Method, control unit and vehicle for obtaining object information related to an object in the surrounding environment of a vehicle

Abstract

The invention relates to a method for obtaining object information of an object (O) in the surroundings (U) of a vehicle (1), wherein the vehicle has a camera (4), and the method comprises at least the following steps: - photographing the surroundings (U) from a first viewpoint (SP1) with the camera (4); - changing the viewpoint (SP) of the camera (4); - photographing the surroundings (U) from a second viewpoint (SP2) with the camera (4); - obtaining object information related to the object (O) by: - ​​selecting at least one first pixel in a first image and at least one second pixel in a second image, wherein the first pixel and the second pixel are selected so that they are assigned to the same object point of the object (O), and - determining the object coordinates of the assigned object points by triangulation (T). According to the invention, it is provided that the change of the viewpoint (SP) of the camera (4) is caused by actuating an active actuator system (8) in the vehicle (1), wherein the active actuator system (8) causes at least one camera (4) to adjust an adjustment travel (W) without changing the driving state (Z) of the vehicle (1).

Description

Technical Field

[0001] The invention relates to a method for acquiring object information about an object in the surroundings of a vehicle, and a control unit and a vehicle for executing the method. Background Art

[0002] It is known from the prior art how the 3D structure of a scene can be determined using a single camera by means of a forward or backward movement of a vehicle equipped with the camera (so-called structure from motion (SfM)) using photogrammetric methods. It is also known from the prior art that the evaluation of the odometry data of the vehicle can support the determination of a baseline between two viewpoints of the camera, which baseline is known or is to be estimated for the triangulation used for depth determination.

[0003] US2018 / 0204072A1 also provides that a camera is fixed to a trailer of the vehicle combination. In addition, a driving dynamics sensor is provided, which outputs odometer data related to the vehicle movement, such as the vehicle speed. The camera data output by the camera is compared with the odometer data, wherein the odometer data is used to compensate for the vehicle movement when processing the camera data to create an image. In this case, the camera data of different cameras can also be combined.

[0004] DE 10 2005 009 814 B4 provides that the camera data are processed together with the odometer data output by the wheel speed sensor in order to determine the yaw rate. DE 60 009000 T2 also provides for image processing taking into account the odometer data of the vehicle in order to support the driver when parking. In DE 10 2015 105248 A1, images from a first camera and images from a second camera are processed together with the odometer data, wherein the cameras can be arranged on the trailer and the tractor of a multi-section vehicle. The images recorded by different cameras and output in the form of camera data are combined together. As a result, a combined image of the surrounding environment is generated, wherein, for example, the bending angle is also taken into account when driving around a corner, which characterizes the viewpoint of the cameras relative to each other. A bird's-eye view can be superimposed on the entire multi-section vehicle in order to display the surrounding environment around the vehicle, so that, for example, parking assistance can be implemented.

[0005] In WO 2016 / 164118, an omnidirectional camera is provided, which captures object points of objects in the surroundings of the vehicle and outputs camera data in dependence thereon. The camera data is processed by means of a control device in the vehicle, including the recorded odometer data, wherein the odometer data, for example from a wheel speed sensor, a positioning sensor or a steering angle sensor, is received via the data bus of the vehicle. The object points of interest in the surroundings of the vehicle are identified by the camera, and the distance to the object assigned to the captured object point is determined by the control device based on the odometer data. For this purpose, multiple images are recorded via a camera, wherein the images are recorded from different viewpoints with overlapping fields of view. By tracking the object points, the depth information of the scene can be estimated by means of triangulation and bundle adjustment. The camera data is also presented to the driver on a display in the form of an image. The image and the determined distance are used to make it easier to maneuver the passenger car as a tractor onto the trailer in order to connect them. Although other objects such as the ground and pedestrians can also be identified, this requires the vehicle to be in full motion, as different viewpoints can only be set for the camera in this way.

[0006] This has the disadvantage that, for example, a person lying on the ground cannot be captured when stationary or at very slow, indistinguishable speeds. Other objects in the vehicle's surroundings cannot be detected by the "motion inference structure" when the vehicle is not moving or is moving only very slowly. Therefore, it is not possible to spatially capture the surroundings of the vehicle or objects with only one camera when stationary, so that neither automatic classification of objects nor distance determination is possible.

[0007] However, safety requirements for unmanned land vehicles require that people on the ground be detected, for example according to ISO 3691-4, and therefore must be identified before departure, i.e. when stationary or at very low speeds. Even at traffic lights or in parking situations, measuring distances or classifying objects or acquiring other spatial object information about the objects is not possible with known systems. Summary of the invention

[0008] The object of the present invention is therefore to specify a method for determining object information of an object, by means of which a spatial observation of the vehicle surroundings can be performed with only one camera even when the vehicle is stationary or at very low speeds. A control unit and a vehicle are also specified.

