CUTTING MACHINE WITH OVERVIEW CAMERA

Description

[0001] The invention relates to a cutting machine with a camera, in particular a cutting machine designed for cutting objects that have a surface with a graphic design and optical registration features. These objects can, in particular, be printed sheets of paper, cardboard or similar materials, plastic films or cloths, or the like.

[0002] Machines of this type are described, for example, in documents EP 1 385 674 B1 and EP 2 488 333 B1. Such a cutting machine has a work surface designed to hold at least one object, and a work assembly movable above the work surface, comprising a blade or other cutting device for cutting objects located on the work surface. Furthermore, a camera unit is positioned relative to the work surface, particularly above it, such that its field of view encompasses the entire work surface ("overview camera"). Based on the positions of the optical registration features in an image from the overview camera, a cutting path can then be defined depending on a selected cutting job.

[0003] The term "cutting" does not necessarily mean complete severing, so a "cutting order" may also include perforating or folding the object or a similar work step that can be carried out with a machine of the appropriate type.

[0004] One objective of the invention is to provide an improved cutting machine.

[0005] In particular, one object of the invention is to provide a cutting machine by means of which cutting orders can be carried out more quickly.

[0006] Another task is to provide a cutting machine that allows cutting orders to be carried out with less personnel or with a higher degree of automation.

[0007] Another task is to provide a cutting machine with an overview camera, which allows cutting paths to be defined more quickly and / or more accurately.

[0008] Another task is to provide such a cutting machine that produces less waste.

[0009] The problem is solved by a device according to claim 1 and a computer program according to claim 7.

[0010] The invention relates to a cutting machine designed for cutting objects with a flat surface, wherein the surface has a graphic design with optical registration features. The cutting machine according to the invention has a work surface designed to receive at least one object, a first camera unit arranged relative to the work surface such that its field of view encompasses the entire work surface, and a working unit movably arranged above the work surface and comprising at least one cutting device for cutting the at least one object.

[0011] Furthermore, a processing unit with circuitry and program code for controlling the cutting machine is provided, comprising a storage unit for storing cutting orders for specific objects. The processing unit includes circuitry and program code for evaluating images from the first camera unit and is designed to recognize the registration features of at least one object in an image from the first camera unit. It is also designed to define a cutting path for the cutting device based on at least one stored order and the positions of the registration features in the image.

[0012] The registration features can be in the form of registration marks specifically designed for use with the cutting machine to determine the object's position and orientation relative to the work surface. The processing unit is then designed to recognize the registration marks on the surface of the at least one object in an image from the first camera unit and to define the cutting path based on the positions of the registration marks.

[0013] In one embodiment, the computing unit is designed to select an order based on recognized registration characteristics or their positions.

[0014] A first aspect of the disclosure concerns a cutting machine in which reference marks are provided on the work surface and in the camera's field of view, with the help of which a more precise positioning of the objects is made possible.

[0015] A second aspect of the disclosure concerns a cutting machine in which a known material thickness of the objects is taken into account when determining the position.

[0016] A third aspect of the revelation concerns a cutting machine in which the camera is designed to capture and superimpose multiple images of the identical scene.

[0017] A fourth aspect of the revelation concerns a cutting machine in which an additional, movable second camera, whose field of view covers a small section of the work surface, is used together with the first camera to determine the position of the objects.

[0018] A fifth aspect of the revelation concerns a cutting machine in which an additional, movably arranged second camera, whose field of view each covers a small section of the work surface, can be used to calibrate the first camera.

[0019] A sixth aspect of the disclosure relates to the invention, namely a cutting machine in which further information is used to determine the position of the registration features on the work surface more quickly and / or more accurately, by means of which a sub-area of ​​the work surface is defined as the region of interest.

[0020] In a cutting machine according to the first aspect, reference features are additionally arranged in a known positioning and distribution relative to the working surface and in the field of view of the first camera, wherein the computing unit is designed to recognize the reference features in the image of the first camera unit and to define the cutting path also based on relative positions of the registration features and the reference features in the image of the first camera unit.

[0021] In one embodiment, the relative positions of the reference features and the working group are known to each other.

[0022] In another embodiment, the computing unit is designed to check the alignment of the first camera unit relative to the work surface based on the positions of a large number of reference features in the image of the first camera unit.

[0023] In a cutting machine according to the second aspect, information about the material thickness of the object to be cut is provided to the computing unit, and the computing unit is designed to define the cutting path also based on the information about the material thickness.

