WO2008064892A1 - Method for determining a position, device and computer program product - Google Patents

Abstract

The invention relates to a method for determining a position of a two-axle trailer (14) having at least one movable wheel axle relative to a tractor vehicle (12), having the steps: generating image data with at least one associated measurement point (R, S, V), detecting at least three predefined measurement points (R, S, V) in the image data, determining the position of each measurement point (R, S, V) relative to the image recording device (16,32) in reference coordinates of a predefined coordinate system, determining an absolute position of each measurement point (R, S, V) with respect to the image recording device (16,32), determining from this an arrangement angle T<SUB>1,2</SUB> between a vehicle longitudinal axis (22) of the tractor vehicle (12) and a steering axis (20) of the trailer (14), and an arrangement angle T<SUB>2,3</SUB> between the steering axis (20) and a trailer longitudinal axis (34) of the trailer (14).

Description

 description

The present invention relates to a method for determining a position of a biaxial trailer having at least one movable axle, a device for determining a position of a biaxial trailer having at least one movable axle and a computer program product.

Today road tractors are dominated by semi-trailers with semi-trailers. In kinematics such vehicles are classified in the category of "general-2-trailer" or short G2T vehicles. Compared to G2T vehicles have trucks with biaxial trailers, in particular with at least one steering axle, kinematically referred to as G3T vehicles, some advantages. These include:

- comparatively low price,

comparatively low wear, in particular with regard to the abrasion of the tires,

- good ratio between payload and dead load.

Experience has shown that G3T vehicles, in particular when driving backwards, are difficult to control, so that rearward maneuvers by a driver demand a high degree of driving skills. Traditionally, inexperienced drivers are quickly overwhelmed when targeted driving maneuvers are necessary when reversing. Object of the invention

It is therefore an object of the present invention to provide a simplified maneuverability of G3T vehicles. This object is achieved by means of the method according to claim 1, the device according to claim 24 and the computer program according to claim 28. Preferred embodiments or embodiments are the subject of the dependent claims.

Device according to one aspect of the invention

One aspect of the present invention relates to an apparatus for determining a position of a trailer having at least one movable axle relative to a towing vehicle, comprising the steps of:

Generating image data of at least one observation element with at least one associated measuring point on the basis of an image recording device,

Detection of at least three predetermined measurement points in the image data,

Determining the position of each measuring point relative to the image pickup device, in reference coordinates of a given coordinate system, in particular in spherical coordinates or

Cylindrical coordinates,

Determining an absolute position of each measuring point relative to the image recording device on the basis of the reference coordinates,

Determine

- _An arrangement _angle Θ _1i2 between a vehicle _{longitudinal axis of} the

Towing vehicle and a steering axle of the trailer and - an assembly angle Θ ₂ , 3 between the steering axle and a Trailer longitudinal axis of the trailer based on the absolute positions of the measuring points relative to the image pickup device.

Advantageously, according to the present invention, the alignment of all three vehicle links is determined in real time. The three vehicle links here are preferably the vehicle longitudinal axis of the towing vehicle, the steering axle of the trailer and the trailer longitudinal axis of the trailer. The vehicle longitudinal axis of the towing vehicle is in particular an axle which is perpendicular to the wheel axles of the towing vehicle when driving straight ahead, i. for example, the front axle and the one or more rear axles of the towing vehicle. Furthermore, the vehicle longitudinal axis of the towing vehicle is arranged in particular centrally between the opposite wheels of a (each) axis. The vehicle's longitudinal axis therefore halves the front axle (s) and the rear axle (s).

The steering axle of the trailer is preferably an axle which is perpendicular to the at least one movably arranged or rotatably arranged wheel axle of the trailer. The movably arranged wheel axle is preferably the front axle or the frontmost wheel axle of the trailer. Further, the steering axis is arranged centrally between the corresponding, opposite wheels of the movable axle of the trailer. The steering axle of the trailer is thus arranged movable relative to the trailer, wherein, in the frame of reference of the earth, the steering axle of the trailer is horizontally and / or vertically movable. The trailer can have more than one moving axle. Each of these movable axles may have a steering axle. The steering axle may for example be parallel or identical to a drawbar of the trailer.

The trailer longitudinal axis of the trailer is, analogous to the vehicle longitudinal axis of the tractor, perpendicular to the front axle and the rear axle of the trailer, when driving straight ahead of the trailer, ie, when the front axle of the trailer and the rear axle of the trailer are parallel. Furthermore, in this state, the trailer longitudinal axis is centered between opposite wheels, ie centered arranged between opposite front wheels and centrally between opposite rear wheels. In other words, the trailer longitudinal axis intersects the center of the front axle and the center of the rear axle of the trailer when the trailer is traveling straight ahead. If, for example, the orientation of the front axle of the trailer and / or the orientation of the rear axle of the trailer is changed, in particular to allow cornering of the trailer, the trailer longitudinal axis is stationary, ie, the orientation of the trailer longitudinal axis to the front axle / or the rear axle is variable.

When driving straight ahead of the team consisting of towing vehicle and trailer in particular the vehicle longitudinal axis of the towing vehicle and the trailer longitudinal axis of the trailer are parallel. The vehicle longitudinal axis and the trailer longitudinal axis, however, can be moved vertically against each other due to different ground clearance of towing vehicle and trailer. The steering axle of the trailer is in straight ahead of the team in plan view of the team parallel to the vehicle longitudinal axis and the trailer longitudinal axis. In particular, the vehicle longitudinal axis, the steering axle of the trailer and the trailer longitudinal axis in plan form a line in this case. In a side view, the vehicle longitudinal axis and the trailer longitudinal axis are parallel, but the steering axis of the trailer is at different heights of the axles of towing vehicle and trailer relative to the road, not parallel to the vehicle longitudinal axis and the trailer longitudinal axis. Rather, the steering axle of the trailer connects the vehicle longitudinal axis and the trailer longitudinal axis.

The vehicle's longitudinal axis, the steering axle and the trailer steering axle need not be the physical axes of the trailer. Rather, the aforementioned axes are geometric axes that are used to describe the kinematics. However, the aforementioned axes can also coincide at least partially with physical axes of the team. For example, the steering axle may be at least partially identical to a longitudinal axis of a drawbar of the trailer.

The vehicle's longitudinal axis and the trailer's longitudinal axis can also be vertical be displaceable. In other words, the vehicle longitudinal axis is not limited to the plane which is formed on the basis of the wheel axles of the towing vehicle. Similarly, the trailer longitudinal axis is not limited to the plane formed by the wheel axles of the trailer. Rather, the vehicle longitudinal axis may be any straight line which is parallel to the plane formed from the wheel axles of the front and rear wheels of the towing vehicle and which is parallel to a plane perpendicular to the wheel axles of the towing vehicle (in straight ahead travel) and the centers of the wheel axles with respect to the distance of opposing wheels comprises. The same applies mutatis mutandis to the position of the trailer longitudinal axis of the trailer.

The position of the steering axle of the trailer is preferably determined by a point of the towing vehicle and a point of the trailer. The point of the towing vehicle is, for example, the point of coupling or coupling of the trailer to the towing vehicle. The corresponding point on the trailer is, for example, the pivot point or the pivot bearing of the front axle on the trailer.

For the purposes of this invention, the term "determining" includes in particular "calculating" a position, for example, the position of the observation element, in particular the absolute position, an angle, in particular an arrangement angle, etc. of the observation element. Additionally / alternatively, the term "determining" may also include an approximation method, a partial or complete extraction of a table, etc.

An "arrangement angle" in the sense of this invention is in particular an angle in three-dimensional space. However, an array angle may also be limited to a plane in three-dimensional space, i. to be a two-dimensional size.

An "absolute position" of a measuring point, of an auxiliary point, etc. is preferably the position of the measuring point in a predetermined coordinate system, such as a Cartesian coordinate system, for example the image recording device at the origin or a coupling point at the origin of the coordinate system. The absolute position relative to this origin is given, and the unit used is, for example, a unit in the metric system, such as meters, centimeters, and so forth.

The absolute position of a measuring point or an auxiliary point therefore differs from the position of the measuring point or the auxiliary point in the image data, although both positions in the coordinate system of the image recording device can be determined. In particular, the image data does not have complete spatial information in three-dimensional space, since the measurement points or auxiliary points are all reproduced in the image plane. The absolute position of a point, on the other hand, is a (true) three-dimensional position specification, in which, in particular, a distance of each measuring point or auxiliary point from the coordinate origin is indicated. For example, this can be done in Cartesian coordinates and the position can be specified with respect to each of the three axes of the Cartesian coordinate system. In this case, the Cartesian coordinate system can be embedded in a coordinate system of the towing vehicle and / or the trailer, wherein the vehicle longitudinal axis can be an axis of the coordinate system and, for example, the two other axes through a plane parallel to the road surface or parallel to a loading area of the towing vehicle and a plane determined perpendicular to the first plane, wherein the two planes must contain the vehicle longitudinal axis. Analogously, the coordinate system can also be formed with regard to the trailer.

Advantageously, the method according to the invention makes it possible to precisely determine the orientation of the three vehicle members, i. the vehicle's longitudinal axis, the towing vehicle, the trailer's steering axle and the trailer's trailer axle. This determination is preferably possible in real time.