[0009] This object is achieved by a method, a control unit and a vehicle for determining object information about an object in the surroundings of a vehicle according to the following. Preferred developments are given in the dependent claims.

[0010] Therefore, a generic method for determining object information about objects in the surroundings of a vehicle is provided, wherein the vehicle has at least one camera, the method comprising at least the following steps:

[0011] - recording the surroundings from a first viewpoint with at least one camera and creating a first image consisting of first pixels based thereon;

[0012] -Change the viewpoint of at least one camera;

[0013] - photographing the surroundings from a second viewpoint with at least one camera and creating a second image consisting of second pixels based thereon, wherein the first viewpoint is different from the second viewpoint due to a change in the viewpoint of the camera at intervals;

[0014] - Obtaining object information related to the objects in the photographed surrounding environment by:

[0015] - selecting at least one first pixel in the first image and at least one second pixel in the second image, wherein the first pixel and the second pixel are selected such that they are assigned to the same object point of an object in the captured surroundings, and

[0016] Assuming a baseline length between the two viewpoints of the camera, determining the object coordinates of the associated object point by triangulation from the first image coordinates of at least one first pixel and the second image coordinates of at least one second pixel.

[0017] According to the invention, it is provided that the viewpoint of at least one camera is changed from a first viewpoint to a second viewpoint by controlling an active actuator system in the vehicle, wherein the active actuator system causes the at least one camera to adjust the adjustment path without changing the driving state of the vehicle. The driving state is understood here to be the movement state of the vehicle as a whole, that is, for example, stationary or driving at a certain vehicle speed. The active actuator system does not change this movement, so the adjustment path is not related to the driving movement of the vehicle, so the active actuator system is different from the drive system or brake system that directly affects the overall movement state of the vehicle.

[0018] Advantageously, the method according to the invention thus makes it possible to determine depth information or object information with only one camera, independently of the driving state of the vehicle. As a result, even when the vehicle is stationary or at such a low speed that no reliable conclusions can be drawn about the movement between two viewpoints from the odometer data, depth information or an approximate 3D position or object coordinates of the respective object point can be determined by triangulation. For this purpose, only the active actuator system has to be driven in a controlled manner, which is independent of the vehicle movement. Under the assumption of a certain baseline length (even if this baseline length cannot be known exactly), at least incomplete depth information, such as the object shape or object contour, can first be obtained by triangulation.

[0019] The actuator system for adjusting the camera between viewpoints is therefore not limited to a stationary or low-speed vehicle. Additional adjustment of the adjustment path can thus also be achieved by the active actuator system during driving. This allows object information to be acquired more flexibly or in different driving states or driving situations.

[0020] Preferably, provision is also made for the baseline length between the two viewpoints to be determined from odometer data of the vehicle, wherein the odometer data are generated at least as a function of the adjustment path and characterize the adjustment of the camera between the two viewpoints. As a result, the baseline length is no longer merely assumed, but is determined in dependence on the actively set adjustment path, thereby making the triangulation more precise. When determining the depth information or object coordinates, if the adjustment path is taken into account in addition to the usual odometer data related to the driving state of the vehicle, this adjustment path is also used in addition to the vehicle movement (if any). As a result, object information or depth information is determined more accurately and more flexibly or in different driving situations.

[0021] Preferably, provision is also made for actuating a camera adjustment system as an active actuator system, which has an adjustment motor and / or a pneumatic cylinder and / or a hydraulic cylinder and / or an electric servo cylinder, wherein at least one camera is directly fastened to the camera adjustment system so that when the camera adjustment system is actuated, the adjustment travel of the at least one camera is adjusted in order to change the viewpoint of the at least one camera. Thus, according to one embodiment, the camera can be adjusted directly without the vehicle or a component of the vehicle having to be moved concomitantly, wherein the camera adjustment system is mounted and oriented accordingly on the vehicle.

[0022] It is also preferably provided that an active air spring system or a chassis adjustment system having air springs (ECAS) as an active actuator system is controlled, wherein the vehicle body is adjusted by an adjustment stroke in terms of its height by controlling the active air spring system or the chassis adjustment system, thereby indirectly adjusting the adjustment stroke of at least one camera fastened to the vehicle body to change the viewpoint of at least one camera.