[0024] In one embodiment, the computing unit is designed to determine the positions of the registration features based on the image from the first camera unit and on information about the material thickness.

[0025] In another embodiment, the material thickness can be determined by the cutting machine itself, in particular by means of a camera.

[0026] In another embodiment, the material thickness is provided together with the corresponding application in the storage unit, in particular as part of the application.

[0027] In a cutting machine according to the third aspect, the computing unit is designed to jointly evaluate at least two images of the work surface taken at different times by means of the first camera unit and to determine the positions of the registration features by evaluating the at least two images.

[0028] In one embodiment, the first camera unit is designed to take at least two images of the work surface with different exposure times at different times, and a high-contrast image is created based on the at least two images.

[0029] In another embodiment, the computing unit is designed to create high-contrast images from a series of exposures of several images from the first camera unit.

[0030] In another embodiment, the first camera unit is designed to capture high-contrast images.

[0031] In a further embodiment, the first camera unit is configured to provide high-contrast images of the work surface, using different exposure times for recording, and the computing unit has a circuit and program code for evaluating the high-contrast images of the first camera unit and is configured to recognize registration features of the at least one object in a high-contrast image.

[0032] According to the fourth and fifth aspects of the disclosed cutting machine, it additionally features an optical sensor unit oriented towards the work surface and arranged to be movable relative to the work surface such that the optical sensor unit can assume a multitude of positions, in each of which a detection range of the optical sensor unit encompasses a portion of the work surface. The processing unit additionally includes a circuit and program code for evaluating data from the optical sensor unit.

[0033] In a cutting machine according to the fourth aspect, the control unit is designed to to capture the positions of at least a multitude of the registration features relative to each other using the image from the first camera unit as relative positions, to determine the positions of a first subset of the multitude of registration features relative to the work surface using the data from the optical sensor unit as absolute positions, and to determine the positions of a second subset of the multitude of registration features relative to the work surface based on the captured relative positions and the determined absolute positions.

[0034] In one embodiment, the optical sensor unit is designed as a second camera unit.

[0035] In another embodiment, the optical sensor unit is designed as part of the working group.

[0036] In another embodiment, the detection areas of the optical sensor unit together encompass the entire working surface.

[0037] In another embodiment, the first camera unit is designed as a line scan camera, with the field of view extending over the entire width of the work surface.

[0038] In another embodiment, the control unit is designed to determine the position of an object on the work surface and / or to define a cutting path based on the positions of registration features relative to the work surface.

[0039] In a cutting machine according to the fifth aspect, the work group and the optical sensor unit are movable relative to the work surface by means of the same traversing mechanism. The cutting machine has a calibration function for the first camera unit, and within the scope of this calibration function, the cutting machine is designed to to determine the positions of a large number of points using the optical sensor unit, to record the same points using the first camera unit, and to calibrate the first camera unit with the positions determined by the optical sensor unit as target positions.

[0040] In one embodiment, the work surface is designed as a calibration work surface, and the multitude of points are optical markings on the calibration work surface.

[0041] In another embodiment, the points of the plurality of points are designed as grid points specifically intended for use with the calibration functionality.

[0042] In another embodiment, the cutting machine is designed, within the scope of the calibration functionality, to calibrate the traversing mechanism and the first camera unit relative to each other.

[0043] In another embodiment, the calibration functionality runs fully automatically after it is started, in particular where the start can be initiated by a user.

[0044] In another embodiment, the optical sensor unit is designed as a second camera unit.

[0045] In another embodiment, the optical sensor unit is designed as part of the working group.

[0046] In another embodiment, the detection areas of the optical sensor unit together encompass the entire working surface.

[0047] In a cutting machine according to the sixth aspect of the invention, an order to cut a specific object includes information about an expected position of the object on the work surface, wherein the control unit is designed to derive expected positions of registration features based on the expected position of at least one object and to define areas around the expected positions as a region of interest in an image of the first camera unit, outside of which no registration features are sought.

[0048] In one embodiment, the first camera unit is designed to image only one or more sub-areas that comprise the areas defined as the region of interest.

[0049] In another embodiment, the first camera unit has a zoom functionality and is designed to zoom in on a partial area.

[0050] In another embodiment, the control unit is designed to evaluate only the at least one area defined as a region of interest in a sub-area.

[0051] In the cutting machines for each of the aforementioned aspects, the registration features can be in the form of "registration marks" specifically designed for use with the cutting machine. These marks are designed to detect the position and orientation of the object relative to the work surface. The processing unit is then configured to recognize these registration marks on the surface of the at least one object in an image from the first camera unit and to define the cutting path based on the positions of the registration marks. The registration marks can, in particular, include geometric shapes.

