Operation assistance method, control unit, operation assistance system and working equipment

Abstract

The invention relates to an operation assistance method (S) for a working device (1) or for a vehicle, in which (S1) an object frame (54) of an object (52) or data characterizing such an object frame (54) is obtained at successive moments in the field of view of the working device (1), (S2) from the object frame (54) of a given object (52) in images taken successively or directly successively, a current change in the scale of the object frame (54) of the corresponding object (52) or a variable derived therefrom is determined, and a current change in the lateral position of the object frame (54) of the corresponding object (52) is determined, (S3) based on the corresponding object frame (54) Determine a future predicted object frame (55) based on a current change in the scaling ratio of an object frame (54) of the image (52) or a parameter derived therefrom and a current change in the lateral position, (S4) determine and evaluate the position of the predicted object frame (55), and / or the ratio of the lateral extension of the predicted object frame (55) to the lateral extension of the detected field of view (50) and / or the lateral extension of the captured image, (S5) based on the result of the evaluation, (i) determine whether the object (52) on which the predicted object frame (55) is based is critical with respect to a possible collision, and / or (ii) control or regulate the operating state of the working device (1).

Description

Technical Field

[0001] The invention relates to an operating assistance system for a working machine or for a vehicle, a control unit for an operating assistance system for a working machine, such an operating assistance system and a working machine, in particular a vehicle. Background Art

[0002] In working devices, in particular in the field of vehicles, operating assistance systems and operating assistance methods are increasingly used, which are intended to check the surroundings of the respective device with respect to a possible collision with an object and are intended to output corresponding warning messages and / or intervene in the operation of the device. In the known systems and methods, for example, relatively complex systems and data structures with three-dimensional data evaluation are used, and / or the effectiveness of the corresponding evaluation of the surroundings is not sufficient for intervening in the operation of the device, for example, for a braking decision. Summary of the invention

[0003] In contrast, the operation assistance method according to the invention has the advantage that a particularly reliable collision prediction can be achieved for the operation of the working device with the aid of relatively simple means. According to the invention, this object is achieved in the following way: an operation assistance method for a working device, in particular a vehicle, is provided, in which:

[0004] (S1) acquiring an object frame of an object in the field of view of the working device at successive times or acquiring data characterizing such an object frame;

[0005] (S2) determining the current scaling change of the object frame of the corresponding object from the object frame of the given object in the images taken successively or directly successively. or scale change and determining a current horizontal position change of an object frame of the corresponding object;

[0006] (S3) determining a future predicted object frame based on a current scaling change, a scale change or a parameter derived therefrom, and a current lateral position change of the object frame of the corresponding object;

[0007] (S4) determining and evaluating a position of the predicted object frame and / or a ratio of a lateral extension of the predicted object frame relative to a lateral extension of the detected field of view and / or a lateral extension of the captured image;

[0008] ( S5 ) Based on the result of the evaluation, (i) a determination is made as to whether the object on which the predicted object frame is based is critical with respect to a possible collision, and / or (ii) an operating state of the working device is controlled or regulated.

[0009] According to the invention, the evaluation of the surroundings of the working device is therefore based on so-called object frames, corresponding predicted object frames and changes in the size of the object frames relative to the detected field of view. These data can be detected essentially in pure two dimensions and can be determined with high precision.

[0010] In principle, the data relating to the object frame can be provided externally, for example from an optical detection unit of a conventional driver assistance system.

[0011] However, in a preferred embodiment of the operation assistance method according to the present invention, it is provided that in or for the step (S1) of obtaining an object frame and / or obtaining data characterizing such an object frame:

[0012] (S1a) optically two-dimensionally and / or monocularly detecting the field of view of the underlying working device by recording images that are temporally successive;

[0013] ( S1 b ) At least one object is determined in the images captured one after the other or in a section of these images and an object frame assigned to the object is determined.

[0014] As already mentioned above, an important aspect of the invention is to predict the detected object frames of objects in the field of view in the future. This prediction can be implemented in different ways.

[0015] In one embodiment of the operation assistance method according to the invention, a future predicted object frame is determined at least for the current last detected image or a section of an image.