Preferred Ausführunqsvarianten the method Preferably

the arrangement angle Θ-i _i2 between the vehicle longitudinal axis of the towing vehicle and the steering axle of the trailer as a function of a position angle Θi the vehicle longitudinal axis of the towing vehicle and a

Position angle Θ _{2 of} the steering axle of the trailer determined on the basis of the absolute positions of the measuring points relative to the image pickup device and

the arrangement angle Θ ₂ , ₃ between the steering axle of the trailer and the trailer longitudinal axis of the trailer as a function of the attitude angle Θ _{2 of} the steering axle of the trailer and a position angle Θ _{3 of} the trailer longitudinal axis of the trailer determined by the absolute positions of the measuring points relative to the image pickup device,

where:

and

A position angle is preferably a variable in the three-dimensional space, which in particular contains a yaw angle, a pitch angle and a roll angle of the respective vehicle member, wherein the attitude angle of an axis in particular the orientation of the axis in three-dimensional space and a

Rotation angle of the axis, called "roll angle" of the axis contains (schematically, a position angle of an axis in Figure 1 is shown).

Preferably, exactly three predefined measuring points R, S, V are detected in the image data, with the arrangement _angles Θ _1j2 and Θ ₂ , ₃ :

and

(O * jc, y Hy /.- _g.) - (* / .- * g ₁ HO-yg,)

Ä '2,3 = arctane

in which

(xκ ₂ , yκ ₂ ) the coordinates of a point K ₂ and (x _P , yp) are the coordinates of a point P in a coordinate system, in particular in a Cartesian coordinate system of a predetermined point K ₁ as a center,

K _{1 is} the intersection of the steering axle with the trailer longitudinal axis of the trailer,

and

K ₂ by rotation of a connecting vector SR of the points S, R by an angle

- /? = (-arctane (-)) h

and by stretching with a factor h + PK ₂

R _s = D (R, - arctznl-YS)

K ₂ = S + (SR _S ) ^^ m

results and for P applies: H

P = S ₊ (SR ₃ ) -

where m is the length of the vector SR and b, h is given dimensions of the at least one observation element.

By way of example, the variables m, b and h are described in more detail in the following description of the figures, in particular with reference to FIG. The vector SR is defined by two preferably opposite measuring point. The parameters b and h are dimensions of the observation element, in particular parameters of the depth and the width of the observation element. In this case, the symbols θ and Θ may have similar, in particular equal, sampling, in particular equal angles, the symbol θ serving in particular to describe an angle and the symbol Θ preferably being an angular triple. For example, the angle Θ _{12 may} include the angles θ _g12 , θ _n i2, and θ _w i ₂ . If, as in the above preferred embodiment, only movements in two-dimensional space are preferably considered, angle Θi, ₂ includes, for example, only the angle θ _g i ₂ , which is referred to as θi ₂ . Analogously, the angle Θi may include the angles θ _g , θ _n i and θ _w i.

The above conditions are exemplary conditions of a planar movement of the towing vehicle and trailer combination. In other words, the team moves in a plane forward or is arranged in a plane. It is also possible that the towing vehicle is disposed relative to the trailer in a different plane therefrom, for example, when one or more tires of the trailer are located at or above an obstacle, such as a curb, pothole, etc., and the towing vehicle is conventionally located is arranged on the roadway. In this case, for example, the roll angle of the vehicle longitudinal axis and the roll angle of the trailer longitudinal axis may differ and the yaw and pitch angle of the two axes, for example, be substantially equal. Likewise, in a uphill or downhill descent of a team around a curve, the roll angle of the two axes may be the same, but the yaw and / or pitch angles differ from each other. Preferably, colored image data, in particular RGB image data, are generated on the basis of the image recording device.

Most preferably, the colored image data is converted into black and white image data.

Further preferably, all pixels of the black-and-white image data are checked to detect the measurement points and identified on the basis of a minimum number and a maximum number of adjacent white image data at least three auxiliary points.

Particularly preferably, seven auxiliary points are identified.

Further preferably, all pixels of the black-and-white image data are checked to detect the measurement points and identified on the basis of a minimum number and a maximum number of adjacent black image data at least three auxiliary points.

Preferably, the minimum number and maximum number of adjacent pixels depends on a nominal number N of pixels per measurement point. The measuring point may in this case correspond to an image of a display device, in particular an infrared diode.

Furthermore preferably applies for the minimum number min and the maximum number max.

max ≡ (JV -3) / 2

and

min = (-V-2) / 3.

The nominal number N of pixels can be determined in an initial step, for example a calibration step, in particular the calibration step described below or in the calibration position, for example from an image automatically and / or manually counted and / or estimated or calculated on the basis of the given geometry and / or the diode properties or arrangement and / or the properties of the image generation device such as, for example, their resolution or their arrangement.

The center of gravity is preferably calculated in each case for the auxiliary points detected on the basis of the adjacent white and / or adjacent black pixels, and the following applies to the position of the respective center of gravity:

Numberpoints Σ * P _x = ^&

number of points

Number of points y _t

P = -1 = 1

number of points

Furthermore, in a plurality of centers of gravity adjacent centers of gravity are preferably detected and the neighboring centers of gravity identified as lying on a track, when the sum of the distances of the individual centers of gravity is equal to the greatest distance between two of the centers of gravity.

In other words, the detected auxiliary points, in particular the centers of gravity of the detected auxiliary points are processed one after the other. To each

Emphasis is given to the nearest center of gravity. This or that focus (s) will become or become the closest

Focuses sought. It then checks if the points are on a

Straights or a stretch lie.

This check is accomplished by comparing the distance, preferably from the first to the third center of gravity, with the distance from the first to the second to the third center of gravity. If both distances, ie the distance between the first and the third center of gravity and the sum of the distances between the first and the second center of gravity and the second and the third center of gravity are the same, the three centers of gravity must inevitably lie on a straight line. Otherwise, an error is output and / or the priorities are discarded, in particular, the priorities are redefined in this case. Alternatively, an error can be output and then the centers of gravity can be assigned in another way to possibly form a leg.

Furthermore, it is preferably checked whether the ratios of the lengths between the individual points correspond to the given geometric relationships from the (actually given) geometry data, in particular the geometric data of the at least one observation element. In other words, the actual geometric relationships of the auxiliary points can be predefined, for example, due to the construction and compared with the detected geometric relationships of the auxiliary points.

Further preferably, at least one leg is identified when four adjacent centroids lie on a common path.

It can also lie more adjacent focal points on the leg or form this.

In other words, if, as stated above, three centroids or auxiliary points are found, an attempt is made to find a fourth centroid or auxiliary point in the same way. If so, it is assumed that these four points form a leg.

Particularly preferably, a leg is identified when the maximum distance between two centers of gravity on one of the distances is greater than a predetermined minimum length.

Further preferably, a ramp is identified when

two thighs are identified,

the two legs have a common point and the common point is the end point of both thighs.

In other words, it is attempted to assemble a ramp in the image data or to determine the actual dimension of the ramp from the legs obtained. Since the legs are preferably determined so that first the closely adjacent centers of gravity or auxiliary points are found and then only the focal points or auxiliary points that are slightly farther apart, it follows that the last detected center of gravity or auxiliary point of a leg of the center the ramp is.

Thus, if two legs are identified which have a common, last center of gravity or auxiliary point, a ramp results from these two legs.

The validity of the ramp may be further preferably checked by determining the length of the legs forming the ramp. The farther a leg is turned away from the image pickup device, the smaller it appears in the image data. Conversely, a thigh facing the image pickup device appears larger than it actually is. It follows that preferably at least one of the two legs must not fall below a certain minimum length. However, if both legs of a detected ramp are smaller than the predetermined minimum length, the ramp is discarded again. The above method is then preferably repeated, that is, a ramp is repeatedly identified until the identified ramp is accepted.

Further preferably, the at least three measuring points R, S, V are determined based on the position of the ramp, wherein one of the measuring points is assigned to a center of the ramp and the other two measuring points are assigned to outer ends of the ramp.

In other words, the measuring point R is preferably assigned to the auxiliary point arranged in the image data at the left end of the ramp. The auxiliary point arranged in the middle of the ramp is preferably the measuring point S assigned and arranged at the right end of the ramp auxiliary points preferably the measuring point V assigned.

The above statements can be carried out analogously for more than three measuring points. For example, a plurality, in particular 8, 9, 10, 15, 20, etc. auxiliary points can be determined and based on the auxiliary points 3, 4, 5, 10, etc. measuring points are determined. For example, the plurality of auxiliary points can be specified by arranging a corresponding number of light-emitting diodes (see below).

Furthermore, two or more ramps can be identified, wherein, for example, the individual ramps can be arranged at a predetermined angle to one another or must be so arranged for construction reasons. For example, a first ramp may be perpendicular to a second ramp. This is particularly advantageous if the arrangement angles are determined in three-dimensional space.

In other words, the legs of the ramp are determined, in particular calculated, with the aid of the aforementioned steps, preferably on the basis of the detected center of gravity, wherein a limb has the following properties:

a leg consists of at least four centers of gravity or auxiliary points;

all four focal points or auxiliary points lie in one line;

the legs forming a ramp have exactly one center of gravity or auxiliary point;

the ramp consists of exactly two such legs;

none of the two legs that form the ramp may be below a minimum length.

