Method and device for automatically determining position and / or motion-related status information of a vehicle

Abstract

Method for automatically determining position, movement and / or attitude-related status information of a vehicle, wherein the position, movement and / or attitude-related status information is determined by means of an optical navigation method by an optical navigation system (1) of the vehicle, in which - digital image data of at least part of the vehicle environment are recorded by means of at least one camera (2) in a recording area (10), - by an image evaluation unit (4) at least two image features contained in the digital image data are recognized, which are recorded by the camera (2) as environmental features located in the vehicle environment of the vehicle, - in an initialization phase, for each detected image feature, a feature position of the environment feature corresponding to the respective image feature is determined in relation to the vehicle, and - After the initialization phase, position-, movement- and / or orientation-related status information of the vehicle is determined from the calculated feature positions of the respective image features and their respective image positions (P1, P2, P3) within the recording area, whereby the recording direction (R1, R2, R3) of the at least one camera (2) is changed by panning and / or tilting the camera (2) relative to the vehicle by means of a control unit, so that digital image data of the vehicle environment in a first recording sub-area (11), within which a first image feature (B1) is contained, and digital image data of the vehicle environment in at least a second recording sub-area (12, 13), within which at least a second image feature (B2) is contained, are recorded by the camera (2) in a first recording direction (R1).wherein the first recording direction (R1) with the first recording sub-area (11) differs from the second recording direction (R2) with the second recording sub-area (12), characterized in that digital image data of the first recording sub-area (11) are recorded by the camera (2) at a first time point and digital image data of the second recording sub-area (12) at a second time point following the first time point, wherein in the time interval between the first time point and the second time point the recording direction of the at least one camera (2) is panned and / or tilted from the first recording sub-area (11) to the second recording sub-area (12), wherein in the time interval between the first and the at least second time point additional position, movement and / or orientation data of the vehicle are determined by means of sensors that are not provided by the optical navigation system (1), wherein the position,movement and / or position-related status information is determined depending on the feature position of the first image feature and its image position (P1, P2, P3) in the first recording sub-area (11), the feature position of at least the second image feature and its image position (P1, P2, P3) in the second recording sub-area (12) and the additional position, movement and / or position data of the vehicle in the time interval between the first time point and the second time point.

Description

[0001] The invention relates to a method for automatically determining position-, movement-, and / or attitude-related status information of a vehicle, wherein the position-, movement-, and / or attitude-related status information is determined using an optical navigation method implemented by an optical navigation system. The invention also relates to a navigation system, in particular an optical navigation system for implementing the method, and to a vehicle, in particular an unmanned vehicle, having such an optical navigation system.

[0002] In the field of unmanned aircraft, among other elements, autonomous self-localization / navigation via the unmanned aircraft's electronics plays a key role in safe operation. This is the only way to ensure that unmanned aircraft—whether in military or civilian applications—can safely participate in air traffic without endangering others. The topic of autonomous navigation plays a major role in almost all unmanned vehicles, such as autonomously driving road, rail, and / or water vehicles, and is a basic prerequisite for enabling the vehicles to move through traffic and navigate accordingly, independently and without human intervention or human intelligence.

[0003] The autonomous navigation of unmanned vehicles relies on sensor data, which can be used to derive position, movement, and / or attitude-related information about the vehicle's current state at a given time. Such sensor data predominantly originates from a GNSS (Global Navigation Satellite System), which can be GPS, Galileo, or Glonass, for example, which can determine the unmanned vehicle's position in space, usually within a few centimeters. Furthermore, movement-related status information can usually be derived from the GNSS sensor data, since the Doppler effect allows a speed calculation and a direction of movement to be derived from the sensor data.

[0004] It has proven advantageous to acquire additional sensor data from other systems of the unmanned vehicle for support or redundancy purposes in addition to the GNSS. From these data, position, movement, and / or attitude-related status information of the vehicle can be derived and merged with the GNSS sensor data. Such systems can be, for example, the inertial measurement unit (IMU) of an unmanned vehicle, in which, using motion and / or acceleration sensors, corresponding movement, acceleration, and / or attitude data of the vehicle can be determined, preferably in three translational and three rotational axes of movement.If, for example, the provision of GNSS sensor data fails due to shadowing or other unforeseeable events, the navigation capability of the vehicle can be maintained at least for a certain period of time using the movement, acceleration and / or attitude data derived from these inertial sensors, since the position can then be estimated during the GNSS failure based on the last known position using the determined inertial data.