[0016] This is achieved in particular by iteratively determining and updating values ​​of a scaling factor of the corresponding object frame or a variable derived therefrom, values ​​of coordinates of the corresponding object frame, values ​​of a translation of the corresponding object frame and / or values ​​of a lateral width of the corresponding object frame over a plurality of time increments up to the prediction time span.

[0017] The prediction time span and / or the time increment can be predetermined and specified. However, it is also possible to make the prediction period and / or the time increment depend on other operating parameters, for example on the inherent speed and / or position of the device (especially the vehicle) itself, or also on the previously predicted speed and / or position of one or more objects in the surroundings of the device (especially the vehicle). Thus, if this is required due to the number of objects in the surroundings and / or due to the relatively high inherent speeds of the device and / or the objects, a closer monitoring in time can be achieved in an advantageous manner. Conversely, the monitoring overhead can be reduced in relatively low traffic volumes or similar situations.

[0018] Against this background, according to a further embodiment of the operation assistance method according to the invention, the following steps can be carried out for each time increment (in particular in the order described):

[0019] (11) Reset or preset the value to be calculated according to the allocation rule:

[0020] Skalierung alt :=Skalierung neu

[0021] BoxTranslationX al :=BoxTranslationX neu

[0022] BoxBreite alt :=BoxBreite neu

[0023] BoxPositionLinks alt :=BoxPositionLinks neu

[0024] BoxPositionRechts alt :=BoxPositionRechts neu

[0025] (12) Update the scaling ratio according to the following allocation rules:

[0026] Skalierung neu :=1 / (2-Skalierung alt )

[0027] (13) Update the horizontal or lateral object frame translation according to the following distribution rule:

[0028] BoxTranslationX neu :=BoxTranslationX alt ×Skalierung alt

[0029] (14) Update the horizontal or transverse object frame width according to the following distribution rule:

[0030] BoxBreite neu :=BoxPositionRechts alt —BoxPositionLinks alt

[0031] (15) Predict the horizontal or lateral box position according to the following assignment rule:

[0032] BoxPositionLinks neu

[0033] :=BoxPositionLinks alt +BoxTranslationX neu -0.5×BoxBreite neu ×(Skalierung neu -1) / Skalierung neu

[0034] BoxPositionRechts neu

[0035] :=BoxPositionRechts alt +BoxTranslationX neu +0.5×BoxBreite neu ×(Skalierung neu -1) / Skalierung neu

[0036] Among them, Skalierung alt 、Skalierung neu Indicates the old or new scale of the object box, BoxTranslationX alt 、BoxTranslationX neu Indicates the old or new displacement of the object box, BoxBreite alt 、BoxBreite neu Indicates the old or new width of the object box, BoxPositionLinks alt 、BoxPositionLinks neu The old or new position of the lower left corner of the object box (52) as the first x coordinate of the corresponding object box, and BoxPositionRechts alt 、BoxPositionRechts neu Indicates the old or new position of the lower-right corner of the object box as the second x-coordinate of the corresponding object box, or their values.

[0037] Alternatively or additionally, the above equation may be replaced or supplemented by the following calculation rule

[0038]

[0039] and

[0040]

[0041] Among them, BoxPositionLinks neu and BoxPositionLinks aktuell or BoxPositionRechts neu and BoxPositionRechts aktuell is the new or current position of the left or right box edge, BoxGeschwindigkeitLinks aktuell or BoxGeschwindigkeitRechts aktuell is the currently measured angular velocity of the left or right frame edge, NormGeschwindigkeit aktuell is the so-called normalized or standardized frame velocity currently measured, The prediction time corresponds to the prediction time step. In particular, the NormGeschwindigkeit is derived from the calculated scaling change of the object frame. aktuell .

[0042] In a specific embodiment of the operation assistance method according to the invention, it is provided that the object on which the predicted object frame is based is determined to be critical with respect to a possible collision, in particular with a criticality of 100%, if the proportion of the width of the predicted object frame of the object to the width of the image on which it is based or the width of a predetermined section of the image exceeds a predetermined first threshold value. This threshold value must be reapplied for each vehicle model and / or each new vehicle model.

[0043] In this context, it is particularly advantageous if the value of the criticality determined for an object is reduced by the proportion by which the object frame predicted for the object lies with respect to its width outside the underlying image or the predefined segment.