[0023] Thus, it is advantageously possible to use an actuator system that is already present in the vehicle and can therefore fulfill a dual function, i.e., to raise and lower the vehicle body, for example for air suspension, stabilization (roll, tilt), etc., and also to enable the camera to be adjusted in a targeted manner to different viewpoints. For this purpose, the air suspension system or the chassis control system is activated only in the corresponding situation, which is possible in all driving situations, in particular also when stationary. In this case, the camera can be freely mounted on the vehicle body in order to move it along with it.

[0024] It is also preferably provided that a component adjustment system is controlled as an active actuator system, wherein, by controlling the component adjustment system, a component of the vehicle, such as a cockpit and / or an aerodynamic component, has its adjustment stroke adjusted, thereby causing at least one camera fastened to the component to indirectly adjust its adjustment stroke in order to change the viewpoint of at least one camera.

[0025] Thus, it is possible to use actuator systems that do not raise and lower or adjust the entire vehicle body, but only individual components or parts. Such adjustment systems are already present in certain vehicles, so that no retrofitting is necessary. The camera then only needs to be fastened to the component.

[0026] The active actuator systems mentioned can be provided individually or in combination with one another, for example in order to increase the variability and to allow a combined adjustment with a possibly extended adjustment path.

[0027] It is also preferably provided that when the active actuator system is activated, the vehicle is in a driving state in which the vehicle has a vehicle speed below a limit speed or the vehicle is stationary. The method is therefore also advantageously suitable for situations in which a conventional odometer cannot be used to reliably determine depth information of an object by means of the provided odometer data (e.g. wheel speed signals, bend angles, steering angles, etc.) because this is inaccurate. In contrast, the method is independent of the driving motion of the vehicle due to the active adjustment of the camera according to the invention.

[0028] It is also preferably provided that object coordinates or object information are determined from at least two images by triangulation for a plurality of object points, and that an object contour and / or an object shape are determined from a plurality of object points. In this case, objects can preferably be divided into object classes based on the object contour and / or the object shape. This makes it possible to recognize and classify objects, in particular stationary objects such as persons, in a simple manner even when they are stationary.

[0029] According to another embodiment, a plurality of cameras are provided and object information about the object is determined independently of one another by adjusting the adjustment path from the difference or the baseline length with the aid of each camera according to the described method. Depth information or object information can thereby preferably be determined from a plurality of sources, thereby increasing reliability. Furthermore, a plausibility check of the object information determined by the plurality of cameras can thereby be made possible.

[0030] Preferably, it is also provided that more than two images are recorded at different viewpoints and pixels assigned to the same object point of an object in the captured surroundings are selected from each recorded image, wherein the object coordinates of the assigned object point are determined by triangulation from the image coordinates of the selected pixels, assuming a baseline length between the respective viewpoints of the cameras. It is thus also possible to track the respective object or the respective object point over a longer period of time, so that depth information or the respective object information can be determined more accurately or more robustly by bundle adjustment, if necessary. In this case, a plurality of pixels can also be combined into one or more feature points, and the temporal correspondence of the feature point or points between the respective images can be determined by triangulation.

[0031] Furthermore, it can be provided that the object information obtained by adjusting the control path of the camera by the active actuator system or by the control path as odometer data is subjected to a plausibility check with the object information obtained from the following odometer data of the vehicle, which are selected from the group consisting of: wheel speed signal and / or vehicle speed and / or steering angle and / or bending angle. Thus, the object information about the object obtained by different movements of the camera can be compared. For example, in the case of passive wheel speed sensors, if the vehicle speed is very low, for example, the reliability of the depth information obtained from the wheel speed can no longer be ensured, so that in addition to the plausibility check, the camera also adjusts the control path by the active actuator system and can thus obtain depth information.

[0032] According to the invention, a control unit and a vehicle having such a control unit for executing the described method are also provided, wherein the vehicle is single-section or multi-section and at least one camera is arranged on a trailer of the vehicle and / or on a tractor (with or without a trailer). BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The present invention is described in detail below based on the embodiments.

[0034] Figure 1 The driving condition of a multi-section vehicle is shown;

[0035] Figure 1a A detailed view of a multi-section vehicle is shown;

[0036] Figure 2a shows an image recorded by a camera;

[0037] Figure 2b showing recording of object points from different viewpoints using a camera; and

[0038] Figure 3 A flow chart of a method according to the invention is shown. DETAILED DESCRIPTION

[0039] Figure 1 The schematic diagram of a multi-section vehicle 1 consisting of a tractor 2 and a trailer 3 is shown, wherein, according to the illustrated embodiment, a camera 4 with a recording area E is arranged on each of the two vehicle parts 2, 3. A tractor camera 42 with a tractor recording area E2 is arranged on the tractor 2, and a trailer camera 43 with a trailer recording area E3 is arranged on the trailer 3. The cameras 4, 42, 43 output camera data KD, KD2, KD3, respectively.