Preferably, the at least one ramp is detected in an initial step and detected after detection of the ramp, a change in the position of the at least one auxiliary point, in particular at least one of the measuring points and again determines the position of the ramp based on this detection.

Particularly preferably, the change in the position of the at least one auxiliary point, in particular at least one of the measuring points is detected by checking (adjacent) pixels in an environment of the at least one auxiliary point, in particular at least one of the measuring points, and the position of the center of gravity of the at least one auxiliary point, in particular at least one measuring point is identified.

In other words, the initially detected coordinates of at least the points R, S, V are used and searched again in a certain radius around them for emphases. If all three measuring points are found again in this way, the coordinates of these points are returned as R, S, V. It can also be assumed that all auxiliary points, not only of the preferably three measuring points R, S, V and a change in the position of the auxiliary points are detected. Based on the changed position of the auxiliary points, the positions of the measuring points can then be redetermined. The auxiliary points are preferably identical to the focal points.

Particularly preferably, if not all measuring points R, S, V are identified, the aforementioned method is carried out again.

In other words, should one or more measurement points not be found, again the measurement points R, S, V are determined by identifying auxiliary points in the image data by detecting centroids, identifying legs therefrom, identifying one or more ramps therefrom, and finally, the measurement points R, S, V are identified.

Particularly preferred is one of the steps or several of the steps, in particular all these steps repeatedly performed.

Further preferably, in an initial step for each auxiliary point in a Calibration position of the image pickup device relative to the observation element determines the position of each auxiliary point in pixels of the image data, wherein the actual position of each auxiliary point is determined by an arrangement of a corresponding infrared diode on the observation element and the position of each infrared diode with respect to the image pickup device in the calibration position is predetermined ,

In other words, an image data record can be generated in a calibration position, for example in a position for straight-ahead driving of the combination, on the basis of the image recording device. Since the auxiliary points are, for example, in particular manually predeterminable or predefined for construction reasons, the actual position of each auxiliary point relative to the image recording device is known in the calibration position. Likewise, the position of each measuring point for the aforementioned reason / reasons is known. In particular, the position of the auxiliary points to each other or the position of the measuring points to each other is known. For example, the distances of the auxiliary points to each other can be measured. Furthermore, the distance between two auxiliary points in pixels or all auxiliary points to each other in pixels can be determined on the basis of the image data. It is therefore possible to establish a relation of the positions of the auxiliary points in the image data to the actual position of the auxiliary points. In other words, a position indication in pixels can be converted into an actual position specification, for example in meters, centimeters, etc., and vice versa.

In particular, it is possible that the auxiliary points correspond to positions of infrared diodes. The auxiliary points or their position, in particular their absolute

Position in calibration position, therefore, in particular by positioning of

Preset infrared diodes. Alternatively or additionally, the auxiliary points can be determined in any other possible way. For example, the auxiliary points can be determined by attaching colored or black-and-white or patterned stickers, etc. In particular, such a sticker may be a conventionally known saddle point with opposite black and white

Be quarter circles.

Particularly preferably, the position of each measuring point relative to the Help points are determined in the image data and can be determined by calibrating the image data, a position of each measuring point relative to the image pickup device.

In other words, in the image data, for example, the distance of measuring points from one another or from measuring points to auxiliary points in pixels can be determined and, based on the calibrated or already predetermined distances of pixels, this distance can be converted into actual distances, for example in meters, centimeters, etc.

Further preferably, the angular displacement of each measuring point relative to the image pickup device is given by

win = arctan (^ ≡ ¹ ).

L ₁ -PK ₂ + CK ₁ This trigonometric relationship is described by way of example in FIG.

Particularly preferred is based on the arrangement angle Θi, ₂ between the vehicle longitudinal axis of the towing vehicle and the steering axis of the trailer and based on the arrangement angle Θ ₂ , ₃ between the steering axis and the trailer longitudinal axis of the trailer, a trajectory of the team consisting of towing vehicle and trailer in particular determined automatically.

Advantageously, the trajectory can be displayed during a reversing a driver, so that the driver detects the reverse, especially the position of the trailer in a reverse drive and can react accordingly.

Further advantageously, an automatic reversing of the team can be made possible, with a driver, for example, only controls the speed.

Device according to one aspect of the invention Another aspect of the present invention relates to an apparatus for determining a position of a trailer having at least one movable axle relative to a towing vehicle

at least one observation element to which at least one measuring point is assigned,

an image recording device which is designed to generate image data of the at least one observation element,

a detection device which is designed to detect at least three predetermined measurement points in the image data,

a determination device which is designed

- The position of each measuring point relative to the image pickup device, in reference coordinates, in particular in Kugelbzw. Cylinder coordinates, to determine - an absolute position of each measuring point with respect to the

To determine image recording device based on the reference coordinates, - to determine an arrangement angle Θ-ι, ₂ between a vehicle longitudinal axis of the towing vehicle and a steering axle of the trailer on the basis of the absolute positions of the measuring points relative to the image pickup device and

~ To determine an arrangement angle Θ ₂ , ₃ between the steering axis and a trailer longitudinal axis of the trailer on the basis of the absolute positions of the measuring points relative to the image pickup device.

Preferred embodiments of the device

Preferably, the observation element comprises a plurality of display devices. For example, such Representation means a diode, in particular an infrared diode, a laser diode or a conventional semiconductor diode which emits yellow and / or green and / or red and / or blue and / or white light, be.

The display devices are preferably infrared diodes, in particular 3, 7, 10, etc., infrared diodes.

Additionally or alternatively, the presentation means may also be a sticker, a color dot, a color area, a conventional incandescent lamp, an acoustic signal transmitter, an RFID transmitter, etc.

Further preferably, the image pickup device comprises a conventional digital camera and / or a conventional analog camera.

Computer product according to one aspect of the invention

Another aspect of the present invention relates to a computer program product, particularly stored or signalized on a computer readable medium, which, when loaded into the memory of a computer and executed by a computer, causes the computer to perform a method according to the invention.

The foregoing description of the aspects of the invention is not limited to the respective aspects. Rather, the statements on the respective aspects apply mutatis mutandis to the other aspects of the invention. In particular, the statements with regard to the method or preferred embodiments of the method mutatis mutandis apply to the device and the computer program product or preferred embodiments thereof.

Figurenbeschreibunq

Hereinafter, preferred embodiments or embodiments of the present invention with reference to accompanying figures described by way of example. Individual elements of the described embodiments or embodiments are not limited to these embodiments or embodiments. Rather, elements of the embodiments or embodiments can be combined with each other arbitrarily and new variants or embodiments are created by it. It shows

Figure 1: a schematic representation of an axis in three-dimensional space; Figure 2: a schematic arrangement of objects and axes;

Figure 3 is a schematic view of a team;

FIG. 4 shows a schematic illustration of an observation element,

FIG. 5a shows a schematic view of an exemplary arrangement;

FIG. 5b shows a schematic illustration of an exemplary arrangement; FIG. 6 is a schematic detail view according to FIG. 3;

FIG. 7 shows a schematic view of an exemplary geometry of a trailer;

FIG. 8 shows a schematic view of an exemplary receptacle on the basis of FIG

Image recording device; FIG. 9 is a schematic view of a special receptacle;

FIG. 10: a schematic view of partial elements;

Figure 11a is a flow chart;

Figure 11b is a flow chart;

FIG. 12 is a schematic view of a geometry; Table 1: an overview of model values;

Table 2: an overview of actual measured values;

Table 3: a visual representation of actual measured values;

Table 4: a visual representation of actual measured values;

Hereinafter, a preferred embodiment of the method will be described with reference to the figures. Here, a mathematical description of objects in three-dimensional space is used. The basis of the description of objects in the room is their location. To describe the position, a reference point of the object in space and the orientation of the object in space is assumed. The reference point of an object O in the coordinate system X is generally described by:

The orientation of an object O in the coordinate system X is described by its yaw, pitch and roll angle:

which are shown by way of example in FIG. The position of an object O in the coordinate system X is then:

L _X (O) = (K _X (O), _X (O))

Furthermore, the orientation of three members G ₁ , G ₂ and G _{3 of} a team of towing vehicle and trailer with at least one movable axle (hereinafter referred to as G3T vehicle) in three-dimensional space can be clearly described by static and dynamic information

Static data: The static specification can be a length L ₁ , i≡ {!, ..., 3} of each vehicle link.

Dynamic data: The yaw, pitch and roll angles of each vehicle link can be used as dynamic data. The yaw, pitch and roll angles are usually represented as follows: / e {l, ..., 3},

where θ _xg (G,) denotes the yaw angle, θ _{X n} (G _t ) denotes the pitch angle, and θ _{X w} (G _t ) denotes the roll angle. The angle O _x (G ₁ ) or the angle triple here is a preferred attitude angle.