[0052] In In one embodiment, the registration features can also include edges of the object.

[0053] In In another embodiment, the computing unit is designed to select a job based on recognized registration characteristics.

[0054] The cutting machine according to the invention is described in more detail below with reference to specific embodiments schematically depicted in the drawings, and further advantages of the invention are also discussed. Specifically, the drawings show: Fig. 1 a generic cutting machine with an overview camera; Fig. 2a-ce an image of the overview camera, cutting contours derived from the image, and a cutting path of the cutting device defined based on the image; Fig. 3 an exemplary embodiment of a cutting machine according to the first aspect of the invention; Fig. 4 an exemplary embodiment of a cutting machine according to the second aspect of the invention; Fig. 5 distortions in an image of the overview camera; Fig. 6 a shadow cast in an image of the overview camera; Fig. 7 an exemplary embodiment of a cutting machine according to the fourth aspect of the invention; Fig. 8 the working assembly of the cutting machine made of Figure 7 from above; Fig. 9a-b two exemplary embodiments of a cutting machine according to the fifth aspect of the invention; and Fig. 10a-b areas defined as Region of Interest according to the sixth aspect of the invention.

[0055] Figure 1Figure 1 shows a cutting machine of the type. As a flatbed cutting machine, it has a table with a flat working surface 10, on which two objects to be cut 40, 40' are placed as an example.

[0056] Above the work surface 10, a work unit 12 with a cutting tool 15, in particular a knife, is arranged. The work unit 12 is movable two-dimensionally relative to the work surface 10 by a motor in order to be able to reach any point on the work surface 10. For this purpose, the work unit 12 is movably attached in the X-direction to a beam 13, which in turn is movably attached to the table in the Y-direction.

[0057] A camera unit (overview camera 20) is positioned above the work surface 10 in such a way that images of the entire work surface 10 can be captured.

[0058] In particular, the cutting machine 1 may also have an oscillating driven cutting tool 15 and / or be designed for cutting multi-walled composite panels, as described for example in EP 2 894 014 B1.

[0059] The cutting machine 1 also has a computing unit 30. This can be designed as an external computer with a data connection to the machine 1, as shown here, or integrated into the machine 1 itself as an internal control unit. The overview camera 20 is designed to provide the computing unit 30 with data from captured images for evaluation.

[0060] The computing unit 30 comprises a processor with computing capacity and algorithms for controlling the cutting machine 1 according to a provided cutting order. The computing unit 30 also has a data memory for storing the cutting orders and, if necessary, other data.

[0061] As a starting point, one or more objects 40, 40' to be cut are placed on the work surface 10. It is either known exactly to which order or orders the objects 40, 40' lying on the work surface 10 are assigned, or it is at least known from which collection of orders this order or these orders originate.

[0062] Using the overview camera 20, an image of the entire work area is captured, and the position of the cutting contours is determined based on this image. This is done by detecting registration features in the graphically designed surface of the objects and their position. The registration features are stored as part of the order data in the respective order and can be either general features of the graphic design or, advantageously, registration marks specifically intended for registration. This is known from the prior art.

[0063] If the relevant order is not yet known, the corresponding order can first be determined using these markers and their position. If there are multiple orders, all relevant orders are determined. Then, the position of the cutting contours on the work surface is determined using the object positions and the relative position of the cutting contours in the order data. This is described in the Figures 2a-c Shown as an example.

[0064] Figure 2a shows a view from the overview camera 20 of the cutting machine 1 Figure 1The image 50 shows the entire working area of ​​the cutting machine, including the work surface 10, on which two objects 40, 40' to be cut are located. The work unit 12, which is preferably moved to the edge of the working area for image capture, can be seen at the top of the image. In this example, the objects to be cut are sheets 40, 40' (e.g., made of paper, cardboard, or plastic) and each has a graphic design 44, 44' with patterns and / or inscriptions on their side facing the camera. In the example shown, one is a crescent-shaped pattern 41, and the other is a heart-shaped pattern 41'. In addition, a number of registration marks 42 are depicted on each sheet 40, 40'. The registration marks 42 can be, in particular, geometric figures, e.g., as shown here, circular dots of a specific diameter.

[0065] Figure 2bThis shows the cutting contours 45, 45' of the sheets 40, 40' to be cut. The respective shape of the cutting contours 45, 45' and their relative position on the respective sheet 40, 40' are stored in the orders. Together with image 50 from Figure 2a The position and location of the cutting contours 45, 45' on the work surface can be determined.