[0005] Another system that can be used for autonomous navigation, particularly for unmanned vehicles, is an optical navigation system that implements an optical navigation method. For example, a classic optical navigation method is known from AJ Davison et al., "MonoSLAM: Real-Time Single Camera SLAM," vol. 29, no. 6, pp. 1052-1067, 2007, which uses camera data from fixed-lens cameras. In optical navigation, which uses visual odometry (VO) or simultaneous localization and mapping (SLAM) as its underlying algorithms, digital images or digital image data are captured using cameras mounted on the vehicle, which are then analyzed using an image processing unit.In this process, image features are identified in the digital images that correspond to environmental features in the vehicle's surroundings and are captured by the camera. These can be, for example, prominent environmental objects such as houses, bridges, towers, and the like. However, the environmental features can also be corners, edges, and / or contrast transitions contained as image features in the digital image data. The image features are particularly designed to be easily recognizable and identifiable from the captured image data, thus serving as a support for the subsequent process.

[0006] Based on the detected image features within the captured images, a feature position can now be derived for each image feature, from which position-, movement-, and / or attitude-related status information of the vehicle can then be determined. Such a feature position can be an absolute or relative position of the feature in the vehicle's surroundings that was included as an image feature in the digital images and recorded and recognized by the camera.

[0007] For example, a feature position can be determined for an image feature by recognizing an image feature in the images based on a known navigation solution or known position, movement, and / or attitude-related status information of the vehicle. Based on the image position of the image feature within the recorded digital image data, a first angle is determined, which defines the position of the feature in the vehicle's surroundings in relation to the vehicle itself. At a later point in time, the image feature is then recognized again in the digital images, and a second angle is determined. This second angle is generally different from the first angle because the vehicle has moved between the first and second recordings.From the angle change over time and the known navigation solution, a feature position of the image feature in the vehicle environment can then be calculated, which can be absolute (e.g. GPS coordinates) or relative to the vehicle, depending on the known navigation solution of the vehicle.

[0008] However, it is also conceivable that the feature position is known, since the image feature is, for example, so-called landmarks whose fixed position is stored in a database.

[0009] If the position of the image features in the vehicle's surroundings is known, angular information relative to the vehicle can be derived from the image position of the individual image features. Triangulation can then be used to determine the vehicle's position. This creates further redundancy in autonomous navigation using an optical navigation method based solely on visual analysis of the vehicle's surroundings.

[0010] In optical navigation, a vehicle's movement (particularly with regard to direction) is detected by the movement of image objects within the recorded images. This allows inferences about the vehicle's movement-related status information based on the image position and, in particular, the displacement or movement of the detected image objects between two consecutive digital images. Examples of optical navigation include visual odometry or SLAM.

[0011] The accuracy of an optical navigation method, particularly when it is based on angular information, correlates with the resolution of the camera and the extent of the recording area (FOV - Field Of View). With increasing resolution of the digital camera, i.e. with an increase in the number of pixels and a decrease in the pixel pitch, the angular resolution with regard to an image feature detected in the digital image data can be increased, so that the accuracy with regard to the angular information is increased. In addition, by increasing the recording area (FOV), the proportion of the vehicle's surroundings that is recorded using the camera for the optical navigation method can be increased, which also improves the accuracy of the optical navigation method because this allows more image features that have a large angle to one another to be detected in the image data.The closer the image features are arranged to each other in the image data, the worse the position determination using triangulation becomes.

[0012] The field of view (FoV) can be enlarged up to 360° to detect as many image features as possible. However, as the FoV size increases, a compromise must be found regarding resolution, as the resolution generally decreases with increasing FoV size, leading to inaccuracies in the calculation of the vehicle's position, movement, and / or attitude-related status information. The lower the resolution, the worse the angular accuracy with respect to the image feature.

[0013] US 2015 / 0 163 993 A1 discloses an autonomous lawnmower robot that navigates using a camera and an optical navigation system. The camera can be designed to be movable to track detected objects.