If it is clear to which (indexed) object and which reference point is based, the position angle ^ _x (G ₁ ) can be written in a simplified way:

The following relationship of the angles to the Cartesian coordinates of a member G ₁ , fe {l, ..., 3}, exists: x _ι = l _ι sin (θ _{n ι} ) cos (θ _{g ι} ) y, = ^ι . sHθ _nι ,) sm (θ _gιl )

Optical measuring systems, as described below by way of example as a preferred embodiment (s) of one or more constituents or as a preferred embodiment (s) of the device according to the invention, include, for example, a device C for optical

Capture of an observation object O. The boundary conditions for the

Functioning of the measuring system are in particular that O is within sight of C, and that C is rigidly coupled to Q ₁ and O is rigidly coupled to Θ ₃ (or O rigidly to G ₁ and C rigid to Θ ₃ ), as example is shown with reference to the schematic drawing of Figure 2. Line O ₁ may represent an exemplary representation of a vehicle longitudinal axis (shown in FIG. 3) of the towing vehicle (shown in Figure 3). Line Θ ₃ may be an exemplary representation of a trailer longitudinal axis (shown in FIG. 3). The line O ₁ may also be an exemplary representation of the trailer longitudinal axis. In this case, the line Θ _{3 may be} an exemplary representation of the vehicle longitudinal axis of the towing vehicle. Line Θ ₂ may be an exemplary representation of a steering axle of the trailer (shown in FIG. 3).

The rigid coupling of C to G ₁ and O to G ₃ is expressed by the relative positions L _C (C) and L _Gj (O).

Due to the fixed coupling of the vehicle links are the angle Θ, _{i + 1} between the

Vehicle links are preferably defined as preferred arrangement angles as follows:

Θ,, _{1 + 1} = Θ, - Θ, ₊₁ i ≡ {!, ..., 2}

The determination of the angles between the vehicle members, i. the arrangement angle can take place according to a preferred embodiment variant in two process steps to be carried out successively:

First step: Detection of the object of observation

In order to recognize the object of observation O as a preferred embodiment of the at least one observation element, which is arranged wholly or partly in the field of view of the camera C, as a preferred image recording device, m excellent points, in particular auxiliary points, are sought on the observation object. Advantageously, it is sufficient if (O ₁ , ..., O _n ) points are found and identified as preferred measurement points, where n ≦ m. In particular, the measuring points are a subset of the auxiliary points. The points (0 ,, ..., O _n ) can be converted into angles, among which the

Camera sees these points. Related to an imaginary plane, called Lensenbene in the further, which goes longitudinally through the lens of the camera, this Points are assigned two angles analogous to polar spherical coordinates. The statements of this application apply mutatis mutandis to the determination of the angle in three-dimensional space, ie in particular as spherical coordinates.

According to a preferred embodiment variant, the lens plane is preferably a plane which in the position of use of the camera is parallel to a plane which is spanned by two axes of the towing vehicle. According to a further preferred embodiment of the present invention (as further described below), the vehicle longitudinal axis, the steering axis and the trailer longitudinal axis are preferably arranged in a plane. The lens plane is preferably parallel to this plane.

For an angle tuple or angle pair C ₁ (in the lens plane of the camera or a projection into the lens plane), under which a point O _{1 is arranged} relative to the camera, the following applies:

C _J = [AJ ^' e {!, .., «}

Thus, the output of recognition of the subject of observation is a tuple (C ₁ , ..., C _n ) of pairs of angles defining a family of n straight lines passing through both the center of the lens and each point O ₁ , every {l, ..., n} go. The points O ₁ may represent one or more light-emitting diodes.

Second process step: calculation of the angles Q _{1 2} and Θ ₂₃

From the n angle pairs (C ₁ , ..., C _n ) triples of angles Q _{1 2} and Θ _{23 are} calculated. This is done in two steps:

First step:

Localization of an observation object or one or more Auxiliary points or measuring points in Cartesian coordinates relative to the camera:

The position of the object of observation, ie the position of the totality of the points (O ₁ ,..., O _n ) relative to the lens plane and the center of the lens, is given by the Cartesian coordinates

and the three degrees of freedom of possible rotations:

By means of a calculation rule f _x (see below), the position of the points (O ₁ ,..., O _n ) in relation to the camera C can be determined from the angle pairs (C ₁ ,..., C _n ):

Z ₁ (C ₁ , ..., CJ = (L ₀ (O ₁ ), ..., L ₀ (O _n )).

Second step:

From the positions (L ₀ (O ₁ ), ..., L ₀ (O _n )), the two angles Θ _{1 2} and Θ _{23 of} the two-axle trailer can be determined by means of a calculation rule / ₂ (see below):

f ₂ (L _c (O _ι ), ..., L _c (O _n )) = (Θ _U2 , Θ _v ).

To further illustrate the invention, a preferred embodiment of the method according to the invention is described below. According to this embodiment, the two-axle trailer preferably moves only planar. Thus, preferably only the yaw angles {θ _gn , θ _{g 23} ) are calculated, the pitch (θ _{n 2 2} , θ _{n Ω} ) and roll angle ( _{w w U} , θ _{w 23} ) are preferably neglected. However, the pitch (θ _nn , θ _{n 23} ) and roll angle (θ _{w l2} , θ _{w 23} ) can be calculated analogously to the yaw angles {θ _{g i2} , θ _{g 2i} ). For ease of understanding, the yaw angles (θ _{g l2} , θ _{g 23} ) will hereinafter be referred to as θ ₁₂ , θ ₂₃ .

A preferred embodiment is shown by way of example in the schematic drawing of FIG. FIG. 3 shows a vehicle 10 with a vehicle 12 and a trailer 14. The basis of the optical surveying system according to the preferred embodiment of FIG. 3 is a camera 16, which is preferably mounted on the rear of the tractor 12. An observation object 18 is arranged on the trailer 14. Advantageously, the subject matter 18 may be mounted in front of or behind a towing device 20, in this example the towing device 20 is identical to a steering axle 20 of the trailer 14. The camera 16 is preferably mounted on the longitudinal axis 22 and in the center of the towing vehicle 12. Alternatively, the camera 16 may also be arranged on the trailer 14 and the observation object 18 on the towing vehicle 12.

The camera 16 is preferably equipped with a filter for infrared light, which filters out the incoming light up to the infrared component. This ensures that less stray light is incident on the camera 16 and the system still works well even in bad weather conditions. However, the camera 16 may also be an optical element which detects electromagnetic radiation of a predetermined wavelength, e.g. UV radiation, etc. Alternatively / additionally, the optical element may also be a sensor or detector of laser radiation, in particular a laser scanner. Here, the laser light may have any predetermined wavelength, e.g. red, blue, green light, etc. The laser may be, in particular, an infrared laser, a UV laser, etc. Preferably, the laser is a conventional diode laser.

The counterpart to the camera 16, the observation object 18 or the Observation element 18, is preferably a pyramid-like ramp 18, which is preferably located on a front side of the trailer 14 substantially at the height of the camera 16.

At this ramp 18, preferably seven infrared diodes 24 are arranged as preferred auxiliary points, which can be recognized at least partially on the basis of the image data of the camera 16. More or fewer infrared diodes 24 can be used. For example, 3, 4, 5, 6, 8, 10, 12, 14, 16, 18, etc. infrared diodes 24 can be used. Preferably, the infrared diodes 24 are arranged at substantially the same distance and symmetrically to one another.

According to a further embodiment of the present invention, in addition to or as an alternative to the infrared diodes 24, further auxiliary points are predefined, e.g. by arrangement of laser diodes and / or acoustic signal generators, in particular in the ultrasonic range, one or more colored stickers, colored surfaces, etc.

For the later calculation of the yaw angles, of these, preferably seven infrared diodes 24, particularly preferably only the two outer auxiliary points (R, V) and the tip (S) are required as preferred measuring points. For this purpose, the angle at which the camera 16 sees the point S is calculated and passed on to the algorithm. The same applies to the points R and V. The measuring points R, S, V are shown in Figure 3 by the reference numeral 26.

Furthermore, a rear axle 28 of the trailer 14 is shown in Figure 3, are arranged on the other LEDs 28. The light-emitting diodes 28 can be imaged, for example, by means of a camera 32. The rear axle 28 may be, for example, a steerable axle and based on the LEDs 30, analogous to the light emitting diodes 24, the orientation of the rear axle 28 relative to other axes, such as a trailer longitudinal axis 34, the steering axle 20 and the vehicle longitudinal axis 22 are determined. For this purpose, a ramp (not shown) similar or identical to ramp 18 may be arranged on the rear axle 28. In FIG. 4, a preferred ramp 18 is shown by way of example as a preferred observation element 18. Ramp 18 has a plurality of diodes 24 as preferred auxiliary points. Three of the diodes 24 are used as preferred measurement points to determine the aforementioned angles. Instead of the ramp 18, the diodes 24 etc. can be arranged directly on the trailer 14 and the towing vehicle 12, respectively. For example, seven infrared diodes 24 may be disposed on the trailer 14. The infrared diodes 24 thus together form an observation element 18 and in each case an auxiliary point. If, for example, only three infrared diodes 24 are arranged, the three infrared diodes 24 together represent an observation element 18 and in each case an auxiliary point and also a measuring point. However, the three diodes 24 can also each represent an observation element 18.