[0044] Alternatively or additionally, it is advantageously provided that the object on which the predicted object frame is based is determined to be non-critical with respect to a possible collision, in particular with a criticality value of 0%, if the predicted object frame is completely outside the underlying image or the predefined segment.

[0045] In order to take into account scenarios that are as realistic as possible when predicting object frames in the future, according to another advantageous extension of the operation assistance method according to the present invention, pedestrians are identified as objects, the position and movement of the pedestrians as objects are checked and evaluated based on a pedestrian model, the acceleration capability of the pedestrians as objects is determined based on the speed determined for the pedestrians, and the criticality of the pedestrians as objects is determined based on the speed and the acceleration capability.

[0046] It is particularly advantageous here if a prediction object frame which is extended and surrounds (or at least laterally or horizontally surrounds) the prediction object frame is generated based on the acceleration capability and is used as a basis for assessing the criticality.

[0047] According to another aspect of the present invention, a control unit is provided for an operating assistance system of a working device, in particular a vehicle.

[0048] The control unit according to the invention is provided for controlling the operating assistance method according to the invention, for enabling the operating assistance method according to the invention to be operated, and / or for enabling the operating assistance system based on which the operating assistance method according to the invention is operated.

[0049] The subject matter of the present invention is also such an operating assistance system for a working device, in particular a vehicle. The operating assistance system is provided for carrying out the operating assistance method according to the present invention. For this purpose, the operating assistance system has in particular a control unit configured according to the present invention.

[0050] Furthermore, the present invention provides a working machine having an operation assistance system according to the present invention.

[0051] The working device is designed in particular as a vehicle, a motor vehicle or a passenger car.

[0052] According to another aspect of the present invention, there is also provided the use of an operation assistance method according to the present invention, the use of a control unit according to the present invention, the use of an operation assistance system according to the present invention and / or the use of a working device according to the present invention, wherein the operation assistance method, the control unit, the operation assistance system and / or the working device are used for pedestrian protection or cyclist protection or adaptive cruise control (ACC) and / or an avoidance system or method. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Embodiments of the present invention are described in detail with reference to the accompanying drawings.

[0054] Figure 1 A schematic top view of a working device according to the invention is shown in the form of a vehicle, in which one embodiment of the operating assistance method according to the invention can be used;

[0055] Figures 2 to 4 shows a schematic side view of different scenes in a field of view, which can be evaluated with the aid of the operation assistance method according to the invention;

[0056] Figure 5 and Figure 6 A flow chart showing an embodiment of the operation assistance method or iterative determination of a predicted object frame according to the present invention is shown. DETAILED DESCRIPTION

[0057] Reference below Figures 1 to 6 The embodiments and technical background of the present invention are described in detail. The same and equivalent elements and components having the same or equivalent functions are represented by the same reference numerals. A detailed description of the represented elements and components is not given in any case.

[0058] The features shown and other characteristics can be isolated from one another in any form and combined with one another as desired without departing from the core of the present invention.

[0059] Figure 1 A schematic top view of a working device 1 according to the invention is shown in the form of a vehicle 1 ′, in which an embodiment of the operating assistance method S according to the invention can be used.

[0060] The vehicle according to the invention consists essentially of a body 2 on which wheels 4 are mounted, which wheels can be driven by a drive unit 20 via a drive train 12 and which wheels can be braked and / or steered via a steering and braking unit 30 and via a corresponding braking and steering system 13 .

[0061] Furthermore, one embodiment of the operating assistance system 100 according to the invention is a component of the vehicle 1 ′ according to the invention (as a working device 1 in the sense of the invention). The operating assistance system 100 comprises a camera unit 40 for monocular imaging of a field of view 50 in the surroundings of the vehicle 1 ′. The field of view 50 comprises a scene 53 with a pedestrian 52 ′ as an object 52.

[0062] The control unit 10 is connected via a control and detection line 11 to the camera unit 40 on the one hand and to the drive unit 20 and the brake and / or steering unit 30 on the other hand for control purposes.

[0063] exist Figure 1 In the embodiment shown in FIG, an image 51 of a scene 53 captured by a camera unit 40 from a field of view 50 is transmitted via a control and detection line 11 to a control unit 10 and is evaluated in the control unit in conjunction with an operation assistance method S according to the invention.

