Methods for monitoring load handling devices on industrial trucks, industrial truck and software program product

Abstract

The invention relates to a method for monitoring at least one load-handling device (14, 15), in particular forks, attached to a lifting device (12) of a forklift truck (10), wherein images (40) are captured with a camera (16) arranged such that it is moved vertically and / or horizontally by the lifting device (12) together with the at least one load-handling device (14, 15), which at least partially show the at least one load-handling device (14, 15). The invention further relates to a forklift truck (10) and a software program product. According to the invention, at least one image area (44, 45) in the images (40) is determined in which the at least one load-bearing means (14, 15) is depicted, and a position of the at least one load-bearing means (14, 15) is determined on the basis of the at least one determined image area (44, 45).

Description

[0001] The invention relates to a method for monitoring at least one load-handling device, in particular forks, attached to a lifting device of an industrial truck. Images are taken using a camera arranged such that it is moved vertically and / or horizontally by the lifting device together with the at least one load-handling device, which images at least partially show the at least one load-handling device. The invention further relates to an industrial truck and a software program product.

[0002] The present invention relates to the field of industrial trucks that have forks for load handling or as load-handling equipment. In a conventional industrial truck, the forks are usually hooked in. In this case, they can be unnoticed by another user and not detected by the vehicle's electronics. Mechanical influences can also change the position of the forks, or if the fork remains in the same position, it can hang crooked. Furthermore, the forks can also bend due to violent damage. If this process goes unnoticed, it can lead to damage, e.g., if a fork hits the outer block of a pallet. If this process is noticed during the handling attempt, a delay and a reduction in productivity occur in order to correct the incorrect fork position. There is currently no sensor technology for this situation.

[0003] Another application for industrial trucks is sensing the fork position, which can be actively adjusted if necessary, for example, using a fork positioner that changes the distance between the forks. According to the current state of the art, this requires additional sensor technology, which can be implemented, for example, using proximity sensors and an inclined plane. Using additional sensors, for example, a gentle end stop during fork adjustment can be realized. However, the use of additional sensors increases manufacturing costs.

[0004] The present invention is based on the object of covering the aforementioned applications of monitoring and determining the position of the load handling device in a robust and safe manner with little additional effort.

[0005] This object is achieved by a method for monitoring at least one load-handling device, in particular forks, attached to a lifting device of an industrial truck, wherein images are taken using a camera which is arranged such that it is moved vertically and / or horizontally by the lifting device together with the at least one load-handling device, which images at least partially show the at least one load-handling device, wherein at least one image area in the images is determined in which the at least one load-handling device is depicted, and a position of the at least one load-handling device is determined based on the at least one specific image area.

[0006] In the context of the invention, the term "position" refers in particular to the position in the camera image. Absolute spatial positions cannot usually be derived from a single two-dimensional image, but are accessible using stereoscopic devices and methods.

[0007] Many industrial trucks are equipped with a camera that images the area in front of the forks, as well as parts of the forks or forks themselves. One example is the Jungheinrich fork camera, which was previously an analog camera. Newer models use digital cameras instead. The method according to the invention uses the digital images already available, thus eliminating any additional system overhead.

[0008] Until now, the analysis of camera images has been limited to displaying them on a screen within the driver's field of vision, which allows the driver to assess whether the load handling device is correctly positioned for a load to be picked up. However, the camera image alone does not provide the user with any reliable information about distances, angles, or fork spacing. Furthermore, it is not possible to determine whether the camera position or the fork position deviates from an original position.

[0009] It is also known that a line superimposed on the digital camera image forms a virtual line laser to further assist the operator in positioning the forks. Such a virtual line laser is superimposed at a specified position in the image and is usually fixed to the load-handling device. However, there is no automatic detection of the forks and no adjustment of the line's display position if the load-handling device is not in the expected position. Therefore, if the fork position changes, the line may be inaccurate.

[0010] The method according to the invention provides a remedy for this. Information about the position of the forks is derived from the digital image of a camera directed at or attached to the load-handling device. Position determination is performed by detecting the image areas in the camera image where the load-handling device is depicted. From the recorded image data, downstream image processing calculates whether the fork position has changed relative to a previously known position. This also makes it possible to determine absolute position information for the load-handling device.

[0011] In general, load handling devices, such as forks, will have a variety of optically significant features due to their shape, structure and surface finish, but also due to their use, including, for example, scratches or other structures, so that a position analysis is also possible within the image areas occupied by the load handling devices.