[0014] US 2016 / 0 291 593 A1 describes an unmanned aerial vehicle that uses a scanner to scan its surroundings in order to avoid collisions with other objects.

[0015] It is therefore an object of the present invention to provide an improved method and an improved system for an optical navigation method with which the accuracy in determining position-, movement- and / or attitude-related status information of the vehicle, in particular UAVs, can be improved.

[0016] The object is achieved according to the invention with the method according to claim 1 and a navigation system according to claim 10.

[0017] According to claim 1, a method is specified for automatically determining position-, movement-, and / or attitude-related status information using an optical navigation method by an optical navigation system of the vehicle. The optical navigation method is a method in which position-, movement-, and / or attitude-related status information of the vehicle is calculated using a recording of a part of the vehicle's surroundings. This can include, in particular, visual odometry and simultaneous localization and mapping techniques.

[0018] Position-, movement-, and / or attitude-related status information of a vehicle is understood to mean status information relating to the position, movement, and / or attitude / orientation of the vehicle, which can be used to infer the position, movement, and / or attitude / orientation of the vehicle. Such position-related status information can, for example, be a position of the vehicle in space, for example by specifying a longitude and a latitude, and in the case of flying objects, possibly also the altitude. The position-related status information can be absolute (within a world coordinate system) or relative (the vehicle represents the origin of its own coordinate system). Such position-related status information can, however, also be a change in position relative to a reference position or a position specification with respect to such a reference position.Motion-related status information is information relating to the movement of the vehicle, such as speed, direction, acceleration, or deceleration. Position-related status information can be, for example, the spatial position of the vehicle, e.g., relative to the earth's vertical, or the orientation of the vehicle relative to its longitudinal axis.

[0019] This position-, movement- and / or location-related status information is particularly suitable for merging with existing navigation systems of other types in order to establish detection redundancy, to support the results of other navigation systems or, if necessary, to close detection gaps with the help of such position-, movement- and / or location-related status information originating from the optical navigation method.

[0020] In the optical navigation method, as generally known from the prior art, digital image data of the vehicle's surroundings are continuously recorded using at least one camera. Digital image data can be the image information that, when put together, forms an image or an image sequence. An image sequence is a chronological sequence or order of recorded images. The camera does not have to record the entire vehicle's surroundings, but only a section or part of the vehicle's surroundings, so that at any given time only a part of the visible hemisphere can be digitally recorded by the camera. With the help of an image evaluation unit, image features contained in the digital image data are recognized. These are environmental features in the vehicle's surroundings and were recorded by the camera when the digital image data was recorded.

[0021] In an initialization phase, for each detected image feature, a feature position of the corresponding environmental feature in the respective image feature is calculated in relation to the vehicle. This can be done, for example, by detecting an image feature in digital image data recorded at two consecutive recording times and then calculating the feature position based on an existing navigation solution. Advantageously, a change in the image position of the respective image feature within the recording area can be calculated with respect to two consecutive recording times and a known navigation solution or known position-, movement-, and / or location-related status information, so that in this initialization phase, a feature position of the environmental features contained and recognized as image features in the digital image data is determined for the known image features.

[0022] The determination of the feature positions of the recognized image features can also be achieved in the initialization phase by knowing the feature positions of the image features and only having to read them out from a database, since the feature positions of the image features have already been calculated and saved in a previous step, for example, or the environmental features that were recognized as image features in the digital images are landmarks that are recognized with the help of the image evaluation unit and the corresponding positions of these landmarks are read out from a database.

[0023] Depending on the feature positions of the known image features and their respective image positions within the recording area, position-, movement-, and / or attitude-related status information of the vehicle can then be determined, which is possible, for example, through triangulation and backprojection. If the positions of the individual environmental features identified as image features in the digital images are known, the vehicle's own position can be derived based on angular information of the individual image features resulting from the image position within the recording area of ​​the digital image data. This makes it possible for the vehicle to fly or move over known terrain and still perform an optical navigation process.