[0066] Optionally, based on image 50, Figure 2a The control unit should also be able to assign the corresponding order to a sheet of 40, 40'.

[0067] Figure 2cFigure 1 illustrates an example of a movement path for the machine's cutting tool, generated by the control unit based on the determined positions of the cutting contours 45, 45'. The work group is moved relative to the work surface such that the cutting tool is first moved from its original position 150 to a first cutting path (dashed line 151). Then, the cutting tool is brought into a cutting position, for example, lowered, and cuts the object along the cutting path (solid line 152).

[0068] Figure 3Figure 1 shows an exemplary embodiment of a cutting machine 1, which, according to the first aspect of the disclosure, has a plurality of reference marks 25 that are fixedly arranged in the field of view of the overview camera 20 and relative to the work surface 10. In the illustrated example, six reference marks 25 are distributed around the perimeter of the work surface. The reference marks 25 can be identified in the images of the overview camera 20, and their positions in the image can be compared with their known, defined positions relative to the work surface 10. The processing unit 30 (shown here integrated into the machine) is thus able to determine, with greater accuracy, the positions of objects 40, 40' on the work surface 10 and the positions of referencing features on the objects 40, 40', respectively, based on the positions of the reference marks 25 in the image.

[0069] It is also possible to verify and, if necessary, correct the correct alignment of the overview camera 20 with respect to the work surface 10 by means of the positions of the reference marks 25 in the image of the overview camera 20.

[0070] Figure 4 Figure 1 shows a cutting machine 1 on whose work surface 10 two objects 40, 40' of different material thicknesses are placed. The first object 40 has a greater material thickness and consists, for example, of multi-layered cardboard or a composite panel. Due to the position of the overview camera 20, distortions occur in the images it captures, which increase towards the edges of the image. However, with a negligible material thickness (e.g., paper), as is the case here with the second object 40', this does not pose a problem for recognizing the objects 40, 40' or their position on the work surface 10, thanks to the flat surface of the work surface 10.

[0071] However, distortions become significant with increasing material thickness and increasing eccentricity in the object's positioning relative to the camera. This is shown in Figure 5 illustrated. This shows the object features 44, 44' of the two objects recognized in the camera image. Figure 4 While the features 44' of the thin paper object 40' are assumed to be in their correct position, the assumed positions of the features 44 of the thick object 40, which, due to the greater material thickness, are located in a plane further away from the work surface 10 than the features 44' of the thin object 40', deviate more from their actual positions with increasing distance from the camera position 21. Thus, the reference marks 42 in the camera image are depicted further away from the image center (dotted circles 49) than they are actually located.

[0072] This can lead, on the one hand, to object 40 not being recognized at all in the image from the overview camera 20, or even being mistaken for a different object and thus cropped incorrectly. On the other hand, it is possible that object 40 is correctly recognized, but because the positions of the reference features are derived incorrectly, an inaccurate or completely incorrect cropping path is calculated. In this case, object 40 is also cropped incorrectly.

[0073] According to the second aspect of the present disclosure, this problem is solved by providing the control unit 30 with information about the material thickness of the object 40 to be cut. The material thickness can, for example, be determined in advance by a camera, queried by a user, or provided as part of the order.

[0074] Using the information about the material thickness, a different distortion in the image of the overview camera 20 can be calculated, thus enabling exact detection and position determination of the object 40 and its registration features.

[0075] Alternatively, the overview camera 20 can be designed to be automatically height-adjustable and moved in the Z-direction depending on the material thickness, so that the distance to the object surface and thus the focus remains constant regardless of the respective material thickness.

[0076] Figure 6Figure 50 shows an image 50 of the work surface 10, taken by the overview camera 20. The work surface 10 is partially in shadow 70. This can be due, for example, to direct sunlight, so that, as in this example, the work group 12 casts a shadow 70 onto the work surface 10. The object 40 to be cut is partially in shadow 70 and partially in the brightly lit area of ​​the work surface 10.

[0077] As a disadvantage, not all contours of the registration features can be captured with sufficient precision in image 50 of the overview camera. According to the third aspect of the invention, the camera therefore captures an HDR image (HDR = High Dynamic Range) of the work surface 10 in order to ensure sufficiently high contrast in both dark and light areas for determining the position of the registration marks 42 in image 50.