[0012] In a further development, the position of at least one load-handling device is monitored to determine whether it has changed from a predetermined target position. If such a change is detected, the operator is notified, for example, possibly with an indication of the target position and / or the correction to be made to the position of the load-handling device.

[0013] In a first alternative of image analysis, an optical flow is determined in images of an image series between which a movement has occurred relative to the surroundings of the industrial truck, and at least one image area in which no optical flow occurs is determined as the image area in which the at least one load handling device is depicted.

[0014] In image processing, optical flow refers to a vector field that specifies the direction and velocity of movement for each pixel in an image sequence. Optical flow can be understood as the velocity vectors of visible objects projected onto the image plane. Optical flow analysis is used, for example, in connection with optical navigation in vehicles and optical computer mice.

[0015] A direct determination of the optical flow is possible using a suitable method, e.g. using the Lucas-Kanade method, the Horn-Schunck method or the Buxton-Buxton method. Such methods are stored in software libraries for image processing such as OpenCV. The determination of the optical flow of an image pair is based on assigning each pixel a motion vector that occurs due to an object movement or a camera movement from one image to the other. This evaluation does not require any knowledge of the movement of the camera with the lifting device or with the industrial truck. In this application, the optical flow can be statistically evaluated over a longer period of time. This way, it can be seen that in the image areas in which the load handling device is depicted, no optical flow occurs, or in the case of false detection, correspondingly less frequently.

[0016] In a further alternative of image analysis, significant image elements in images of an image series are determined and tracked across subsequent images of the image series between which a movement relative to the surroundings of the industrial truck has taken place, and at least one image area in which one or more of the tracked significant image elements are not detected during tracking is determined as the image area in which the at least one load handling device is depicted.

[0017] By determining and tracking significant pixels along a sequence of images using a suitable method, such as SIFT (Scale-Invariant Feature Transform) or SURF (Speeded-Up Robust Features), as implemented in OpenCV, for example, significant pixels can be described in a scale-invariant form, so that corresponding pixels can be found again in a subsequent image with a very high probability. If the camera or object moves accordingly, the points are then detected at a correspondingly different position in the image. A subsequent comparison then provides the information that corresponding points have been found. It can be observed that, for example, during a lifting process, certain pixels disappear behind the forks (occlusion) and then reappear. From this information, the image areas in which the forks are located can be determined.

[0018] In a further alternative of the image evaluation, a disparity analysis is carried out with regard to different images of an image series, between which the position of the camera was changed relative to the surroundings of the industrial truck, and at least one image area is determined as the image area in which the at least one load handling device is depicted, in which the disparity analysis results in a vanishing disparity and / or an apparent infinite distance.

[0019] Disparity analysis is well known in the field of three-dimensional vision. Between two two-dimensional images taken with two cameras spaced apart by a certain base length, or between consecutive images taken with one camera moved by the base length, objects visible in both images exhibit a shift characteristic of the base length and the distance from the object to the camera, known as disparity. If the base length is known, the distance of the object imaged with a specific pixel from the camera can be calculated from the observed disparity.

[0020] An approach for generating a point cloud from two two-dimensional images and the known intrinsic motion of the camera between the recordings for an industrial truck is generally described, for example, in the applicant's European patent application EP 3 981 731 A1. When the approach based on disparity analysis is used in the context of the present task, a disparity of approximately zero and consequently a seemingly infinite distance results for the image areas with imaged forks, which remain essentially stationary relative to the camera. In this respect, the forks elude classical absolute location determination in such a case, since they are stationary relative to the fork camera.

[0021] Compared to the optical flow determination alternative described above, disparity analysis has the advantage that, due to the known inherent movement of the camera caused by vertical or horizontal movement of the lifting device, by cornering of the industrial truck, or a combination of these movements, the search space for correspondences is greatly restricted and thus the required computational effort is significantly reduced. For the purpose stated here, the exact base length for the disparity analysis does not even need to be known; rather, knowing the direction of the disparity is sufficient, since the focus is on detecting image areas with vanishing disparity and not on determining the absolute distance to objects in the vicinity of the industrial truck.