[0024] According to the invention, it is now provided that during the recording of the digital image data for the optical navigation method as described above, the recording direction of the at least one camera is changed by panning and / or tilting (moving) the camera relative to the vehicle by means of a control unit, so that in a first recording direction, digital image data of a part of the vehicle surroundings in a first recording sub-area, within which at least one first image feature is contained, and in at least one second recording direction, digital image data of a part of the vehicle surroundings in a second recording sub-area, within which at least one second image feature is contained, are recorded by the camera.The first recording direction with the first recording sub-area is different from the second recording direction with the second recording sub-area, allowing the camera to record several different recording sub-areas within the entire vehicle environment. The overall recording area is thus no longer fixed, as is the case with permanently installed cameras, but is variably divided into recording sub-areas within the vehicle environment, thus overcoming the disadvantages known from the prior art, particularly with unmanned aerial vehicles at high altitudes.

[0025] The camera's total range of motion defines a tilt and pan range that corresponds to the entire recording area. Within the total recording area, a sub-recording area is formed depending on the recording direction and the camera's FoV. The entire recording area (defined by the tilt and pan range) can never be completely recorded at the same time.

[0026] This makes it possible to use high-resolution cameras, which, due to their high resolution, exhibit very high angular accuracy, thus increasing the accuracy in determining the position, movement, and / or attitude-related status information of the vehicle. Furthermore, the accuracy in the initialization phase, in which the feature position of the image features is determined, can be increased, which in turn increases the accuracy in determining the position, movement, and / or attitude-related status information of the vehicle. Overall, the accuracy of the optical navigation method can thus be increased. It is also possible to achieve a larger FoV with a constant camera resolution compared to the state of the art.

[0027] This also makes it possible to increase the angular resolution of a camera with any resolution by using a small FoV and at the same time increase the extent of the FoV.

[0028] Cameras with digital image sensors featuring a very small pixel pitch can now also be used, enabling high angular resolution and thus, generally, a high pixel count. At the same time, however, the recording area is reduced, so that overall computing power is not increased. The reduced recording area with high angular resolution is compensated for by changing the recording direction, thus still allowing the hemisphere of the vehicle's surroundings within the camera's pan and tilt range to be covered.

[0029] According to the invention, the digital image data of the first recording sub-area are recorded by the camera at a first point in time and digital image data of the second recording sub-area are recorded at a second point in time following the first point in time, wherein in the time period between the first point in time and the second point in time the recording direction of the at least one camera is panned and / or tilted from the first recording sub-area to the second recording sub-area. The image features are thus recorded one after the other by the camera. It is of course conceivable that the environmental features in the vehicle's surroundings are distributed in such a way that each image feature corresponding to an environmental feature lies in its own recording sub-area. The camera must therefore pan and / or tilt back and forth between the individual recording sub-areas in order to be able to record the individual recording sub-areas accordingly.

[0030] In the period between the first and at least the second point in time, additional position and / or movement data of the vehicle are determined using sensors that are not provided by the optical navigation system. Such position and / or movement data can originate, for example, from a GNSS or an inertial sensor system, whereby changes in the position and / or movement of the vehicle can be detected in the period between the first and second point in time.

[0031] The position-, movement- and / or location-related status information is now determined as a function of the feature position of the first image feature and its image positions in the first recording sub-area, the feature position of at least the second image feature and its image positions in the second recording sub-area and the additional position and / or movement data of the vehicle in the time period between the first time and the second time, so that movements of the vehicle in the time period between the first time and the second time do not have a negative effect on the accuracy in determining the position-, movement- and / or location-related status information.

[0032] For example, it is conceivable that, in the time period that the camera requires for panning and / or tilting from the first recording sub-area to the second recording sub-area, changes in position, movement and / or attitude of the vehicle are calculated from the position, movement and / or attitude data of the vehicle in order to be able to determine a change in position, movement and / or attitude of the vehicle within this time period or at the end of the time period between the first and second time points.Based on this change in position, movement, and / or attitude of the vehicle within this time period, an estimated image position of the first image feature in the first recording sub-area at the second time can then be calculated, for example, so that the estimated image position of the first image feature in the first recording sub-area corresponds to the image position as if the first image feature were recorded and recognized in the first recording sub-area at the second time. This can prevent a navigation error resulting from the panning and / or tilting period, which is due to the panning and / or tilting of the camera from the first recording sub-area to the second recording sub-area.