To simulate the geometry of a two-axle trailer, a model of such was exemplified on a 100 cm x 150 cm base plate

Set up two-axle trailer, as shown schematically in Figure 5a and is shown by way of example in Figure 5b. As shown in Figures 5a, 5b, two pivot points 36, 38 are arranged, on which two small rotatable

Wood panels were attached. On the back, a 80 cm x 80 cm black wooden panel was placed vertically, which is to simulate the measuring area at the front of the trailer. This is the plate

Observation object 18 in the form of a diode ramp 18, attached. The camera

16 was mounted at the other pivot point exactly at the height of the ramp 18

(see Figure 5a).

As stated above by way of example, there are preferably seven infrared diodes 24 on this ramp 18, which are mounted at a lateral distance of 10 cm and 7 cm apart in height. This results in that the diodes 24th

Λ / 7 ² +10 ² = Λ / Ϊ49 = 12, 21cm (1)

away from each other. Furthermore, the ramp 18 is 60 cm wide and 21 cm high. Further dimensions of the exemplary model calculation are given in Table 1. Other important or necessary variables for the calculation described below are shown in FIG. 6, which is a detailed view of a section of FIG. 3. Identical components of Figure 3 and Figure 6 are therefore provided with identical reference numerals. In particular, FIG. 6 shows geometric relations of the observation device 18, for example in relation to the towing vehicle 12, in particular the vehicle longitudinal axis 22 of the towing vehicle 12 and / or the trailer 14, in particular the steering axle 20 and / or the trailer longitudinal axis 34. The ramp 18 in particular has a three-dimensional shape FIG. 6 shows the projection of the ramp 18 into the aforementioned plane. The ramp 18 thus has a depth h and a width 2b. The depth h is preferably defined by the distance of the foremost diode 24, ie the diode 24 closest to the towing vehicle 12 and the rearmost diode 24. In the example shown, the depth h equals the distance of the points S and P. The width 2b is equal to the distance of the two outermost diodes 24. In this example, the width 2b is equal to the distance of the points V and R from each other.

FIG. 6 also shows the coupling points 40, 42, which preferably determine the position of the steering axle 20. Furthermore, the arrangement angles Θi, ₂ and Θ ₂ , ₃ are shown, which are determined by the position of the vehicle longitudinal axis 22, the steering axis 20 and the trailer longitudinal axis 34.

Since preferably only the movement in a planar plane is considered, the combination can be identified with its orthogonal projection onto this plane. The points R, S, V, K _ι , K ₂ are consequently points in the (predefined) plane or projections of the points into this plane. The level can be:

a plane spanned by the axles of the towing vehicle,

a plane that is spanned by the wheel axles of the trailer,

a plane parallel to a horizontal plane in the frame of reference of the earth, a plane parallel to the road surface,

a combination of the aforementioned levels, etc.

In the following description, the term "movement" does not mean a movement in the sense of forward and / or reverse travel, but the change of the position of the trailer in relation to the vehicle.

In this case, the point K _x is preferably defined as the reference point 40 of the movement and receives the coordinates in the coordinate system O, y) chosen by itself

(0,0). The point ^ ₁ is therefore a fixed point in the example selected

Coordinate system. The connecting straight line from K ₁ to the camera 16 defines the x-axis. Thus, the camera 16 also has fixed coordinates in the exemplary coordinate system. It is also possible that the camera 16 defines the origin of the coordinate system.

The coordinates of the points K ₂ , R, S, V change with the movement of the

Traction vehicle 12 and the team consisting of towing vehicle 12 and trailer 14. While driving a clutch K ₂ moves on a circle around the center K ₁ . The radius corresponds to the length of the connection L ₂ between the clutches K ₁ , K ₂ . Likewise, each vertex of the triangle moves on its circle around the center K ₂ . Their fixed connections to K ₂ clearly define the radii of these points. This is illustrated by way of example in FIG.

However, not every point on these circles can be assumed, as the technical requirements, such as the combination parameters, the camera opening angle, etc. do not allow certain constellations.

From the images of the points R, S, V, the angles are determined under which the camera 16 sees the dots, that is, under which the dots are imaged in the image data. These are called recording angles below and by C _R , C _S , C _V denotes. The recording angles are preferably angles in spherical coordinates or in polar coordinates in the coordinate system of the camera 16 or the reference point K ₁ .

The algorithm described below according to a preferred embodiment of the present invention, as stated above, consists essentially of two steps:

Based on the recording angle of the camera C _R , C _S , C _V , the positions of the measuring points R, S, V are determined.

Based on the positions of the measuring points R, S, V, the two yaw angles <9 ₁₂ , <9 _{23 are} determined. The yaw angles θ _u , θ ₂₃ serve to determine the relative position of the

Towing vehicle 12 and the trailer 14 represent.

The acceptance angles C _R , C _C , C _V correspond exactly to only one position of R, S, V, ie the triangle (S, V, R) can only be placed in a single way between the rays, so that the corners on the Rays are appropriate, which is proved in the following.

Suppose the triangle could be placed so that the point S - seen from point C - lies between R and V, as exemplified in FIG. Furthermore, it is assumed that two different possibilities R, S, V and R \ S ', V exist to position this triangle. If the point S 'is between S and C, because of the fixed length of the side S, R (or S', R '), the angle of the triangle between the side S', R 'and the S-beam at the Corner S 'smaller than the corresponding angle at S. Accordingly, this is also the case on the side S ', V'. Therefore, for the angles:

Z (V'S'R ') <Z (VSR) (2)

But since the triangle is fixed, that is, all sides and all angles are the same this is not possible. If S 'is on the other side of S, the corresponding angle becomes larger, which is not possible. This means that the possible position of the triangle is unique. In all other cases one will proceed in the same way.

Hereinafter, a preferred embodiment of the method according to the invention is described to determine the position of a trailer 14 relative to the towing vehicle 12, in particular the vehicle longitudinal axis 22, the steering axle 20 and the trailer longitudinal axis 34 to each other. In this case, the calculation rule f _{x is} formulated in the first step and in the second step

Calculation rule / ₂ formulated. In that regard, the following description of the first step is a definition of the calculation rule Z ₁ and the following description of the second step is a definition of the calculation rule f ₂ .

First step: The coordinates of R, S, V are determined from C _R , C _S , C _V.

Let AB denote the distance between two selected points A and B. The searched three points R, S, V fulfill the following conditions:

R, S, V are on rays emanating from C and through the

Recording angles are determined.

The distances between them are fixed:

RV = I b (3)

RP = PV = b (4)

SP = h (5)

(6) Z (K ₂ SR) = Z (K ₂ SV) = p = arctane - (7]

All possible recordings are divided into four different cases, which are described individually below:

C _V <C _S <C _R

C _V = C _S <C _R or C _V <C _S = C _R

C _S <C _V <C _R

C _V <C _R <C _S

Case 1: C _V <C _S <C _R

The corresponding angles in FIG. 8 are determined by the recording angles:

C _x = C _n -C ₅ (8)

The angle at S within the triangle (RSV), which is 2-p, is divided by the beam through S into two angles, S ₁ and S ₂ :

S _λ + S ₂ = 2-p (10)

The following applies:

Z (CRS) = R ₀ = S ₁ -C ₁ (11)

Z (CVS) = V ₀ = S ₂ -C ₂ (12) V ₀ = IPC ₁ - C ₂ -R ₀ (13)

Furthermore, let D be defined by:

D = 2-P- (C ₁ + C ₂ ) (14)

i.e.

V ₀ = D -R ₀ (15)

The triangle (VCS) and the triangle (SCR) have the angle C ₁ and C ₂ at point C, respectively.

Opposite these angles, both triangles have a side of equal length m and a common unknown side of length CS. The following applies:

m CS

(16) sin C ₁ sin R ₀

and: m CS

(17) s is C c, ₂ s is VV _n0

From these equations follows:

sin R ₀ sin V ₀ CS = m - = m ^ (18) sin C, sin C,

With the help of (15):

sin R ₀ = ^^ - Sm (D -R ₀ ) (19) sin C ₂

or: sin V ₀ = ^ .. SiXi (DV ₀ ) (20

SUiC ₁

After transformation and the division by cos R ₀ we get:

tan ^ = sin C ₁ SInD _{2χ sin C ₂ + sin C ₁ ^■ cos D

or correspondingly: sin C ₂ -SmD sin C ₁ + sin C ₂ • cos D

Now the distance CS with the known angle R ₀ (or V ₀ ) can be calculated:

SmR ₀ CS = m - (23)

Sm (C ₁ )

or:

CS = m - ^^ - (24)

SUi (C ₂ )

Let p _{c give} the normalized direction vector determined by C ₅ . The searched points:

S = C + CS -p _{c $} (25)

Let an auxiliary point R _s be defined such that the distance SR _{5 has} the length m and is rectified by p _c :

R _s = S + mp _Cs (26)

By corresponding rotations of this point R ₅ , the points R and V Are defined. Let D (A, a, B) denote a rotation of the point A about the angle a about the point B. Then:

R = D (R _5, S, S) (27)

V = D (R -2 -p, S) (28)

or

V = D (R _s , -S ₂ , S) (29)

R = D (V, 2 p, S) (30)

Thus, the searched points R, S, V are calculated.

Case 2: C _V = C _S <C _R or C _v <C ₅ = C _R

As a special case, it should be emphasized if C _R = C ₅ (or C _v = C ₅ ), as shown by way of example in FIG. 9.