[0040] like Figure 1 As shown, the vehicle 1 can be implemented as a multi-section type, for example as a truck train with a truck and a drawbar trailer or a turntable trailer or as a semitrailer train with a tractor and a semitrailer. However, in principle, the vehicle 1 can also be a single-section type, such as Figure 1a The orientation of the camera 4 is selected depending on the respective application.

[0041] The respective camera data KD, KD2, KD3 are generated depending on the surrounding environment U around the vehicle 1, and the respective recording areas E, E2, E3 point to the surrounding environment. From the camera data KD, KD2, KD3, an image B consisting of pixels BPi with image coordinates xB, yB can be created (see Figure 2a ), wherein each pixel BPi is assigned an object point PPi in the surroundings U (see Figure 2b ). The object point PPi belongs to an object O which is located in the surroundings U and can be assigned certain absolute object coordinates xO, yO, zO in space. Depending on the viewpoint SP of the respective camera 4, 42, 43, the object point PPi of the object O is imaged at different pixels BPi in the image B or has different image coordinates xB, yB in the image B.

[0042] The camera data KD, KD2, KD3 of the respective cameras 4, 42, 43 are transmitted to a control unit 5, which is designed to determine object information OI by means of triangulation T generally known to a person skilled in the art, based on the camera data KD, KD2, KD3 and selected odometer data DD of the vehicle 1, which are related to the current driving situation of the vehicle 1 or of the sub-vehicles 2, 3 and therefore also characterize the movement of the camera 4. The object information OI in this case specifies in particular the spatial characteristics of the respective object O in the surroundings U recorded by the cameras 4, 42, 43.

[0043] For example, object information OI may be considered

[0044] - the absolute object coordinates xO, yO, zO of the object O in space (world coordinates), and / or

[0045] - a distance A between a reference point PB, for example the rear side 1a of the vehicle 1 (in the case of a single-section vehicle 1) or of the trailer 3 (in the case of a multi-section vehicle 1) or between the current viewpoint SP of the respective camera 4, 42, 43 and the identified object O or an object point PPi on the object O, or a variable associated therewith, and / or

[0046] - object shapes OF or object contours OC, which are divided, for example, into n different object classes OKn, and / or

[0047] - Object dynamics OD, i.e. the temporal motion of the captured object O in space.

[0048] The object information OI is hereby determined using the Structure from Motion (SfM) method, in which in substeps ST1, ST2, ST3 (see Figure 3 ), the object O is recorded by the camera 4 from at least two different viewpoints SP1, SP2 (see Figure 2b In a further step ST4, depth information related to the object O or the respective object information OI can be obtained by triangulation T. Figure 2b Here, image coordinates xB, yB are determined for at least one first pixel BP1i in the first image B1 and at least one second pixel BP2i in the second image B2, which are each assigned to the same object point PPi (ST4.1).

[0049] In order to simplify the process, a certain number of pixels BP1i, BP2i in the respective images B1, B2 can be combined into feature points MP1, MP2 (see Figure 2a ), wherein the pixels BP1i, BP2i to be combined are selected so that the respective feature points MP1, MP2 are assigned to certain, clearly localizable features on the object O (see Figure 2b ). The feature M may be, for example, a corner ME or an edge MK on the object O, which may be extracted from the entire image B1, B2 respectively and their pixels BP1i, BP2i may be combined into feature points MP1, MP2.

[0050] Approximately, at least the object shape OF or the object contour OC can be estimated by triangulation T from the image coordinates xB, yB of the individual pixels BP1i, BP2i or feature points MP1, MP2, which are assigned to the same object point PPi or the same feature M in at least two images B1, B2. To this end, the image coordinates xB, yB of multiple pixels BP1i, BP2i or multiple feature points MP1, MP2 can be triangulated T to obtain spatial object coordinates xO, yO, zO that are not necessarily located on the object O in space.

[0051] Without knowing the exact baseline length L, i.e. the distance between the different viewpoints SP1, SP2 of the camera 4, the object coordinates xO, yO, zO are obtained in unscaled form by the triangulation T. Therefore, only the unscaled object shape OF or the object contour OC can be obtained from the object coordinates xO, yO, zO obtained in this way, but this is sufficient to determine the shape or contour. In this case, an arbitrary baseline length L can be assumed for the triangulation T.