[0022] In practice, vibrations during operation of the industrial truck repeatedly cause slight vibrations of the forks, so that the disparity in the area of ​​the forks is not always exactly zero. However, such disparities are small compared to those of the background and will disappear when averaging. Thus, in embodiments, an upper limit for the magnitude of the disparity can be used, which separates the small disparities that occur in practice and cancel each other out on average in the image areas to be determined from the significantly larger disparities for objects in the vehicle's surroundings. This leads to the desired goal of determining the image areas in which the load handling devices are shown.

[0023] In embodiments, two or more of the optical flow determination, the tracking of significant image elements, and the disparity analysis can be combined to determine the image area in which the at least one load-carrying device is depicted. This makes the detection of these image areas more robust.

[0024] Furthermore, the determination of the image area in which the at least one load-handling device is depicted can be repeated over a plurality of image series and / or images and statistically evaluated. This measure also further increases the robustness of the method.

[0025] In a further development of the method, at least one expected image area in which the at least one load-handling device is located or may be located is preset or learned over a predetermined period, wherein subsequent determinations of the at least one image area in which the at least one load-handling device is located are compared with the at least one expected image area. This measure is particularly suitable for monitoring the position of the load-handling device against changes, for example due to damage or incorrect suspension or attachment of the load-handling device. Thus, if the determined at least one image area deviates from the expected image area, an operator is advantageously notified and, in particular, instructed to check the position of the at least one load-handling device.

[0026] In a further development, in a load handling device which comprises two tines whose distance from one another is adjustable, an image area is determined for each of the two tines, in which area one of the two tines is depicted, an actual distance between the two tines is determined from the image areas, which is in particular displayed and / or in particular the actual distance is controlled towards a target distance, in particular braking before an end stop. These measures can be used for industrial trucks with a fork adjustment device which can adjust the distance between the forks. The image analysis for the current fork spacing can be used both for monitoring and as input for controlling the fork adjustment, so that additional sensors for this purpose are unnecessary.Although the prongs are no longer strictly stationary in the image in this case, their adjustment direction and adjustment speed are predictable, namely only in the horizontal direction, so that this can be taken into account in image evaluation, for example with optical flow, tracking of significant image elements or disparity analysis.

[0027] In a further development of the method, starting from a current position of at least one load handling device, one or more movement lines or movement curves are superimposed on the recorded images and displayed on a screen, along which the at least one load handling device can move depending on a current control position of the industrial truck. This is similar to the familiar lines or curves in reversing cameras of other vehicles and can be used to graphically display, based on the current position of the load handling device known from the image analysis, for example the tips of the forks, where these would end if the vehicle continued to travel in the current control position. This makes load handling processes easy for an operator to plan and control in advance.

[0028] Likewise, a line between the fork tips determined from the image analysis can be projected over the image as a virtual line laser, making it easier for the operator to control the system. A virtual line laser generated in this way is accurate even if a fork is not in the expected position.

[0029] The object underlying the invention is also achieved by an industrial truck with a lifting device, a camera which is arranged on the industrial truck so that it can move with the lifting device such that at least one load-handling device attached to the lifting device is at least partially shown in images from the camera, and an evaluation unit which is connected to the camera and is designed to evaluate images from the camera, in which the evaluation unit is designed to determine at least one image area in the images in which the at least one load-handling device is depicted, and to determine a position of the at least one load-handling device based on the at least one specific image area. In particular, the evaluation unit is designed to carry out a method as described above according to the present invention.

[0030] Furthermore, the object underlying the invention is also achieved by a software program product with program code means configured to execute a previously described method according to the invention when executed on an evaluation unit of a aforementioned industrial truck. Thus, the industrial truck and the software program product also implement the features, properties, and advantages mentioned for the method.

[0031] Further features of the invention will become apparent from the description of embodiments of the invention together with the claims and the accompanying drawings. Embodiments of the invention may incorporate individual features or a combination of several features.

[0032] Within the scope of the invention, features marked with “in particular” or “preferably” are to be understood as optional features.

[0033] The invention is described below, without limiting the general inventive concept, using exemplary embodiments with reference to the drawings, whereby express reference is made to the drawings for all details of the invention not explained in more detail in the text. They show: Fig. 1 an industrial truck with a forklift camera picking up pallets from the rack, Fig. 2 an image of a fork camera, Fig. 3 an image of a fork camera with a virtual line laser displayed, Fig. 4 an image sequence to explain the method of tracking significant image elements, Fig. 5 images to explain the procedure for determining an optical flow, Fig. 6 an image of a fork camera with a virtual line laser displayed, Fig. 7 an image of a fork camera with superimposed movement lines and Fig. 8 an image of a fork camera with displayed movement curves.