Here S and another point, here without restriction of generality R, lie on the same ray:

C _R = C _S (31)

C ₁ = QC ₅ = O (32)

C ₂ = C _S -C _V = Z (RCV) (33)

The exact position of the triangle, ie the coordinates of R, S, V, can be easily calculated. The following applies: CS m

(34) sin (π- (C ₂ + (π- 2 p)) sin C ₂

from this: m - sin (2 - /? - C ₂ )

CS = - '2 / (35) sin C,

and the searched points are calculated according to the first case.

Analogously, it is possible to proceed for C _S = C _V , ie if S and V lie on the same beam.

Case 3: C _V <C _R <C _s

Since the ramp is preferably made of wood, this case can not occur in the embodiment described above. The top of the ramp S would obscure the point R and this is no longer perceptible to the camera (as shown by way of example in Figure 10). Thus, the above embodiment is limited by the geometry of the ramp to a certain angle.

However, the ramp can at least partially consist of a partially transparent or completely transparent material. This may preferably be limited to the light of the wavelength of the diodes. For example, if the diodes are infrared diodes, the ramp may be made of a material that is at least partially transparent to infrared light. Advantageously, now with the help of the following two cases higher degrees can be detected.

The angles defined in (8) and (9) have different signs here:

C ₁ <0 (36)

C ₂ > 0 (37) m CS

(38;

SUi (C ₁ ) Sm (TJ ₀ )

m CS

(39;

SUi (C ₂ ) SUi (V ₀ )

That means:

A ₀ <0 (40)

V ₀ > 0 (41)

By (14) results:

sin (C) - sin (D)

In (R ₀ ) = - (42) sin (C ₂ ) + sin (C,) cos (D)

and

SUi (C ₂ ) -sin (Z)) tan (V ₀ ) = - (43) sin (C _j ) + SUi (C ₂ ) ^• cos (D)

In case 1, the desired size CS could be determined either from the formula for R ₀ or from the formula for V ₀ , in particular be calculated. In order for this to be possible in this case as well, the following must apply:

tan (i? o) <O (44)

and tan (V _o )> 0 (45)

In formula (42), the counter is negative, i.

tan (/? ₀ ) <0 (46) However, this only applies if

SUi (C ₂ ) + SiIi (C ₁ ) - cos (D)> 0 (47, ie sin (C ₂ )> - sin (C,) ^■ cos (D) (48) But because

C ₂ > -C ₁ > 0 (49;

SUi (C ₂ )> sin (-C _j )> sin (-C,) ^• cos (D) (50) and thus Im (R ₀ ) KO (51)

In formula (43) the counter is positive, ie tan (V ₀ )> 0 (52)

However, this only applies if

Sm (C ₁ ) + sin (C ₂ ) -cos (D)> 0 (53) ie sin (C ₂ ) -cos (£>)> -Sm (C ₁ ) (54) respectively.

This means that (52) holds only if (55) is satisfied. However, there are no D that meet this condition. Therefore, in the case

C _V <C _R <C _s (56;

the desired size CS can be calculated with the help of R ₀ . The subsequent calculation of the points R, S, V is analogous to Case 1.

Case 4: C _S <C _V <C _R

The procedure here is analogous to Case 3, except that here the desired size CS must be calculated from V ₀ .

2nd step: _Calculate the angles θ _l2 , θ ₂₃ from the points R, S, V

First, the point K _{2 is} calculated by the vector SR, which has the length m, around the fixed angle

- / 7 = (-arctane (-)) (57) h

is rotated around the point 5 and then stretched by the factor (h + PK ₂ ).

R _s = D (R, - aτctanl -), S) (58)

K - S ₊ (SR ₅ ) ^^^ (59) m Similarly, the point P is found as follows:

P = S + (SR _S ) -] (60 ^'

Let (χ _A , y _A ) denote the coordinates of the point A. Thus, the sought angles result from:

θ _n = arctan Vx _t (61:

θ _2i = arctane | (62;

[(0- _{Xκ2). (Xp} - _Xκ2) - (y _p - _yK2). (0-y _K2)

where θ _{23 is} calculated as the angle between the vectors (K ₂ K ₁ ) and (K ₂ P).

Similarly, with an inverse algorithm, the points R, S, V and the angles C _S , C _R , C _V can be calculated from the angles θ _l2 , θ ₂₃ .

From the yaw angles θ _ι2 , θ ₂₃ , the coordinates of the points S, R, V as well as those of P, K _{2 can be} calculated.

x _Ki = K ₁ K ₂ cos θ _n (63)

y _K2 = K ₁ K ₂ -sin θ _n (64)

Let P ^ denote a provisional position of the point P. Only θ _n is considered (O ₂₃ = O). P ₀ = (L ₂ -PK ₂ W (65;

P, = D (P ₀ A ₂ ^ ₁ ) (66; P ₀₁₂ has the following coordinates: x = (K ₁ K ₂ -PK ₂ ) - COS (O ₁₂ ) (67; y = (K ₁ K ₂ -PK ₂ ) -Sm (O ₁₂ ) (68;

The point P, with coordinates (x, y) is as rotation of P ₀₂ by the angle <9 ₂₃ um

Point K ₂ emerged:

P = D (P ^ O ₂₃ , K ₂ ) (69]

Analogous to the point S with coordinates (χ _s , y _s ) is first a corresponding

_Calculated point S _θi2 :

S ^ = D ((K ₁ K ₂ -PK ₂ -h, 0), θ ₁₂ , (0,0)) (70)

and subsequently

S = D (S ^ K ₂ ) (71)

For the point R with coordinates (χ _R , y _R ), a corresponding one is also first

_Calculated point R _θ2 :

R ₀₁₂ = D ((K _λ K ₂ -PK ₂ , b), θ _n , Φ, ^Q )) (72)

and subsequently

R = D (R ₀₁₂ ^ K ₂ ) (73) The same applies to the point V with coordinates (χ _y , y _v ), for which a corresponding point V ^ is first calculated:

V ₀₁₁ = Di (K ₁ K ₂ - PK ₂ ^ b) A ₂ MO)) (74)

and subsequently

V = D (V _θιi , θ ₂₃ , K ₂ ) (75)

For these points (R, S, V), the angles at which the camera sees the points can be calculated. Let C _R , C _S , C _V denote the corresponding angles:

tan (C ") = y _R (76)

CK _x + x _R

tan (Q) = ^ (77)

⁵ CK ₁ + x ^V _s'

tan (Q) = - p- (78)

CK ₁ + X _y

FIG. 11a shows a flow chart for an overview of the process steps that are taking place. In step S1, image data is generated by an image pickup device 16 (shown in the previous figure 6), which is also referred to as "fetch image." In step S2, the measurement points necessary for further determination of the arrangement angles are detected on the basis of the image data Measurement points in the image data, that is to say based on the actual position of the measurement points in pixels in the image data, are calculated in step S3 the actual positions of the measurement points, ie for example the infrared diodes 24, in particular in the frame of reference of the camera These positions are referred to as acquisition angles and are preferably as cylindrical coordinates or spherical coordinates, in particular polar spherical coordinates. The reference system of the camera can be easily transformed to the reference system of the towing vehicle, for example. Finally, in step S4, the arrangement angle of the vehicle longitudinal axis relative to the steering axis and the arrangement angle of the steering axis to the trailer longitudinal axis are determined, in particular calculated, on the basis of the acceptance angles. These angles can be output by means of a display, a computer interface, a radio transmission, etc., for example. These angles can also be used to determine a trajectory of the team. Furthermore, based on an angle change, a change in the trajectory can be determined and in particular displayed. The arrangement angle or change thereof can also be used to automatically control the reversing of a trailer.

However, the exact processing steps according to a preferred embodiment of the present invention, which preferably occur in a computer, are far more complex than just fetching and outputting angles. Thus, in the following paragraphs, referring to the flow chart shown in Fig. 11b, a further preferred embodiment of the present invention will be described by way of example, in particular how to determine the yaw angles as preferred arrangement angles from an RGB image of the camera.

In an initial step, the image data is generated on the basis of the image recording device. These image data are preferably color image data, in particular RGB image data.

Preferably, the preferred embodiment of the present invention described below is implemented as a computer program product under Linux, for example SuSe Linux 10 in the C ++ language, for example with the GCC 4.0 compiler. This is preferably realized with the help of the QT library, in particular with QT 3.3.

Advantageously, the Linux standard library Video4Linux (as v4l referred to) are used. This driver preferably reads the images in step S1, as shown in FIG. 11a, from a frame grabber card as an exemplary component of the image recording device, to which the camera is connected as an exemplary, further component of the image recording device. The synchronization of the software with the camera is preferably also done via this driver, which finally gives a complete RGB image from the camera.

Since the above-described preferred method is based on infrared diodes as exemplary auxiliary points or measuring points, it is not necessary to work with RGB images all the time. Therefore, in a first step of this step S2 (shown in FIG. 11a), in particular in step S20, a gray scale image is obtained from the RGB image

K-Ptxel, ⁺ (-Tpixel, ⁺ "pixels,, -., -.> Gray _p . = '-' ^■ - (79,

determined, in particular calculated, which advantageously only ^■ * ^■ occupies the memory.