[0052] In order to be able to determine the above-mentioned additional object information OI by triangulation T, the actual baseline length L is additionally taken into account. Figure 2b If the relative positioning between the different viewpoints SP1, SP2 of the camera 4 at which the two images B1, B2 are recorded and thus the baseline length L is known or acquired, the absolute object coordinates xO, yO, zO (world coordinates) of the object O or of the object point PPi or of the feature M can also be acquired by triangulation T. From this, the distance A between the reference point PB and the identified object O or the object point PPi on the object O can also be acquired, wherein the coordinates of the reference point PB in world coordinates are derived directly from geometric observations.

[0053] In this way, if the exact object coordinates xO, yO, zO of a plurality of object points PPi or features M are known, a scaled object contour OC or a scaled object shape PF compared to the above situation can be estimated by the control device 5. From the object contour OC, a classification of the object O into a specific object class OKn can also be achieved. In this case, if the object points PPi are observed in a time-resolved manner, the object dynamics OD, for example the direction of movement R of the object point PPi and / or the object velocity vO, can also be taken into account.

[0054] For example, objects O identified as people may be classified into the first object class OK1. Objects O identified as stationary, such as signs, loading ramps, houses, etc. may be classified into the second object class OK2. Objects O identified as moving, such as other vehicles, may be classified into the third object class OK3.

[0055] In order to determine the determined object information OI more precisely, it can additionally be provided that more than two images B1 , B2 are recorded and evaluated by triangulation T as described above and / or that a bundle adjustment BA is additionally carried out.

[0056] As already described, for the SfM method, the object O is observed by the camera 4 from at least two different viewpoints SP1, SP2, such as Figure 2b For this purpose, in substep ST2, the camera 4 is brought to different viewpoints SP1, SP2 in a controlled manner and, in the case of scaling, it is determined from the odometer data DD what baseline length L (ST4, ST4.2) results from this movement between the two viewpoints SP1, SP2. Different methods can be used for this purpose:

[0057] If the entire vehicle 1 is in motion, the movement of the camera 4 is already derived from this. This means that the vehicle 1 is actively moving as a whole, for example by the drive system 7, or passively moving, for example by going downhill. If during this movement at least two images B1, B2 are recorded by the camera 4 within a time offset dt, the baseline length L can be determined with the aid of the odometer data DD, from which the vehicle movement and therefore the camera movement can be derived. The odometer can determine the two viewpoints SP1, SP2 associated with the images B1, B2.

[0058] For example, vehicle 1 (see Figure 1 ) can be used as odometer data DD. From this wheel speed signal, it can be determined, depending on the time offset dt, how far the vehicle 1 or the camera 4 has moved between the viewpoints SP1, SP2, thereby obtaining the baseline length L. However, it is not necessary to use a vehicle odometer, that is, to use motion sensors on the vehicle 1 to evaluate the vehicle movement. In addition or as an alternative, a visual odometer can also be used. In the case of a visual odometer, the camera positioning can be continuously determined from the camera data KD of the camera 4 or the information in the captured images B, B1, B2, as long as, for example, the object coordinates xO, yO, zO of a certain object point PPi are known at least at the beginning. The odometer data DD can therefore also contain a correlation with the camera positioning determined in this way, because the vehicle movement between the two viewpoints SP1, SP2 can be derived from this, or the baseline length L can also be directly derived.

[0059] In order to determine the base length L more precisely by odometer while the vehicle 1 is in motion, further odometer data DD available in the vehicle 1 can be used. For example, the steering angle LW and / or the yaw rate G can be used in order to also take into account the rotational movement of the vehicle 1. In the case of a two- or multi-section vehicle 1, the kink angle KW between the tractor 2 and the trailer 3 can also be taken into account in addition in order to take into account the exact dynamics of the trailer 3, in particular during maneuvering operations.

[0060] If the single-section vehicle 1 or the multi-section vehicle 1 with its vehicle parts 2, 3 is not moving, or the movement within the time offset dt is so small that the odometer data DD is too inaccurate to reliably determine the baseline length L, then in step ST2 the camera 4 can also be moved by the active actuator system 8. The movement of the camera 4 caused by the actuator system 8 is different from the previously observed movement of the vehicle 1, in particular the actuator system 8 only moves the camera 4 or the vehicle section connected to the camera 4. The movement of the vehicle 1 as a whole or the driving state Z of the vehicle 1 is not changed as a result, so that the stationary vehicle 1 continues to remain stationary SS when the actuator system 8 is actively actuated.

[0061] The actuator system 8 is controlled by the control unit 5 via the actuator signal SA. This can occur, for example, when the control unit 5 detects that the odometer data DD, which characterize the movement of the entire vehicle 1, i.e. the wheel speed signal SR and / or the steering angle LW and / or the yaw rate G and / or the camera data KD, are not accurate or detailed enough to determine the baseline length L. This can be the case when it is detected that the vehicle 1 is stationary SS or the vehicle speed v1 is below the speed limit value vt.