[0078] Several methods exist for capturing HDR images. For example, two images taken in quick succession with different exposure times can be superimposed. Alternatively, only one image is captured, with the overview camera designed to select the exposure time for each pixel or for specific pixel areas depending on the brightness of the respective image area.

[0079] Taking multiple images of the same scene can be advantageous – even under uniform lighting – for reducing artifacts and image noise, and thus for more precise contour determination, enabling a more accurate and faster positioning of the registration marks 42 in image 50. For example, pixels in the peripheral areas of the registration marks 42 can be assigned a brightness value averaged from the values ​​of the multiple images.

[0080] According to the fourth, in the Figures 7 and 8In the illustrated aspect of the invention, the cutting machine 1 has, in addition to the overview camera 20, a further camera 60. This second camera is also directed towards the work surface 10. It has a significantly smaller recording area 62 than the overview camera 20, but is arranged to be movable relative to the work surface 10, so that images of the entire work surface 10 can preferably be recorded. The second camera 60 is preferably mounted as a beam camera on the same beam 13 as the working group 12. In particular, it can be designed as part of this working group 12. Figures 7 and 8 They show an exemplary embodiment of the cutting machine 1.

[0081] In Figure 7An exemplary embodiment of the cutting machine 1 is shown, wherein a second camera unit 60 is provided in the working group 12, which is designed to capture images in the direction of the work surface 10. Its image area 62 at each position comprises only a small part of the work surface 10. In this embodiment, the overview camera 20 is also designed to capture images of the entire work surface 10.

[0082] In Figure 8 The work group 12, movably mounted on beam 13, with knife 15 and beam camera 60, is shown from above. The position of the relatively higher overview camera 20 is also shown. The work group 12 is positioned here such that two registration marks 42 of an object 40 to be cut are within the field of view 62 of the beam camera 60.

[0083] A detailed image captured by the bar camera 60 can now be compared with the overall image previously captured by the overview camera 20. This allows the positions of the registration marks 42 to be verified or determined relative to the work surface 10. First, an image is captured with the overview camera 20. Using this image, the relative positions of the registration marks 42 are determined, i.e., their arrangement relative to each other. Then, the bar camera 60 moves to one or more registration marks 42 and determines their position(s) with high accuracy.

[0084] To verify the registration mark positions, the positions determined with the overview camera 20 are compared with the positions determined with the bar camera 60.

[0085] To determine the registration mark positions on the work surface 10, the positions of all registration marks 42 are determined with high accuracy by transforming the positions determined in the image of the overview camera 20 by the positions determined in the image of the bar camera 60.

[0086] According to the fifth aspect of the revelation, such an additional camera 60 can also be used to calibrate the overview camera 20. This is stated in the Figures 9a and 9b The cutting machine 1 features a calibration function controlled by the computing unit 30. Within this functionality, after startup, the bar camera 60 can automatically determine 10 positions of a large number of grid points with high accuracy across the entire work surface. This can be done as shown in Figure 9aThe work surface itself is depicted as a calibration work surface 18, i.e., it itself has the corresponding grid points, or alternatively, as shown in Figure 9b As shown, for calibration a calibration sheet 48 is placed on the work surface 10, which has the grid points.

[0087] The positions of the grid points determined by the bar camera 60 are stored as target positions. Subsequently, the same grid points are recorded by the overview camera 20. Using the target positions and comparing them with the positions of the grid points in the image from the overview camera 20, the overview camera 20 and the bar camera 60 can be calibrated relative to each other. If the bar camera 60 is located in the same workgroup 12 as the cutting tool 15, errors in the drive system of the workgroup 12 can also be advantageously compensated for.

[0088] According to the sixth aspect of the disclosure, i.e., according to the invention, a region of interest (ROI) can be selected even before the overview camera takes the image; this ROI is the only area of ​​interest for determining the position of the registration features. This is described in the Figures 10a and 10b illustrated.

[0089] Cutting jobs are provided with specific additional information that allows the work area 10 to be limited to the ROI (Reference Area). This includes, in particular, the expected positions of the objects 40, 40' to be cut and their dimensions. Either only an image of the selected areas is captured, or only the corresponding areas of the overall image are evaluated. This saves processing and storage capacity and speeds up the process. Furthermore, it prevents printed images from being mistakenly interpreted as registration marks. If the overview camera only captures an image of the ROI, it can also be configured to zoom in on the relevant area, thus achieving a higher resolution.