[0064] Therefore, according to the present invention, an object frame 54 is determined in each image or frame 51 in conjunction with a pedestrian 52' as an object 52, and parameters regarding position changes and scale changes, parameters regarding the angular velocity of the frame edge and / or parameters regarding variables derived therefrom of the object frame 54 are determined from the object frames of the images 51 that are directly successive in time, and the following prediction is used as the basis: the prediction is used to describe the predicted frame 55 of the object 52 based on the iterative method I.

[0065] Figures 2 to 4A schematic side view shows different scenes 53 in a field of view 50 which can be evaluated with the aid of the operating assistance method S according to the invention.

[0066] exist Figure 2 In the case shown in FIG. 5 , a pedestrian 52 ′ as an object 52 is located in the field of view 50 . Figure 1 The camera unit 40 shown in FIG. 5 is shown in an image or frame 51. The image or frame 51 is optionally delimited by corresponding segments or parts 51'.

[0067] First, an object frame 54 is derived for the pedestrian 52'. Then, the change in scale and the measure of displacement or translation, the angular position, the angular velocity of the frame edge and / or the quantities derived therefrom are determined for the object frame 54 in the temporally preceding image or frame 51. Then, in the above-described iterative method I, a prediction is provided from these quantities over a plurality of time increments for the elapsed prediction time span in terms of the expected prediction object frame 55. In this way, a prediction object frame 55 can be extrapolated from a given object frame 54 in terms of position by means of the lower right corner, the lower left corner and the width 55b for the future prediction time span.

[0068] For evaluation, width 55b of object frame 55 predicted at a future prediction time span is then compared with width 51b of section 51' of image 51. If the ratio between them exceeds a predefined first threshold value, object 52 is considered to have a criticality of 100%.

[0069] That is, according to the present invention, the critical value can already be used to warn the user of the working device 1 (in particular the driver of the vehicle 1 ′), or the critical value can be used to directly intervene in the operating method of the working device 1. However, it is also conceivable that if the object 52 is a pedestrian 52 ′, other aspects of the object 52 (such as predicted acceleration characteristics, etc.) can be influenced in a realistic manner.

[0070] For this purpose, the current speed and magnitude of the pedestrian 52' as the object 52 can be derived and used as an input parameter of the pedestrian model. The pedestrian model then outputs corresponding values ​​for the expected acceleration or for the expected acceleration characteristics. These values ​​can be used to Figure 3 and Figure 4 An enclosing or surrounding object frame 56 and its right and left segments 56 r and 56 I are constructed, which extend beyond the originally predicted object frame 55 and represent the uncertainty range of the positive or negative acceleration characteristics of the pedestrian 52 ′ as the object 52 .

[0071] Figure 5 and Figure 6A flow chart shows one specific embodiment of the operation assistance method S according to the invention or the iterative determination I of the predicted object frame 55 , which has already been discussed above in conjunction with the general description of the invention, for example.

[0072] In this context, it should also be mentioned that the iterative method I essentially constitutes step S3 of an implementation of the operation assistance method S according to the invention, wherein in step I6 it is checked whether the predicted time span has been reached by traversing the time increment and / or whether there are other termination conditions for the iteration.

[0073] In this case, alternative or additional termination conditions may be identified, for example, in conjunction with the fact that the width of the predicted object frame 55 relative to the image 51 or the segment 51 ′ exceeds a second threshold value, wherein the second threshold value is greater than the first threshold value.

[0074] These and other features and characteristics of the present invention are further elucidated by the following statements:

[0075] The present invention describes measures for determining a criticality criterion for a collision warning system (for example as part of an operating assistance system of a work machine, in particular a vehicle) solely based on measurement data of a monocular video system.

[0076] For this purpose, the collision indicators commonly used are "time to collision" (TTC) and "time to brake" (TBB). These indicators provide information on when a collision will occur or when a braking procedure must be taken to avoid a collision. Based on the data of a monocular video camera and mainly on the basis of the scaling changes of the object frame, the parameters TTB and TTC can be reliably calculated without the need to determine the distance, relative speed and relative acceleration.