[0062] When the actuator system 8 is driven, the camera 4 moves directly or indirectly and is thereby brought to different viewpoints SP1, SP2, so that the surrounding environment U can be imaged in at least two different images B1, B2. Therefore, the SfM method can be implemented as described above. In order to know the baseline length L here, the control unit 5 uses the adjustment stroke W, which is adjusted by the actuator system 8 so that the camera 4 is adjusted between the two viewpoints SP1, SP2. The adjustment stroke W is transmitted to the control unit 5 by the actuator system 8. The control unit 5 can also additionally take into account the adjustment stroke W of the actuator unit 8 in the odometer data DD in order to know the baseline length L.

[0063] Consider different systems in the vehicle 1 as actuator systems 8, which can be Figure 1, the single-section vehicle 1 is shown by way of example and schematically, but can also be used on the sub-vehicles 2, 3 of the multi-section vehicle 1. For example, the camera 4 can be arranged on a camera adjustment system 9, which has one or more adjustment motors 9a or pneumatic cylinders 9b or hydraulic cylinders 9c or electric servo cylinders 9d or similarly acting actuators, wherein the camera adjustment system 9 is fastened to the vehicle 1 so that the recording area E is oriented as desired. In this case, the camera 4 can be brought to different viewpoints SP1, SP2 by adjusting one or more adjustment motors 9a, pneumatic cylinders 9b, hydraulic cylinders 9c, servo cylinders 9d by a certain adjustment stroke W during actuation.

[0064] Another possible solution for the active actuator system 8 is an active air spring system 10 (ECAS, electronically controlled air suspension), which in a single-section vehicle 1 or in a multi-section vehicle 1 in the tractor 2 or in the trailer 3 ensures that the vehicle body 11 can be adjusted in height H, i.e. raised or lowered, in relation to the axles 1b, 2b, 3b of the vehicle 1 or of the tractor 2 or of the trailer 3 via air springs 10a implemented as spring bellows. For this purpose, the pressure in the air springs 10a can be adjusted in a targeted manner. This can be used to achieve an optimal suspension independently of the road conditions or load conditions, to dynamically compensate for changes in the axle load distribution, to avoid shaking or pitching during cornering, or to adjust the height H of the vehicle body 11 when the tractor 2 is connected to the trailer 3 and, for example, on a loading ramp during loading and unloading.

[0065] If the respective cameras 4, 4a, 4b are arranged on the body 11 of the vehicle 1 or of the tractor 2 or of the trailer 3, the active air suspension system 10 can be activated by the control unit 5 to adjust the camera 4, preferably by adjusting the adjustment path W at the height H, so as to position the camera at two different viewpoints SP1, SP2. Since the adjustment path W is known to the active air suspension system 10 and / or can be measured, it can also be transmitted to the control unit 5, so that it can take into account the adjustment path W caused by the active air suspension system 10 in the odometer data DD, so as to determine the baseline length L.

[0066] In this way, when the vehicle 1 is at a standstill SS, the control unit 5 can command the respective camera 4 to adjust via the active air suspension system 10, so that the respective object information OI of at least one object point PPi can also be determined by triangulation using the SfM method. In principle, the control unit 5 can already predetermine the adjustment travel W as a target value, which the active air suspension system 10 is to set by changing the pressure in the air spring 10a. For example, the adjustment travel W (or the baseline length L) can also be disregarded if the adjustment travel W is not measured or cannot be measured, so that the unscaled object shape OF or object contour OC can be determined by triangulation T.

[0067] In addition to the active air spring system 10, any similarly acting active chassis adjustment system 12 can also be used as a further active actuator system 8, which can adjust the height H of the vehicle body 11 and thus can position the camera 4 arranged thereon in a targeted manner at two different viewpoints SP1, SP2. However, the following component adjustment system 13 can also be used as an active actuator system 8: this component adjustment system can only raise or lower the adjustment travel W of the part or component of the vehicle body 11 to which the camera 4 is fastened, such as the driver's cabin 14. Aerodynamic components 15, such as aerodynamic blades or spoilers, can also be considered as further components, on which the camera 4 can be mounted and can be actively adjusted so that the camera 4 can be adjusted in a targeted manner by the adjustment travel W.

[0068] There are therefore a number of possibilities for actively and purposefully positioning the camera 4 at different viewpoints SP1 , SP2 in order to record two images B1 , B2 of the object O and thereby determine respective object information OI (scaled or unscaled) for one or more object points PPi.