[0034] In the drawings, identical or similar elements and / or parts are provided with the same reference numbers, so that a repeated presentation is omitted.

[0035] Fig. 1 shows an industrial truck 10 equipped with a lifting device 12, also called a mast. A height-adjustable load-handling device 14 is attached to the lifting device 12. The load-handling device 14 has a fork camera 16 on one of its forks or forks, the two-dimensional image area 18 of which extends beyond the fork tips. In the schematic diagram, the image area 18 extends symmetrically to the fork tip, corresponding to a camera attached to the side of the fork. In principle, it is also possible to provide a fork camera mounted below the fork, the image area 18 of which is limited below the fork. Alternatively, a camera could also be attached to a movable part of the lifting device 12, which is connected to the load-handling device 14.

[0036] The industrial truck 10 is located in front of a rack 20 with a load 22 to be picked up, which rests on a schematically illustrated pallet 24 as the load carrier. By lifting the load-handling device 14 by a distance 26 (vertical double arrow), the image area 18 transitions into an image area 18'. The image areas 18 and 18' overlap in an overlap area 28. For the overlap area 28, two images of the same object from different perspectives are thus available, which allow a distance evaluation, for example using a disparity analysis. Using the distance evaluation 30, for example, the distance of the camera 16 from the front of the rack can be calculated. The distance of the fork tip from the rack front can thus be deduced from the distance of the camera 16 from the fork tip.

[0037] Since the fork camera 16 moves together with the load handling device 14, the position of the fork tip does not change in the two images. This results in a disparity of zero, i.e., a seemingly infinite distance, so that those image areas that exhibit a vanishing disparity or a seemingly infinite distance can be equated with the forks.

[0038] In Fig. 1 shows a comparatively large travel path, in particular a relatively large distance 26, which leads to a relatively small overlap area 26. It is preferred if smaller distances 26 are used and the method is carried out iteratively. In this way, even larger travel paths, as in Fig. 1, is handled iteratively by the sequence of several smaller travel paths.

[0039] Fig. Figure 2 schematically shows an image 40 of a forward-facing fork camera 16, which is attached to the inside of a right fork 15 of a pair of forks 14, 15. Due to the resulting perspective distortion, the left fork 14 is depicted as narrow, while the right fork 15 widens significantly toward the edge of the image. The forks or load handling devices 14, 15 divide the image into three image areas, namely the image areas 44, 45 with the load handling devices 14, 15 and the background 42. Each movement of the industrial truck 10 or the lifting device 12 results in the image in the background 42 changing, while essentially no change occurs in the image areas 44, 45. Although changes in the lighting may occur, these have no effect on an offset of the corresponding image points in the image areas 44, 45.A vertical movement of the lifting device 12 leads in this way to an opposite vertical movement of the structures in the background 42, while a horizontal movement of the load handling device 14, 15 or driving around a curve with the industrial truck 10 leads to a horizontal movement of the background structures in the image 40.

[0040] Fig. 3 shows schematically an image 40 of a fork camera 16 as in Fig. 2, but in this case with a virtual line laser 44, which, as usual, is displayed at a predetermined position in the image, independent of the actual position of the load-carrying devices. As in Fig. As can be seen in Figure 3, the line laser 44 is shown horizontally in Figure 40, but the ends of the fork tines 14 and 15 are not at the same height. This can be caused, for example, by incorrect suspension of one of the two fork tines or by one of the two fork tines being bent in a load handling accident. Fig. In the situation shown in Figure 3, the line laser 44 is not optimally aligned with the actual position of the load handling equipment. If a user of the industrial truck 10 orients himself using the line laser 44, he risks hitting obstacles with the tips of the forks and causing a load handling accident.

[0041] Fig. Figure 4 shows an image sequence to explain the method of tracking image features, so-called significant image elements 46. In the images 40 of the image sequence, a significant image element 46 is detected in the background 42 of the respective image 40. The images 40 originate from a fork camera 16 mounted on the right fork. The lifting device 12 of the industrial truck 10 performs a lifting movement, which causes elements in the background 42 to move vertically downward in the image 40. This is indicated by the arrows pointing vertically downward.