[0077] Another collision indicator is the value of "Constant Bearing (CB)", which originally originates from shipping and indicates whether the person in continuous intrinsic motion or own motion and in object motion is on a collision trajectory with the object 52. The CB can also be calculated purely based on monocular video data, i.e. based on a two-dimensional database.

[0078] The existing technology for calculating the criticality level is to use a 3D based world coordinate system.

[0079] The basis of this method is the use of three-dimensional data or 3D data, for example in the sense of distances, relative velocities and relative accelerations (which can only be determined with limited quality by estimation using a single camera or a monocular camera).

[0080] The CB concept is difficult to understand, difficult to parameterize, unable to make any predictions about the future, and unable to use pedestrian motion models.

[0081] The concepts of TTC / TTB alone are not sufficient for braking decisions, since only the time aspect is taken into account, but not whether an object is located on the collision trajectory. The concepts of TTC and TTB cannot, for example, draw conclusions as to whether an object is being driven past.

[0082] The new two-dimensional or 2D-based criticality calculation scheme based on the present invention is purely or essentially based on measured two-dimensional or 2D data or 2D signals, in particular based on the determination of so-called frame coordinates of an object frame and parameters of scaling changes, frame translations, angular positions and / or angular velocities, which describe scale changes, size changes, movements or displacements of corresponding object frames in images, frames, or corresponding segments of images or frames, and which parameters in combination with the frame coordinates can be present or determined with high signal quality.

[0083] The solution according to the invention also includes a forecast, prediction or prediction of the future (more precisely, with regard to the position of the corresponding object frame 54 and its size / width or size change / width change) and thus allows the use of pedestrian movement models according to the invention.

[0084] Therefore, a pedestrian movement model can be used to predict the staying area of ​​pedestrian 52'. In this case, a model of the acceleration ability of pedestrian 52' in various movement states (e.g., standing, walking, running, sprinting) can be used to draw conclusions about where pedestrian 52' may be in the future. The value of the criticality is calculated from the overlap of the predicted or predicted own staying area of ​​pedestrian 52' and the predicted or predicted possible staying area 56.

[0085] Compared to the pure CB concept, the solution according to the present invention is easier to understand and parameterize, and experimental evaluations show that it can provide better results than conventional 3D-based methods based on 3D data estimated with the help of a CB implementation and a single camera.

[0086] A forecast or prediction is usually also performed for each captured image or frame. In this case, a target time span (e.g., 2 seconds) is set for each captured image or frame. Thus, a forecast is made for the future target time span. In this case, the target time span is divided into a plurality of, for example, equal time increments. However, the target time span and / or the time increments can also be designed to be variable and be dependent on, determined, selected, and determined by other operating parameters.

[0087] The following process is carried out for each recorded image 51 or frame:

[0088] - Detecting the lateral box coordinates of the corresponding object box 54 of a given object 52 in the field of view 50 .

[0089] - Detecting a change in the scaling of the object frame 54 of the object 52 relative to one or more previous images or frames.

[0090] - Detecting frame motion, frame displacement or frame translation (in pixels), angular position and / or angular velocity of a frame edge of an object frame 54 of a given object 52 in the field of view 50 in a lateral direction relative to a previous image, frame, portion or segment.

[0091] The process already described above with the aid of methods S and I for n time increments up to the target time span (as forecast time span) is used as the basis for the concrete forecast or prediction.

[0092] In each of the n periods (ie for each time increment until a predefined target time span is reached), the predicted frame width is compared with the image width of the camera image 51 or with the width of a predefined or predetermined section or segment 51 ′.

[0093] The ratio between the frame width and the image width or the segment width in the image can be used as the basis for the criterion for determining the criticality. As long as the frame width 55b occupies a certain width 51b of the image 51 or the segment 51' and thus exceeds a threshold value (e.g. 35%), the criticality is considered to be 100% and the criticality is set to 100%.

[0094] If, on the other hand, for example the predicted or anticipated frame 55 lies partially or completely outside the image 51 or the section 51 ′, the situation is not critical and the criticality is set to 0%.

[0095] Based on the position and / or speed of the pedestrian 52', the pedestrian's acceleration capability can be determined with the aid of a pedestrian model. In this case, a second expanded object frame 56 can be expanded for the pedestrian 52' as object 52, which has a possible lateral stay area of ​​the pedestrian 52' in the image 51 or the section 51' of the image 51 taking into account the pedestrian's acceleration capability.