[0069] List of reference numerals (part of the specification)

[0070] 1 Vehicle

[0071] 1a Rear side of vehicle 1

[0072] 1b Axle of vehicle 1

[0073] 2 Tractor

[0074] 2b Axle of tractor 2

[0075] 3 Trailer

[0076] 3b Axle of trailer 3

[0077] 4 Cameras

[0078] 42 Tractor Camera

[0079] 43 Trailer Camera

[0080] 5 Control unit

[0081] 6a Active wheel speed sensor

[0082] 6p Passive Wheel Speed ​​Sensor

[0083] 7 Drive system

[0084] 8 Active Actuator Systems

[0085] 9 Camera adjustment system

[0086] 9a Adjust the motor

[0087] 9b Pneumatic Cylinder

[0088] 9c Hydraulic Cylinder

[0089] 9d Electric Servo Cylinder

[0090] 10 Active Air Spring System (ECAS)

[0091] 10a Air spring

[0092] 11 Body

[0093] 12 Chassis adjustment system

[0094] 13 Component adjustment system

[0095] 14 Cab

[0096] 15 Aerodynamic components

[0097] A Spacing

[0098] B-image

[0099] B1 First Image

[0100] B2 Second image

[0101] BA beam adjustment

[0102] BPi Pixel

[0103] BP1i First Pixel

[0104] BP2i Second Pixel

[0105] DD Odometer Data

[0106] dt Time offset

[0107] E Camera's shooting range

[0108] E2 The first capture area of ​​the tractor camera

[0109] E3 Trailer camera's second shooting area

[0110] G Yaw rate

[0111] H Height of the vehicle

[0112] Camera data for KD cameras

[0113] KD2 tractor camera first camera data

[0114] KD3 trailer camera second camera data

[0115] L Baseline length

[0116] LW Steering angle

[0117] M Features

[0118] MP1, MP2 feature points

[0119] ME corner (as a feature)

[0120] MK Edge (as a feature)

[0121] O Object

[0122] OC Object Outline

[0123] OD Object Dynamics

[0124] OF Object Shape

[0125] OI Object Information

[0126] OKn The nth object category

[0127] PB reference point

[0128] PPi Object Point

[0129] R Movement direction

[0130] SA Actuator signal

[0131] SP Camera 4's viewpoint

[0132] SP1 Camera's First View

[0133] SP2 Second viewpoint for camera

[0134] SR wheel speed signal

[0135] SS Still

[0136] T Triangulation

[0137] U The surrounding environment of vehicle 1

[0138] v1 Vehicle speed

[0139] vO Object speed

[0140] vt velocity boundary value

[0141] W Adjustment stroke

[0142] Z Driving status

Claims

1. A method for obtaining object information (OI) of an object (O) in a surrounding environment (U) of a vehicle (1), wherein: The vehicle (1) has at least one single camera (4), and the method comprises at least the following steps: - recording the surroundings (U) from a first viewpoint (SP1) with the at least one single camera (4) and creating therefrom a first image (B1) (ST1) consisting of first pixels (BP1i); - changing the viewpoint (SP) of the at least one single camera (4) (ST2); - photographing the surroundings (U) from a second viewpoint (SP2) with the at least one single camera (4) and creating a second image (B2) consisting of second pixels (BP2i) therefrom (ST3); - Obtaining object information (OI) related to the object (O) in the photographed surrounding environment (U) by the following method (ST4): - selecting at least one first pixel (BP1i) in the first image (B1) and at least one second pixel (BP2i) in the second image (B2), wherein the first pixel (BP1i) and the second pixel (BP2i) are selected so that they belong to the same object point (PPi) of the object (O) in the captured surroundings (U) (ST4.1), and - determining the object coordinates (xO, yO, zO) of the associated object point (PPi) by triangulation (T) from the first image coordinates (xB, yB) of the at least one first pixel (BP1i) and the second image coordinates (xB, yB) of the at least one second pixel (BP2i), assuming a baseline length (L) between the two viewpoints (SP1, SP2) of the single camera (4), The active actuator system (8) in the vehicle (1) is controlled to change the viewpoint (SP) of the at least one single camera (4) from the first viewpoint (SP1) to the second viewpoint (SP2) (ST2), wherein the active actuator system (8) causes the at least one single camera (4) to adjust the adjustment travel (W) without changing the driving state (Z) of the vehicle (1), wherein an active air spring system (10) or a chassis adjustment system (12) having an air spring (10a) is controlled as the active actuator system (8), wherein the vehicle body (11) is adjusted in terms of its height (H) by controlling the active air spring system (10) or the chassis adjustment system (12), thereby causing at least one single camera (4) fastened to the vehicle body (11) to indirectly adjust the adjustment travel (W) to change the viewpoint (SP) of the at least one single camera (4).