[0033] It is of course also conceivable that, based on the change in position, movement and / or attitude of the vehicle within the time period between the first and the second time, an estimated image position of the second image feature in the second recording sub-area is calculated for the first time, so that, starting from the image position of the recorded second image feature in the second recording sub-area, the estimated image position is back-projected to the first time.

[0034] In both cases, at the second point in time, the image positions of the first and second image features are present as if they were recorded and recognized simultaneously at that point in time.

[0035] The inventors have recognized that the panning and / or tilting of the camera from one recording area to the second recording area takes place within a time span that can be bridged using sensor data from an inertial sensor system or other sensor systems, so that no or only minimal loss of accuracy occurs.

[0036] Of course, it is conceivable that additional third, fourth, fifth, etc., recording sub-areas, each with different recording directions, could be used for the process. Here, too, further sensor data can be acquired between the individual points in time at which the digital image data for the respective recording sub-area were acquired. The first and second recording sub-areas with the first and second recording directions are not to be understood as exhaustive or restrictive.

[0037] In an advantageous embodiment, the feature positions of the respective image features in relation to the vehicle are further calculated taking into account the respective recording direction with which the digital image data was recorded in the respective recording sub-area in which the respective image feature is contained. The offset of the image feature within the digital image data from the image center, which then represents the angle in the respective dimensions, is corrected by the offset of the recording direction from a reference axis or reference axis, for example the longitudinal axis of the vehicle, so that an overall image position in relation to the reference axis of the vehicle is obtained. This can be done, for example, by feeding back the pan and / or tilt state of the camera to the optical navigation system.It is also conceivable, however, that the entire pan and tilt range of the variable camera is initially recorded with a very coarse resolution using another camera, whereby the partial recording area defined by the recording direction is detected within the overall recording and then the displacement of the partial recording area into the overall recording area is detected.

[0038] In a further advantageous embodiment, the focal length of a camera lens of the at least one panning and / or tilting camera is changed during the recording of the digital image data, whereby it is possible to zoom in in a recording direction within the recording sub-area and, after zooming in, to zoom out again. This is particularly advantageous, for example, when the vehicle is a flying object with great altitude, so that a focal length is set depending on the altitude, with which the recording sub-area is zoomed in, thereby reducing the actual dimensions of the recording sub-area. This can increase the angular resolution, which is particularly advantageous for distant environmental features.

[0039] In a further advantageous embodiment, new image features for carrying out the optical navigation method are searched for within a panning and / or tilting range of the camera, wherein the camera is panned or tilted in a search direction of a recording search area that is selected depending on the recording directions of the recording sub-areas of already known image features. It is thus advantageous if new image features are searched for within recording search areas that are a large distance from the recording sub-areas of already known image features. It is particularly advantageous if the search direction of the recording search area is selected depending on a variance in the condition of the vehicle. The search direction is thus limited to those directions in which the greatest uncertainty exists with regard to determining the position-, movement-, and / or attitude-related status information.

[0040] The invention is explained in more detail by way of example with reference to the accompanying drawings. They show: Fig. 1 - Schematic representation of the optical navigation system according to the invention; Fig. 2 - Illustration of an embodiment for the calculation.

[0041] Fig. Figure 1 schematically shows an optical navigation system 1 having a panning and / or tilting camera 2 and a computing unit 3 connected thereto for calculating the position, movement, and / or attitude-related status information of a vehicle (not shown). The computing unit 3 has an image evaluation unit 4, which is signal-connected to the camera 2 and is designed to analyze the recorded image data. The image evaluation unit is further connected to a navigation unit 5, which is designed to calculate the position, movement, and / or attitude-related status information of the vehicle.

[0042] The optical navigation system 1 can be connected to another navigation system 6 of the vehicle or be a component of this navigation system 6, wherein the navigation system 6 is connected to a GNSS sensor 7 for acquiring position, movement and / or attitude data using a satellite-based navigation system. Furthermore, the navigation system 6 is also connected to the vehicle's internal inertial system 8, which in the exemplary embodiment of the Fig. 1 is depicted in the form of a vehicle bus. The inertial system 8 can also be connected to the optical navigation system 1, as will be explained later.