[0096] In order to determine the speed of pedestrian 52 ′ as object 52 and to convert the acceleration capability into pixel coordinates or angle coordinates, an assumption can be made about the size of pedestrian 52 ′ as object 52 . For example, a conventional size of 1.75 m can be used as a basis.

[0097] - The value of the previously calculated criticality can also be reduced or minimized by the proportion by which the last predicted or predicted frame 55 is located outside the image 51, the frame, or the corresponding segment 51' or a part thereof. For example, if 50% of the frame width 55b (i.e. if the possible stay area represented by the object frame 55 taking into account the pedestrian model) is located outside the image 51 or the frame or the given part or segment 51', the criticality is correspondingly reduced. The further away the future prediction is, the greater the criticality is reduced (due to the potentially larger stay area located outside the image 51 or the segment 51'). This behavior is desirable because the uncertainty about the characteristics of the pedestrian 52' increases with increasing time predictions into the future.

[0098] If the predicted object frame 55 of a pedestrian as object 52 reaches a critical size (eg reaches a threshold value: 35% of the frame or image width), the prediction is terminated in this cycle of iteration I, since a collision is already possible at this point in time.

Claims

1. An operation assistance method (S) for a working device (1), in which: (S1) acquiring an object frame (54) of an object (52) or data representing such an object frame (54) at successive moments in the field of view of the working device (1); (S2) determining, from an object frame (54) of a given object (52) in images captured successively or directly successively, a current change in the scale of the object frame (54) of the corresponding object (52) or a variable derived therefrom and determining a current change in the lateral position of the object frame (54) of the corresponding object (52) in the images; (S3) determining a future predicted object frame (55) based on a current scale change of an object frame (54) of a corresponding object (52) in the image or a parameter derived therefrom and a current lateral position change; (S4) determining and evaluating the position of the predicted object frame (55) and / or the ratio of the lateral extension of the predicted object frame (55) to the lateral extension of the detected field of view (50) and / or the lateral extension of the captured image or a section of the image; (S5) based on the result of the evaluation, (i) determining whether the object (52) on which the predicted object frame (55) is based is critical with respect to a possible collision, and / or, (ii) controlling or regulating the operating state of the working device (1), Where the following steps are performed for each time increment: (11) Reset or preset the value to be calculated according to the allocation rule: Scaling alt :=Scaling neu BoxTranslationX alt :=BoxTranslationX neu BoxBreite alt :=BoxBreite neu BoxPositionLinks alt :=BoxPositionLinks neu BoxPositionRight alt :=BoxPositionRight neu (I2) Update the scaling ratio according to the following allocation rule: Scaling neu :=1 / (2-scaling alt ) (I3) Update the horizontal object box translation according to the following distribution rule: BoxTranslationX neu :=BoxTranslationX alt ×Skalierung alt (I4) Update the horizontal object box width according to the following allocation rule: BoxWidth neu :=BoxPositionRight alt -BoxPositionLeft alt (I5) Predict the horizontal box position according to the following assignment rule: BoxPositionLeft neu :=BoxPositionLeft alt +BoxTranslationX neu -0.5×BoxWidth neu ×(Scaling neu -1) / Scaling neu BoxPositionRight neu :=BoxPositionRight alt +BoxTranslationX neu +0.5×BoxWidth neu ×(Scaling neu -1) / Scaling neu Among them, SkalierUng alt 、Skalierung neu Indicates the old or new scale of the object box (54), BoxTranslationX alt 、BoxTranslationX neu Indicates the old or new displacement of the object box (54), BoxBreite alt 、BoxBreite neu Indicates the old or new width of the object box (54), BoxPositionLinks alt 、BoxPositionLinks neu Indicates the old position or the new position of the lower left corner of the object box (54) as the first x coordinate of the corresponding object box (54), and BoxPositionRechts alt 、BoxPositionRechts neu The old position or the new position of the lower right corner of the object frame (54) is represented as the second x-coordinate of the corresponding object frame (54), or their values ​​are represented.

