x _P , x _L ) describes, and the merged object feature from the further object feature (x _P , x _L ) is determined based on the innovation function (h). A method according to claim 4, characterized in that for determining the merged object feature from the further object feature (x _P , x _L ) a linear mapping rule is determined, which comprises the innovation function (h) and a gain factor (K) for the innovation function (h) , where the amplification factor (K) depends on a spatial uncertainty of the further object feature (x _P , x _L ) and a spatial uncertainty of the fuzzy feature (x _U ). A method according to claim 5, characterized in that the gain factor (K) is determined such that a spatial uncertainty of the merged object feature is minimal. Method according to one of claims 3 to 6, characterized in that as the further object feature (x _P, x _L) is a point feature (x _P) or a line feature (x _L) is determined. Method according to one of claims 4 to 7, characterized in that for determining the innovation function (h) a measurement model (a _P , a _L ) is determined which a distance (d) between the position (x _S1 ) of the first sensor ( 6 ) or the position (x _S2 ) of the second sensor ( 6 ) on the one hand and the point feature (x _P ) or the line feature (x _L ) on the other hand describes. A method according to claim 8, characterized in that if the further object feature (x _P , x _L ) is the point feature (x _P ), to determine the measurement model (a _P ) a Messellipse (x ' _E ) is determined so that the point feature (x _P ) lies on the scale ellipse (x ' _E ). A method according to claim 8, characterized in that if the further object feature (x _P , x _L ) is the line feature (x _L ), for determining the measurement model, a measurement ellipse (x ' _E ) is determined such that the line feature (x _L ) tangent to the measuring ellipse (x ' _E ) runs. Method according to one of Claims 8 to 10, characterized in that the innovation function (h) is determined on the basis of a difference between the length of the major semiaxis (a) of the ellipse (x _E ) and the measurement model (a _P , a _L ). Control device ( 5 ) for a driver assistance system ( 2 ) of a motor vehicle ( 1 ) which is designed to carry out a method according to one of the preceding claims. Driver assistance system ( 2 ) for a motor vehicle ( 1 ) with a control device ( 5 ) according to claim and with and with a first sensor ( 6 ) and a second sensor ( 6 ) Driver assistance system according to claim 13, characterized in that the first sensor ( 6 ) and the second sensor ( 6 ) Are ultrasonic sensors. Motor vehicle ( 1 ) with a driver assistance system ( 2 ) according to claim 13 or 14.

Images