[0043] The navigation system 6 can be configured to combine the position, movement, and / or attitude data from the GNSS 7 and the inertial system 8 with the position-, movement-, and / or attitude-related status information from the optical navigation system 1, for example, to independently and completely automatically navigate an unmanned vehicle. The position-, movement-, and / or attitude-related status information from the optical navigation system 1 can be used, for example, when the vehicle cannot access other navigation solutions.

[0044] Returning to the optical navigation system 1, the camera 2 is designed to be pivotable and / or tiltable, whereby in the exemplary embodiment the Fig. 1 shows a panning and tilting of the camera relative to a reference axis X. By panning and / or tilting the camera 2, i.e. a relative movement relative to the reference axis X, digital image data in the form of images or image sequences can be recorded within the recording range 10 of the camera 2. The recording range 10 is also simultaneously the tilt / pan range 20 of the camera. The actual range within which the camera can record digital image data at any one time is, however, smaller than the overall recording range 10 or the tilt / pan range 20, so that corresponding partial recording ranges 11, 12, 13 are created within the overall recording range 10, in which partial recording ranges can be recorded at a time without the camera having to be panned and / or tilted.However, if digital image data is to be recorded from a first recording sub-area 11 into a second recording sub-area 12, the camera 2 must be panned and / or tilted from the first recording sub-area 11 to the second recording sub-area 12. However, without panning and / or tilting, the camera 2 cannot record digital image data in two different recording sub-areas 11, 12, 13 at the same time.

[0045] In a first recording direction R1, in the embodiment of the Fig. 1 defines a first recording sub-area 11, within which an environmental feature is located in the real surroundings of the vehicle, which is contained in the recorded digital image data within the first recording sub-area 11 in the form of a first image feature B1. In the first recording sub-area 11, the recognized image feature, which was recognized by the image evaluation unit 4 in the digital image data of the first recording sub-area 11, is located in the upper left half of the image.

[0046] Subsequently, the camera 2 is panned to the second recording sub-area 12 shown in dashed lines, wherein the image feature B2 to be recognized is contained in the lower right half of the image in the second recording sub-area 12. The camera is then panned or tilted in the same way in order to recognize a corresponding third image feature B3 within the third recording sub-area 13.

[0047] For example, if the feature positions of image features B1, B2, and B3 were determined in relation to their real-world environmental features during a previous initialization phase, position-, movement-, and / or attitude-related status information for the vehicle can now be determined based on the recognition of the individual image features B1, B2, and B3. To do this, the image position is first determined for each image feature B1, B2, and B3.

[0048] With respect to the first recording sub-area 11, the image position B1 can be represented as an angular offset between a camera reference axis (here the image center) as the initial recording direction R1 and the line of sight to the image feature B1. Based on a pan and / or tilt angle α1 of the camera, which is established when the camera 2 is panned and / or tilted between the reference axis X of the vehicle and the first recording direction R1 of the camera 2, an actual angle between the reference axis X of the vehicle on the one hand and the line of sight to the image feature B1 on the other hand can then be calculated together with the angular offset P1. This results in an image position of the image feature B1 in both the horizontal and vertical offset within the entire recording area 10 of the pan and / or tilt range 20 of the camera 2, so that the angle at which the image feature B1 is to be observed with respect to the vehicle can be determined.

[0049] Analogously, the image position of the second image feature B2 and the third image feature B3 is determined. Here, too, angle information P2 is calculated between the second recording direction R2 of camera 2 and the line of sight between the camera and the image feature B2. A corresponding angle to the image feature B2 relative to the entire vehicle can then be determined from an angular offset of the recording direction R2 (camera reference axis) to the reference axis X of the vehicle. The image position for the third feature B3 is then determined in the same way.

[0050] Of course, other recording areas and image features are conceivable; however, for the sake of clarity and to illustrate the functional principle, they are not shown. In principle, three image features are sufficient for a back projection to determine the position-, movement- and / or attitude-related status information, i.e., for example, to determine the own position based on the position of the environmental features corresponding to the respective image features B1 to B3 and the associated angle information. However, to increase the accuracy of the system, it is advantageous if more than three image features are recognizable. With fewer than three image features, a position indication is possible with a certain error variance; however, a sufficient position indication is possible if the altitude of the flying object is known, which generally rarely changes abruptly.