[0090] In Figure 10a Image 50 of the entire work area 10 is shown, as it was taken by the overview camera (see image 50). Figure 2aBased on information about the expected location of objects 40, 40' on the work surface 10, the processing unit defines areas 52, each encompassing an expected position of the relevant registration marks 42. The search for registration marks 42 is conducted only within these areas 52, ensuring that only positions of registration marks 42 located within these areas 52 are identified. This advantageously saves not only processing power and time but also prevents potential misinterpretations of graphic design features 41, 41' as registration features.

[0091] In Figure 10b This is image 50 from the overview camera. Figure 10aThe process has been limited to two ROI areas and therefore only includes area images 51 and 51'. In each area image 51, 51', an object 40, 40' to be cut is at least partially depicted, so that the registration marks 42 are visible, allowing a cutting path to be generated in each case. Due to the smaller image area that needs to be evaluated, the relative positions of the registration marks 42 can be captured more quickly, thus accelerating the process.

[0092] It is understood that these figures only schematically represent possible embodiments. The various approaches can also be combined with each other as well as with devices or methods of the prior art.

Claims

1. Cutting machine (1) designed for the cutting of objects (40, 40') which have a flat surface, wherein the surface has a graphical design (44, 44') with optical register features, the cutting machine having - a working surface (10), which is designed to receive at least one object (40, 40'), - a first camera unit (20), which is arranged relative to the working surface (10) in such a way that the field of vision of the camera comprises the entire working surface (10), - a working group (12), which is arranged movably above the working surface (10) and has at least one cutting device (15) for cutting the at least one object (40, 40'), - a computing unit (30) with a circuit and program code for controlling the cutting machine (1), which computing unit comprises a memory unit for storing instructions for the cutting of certain objects (40, 40'), wherein the computing unit (30) • has a circuit and program code for analysing images (50) of the first camera unit (20) and is designed to recognise register features of the at least one object (40, 40') in an image (50) of the first camera unit (20), and • is designed to define a cutting path (45, 45') for the cutting device (15) in accordance with at least one stored instruction and on the basis of positions of the register features in the image (50), characterised in that the control unit (30) is designed, - to derive anticipated positions of register features on the basis of the anticipated position of at least one object (40, 40') on the working surface (10), and - to define, in an image (50) of the first camera unit (20), areas (52) around the anticipated positions as region of interest, beyond which no register features are searched for, wherein during operation of the machine an instruction for cutting a specific object (40, 40') comprises information regarding an anticipated position of the object on the working surface.

2. Cutting machine (1) according to claim 1, characterised in that the first camera unit (20) is designed to show only one partial area or a plurality of partial areas (51, 51') comprising the areas (52) defined as region of interest.

3. Cutting machine (1) according to claim 2, characterised in that the first camera unit (20) has a zoom function and is designed to zoom in on a partial area (51, 51').

4. Cutting machine (1) according to claim 2 or 3, characterised in that the control unit (30) is designed to analyse, in a partial area (51, 51'), only the at least one area (52) defined as region of interest.

5. Cutting machine (1) according to any one of the preceding claims, characterised in that during operation of the machine the register features are present in the form of register marks (42) which are designed specifically for use with the cutting machine (1) so as to make a position and orientation of the object relative to the working surface (10) detectable, wherein the computing unit (30) is designed - to recognise the register marks (42) on the surface of the at least one object (40, 44') in an image (50) of the first camera unit (20) and - to define the cutting path (45, 45') also on the basis of the positions of the register marks (42).

6. Cutting machine (1) according to any one of the preceding claims, characterised in that the control unit is configured so that during operation the register marks comprise edges of the object (40, 40').

7. Computer program product with program code, which is stored on a machine-readable carrier, for controlling the cutting machine (1) according to any one of the preceding claims, wherein, if the program is executed on the computing unit (30) of the cutting machine (1), at least the following steps are initiated: - associating the at least one object (40, 40') with at least one stored instruction for cutting, which instruction comprises information regarding an anticipated position of the object on the working surface; - recording an image (50) of the working surface using the first camera unit (20) of the cutting machine (1); - recognising register features of the at least one object (40, 40') in the image (50), wherein anticipated positions of register features are derived on the basis of the anticipated position of at least one object (40, 40') and wherein in the image (50) of the first camera unit (20) areas (52) around the anticipated positions are defined as region of interest, beyond which no register features are searched for; - defining at least one cutting path (45, 45') on the basis of the instruction and positions of the register features (42) in the image (50); and - controlling the cutting device (15) of the cutting machine (1) for cutting the at least one object (40, 40') along the at least one cutting path (45, 45').