[0042] A significant image element 46 in the background 42 of the first image 40 is in the left image of the Fig. 4 is indicated by a star. Such a significant image element 46 can be identified, for example, by edge detection or similar known image analysis methods and described in a scale-invariant form using a suitable method, for example SIFT (Scale-Invariant Feature Transform) or SURF (Speeded-Up Robust Features), as implemented, for example, in OpenCV. Typically, a plurality of significant elements 46 are detected, which are, in particular, encoded in a scale-invariant manner and can therefore be identified again in subsequent images 40 of the image sequence.

[0043] The Fig. The image sequence shown in Figure 4 shows that during the lifting movement, the significant element 46 temporarily disappears behind the left fork in the image area 44 (middle image) and reappears below the fork as the lifting movement continues. The points in the image where significant elements 46 disappear and reappear mark the edges of the image areas 44 and 45, in which the left and right fork tines 14, 15 are visible. With a sufficient number of different significant elements 46, the image areas 44, 45 can be determined with high accuracy.

[0044] Fig. Figure 5 shows images 40 to explain the method for determining an optical flow in the images 40 of a forked camera 16. In this case, the images 40 are evaluated for an optical flow 50 that occurs between consecutive images 40 of an image sequence. Various methods exist for this, for example, the Lucas-Kanade method, the Horn-Schunck method, or the Buxton-Buxton method. The result of the optical flow analysis is shown in Fig. 5 in two examples. The left-hand illustration shows the result of a lifting movement of the lifting device 12 of the industrial truck 10. By raising the fork camera 16 on the right fork, the background appears to move downwards, which is symbolized by a vector field with vertically downward-pointing arrows. In areas 44 and 45 of the left and right forks, no movement occurs, as the forks move together with the fork camera 16. Due to the lack of optical flow in areas 44 and 45, the motion vectors there have a length of zero, which is indicated by the dots in this area.

[0045] In the right picture 40 of the Fig. 5, a lifting movement is added to a horizontal movement, for example, by a horizontal movement of the load-handling elements on the lifting device 12 or by the industrial truck 10 describing a curve in its movement. The motion vectors are accordingly aligned obliquely to the background, thus representing a linear combination of horizontal and vertical movement directions. For the left and right forks 14, 15 and the corresponding image areas 44, 45, a vanishing optical flow results, which is used to identify the image areas 44, 45 with the load-handling devices.

[0046] Fig. 6 shows an image 40 of a fork camera 16 with a virtual line laser 44 displayed. In contrast to Fig. 3, in this case, an analysis of the positions of the forks 14, 15 has taken place, and the line laser 44 is not displayed at a predetermined location in image 40, but is generated and displayed between the upper front corners of the forks 14, 15 or image areas 44, 45. Thus, the operator of the industrial truck 10 has correct information about the position of the forks and, with the aid of the correctly displayed line laser 44, can correctly perform load handling operations without risking hitting a load to be picked up or a rack with a bent fork or an incorrectly attached fork.

[0047] In the Fig. 7 and Fig. 8 shows two further exemplary embodiments of images 40, as they can be displayed to an operator of an industrial truck 10. In both cases, these are lines or auxiliary lines superimposed on the respective image 40, which make the expected path of the forks 14, 15 visible, so that a driver of the industrial truck 10 can estimate where the tips of the forks 14, 15 will be when continuing a current movement. In the case of Fig. 7 is a straight-line movement forward, indicated by movement lines 52, which take into account the perspective alignment of the camera position and perspective, while in Fig. 8 movement curves 54 corresponding to a currently set left turn of the industrial truck 10 are shown.

[0048] All mentioned features, including those revealed solely in the drawings as well as individual features disclosed in combination with other features, are considered essential to the invention, both individually and in combination. Embodiments according to the invention may be fulfilled by individual features or a combination of several features. List of reference symbols 10 industrial trucks 12 Lifting device 14 load handling equipment 15 load handling equipment 16 Camera 17 Evaluation unit 18, 18' image area 19 screen 20 shelves 22 Last 24 pallets 26 route 28 Overlap area 30 distance 40 images 41 shelf 42 Background 44 Image area 45 image area 46 significant image element 48 Direction of movement 50 optical flow 52 movement lines 54 movement curves QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] EP 3 981 731 A1

[0020]