2. The operation assistance method (S) according to claim 1, wherein in the step (S1) of obtaining the object frame (54) and / or obtaining data representing the object frame (54) or for the step of obtaining the object frame and / or obtaining data representing the object frame, (S1a) optically two-dimensionally and / or monocularly detecting the field of view (50) of the underlying working device (1) by recording images that are sequential in time, (S1b) In the images recorded successively, at least one object (52) is determined and an object frame (54) assigned to the object (52) is determined.

3. An operation assistance method (S) according to any one of the preceding claims, in which a future predicted object frame (55) is determined at least for the current last detected image by iteratively determining and updating a value of a scaling factor of the corresponding object frame (54), a value of coordinates of the corresponding object frame (54), a value of a translation of the corresponding object frame (54), and a value of a lateral width of the corresponding object frame (54) over multiple time increments up to a prediction time span.

4. The operation assistance method (S) according to claim 3, in which steps (I1) to (I5) are performed for each time increment in the order described.

5. The operation assistance method according to claim 3 , wherein the following calculation rule is implemented to determine the new frame position: in, BoxPositionLinks neu and BoxPositionLinks aktuell or BoxPositionRechts neu and BoxPositionRechts aktuell is the new or current position of the left or right box edge, BoxGeschwindigkeitLinks aktuell or BoxGeschwindigkeitRechts aktuell is the currently measured angular velocity of the left or right frame edge, NormGeschwindigkeit aktuell is the so-called normalized frame velocity currently measured, is the prediction time corresponding to the prediction time step, wherein the NormGeschwindigkeit is derived from the calculated scaling change of the object frame aktuell .

6. An operation assistance method (S) according to claim 1 or 2, in which the object (52) on which the predicted object frame (55) is based is determined to be critical with respect to a possible collision if the proportion of the width of the predicted object frame (55) to the width of the image on which it is based or the width of a predetermined section of the image exceeds a predetermined first threshold value.

7. An operation assistance method (S) according to claim 6, in which the value of the criticality determined for the object (52) is reduced by the proportion by which the object frame (55) predicted for the object (52) lies outside the underlying image or the predetermined section in terms of its width.

8. An operation assistance method (S) according to claim 1 or 2, in which the object (52) on which the predicted object frame (55) is based is determined to be non-critical with respect to a possible collision if the predicted object frame (55) is completely outside the image on which it is based or a predetermined section of the image.

9. An operation assistance method (S) according to claim 6, in which the object (52) on which the predicted object frame (55) is based is determined to have a criticality of 100% with respect to a possible collision if the proportion of the width of the predicted object frame (55) to the width of the image on which it is based or the width of a predetermined section of the image exceeds a predetermined first threshold value.

10. An operation assistance method (S) according to claim 8, in which the object (52) on which the predicted object frame (55) is based is determined to have a criticality of 0% with respect to a possible collision if the predicted object frame (55) is completely outside the image on which it is based or a predetermined section of the image.

11. A control unit (10) for an operation assistance system (100) of a working device (1), the control unit being configured to control and / or operate an operation assistance method (S) according to any one of the preceding claims, and / or the control unit being configured to operate the underlying operation assistance system (100) in accordance with the operation assistance method (S) according to any one of the preceding claims.

12. The control unit (10) according to claim 11, the working device being a vehicle.

13. An operation assistance system (100) for a working device (1), the operation assistance system being configured to carry out an operation assistance method (S) according to any one of claims 1 to 10.

14. The operation assistance system (100) according to claim 13, wherein the working device is a vehicle. 15 . The operating assistance system ( 100 ) according to claim 13 , comprising a control unit ( 10 ) according to claim 11 or 12 .

16. A working machine (1) comprising the operation assistance system (100) according to any one of claims 13 to 15. 17 . The working device ( 1 ) according to claim 16 , which is designed as a vehicle, a motor vehicle or a passenger car.

18. An application of an operation assistance method (S) according to any one of claims 1 to 10, a control unit (10) according to claim 11 or 12, an operation assistance system (100) according to any one of claims 13 to 15 and / or a working device (1) according to claim 16 or 17, wherein the operation assistance method, the control unit, the operation assistance system and / or the working device are used for pedestrian protection, cyclist protection, adaptive cruise control and / or avoidance systems or methods.

Images