Based on a threshold value, the black and white image is then calculated by deciding for each pixel i:

white _Puel > Threshold (80)

black _Pιxel ≤ threshold (81]

This value is preferably set or chosen so high that advantageously as many noise pixels disappear. On the other hand, this value must also be so low that the points of the ramp, i. the auxiliary points or the measuring points are still clearly recognizable.

In the next step S21, a recursive white point search function is applied to the black and white image. In particular, the picture becomes preferably gone through pixel by pixel and checked whether the current pixel

has been studied before or not. If it has already been examined, the next step is taken, otherwise it is checked if this pixel

is black or white. If this is black, the next pixel is transferred. However, if this is white, it will be unclear whether there is also a white pixel in the adjacent environment (i.e., the 8 closest neighbors). If this is the case, this function calls itself again.

In this recursion, the contiguous white pixels are preferably counted. Since the infrared diodes to be detected reach a maximum and a minimum size, two parameters enter into the function here. With the aid of this it is then decided whether the detected area is an infrared diode, i. an auxiliary point or a measuring point is or is not:

min ≤ size ≤ max. (82 '

Each diode is assigned a nominal number of N pixels, which nominal number of pixels may depend on the following constraints:

the geometry of the diodes and / or

- the distance of the diodes from the camera and / or

the focal length of the lens of the camera and / or

the pixel resolution of the diode and / or

Size of the ramp and / or

Geometry of the ramp. In particular, the pixel resolution depends on all these boundary conditions.

Preferably, the value N is determined in advance or determined. This can be done by means of a

Calibration routine or method done. For example, before the first use of the method according to the invention or the device according to the invention, when the diodes and the camera are in the operating position, an image in which the position of the diodes is known can be evaluated. Thus, based on this image, the nominal number N of pixels per diode can be determined. Alternatively or additionally, the nominal number N of pixels per diode can also be determined, in particular calculated, on the basis of geometrical and optical considerations. In this case, N represents a theoretical value.

Since the diodes are at different distances from the camera and the ramp moves during different maneuvers, this value fluctuates. It has been found that based on these variations in conventional geometries of towing vehicles with biaxial trailers, the maximum and minimum sizes of the diodes to be detected are given by:

max ≡ (W - 3) / 2

and

min = (τV - 2) / 3.

In particular:

max = (7V -3) / 2

and

min = (N - 2) / 3.

Consequently, for equation (82):

(N - 2) / 3 ≤ size ≤ (N - 3) / 2. (82 ') Preferably, the value of min is at least about 2, at least about 9, at least about 16, at least about 36, further preferably at least between at least about 16 and at least about 36 pixels, more preferably at least about 4 pixels.

Preferably, the value of max is at least about 36, at least about 49, more preferably between about 30 and about 50 pixels, most preferably at most about 250 pixels.

Preferably, the value of N is between about 25 and about 100, more preferably between about 30 and about 50 pixels, more preferably between about 36 and about 49 pixels. Further preferably, the value of N is at most about 250 pixels, more preferably less than about 150 pixels.

For a given number of N, an algorithm calculates for an amount of pixels, i. searched for contiguous pixels of size "size", where (about) applies,

min ≤ size ≤ mαx, (82)

which can be carried out in the aforementioned first method step.

In other words, the method may further preferably comprise an initial step in which the nominal number N pixels per diode is determined. The nominal number N can in this case be measured in a calibration step or calibration step and / or by means of geometric and / or optical considerations on the basis of

the geometry of the at least one diode and / or

the distance of the at least one diode from the (at least one) camera and / or the focal length of the lens of the (at least one) camera and / or

the pixel resolution of the diode (in the image data) and / or

the size of the ramp and / or

the geometry of the ramp and / or

- The arrangement of the ramp, in particular relative to the (at least one) camera

estimated and / or theoretically determined, in particular calculated. Consequently, due to various geometric arrangements and / or the diode (s) and / or camera (s) used, various real systems or methods may also have different nominal numbers N of pixels. The calibration step may be identical or part of the above-described calibration step.

For all the points that could be considered, the so-called focus is formed. That means:

number of points

X ₁

P _x = ^& (83)

number of points

number of points

Σ 1 = 1 ^y ,

P ^y = - (84)

number of points

These values are considered as possible infrared diode point, i. assumed as a possible auxiliary point or measuring point.

In the subsequent step S22, it is selected whether a so-called "large process" or a so-called "small process" is to be performed. The .large method includes in particular the steps S23, S24, S25. In particular, the small method includes steps S26, S27, S25.

Hereinafter, the large method will be described first. From the center points obtained in step S21, the legs of the ramp are determined, in particular calculated. A leg preferably has the following properties:

it consists of four points

all four points are in line

the legs forming a ramp have exactly one point in common

- The ramp consists of exactly two such legs

one of the two legs, which should form the ramp, must not fall below a minimum length.

The leg recognition preferably proceeds as follows. The found priorities are processed one after the other. For each point the nearest point is searched. In turn, these points are searched for nearest points. It is then checked whether the points lie on a straight line.

This is done by comparing the distance from the first to the third point with the distance from the first to the second to the third point. If both are equal, the three points must lie on a straight line. Furthermore, it is checked whether the ratios of the lengths between the individual points, the predetermined ratios of the geometry data correspond.

If three points are found in this way, an attempt is made to find a fourth point in the same way. If this step is successful, it is assumed that these four points form a leg. From the legs thus obtained, an attempt is made to assemble them into a ramp. Since the thighs are always determined in such a way that first the closely lying points are found and then only the points which are a little farther apart, it follows that the last found point of a leg is the center of the ramp. So if two legs are found that have the same last point, then these two give rise to a ramp.

The validity of this ramp is checked by measuring the length of the legs. In particular, it is checked in step S24 whether the detected leg is actually a leg. To do this, check the length of the thigh. The further a leg is turned away from the camera as a preferred image pickup device, the smaller appears on the camera image than preferred image data. Conversely, the thigh turned towards the camera appears larger than it actually is. It follows that at least one of the two legs must not fall below a certain minimum length.

If both legs of a detected ramp are less than the minimum length, the detected ramp is discarded and searched for another, i. the method described above at least partially performed again. In particular, the aforementioned method can be carried out repeatedly or partially completely until one or more ramps are detected or determined.

Finally, when a ramp has been detected, the coordinates of the right (V), left (R) and middle (S) infrared diodes viewed from the camera are preferably returned in Cartesian coordinates, cylindrical coordinates or spherical coordinates (step S25). Subsequently, the steps S3 and S4 can be performed. Steps S20 to S25 are preferably substeps of step S2.

The above-described exemplary method for detecting the ramp, ie the "large method", assumes that from image to image great changes in the position of the points result .Before points and ramp for each image preferably completely restarted. This method therefore requires a large part of the computing capacity of the processor and is unsuitable for less powerful systems.

Therefore, the preferred embodiment of the method according to the invention described below, i. the "small process" makes use of the fact that in reality the movement of the trailer from picture to picture, due to the high frequency of the camera, is only very small, so the small method is carried out in step S22, if before the large method has already been carried out and / or measuring points, in particular the three measuring points R, S, V. In this case, the coordinates of the points R, S and V recognized by the large method are used and re-centered within a certain radius around them If all the measuring points, ie according to this embodiment, the three points R, S, V are found again in this way (step S27), then the coordinates of these (new) points are determined as the coordinates for R, S, V is returned (step S25). If it is determined in step S27 that one or more of the points are not found, the large method is preferably performed, i.e. a the shift was too big to be able to say clearly that these are the old points. Consequently, after step S27, i. the failed detection of all points, step S23 is executed.

For the conversion of the found R, S, V points into the relative angles from the point of view of the camera, the pixels must preferably first be converted at intervals to the perpendicular of the camera. This is done according to a preferred embodiment with the aid of a table which, for example, previously, in an initial step, in which the preferred device is adapted to the team, is created.

Here it is first determined at which pixel the diode is in the zero position (θ _n = <9 ₂₃ = 0) first 10 cm, then 20 cm, etc. away from the solder of the camera.

The same applies to the other side, where you work with negated values. The pixel values are then interpolated and converted, for example, into centimeters. This is shown by way of example in FIG. After the pixel value has been converted to the corresponding distance to the solder, the angle of the point can be calculated using the tan function of the recording:

win = arctan (^^) (85]

L ₂ - PK ₂ + CK ₁

as shown by way of example in FIG. 12.

With the aid of the exemplary laboratory setup, as it is e.g. As shown in Figures 5a, 5b, various measurements were taken to examine the correctness of the previously described algorithm. On the basis of which test values were created, as shown in Table 2.

It should be noted that manual predetermining of an angle is very difficult. In particular, an angle can not be set exactly to the second decimal place substantially. It is even difficult to specify the first decimal place exactly. Therefore, the values that have a deviation of 0.5 °, ie ± 0.25 ° from the actual value, are nevertheless to be assumed to be correct values.

For a better overview, in Table 3 and Table 4, the individual deviations of the angles θ _n and θ ₂₃ are plotted against each other. In this case, the darker the field is represented, the greater the deviation of the individual angle.