2. The method according to claim 1, characterized in that: A baseline length (L) between two viewpoints (SP1, SP2) is obtained from odometer data (DD) of the vehicle (1), wherein the odometer data is generated at least dependent on the adjustment stroke (W), and the odometer data (DD) characterizes the adjustment of the single camera (4) between the two viewpoints (SP1, SP2).

3. The method according to claim 1 or 2, characterized in that: A camera adjustment system (9) is driven as an active actuator system (8), wherein the camera adjustment system has an adjustment motor (9a) and / or a pneumatic cylinder (9b) and / or a hydraulic cylinder (9c) and / or an electric servo cylinder (9d), wherein the at least one single camera (4) is directly fastened to the camera adjustment system (9), so that when the camera adjustment system (9) is driven, the at least one single camera (4) adjusts the adjustment stroke (W) to change the viewpoint (SP) of the at least one single camera (4).

4. The method according to claim 1 or 2, characterized in that: A component adjustment system (13) as an active actuator system (8) is controlled, wherein the component of the vehicle (1) is adjusted by adjusting the adjustment stroke (W) by controlling the component adjustment system (13), so that at least one single camera (4) fastened to the component indirectly adjusts the adjustment stroke (W) to change the viewpoint (SP) of the at least one single camera (4).

5. The method according to claim 4, characterized in that The components of the vehicle (1) are a cab (14) and / or an aerodynamic component (15).

6. The method according to claim 1 or 2, characterized in that: When the active actuator system (8) is activated, the vehicle (1) is in a driving state (Z) in which the vehicle (1) has a vehicle speed (v1) below a limit speed (vt) or the vehicle is at rest (SS).

7. The method according to claim 1 or 2, characterized in that: Object coordinates (xO, yO, zO) are determined for a plurality of object points (PPi) by triangulation (T) from at least two images (B1, B2), and an object contour (OC) and / or an object shape (OF) are determined from the plurality of object points (PPi).

8. The method according to claim 7, characterized in that The object (O) is classified into object classes (OKi) based on the object contour (OC) and / or the object shape (OF).

9. The method according to claim 1 or 2, characterized in that: A plurality of individual cameras (4) are provided and object information (OI) related to an object (O) is acquired independently of one another by adjusting the adjustment path (W) with the aid of each individual camera (4).

10. The method according to claim 9, characterized in that A plausibility check is performed on the object information (OI) acquired by the plurality of single cameras (4).

11. The method according to claim 1 or 2, characterized in that: More than two images (B) are recorded at different viewpoints (SP) and pixels (BPi) assigned to the same object point (PPi) of an object (O) in the captured surroundings (U) are selected from each recorded image (B), wherein object coordinates (xO, yO, zO) of the assigned object point (PPi) are determined from the image coordinates (xB, yB) of the selected pixels (BPi) by triangulation (T) assuming a baseline length (L) between the respective viewpoints (SP) of the single cameras (4).

12. The method according to claim 1 or 2, characterized in that: A bundle adjustment (BA) is performed based on at least two recorded images (B1, B2).

13. The method according to claim 1 or 2, characterized in that: Combining a plurality of first pixels (BP1i) in the first image (B1) into a first feature point (MP1) and combining a plurality of second pixels (BP2i) in the second image (B2) into a second feature point (MP2), wherein the first feature point (MP1) and the second feature point (MP2) are selected so that they belong to the same feature (M) of an object (O) in the captured surroundings (U), In which, assuming a baseline length (L) between the two viewpoints (SP1, SP2) of the single camera (4), the object coordinates (xO, yO, zO) of the assigned feature (M) are obtained by triangulation (T) from the first image coordinates (xB, yB) of the first feature point (MP1) and the second image coordinates (xB, yB) of the second feature point (MP2).

14. The method according to claim 1 or 2, characterized in that: A plausibility check is performed on the acquired object information (OI) resulting from the adjustment of the adjustment path (W) by the single camera (4) using the active actuator system (8) and the object information (OI) resulting from the following odometer data (DD) of the vehicle (1), the odometer data being selected from the group consisting of the following data: wheel speed signal (SR) and / or vehicle speed (v1) and / or steering angle (LW) and / or bending angle (KW).

15. A control unit (5) for executing the method according to any one of claims 1 to 14.

16. Vehicle (1) having a control unit (5) according to claim 15, wherein: The vehicle (1) has a single camera (4) and the vehicle (1) is a single-section or multi-section vehicle, and at least one single camera (4) is arranged on a trailer (3) and / or a tractor (2) of the multi-section vehicle (1).

Images