[0051] From the combination of the determined and calculated values, a position-, movement- and / or location-related status information Z of the vehicle can be calculated by Z=Projection{P1+α1,P2+α2,P3+α3} calculate. By using a panning and / or tilting camera 2 with normal resolution, it is possible to achieve a high angular resolution with a large FoV. This allows the respective object angles P1 to P3 to be detected with very high resolution, since the respective image objects B1 to B3 can be detected very precisely in the high-resolution digital image data. Furthermore, by reducing the recording areas 11 to 13 of the camera 2 compared to the panning range 20, in particular by significantly reducing them, the total number of pixels can be kept to a level that can be calculated in real time by the image evaluation unit 4.

[0052] Fig. Figure 2 shows, in a highly simplified manner, the case in which the optical navigation system is connected to an inertial data system 8 of the vehicle in order to compensate for temporal recognition gaps. At time T1, the image feature B1 of the first recording sub-area 11 (see Fig. 1) is detected and the corresponding angle information P1 is calculated. At time T2, the second image feature B2 is detected in the second recording sub-area 12, and the corresponding angle information P2 is determined. In the time period between T1 and T2, in which the camera 2 was panned and / or tilted from the first recording sub-area 11 to the second recording sub-area 12, additional inertial sensor data S1 is determined from the inertial sensor system 8 of the vehicle, which contains position and / or movement data of the vehicle within this time period between T1 and T2.

[0053] At a time T3 following the time T2, the third image feature B3 is then determined in the third recording sub-area 13, wherein inertial data are also acquired from the inertial data system 8 in the time period between T2 and T3.

[0054] At time T3, the system therefore has three pieces of angle information P1 to P3 relating to the three image features, from which the position-, movement-, and / or attitude-related status information can then be determined. However, at time T3, the angle information P1 and P2 determined at time T1 and time T2 are outdated with respect to the digital images captured at the respective times, which results in greater uncertainty in the determined position-, movement-, and / or attitude-related status information.

[0055] Therefore, it is advantageous that based on the inertial sensor data S1 and S2 and the feature positions of the image features B1 and B2 for the respective time periods, the angle information P1 and P2 are forwarded and thus estimated first angle information P1' and P2' is determined.

[0056] The position-, movement- and / or attitude-related state information of the vehicle is then determined from the estimated first angle information P1' (as image position of the first image feature) and from the estimated second angle information P2' (as image position of the second image feature) and the angle information P3 (as image position of the third image feature), whereby the uncertainty due to the time period between T1 and T2 or T3 can be reduced. List of reference symbols 1 optical navigation system 2 cameras 3 Computing unit 4 Image output unit 5 Navigation unit 6 Navigation system 7 GNSS 8 Inertial sensor system 10 Total recording area 11 first recording area 12 second recording area 13 third recording area B1 first image object B2 second image object B3 third image object R1 first recording direction R2 second recording direction R3 third recording direction P1 first angle information or image position P2 second angle information or image position P3 third angle information or image position α1 first camera angle α2 second camera angle α3 third camera angle 20 swivel range