LIST OF REFERENCE NUMBERS

O object

C device

10 team

12 towing vehicle

14 followers

16 camera

18 observation object / ramp

20 towbar / steering axle

22 longitudinal axis of the towing vehicle / vehicle longitudinal axis

24 infrared diodes

26 measuring points

28 rear axle

30 IR diodes

32 camera

34 Trailer longitudinal axis

36 pivot point

38 pivot point

40 Coupling / coupling / coupling point K1

42 Coupling / coupling / coupling point K2

Claims

Applicant: University of Koblenz-Landau "Method for determining a position, device and computer program product" Our mark: K 3466WO - ds / hy / edPatentansprüche

A method of determining a position of a biaxial trailer (14) having at least one movable axle relative to a towing vehicle (12), comprising the steps of:

Generating image data of at least one observation element (18) with at least one associated measuring point (R, S, V) on the basis of an image recording device (16, 32),

Detection of at least three predetermined measuring points (R, S, V) in the image data,

Determining the position of each measuring point (R, S, V) relative to the image recording device (16, 32), in reference coordinates of a predetermined coordinate system, in particular in spherical coordinates or cylindrical coordinates,

Determining an absolute position of each measuring point (R, S, V) with respect to the image recording device (16, 32) on the basis of the reference coordinates,

Determine

- An arrangement angle Θi, ₂ between a vehicle longitudinal axis (22) of the towing vehicle (12) and a steering axis (20) of the trailer (14) and

- An arrangement angle Θ ₂ , ₃ between the steering axis (20) and a trailer longitudinal axis (34) of the trailer (14) based on the absolute positions of the measuring points (R, S, V) relative to the image pickup device (16, 32).

2. The method according to any one of the preceding claims, wherein

the arrangement angle G ₁ , ₂ between the vehicle longitudinal axis (22) of the towing vehicle (12) and the steering axis (20) of the trailer (14) as a function of a position angle G _{1 of} the vehicle longitudinal axis (22) of the towing vehicle (12) and a position angle Q ₂ Steering axis (20) of the trailer (14) based on the absolute positions of the measuring points (R, S, V) relative to the image pickup device (16, 32) is determined and

the arrangement angle Θ ₂ , ₃ between the steering axis (20) of the trailer (14) and the trailer longitudinal axis (34) of the trailer (14) as a function of the position angle Q _{2 of} the steering axle (20) of the trailer (14) and a position angle Θ ₃ of Trailer longitudinal axis (34) of the trailer (14) based on the absolute positions of the measuring points (R, S, V) relative to the

Image recording device (16, 32) is determined,

where:

and

3. The method of claim 1 or 2, wherein exactly three predetermined measuring points (R, S, V) are detected in the image data and applies to the arrangement angles Θi, ₂ and Θ ₂ , ₃ :

<9 ₁₂ = arctane

and (0- XJC) ^• (?,> - 3V ₂ ) - (* /> -x _Cl ) - (0-yyr.) Θ ₂₃ = arctan | φ-x) - (x _P -x _K ) - (y _p -y) - (Qy)

in which

(xκ ₂ , yκ ₂ ) the coordinates of a point K ₂ and (x _P , y>) are the coordinates of a point P in a coordinate system, in particular in a Cartesian coordinate system of a predetermined point K ₁ as a center,

K _{1 is} the intersection of the steering axle with the trailer longitudinal axis of the trailer,

and

K ₂ by rotation of a connecting vector SR of the points S, R by an angle

- /? = (-arctane (-)) h

and by stretching with a factor h + PK ₂

R _s = D (R, - arctanl-), S)

h + PK,

K ₂ = S + (SR _S ) m

results and for P applies:

P = S + (SR _s ) (- \,

U.

where m is the length of the vector SR and b, h is given dimensions of the at least one observation element.

,

4. The method according to any one of the preceding claims, wherein based on the image recording device (16, 32) colored image data, in particular RGB image data are generated.

5. The method according to any one of the preceding claims, wherein the color image data are converted into black and white image data.

6. The method according to any one of the preceding claims, wherein for detecting the measuring points, all the pixels of the black and white image data are checked and based on a minimum number and a maximum number of adjacent white pixels at least three auxiliary points are identified.

7. The method according to any one of the preceding claims 1 to 5, wherein for detecting the measuring points all pixels of the black and white image data are checked and based on a minimum number and a maximum number of adjacent black pixels at least three auxiliary points are identified.

8. The method of claim 6 or 7, wherein the minimum number and the maximum number of adjacent pixels depends on a nominal number N of pixels per measurement point.

9. The method according to claim 8, wherein the following applies to the minimum number min and the maximum number max:

max - = (N - 3) / 2

and

min = (.V - 2) / 3.

10. The method according to claim 6, wherein the center of gravity is calculated for the auxiliary points detected on the basis of the adjacent white and / or adjacent black pixels and applies to the position P _x , P _{y of} the respective center of gravity: AnzahlPutύae

Σ

P = 1 = 1 *,

number of points

Anza MhlPPuunnkkttee

Σ ^y ,

P = ^& . y number of points

11. The method according to any one of the preceding claims, wherein at a plurality of focal points adjacent centers of gravity are detected and the adjacent centers of gravity are identified as lying on a track, when the sum of the distances of the individual centers of gravity is equal to the largest distance between two of the centers.

12. The method of claim 11, wherein at least one leg is detected when four adjacent centroids lie on a common path.

13. The method of claim 12, wherein a leg is identified when the maximum distance of two centroids on one of the distances is greater than a predetermined minimum length.

14. The method of claim 12 or 13, wherein a ramp (18) is identified when two legs are identified, the two legs have a common point and the common point is an endpoint of the two routes.

15. The method of claim 13, wherein the at least three measuring points (R, S, V) are determined based on the position of the ramp (18), wherein one of the measuring points

(S) is associated with a center of the ramp (18) and the two other measuring points (R, V) are assigned to outer ends of the ramp (18).

16. The method of claim 15, wherein the ramp (18) is detected in an initial step and after detection of the ramp (18) a change in the position of the at least one auxiliary point is detected and based on this detection again determines the position of the ramp (18) becomes.

17. The method of claim 16, wherein the change in the position of the at least one auxiliary point is detected by checking the pixels in an environment of the at least one auxiliary point and identifying the position of the center of gravity of the at least one auxiliary point.

18. The method according to claim 17, wherein, if not all measuring points (R, S, V) are identified, the method according to one of claims 8 to 15 is carried out.

19. The method according to any one of claims 8 to 18, wherein a step or steps are repeatedly performed.

20. The method according to any one of the preceding claims, wherein in an initial step for each auxiliary point in a calibration position of

Image pickup device (16, 32) relative to the observation element (18) the position of each auxiliary point in pixels of the image data is determined, the actual position of each auxiliary point by an arrangement of a corresponding infrared diode (24, 30) on the observation element (18) predetermined is and the position of each infrared diode (24, 30) relative to the image pickup device (16, 32) is predetermined in the calibration position.

21. A method according to claim 20, wherein the position of each measuring point (R, S, V) relative to the auxiliary points in the image data is determined and by calibrating the image data the one position of each measuring point (R, S, V) with respect to the image pickup device (16, 32) is determined.

22. The method of claim 21, wherein the angular displacement of each measuring point (R ₁ S, V) relative to the image pickup device (16, 32) is given by

win = arctan (^^).

L ₁ -PK ₂ ^ CK _x

23. The method according to any one of the preceding claims, wherein on the basis of the arrangement angle O ₁₁₂ between the vehicle longitudinal axis (22) of the towing vehicle (12) and the steering axis (20) of the trailer (14) and on the basis of the arrangement angle Θ ₂ , ₃ between the steering axis (20 ) and the trailer longitudinal axis (34) of the trailer (14) a trajectory of the team consisting of towing vehicle (12) and trailer (14) in particular automatically determined.

24. An apparatus for determining a position of a trailer (14) with at least one movable wheel axle relative to a towing vehicle (12)

at least one observation element (18), to which at least one measuring point (R, S, V) is assigned,

an image recording device (16, 32) which is designed to generate image data of the at least one observation element (18),

a detection device which is designed to detect at least three predetermined measurement points (R, S, V) in the image data,

a determining device which is designed

to determine the position of each measuring point (R, S, V) relative to the image recording device (16, 32), in reference coordinates, in particular in spherical or cylindrical coordinates,

to determine an absolute position of each measuring point (R, S, V) relative to the image recording device (16, 32) on the basis of the reference coordinates,

- An arrangement angle Θi, ₂ between a vehicle longitudinal axis (22) of the towing vehicle (12) and a steering axis (20) of the trailer (14) based on the absolute positions of the measuring points (R, S, V) relative to the image pickup device (16, 32) to determine and

- An arrangement angle Θ ₂ , ₃ between the steering axle (20) and a trailer longitudinal axis (34) of the trailer (14) based on the absolute Positions of the measuring points (R, S, V) relative to the image pickup device (16, 32) to determine.

25. The apparatus of claim 24, wherein the observation member (18) comprises a plurality of display means (24, 30).

26. The apparatus of claim 25, wherein the display means (24, 30) are infrared diodes, in particular three, preferably seven infrared diodes.

27. Device according to one of claims 24 to 26, wherein the image pickup device (16, 32) comprises a digital camera and / or an analog camera.

A computer program product which, when loaded into the memory of a computer and executed by the computer, performs a method according to any one of claims 1 to 23.