Claims

[1] Method for monitoring at least one load-handling device (14, 15), in particular forks, attached to a lifting device (12) of an industrial truck (10), wherein images (40) showing at least one load-handling device (14, 15) at least partially are recorded using a camera (16) which is arranged such that it is moved vertically and / or horizontally by the lifting device (12) together with the at least one load-handling device (14, 15), wherein at least one image area (44, 45) in the images (40) is determined in which the at least one load-handling device (14, 15) is depicted, and a position of the at least one load-handling device (14, 15) is determined on the basis of the at least one specific image area (44, 45). [2] Method according to claim 1, characterized by that the position of the at least one load-carrying means (14, 15) is monitored to determine whether it has changed from a predetermined target position. [3] Method according to claim 1 or 2, characterized by that an optical flow (50) is determined in images (40) of an image series between which a movement relative to the surroundings of the industrial truck (10) has taken place, and at least one image area (44, 45) in which no optical flow occurs is determined as the image area (44, 45) in which the at least one load-handling means (14, 15) is depicted. [4] Method according to one of claims 1 to 3, characterized bythat significant image elements (46) in images (40) of an image series are determined and tracked across subsequent images (40) of the image series, between which a movement relative to the surroundings of the industrial truck (10) has taken place, and at least one image area (44, 45) in which one or more of the tracked significant image elements (46) are not detected during the tracking is determined as the image area (44, 45) in which the at least one load-handling means (14, 15) is depicted. [5] Method according to one of claims 1 to 4, characterized bythat a disparity analysis is carried out with regard to different images (40) of an image series, between which the position of the camera (16) relative to the surroundings of the industrial truck (10) has been changed, and at least one image area (44, 45) is determined as the image area (44, 45) in which the at least one load-handling means (14, 15) is depicted, in which the disparity analysis results in a vanishing disparity and / or an apparent infinite distance. [6] Method according to one of claims 3 to 5, characterized by that two or more of the determination of the optical flow (50), the tracking of significant image elements (46) and the disparity analysis are combined to determine the image area (44, 45) in which the at least one load-carrying means (14, 15) is imaged. [7] Method according to one of claims 3 to 6, characterized bythat the determination of the image area (44, 45) in which the at least one load-carrying means (14, 15) is imaged is repeated over a plurality of image series and / or images (40) and statistically evaluated. [8] Method according to one of claims 1 to 7, characterized by that at least one expected image area in which the at least one load-handling means (14, 15) is located or may be located is preset or learned over a predetermined period, wherein subsequent determinations of the at least one image area (44, 45) in which the at least one load-handling means (14, 15) is located are compared with the at least one expected image area. [9] Method according to claim 8, characterized bythat if the determined at least one image area (44, 45) deviates from the expected image area, an operator is notified, in particular is instructed to check the position of the at least one load-handling means (14, 15). [10] Method according to one of claims 1 to 9, characterized by in that, in a load-handling device (14, 15) which comprises two tines whose distance from one another is adjustable, an image area (44, 45) is determined for each of the two tines, in which area one of the two tines is depicted, an actual distance between the two tines is determined from the image areas (44, 45), which is in particular displayed and / or wherein in particular a control of the actual distance towards a desired distance takes place, in particular a braking before an end stop. [11] Method according to one of claims 1 to 10, characterized bythat, starting from a current position of the at least one load-handling means (14, 15), one or more movement lines (52) or movement curves (54) are superimposed on the recorded images (40) and displayed on a screen (19), on which screen(s) the at least one load-handling means (14, 15) can move depending on a current control position of the industrial truck (10). [12] Industrial truck (10) with a lifting device (12), a camera (16) which is arranged on the industrial truck (10) so as to be movable with the lifting device (12) that at least one load-handling means (14, 15) attached to the lifting device (12) is at least partially shown in images (40) of the camera (16), and an evaluation unit (17) which is connected to the camera (16) and is designed to evaluate images (40) of the camera (16), wherein the evaluation unit (17) is designed to determine at least one image area (44, 45) in the images (40) in which the at least one load-handling means (14, 15) is depicted, and to determine a position of the at least one load-handling means (14, 15) on the basis of the at least one specific image area (44, 45). [13] Industrial truck (10) according to claim 12, characterized by that the evaluation unit (17) is configured to carry out a method according to one of claims 1 to 11. [14] Software program product with program code means which are designed to execute a method according to one of claims 1 to 11 when executed on an evaluation unit (17) of an industrial truck (10) according to claim 12 or 13.

Images