Claims

[1] Method for automatically determining position-, movement- and / or location-related status information of a vehicle, wherein the position-, movement- and / or location-related status information is determined by means of an optical navigation method by an optical navigation system (1) of the vehicle, in which - digital image data of at least part of the vehicle environment are recorded by means of at least one camera (2) in a recording area (10), - at least two image features contained in the digital image data are detected by an image evaluation unit (4), which are recorded by the camera (2) as environmental features located in the vehicle's surroundings, - in an initialization phase, for each detected image feature, a feature position of the environmental feature corresponding to the respective image feature is determined in relation to the vehicle, and - after the initialization phase, position-, movement-, and / or location-related status information of the vehicle is then determined from the calculated feature positions of the respective image features and their respective image positions (P1, P2, P3) within the recording area, wherein the recording direction (R1, R2, R3) of the at least one camera (2) is changed by panning and / or tilting the camera (2) relative to the vehicle by means of a control unit, so that in a first recording direction (R1), digital image data of the vehicle surroundings are recorded in a first recording sub-area (11), within which a first image feature (B1) is contained, and in at least a second recording direction (R2), digital image data of the vehicle surroundings are recorded in at least one second recording sub-area (12, 13), within which at least one second image feature (B2) is contained,wherein the first receiving direction (R1) with the first receiving section (11) is different from the second receiving direction (R2) with the second receiving section (12), , characterized bythat digital image data of the first recording sub-area (11) are recorded by the camera (2) at a first point in time and digital image data of the second recording sub-area (12) are recorded at a second point in time following the first point in time, wherein in the time period between the first point in time and the second point in time, the recording direction of the at least one camera (2) is pivoted and / or tilted from the first recording sub-area (11) to the second recording sub-area (12), wherein in the time period between the first and the at least second point in time, additional position, movement and / or attitude data of the vehicle are determined by means of sensors that are not provided by the optical navigation system (1), wherein the position-, movement- and / or attitude-related status information is determined as a function of the feature position of the first image feature and its image position (P1, P2, P3) in the first recording sub-area (11),the feature position of the at least second image feature and its image position (P1, P2, P3) in the second recording sub-area (12) and the additional position, movement and / or attitude data of the vehicle in the time period between the first time and the second time is determined., [2] Method according to claim 1, characterized by that in the initialization phase, for each recognized image feature, a feature position of the environmental feature corresponding to the respective image feature is calculated as a function of a change in the image position (P1, P2, P3) of the respective image feature within the recording area with respect to two consecutive recording times and at least one known position-, movement- and / or location-related state information. [3] Method according to claim 1 or 2, characterized bythat the feature positions of the respective image features in relation to the vehicle are further calculated taking into account the respective recording direction (R1, R2, R3) with which the digital image data was recorded in the respective recording sub-area (11, 12, 13) in which the corresponding image feature is contained. [4] Method according to one of the preceding claims, characterized by that the focal length of a camera lens of at least one camera (2) is changed during the recording of the digital image data. [5] Method according to claim 4, characterized by that the focal length of the camera lens is changed depending on a flight altitude of the vehicle designed as a flying object and / or depending on a distance of an environmental feature contained as an image feature in the respective recording sub-area (11, 12, 13) with respect to the vehicle. [6] Method according to one of claims 1 to 5, characterized bythat a change in position, movement and / or attitude of the vehicle in the time period between the first time and the second time is calculated based on the determined additional position, movement and / or attitude data of the vehicle and an estimated image position (P1, P2, P3) of the first image feature in the first recording sub-area (11) is calculated for the second time as a function of the calculated change in position, movement and / or attitude, which is used as the image position (P1, P2, P3) of the first image feature for determining the position-, movement- and / or attitude-related state information. [7] Method according to one of claims 1 to 5, characterized bythat a change in position, movement and / or attitude of the vehicle in the time period between the first time and the second time is calculated based on the determined additional position, movement and / or attitude data of the vehicle and an estimated image position (P1, P2, P3) of the second image feature in the second recording sub-area (11) is calculated for the first time as a function of the calculated change in position, movement and / or attitude, which is used as the image position (P1, P2, P3) of the second image feature for the determination of the position-, movement- and / or attitude-related state information. [8] Method according to one of the preceding claims, characterized bythat new image features are searched for within a panning and / or tilting range (20) of the camera (2) for carrying out the optical navigation method, wherein the camera (2) is panned and / or tilted in a search direction of a recording search range, which is selected as a function of the recording directions of the recording sub-ranges of already known image features. [9] Method according to claim 8, characterized by that the search direction of the recording search area is selected depending on a variance in the condition of the vehicle. [10] Navigation system with an optical navigation system (1) for automatically determining position-, movement- and / or location-related status information of a vehicle, wherein the optical navigation system (1) is set up to carry out the method according to one of the preceding claims and has at least one camera (2) which is designed to be pivotable and / or tiltable relative to the vehicle, an image evaluation unit (4) connected to the camera (2) and a control unit for controlling the camera (2) for pivoting and / or tilting the camera (2). [11] Navigation system according to claim 10, characterized by that the panning and / or tilting camera (2) has a camera lens with variable focal length, wherein the focal length can be adjusted by the control unit. [12] Vehicle with a navigation system, the navigation system comprising, inter alia, an optical navigation system (1) according to one of claims 10 or 11. [13] Vehicle according to claim 12, characterized by that the vehicle is an aircraft, in particular an unmanned aircraft.

Images